<%BANNER%>

Negation Comprehension and Aging

Permanent Link: http://ufdc.ufl.edu/UFE0009342/00001

Material Information

Title: Negation Comprehension and Aging The Role of Inhibition and the Impact of Compensation during Reading
Physical Description: 1 online resource (89 p.)
Language: english
Creator: Margolin, Sara Jill
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: aging, comprehension, inhibition, negation, reading
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present research investigated the influence of several factors on young and older adults' reading comprehension of real-world texts. The first experiment determined the effects of negation on comprehension of written sentences and examined the level of text representation at which negation most severely impacts comprehension. The second experiment assessed the effectiveness of a specific compensatory strategy, the use of a negation symbol, in comprehending negation. Using online, activation, and content measures of comprehension, participants read sentences modified from fiction novels. Participants read sentences one word at a time at their own pace, then immediately named a probe word that was verbatim from the sentence, related to the sentence, or unrelated. Participants then answered a comprehension question about the sentence they had read. The results showed that negation affected comprehension in both age groups, such that sentence reading times were faster for negative sentences, probe naming times were slower following negative sentences, and comprehension accuracy was lower for negative sentences, relative to non-negative sentences. Negation equivalently slowed verbatim and related probes' naming times, suggesting that negation had similar effects on the situation model and surface levels of representation. However, these effects of negation were not exacerbated with age, although older adults' comprehension overall was poorer than young adults. With respect to compensation, providing the negation symbol during reading did not facilitate the processing of negation specifically but instead encouraged readers to focus more attention on all of the sentences. These findings suggest that age and inhibition are less important in influencing negation comprehension than working memory. Furthermore, the lack of age deficits in negation comprehension has positive implications for older adults' functioning in a world where negation is present in many contexts, such as prescription labels and road signs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sara Jill Margolin.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Abrams, Lise.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0009342:00001

Permanent Link: http://ufdc.ufl.edu/UFE0009342/00001

Material Information

Title: Negation Comprehension and Aging The Role of Inhibition and the Impact of Compensation during Reading
Physical Description: 1 online resource (89 p.)
Language: english
Creator: Margolin, Sara Jill
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: aging, comprehension, inhibition, negation, reading
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present research investigated the influence of several factors on young and older adults' reading comprehension of real-world texts. The first experiment determined the effects of negation on comprehension of written sentences and examined the level of text representation at which negation most severely impacts comprehension. The second experiment assessed the effectiveness of a specific compensatory strategy, the use of a negation symbol, in comprehending negation. Using online, activation, and content measures of comprehension, participants read sentences modified from fiction novels. Participants read sentences one word at a time at their own pace, then immediately named a probe word that was verbatim from the sentence, related to the sentence, or unrelated. Participants then answered a comprehension question about the sentence they had read. The results showed that negation affected comprehension in both age groups, such that sentence reading times were faster for negative sentences, probe naming times were slower following negative sentences, and comprehension accuracy was lower for negative sentences, relative to non-negative sentences. Negation equivalently slowed verbatim and related probes' naming times, suggesting that negation had similar effects on the situation model and surface levels of representation. However, these effects of negation were not exacerbated with age, although older adults' comprehension overall was poorer than young adults. With respect to compensation, providing the negation symbol during reading did not facilitate the processing of negation specifically but instead encouraged readers to focus more attention on all of the sentences. These findings suggest that age and inhibition are less important in influencing negation comprehension than working memory. Furthermore, the lack of age deficits in negation comprehension has positive implications for older adults' functioning in a world where negation is present in many contexts, such as prescription labels and road signs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sara Jill Margolin.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Abrams, Lise.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0009342:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110320_AAAADE INGEST_TIME 2011-03-20T19:06:02Z PACKAGE UFE0009342_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 684 DFID F20110320_AABVQM ORIGIN DEPOSITOR PATH margolin_s_Page_08.txt GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
2af4bfc5b2333102966777d14824c10a
SHA-1
06aa543f36cdc436b5cf36cb91e9911f7a47dd27
25610 F20110320_AABVPY margolin_s_Page_72thm.jpg
58e3b75949308d819508a34b83b4b121
c78e1022528693e656597541b4d2f6cfde55e335
40478 F20110320_AABVRB margolin_s_Page_47thm.jpg
af1c14c2fc58fff0585798a3a3d38265
9beee666af0fab79255214ca18297dd97c121f92
57012 F20110320_AABVQN margolin_s_Page_19.pro
02aa82d1a7e66c798d3ab5014bf42ba9
5e0708822940bf603fffe13970dd94d5600cf260
26756 F20110320_AABVPZ margolin_s_Page_15thm.jpg
40c2a0135e9f95ea444b7b5d56d93bd3
dc29f4ecea6ce266d51db62f9c4a5c1c3dd202c6
27217 F20110320_AABVRC margolin_s_Page_16thm.jpg
4133afcdb148fd730e97394f919155f5
90fa41db1f30c821284e555fbbff9ffdd6a1d670
156229 F20110320_AABVQO margolin_s_Page_45.jpg
6b6242cfd6734c06cf145603a0c163f8
24b3adbe864e08c991585288afe6289cff3c5396
25810 F20110320_AABVQP margolin_s_Page_80thm.jpg
4d86284f992d5c035572b46d591030cc
2340c527353fd58e02b685fa6b573794ac6891e6
1053954 F20110320_AABVRD margolin_s_Page_75.tif
4840f89b9b73ea3931813923c0b80de2
fe0200654157cc12ebaf753b63719e6a9fedd588
5047 F20110320_AABVQQ margolin_s_Page_02.jp2
1b2880aae11450eb5c2f5c2179a35f5d
479e81fe637eeeb216bb3b85b51bd9e222cf540b
6516 F20110320_AABVRE margolin_s_Page_01thm.jpg
69bddda3710e3aa28ffdfcbc45778785
3159a06a2699d88b1359d2c0a31267a41cf6e3ec
F20110320_AABVQR margolin_s_Page_76.tif
485c02e5174773871999925870c46576
84e36937968189515ed904299843e587dbfc6d64
50572 F20110320_AABVRF margolin_s_Page_35.pro
29ec488a113b30394d400936ce56c7d0
4936ef4548398a7f305623549e05da81045f850e
55434 F20110320_AABVQS margolin_s_Page_16.pro
08db9dd3892d459d3be9354dc4b1fb24
438001b7ac03ed5c31d7c54e85a4644bf354e767
48942 F20110320_AABVRG margolin_s_Page_51.pro
1302a4351ed5d5ce83814a903f7123ef
be524f499e09b65c6d6b3ce3f1f34bcc2349e488
52745 F20110320_AABVQT margolin_s_Page_71.pro
de55898cc1a40e5631eba49df604e6c6
818d2d01aace64740ed1903b7112256209bd54bf
56986 F20110320_AABVRH margolin_s_Page_42.pro
63ea744edfed257ca9f5b09595448652
50b3a9385d1a802671a12cb19867c83bce505f56
58060 F20110320_AABVQU margolin_s_Page_30.pro
f6d3a04a6ead83c9076dc865386b38cc
643ef3ed784649342fff52223267c00cdef0be72
123944 F20110320_AABVRI margolin_s_Page_30.jp2
550981d9b5948ef462d6a49584617d7b
f80bb3cf4444dd008672d3fb1eb81610aa47e984
53550 F20110320_AABVQV margolin_s_Page_72.pro
dae28df47aee5a6950d9cb54a2cb2a25
485b40b8eea3c488b4fd32155d4e9480dfb58572
36132 F20110320_AABVRJ margolin_s_Page_67.pro
06b65842bd07d63da68ebe729cf995f4
3af0673e49eb4856aed2b60b6f1d3293acb65ebc
84112 F20110320_AABVQW margolin_s_Page_72.QC.jpg
80ee8e78d3e3f4d7d2cd86c66b61bfca
38b25c769e47717a16ec5bd7e396c536450c877f
88807 F20110320_AABVRK margolin_s_Page_39.QC.jpg
414928a7c3b5474df5e58e0ca9f16548
c212f98a35b393afab9c607acf128ec272fe2d64
F20110320_AABVQX margolin_s_Page_25.tif
b0853c4141fe1874f271d319e1927256
264842ec3a23b99079b151276a8b4f4a9e2f4c0b
85645 F20110320_AABVSA margolin_s_Page_73.QC.jpg
901647bec966c91a6d35c47469d00f40
7ae6c1d4dbb08166151f8a953f67322162cbb58f
26598 F20110320_AABVRL margolin_s_Page_33thm.jpg
32105385cec7f38247b72228bafe2009
65835c79647c22868ad3741527c85e1c9bc83dde
47668 F20110320_AABVQY margolin_s_Page_08.QC.jpg
9c18a8cc5a338b527cce8005acacd9e7
c340a158f32454f1ffdc717d1f21a0ea4729058f
86419 F20110320_AABVSB margolin_s_Page_19.QC.jpg
b4574ad983cce1a388f9015768cbce08
3868bdb2ea501bb34d721ede087cbf248edfaa25
2179 F20110320_AABVRM margolin_s_Page_15.txt
39b7ea9b6c431e8c615a45324aa51053
31df537e25abd9ad033bca4629a71c7e4258036c
224643 F20110320_AABVQZ margolin_s_Page_24.jpg
c1895d985ad0c8afc46c2c9af7890b5c
b36b69750c90645ee52809f4577b16f817905b67
24575 F20110320_AABVSC margolin_s_Page_26thm.jpg
6492898dc9f92800075fb752f651f93a
ba67cbda3755ac25eaf60deb87050b7b51397c5a
26999 F20110320_AABVRN margolin_s_Page_40thm.jpg
9d2fd798e205e889b8307127f347bc83
76b5a978cc98cd545a6681bdd64ab3b34daa0e57
26570 F20110320_AABVSD margolin_s_Page_63thm.jpg
e2c8c08b3122b0e5d74089a4517df988
4d57713d8d3552115e910d0d142b11e1e91fdd5e
81510 F20110320_AABVRO margolin_s_Page_83.QC.jpg
9bfd7644e08f9a99beb51d78df681fae
7171adbc9711074dfbcb556b50abddbb858f4310
26057 F20110320_AABVRP margolin_s_Page_36thm.jpg
9c6e8bcd2f7ff133b0dee2b770052afa
a2f018b7cd1803fd40d8f7ab453af98852658e81
2107 F20110320_AABVSE margolin_s_Page_72.txt
51ca0dd341e9e5ae4da98cd893644e70
c2c8122b73a9b378464afdea4266ddaabdd7334e
26349 F20110320_AABVRQ margolin_s_Page_01.jp2
b9685f8ef24300bbf528e76cac3efcb5
a92bcbffafd686524ea609faf6babdc2acd02040
2389 F20110320_AABVSF margolin_s_Page_66.txt
ba31828dad6b695311fcff73bd6dca84
cbf02fc4e14d0d393b472572b4e257045f14b703
112854 F20110320_AABVRR margolin_s_Page_78.jp2
1cde52fc47cb29653a3ff8905e2224ae
681b369e407e1d03478af5c9bf08725b2b95fa88
71004 F20110320_AABVSG margolin_s_Page_04.jpg
14949a1a3450600e9a9036b47151a80a
7d1c3068d0130e165e26b7f39c49f360c09cb429
79633 F20110320_AABVRS margolin_s_Page_06.QC.jpg
68bb24ed0270a761e5a3adfb5e3bc534
c0661c611f18ffe7d45eed467955a7dc12d807ab
2578 F20110320_AABVSH margolin_s_Page_87.txt
f936c403f74ca4a7b3804f03b6b22d1c
51f37fae2e11c007647e8ad5a742ec04362c621a
56971 F20110320_AABVRT margolin_s_Page_60.pro
507bcd8d1dc7ca04eef18e1817566f9c
9cd849211183e0cdbea84ea942168ef18acfd6a0
208281 F20110320_AABVSI margolin_s_Page_52.jpg
5d8fe9e6534f1ac0ba23b1fabb64cc36
490bcc3917297bc913dbb72218a6eacccb8865ef
58333 F20110320_AABVRU margolin_s_Page_22.jp2
3c96d5b57015ba4ef062713d29ff5ece
0f129ea2572bbd0669f9825020f25e024b9dd93a
58669 F20110320_AABVSJ margolin_s_Page_44.QC.jpg
7db216a1ceaea0e86c804aea666a9827
97686374cc491e4d4a477baabaf4a8349ae920d9
84026 F20110320_AABVRV margolin_s_Page_71.QC.jpg
6ddf785e886e4de84f5ab5e38baa9721
28ccc53f48943342256d894bc1a2c5d670f84699
F20110320_AABVSK margolin_s_Page_66.tif
03d7c922faee70e69143046475ac92bd
f245a9d00969af7083885a903590ac1138f287b9
120972 F20110320_AABVRW margolin_s_Page_74.jp2
409eabea7394fc0b1759009b14856226
39f2eb6b30799b580fa348acd7918608bfd336b5
74465 F20110320_AABVTA margolin_s_Page_26.QC.jpg
1aa01bd240fb03d6077e1c626a4d0225
a9e01a398f32305daafd4455edd871809c6bc65c
2211 F20110320_AABVSL margolin_s_Page_31.txt
8d9d67bc3790215800ae61fbc3939603
aa100e023972ca08996b61c446c023bcf742fbca
85474 F20110320_AABVRX margolin_s_Page_36.QC.jpg
d6fdb71bd536dbffe7fc8f2ba96d9e66
81d8a19bd42180922256d79949dea8cd3230b571
224763 F20110320_AABVTB margolin_s_Page_63.jpg
e12fcc08c8da596e5d004c3ebf95f9ef
cbb224cc576569713e23b5ba1bbbd9a26c155b32
55126 F20110320_AABVSM margolin_s_Page_15.pro
12b7c89fc109c1ea08b538efbe298390
8ddcde0ee23a66779960b97d6c55880a48986666
57994 F20110320_AABVRY margolin_s_Page_38.pro
f52d495680ad43a16fd3502b7e9d7e68
0baa9ac3f5711ba8a15f1c98c7f2fbba9e8629da
33354 F20110320_AABVTC margolin_s_Page_03.jpg
1bb6d5bcd6b6ee082e8a8d2263bc7afd
43495b9c54ff39e765b647f0b5a78cc458f4208f
121751 F20110320_AABVSN margolin_s_Page_59.jp2
350b5fd1652d789f5eaa5c12c9b900f1
a3047a7b92a002f69262ad51624e3ccde3a30c58
F20110320_AABVRZ margolin_s_Page_56.tif
b18879aeb2302f8c10187198276b00e2
ba93a7f0de0fb79f32ace4e5a4cd7816ea821b3b
F20110320_AABVTD margolin_s_Page_21.tif
710b277c00bee6e43fef81fb54dd64a9
8e10165b0612a0a73850e6514bde08b8fb8e97f4
2061 F20110320_AABVSO margolin_s_Page_33.txt
0b2b472b97e2d6d21ecd41907f155613
16a4ee608eff78a34c2e66396e6121ce96500846
70011 F20110320_AABVTE margolin_s_Page_89.jp2
b68854f7ab014d6d01cbaa24815c4ca7
62a60fafef7f2c3c010571ab899d77a24b9e3703
F20110320_AABVSP margolin_s_Page_65.tif
a7035f4cdf98bc50634d3d5815c63f17
999b91ee60843820b490e179858a34352770da4a
231973 F20110320_AABVSQ margolin_s_Page_59.jpg
891bbe3904652c80622d0160ed675a23
1ef89b9b08d57a59373c7b7a114a63adee62765e
26439 F20110320_AABVTF margolin_s_Page_48thm.jpg
fc154dbc264a8695a23ec72ce08daa0f
a41952befad87bc13f88726033ad146c6bf29f7b
88064 F20110320_AABVSR margolin_s_Page_16.QC.jpg
104bcaf6991212f16c8cfd99da1a5362
bd667caf9cc03d5871d9dd5eefef7043dc9162c8
2161 F20110320_AABVTG margolin_s_Page_17.txt
82755a06b83de900d51263c881ff190a
16a56b23a0d9d56f0fe5d7c7e43e754e0386c898
2477 F20110320_AABVSS margolin_s_Page_86.txt
fd33faabde80bb120762cf295563809e
544f9dc3a4b7a4e3d2098aa3c0f73bde96286862
225775 F20110320_AABVTH margolin_s_Page_32.jpg
70ace4b9e0b863d6b19c9cceabcdf0a8
1b5e60e791c74c9f158b39e0a400c3745f62c5e4
F20110320_AABVST margolin_s_Page_12.tif
d3bcbe5a4d1243702d5ef0e03c9cf4ce
0dc0eaeb5015689afa608a5dea17c47d9c24b7fd
2187 F20110320_AABVTI margolin_s_Page_13.txt
f67b004039e64f0b10aa0df0fcce4583
3c0d7ce5b33c9bccfcd6687291716b402c58e3e5
33239 F20110320_AABVSU margolin_s_Page_08thm.jpg
b1c9c93598764d71d80b910ef1c6dd1d
dc31292cf2141c9af93a1b2dc1699d08fddb94c8
50293 F20110320_AABVTJ margolin_s_Page_52.pro
c1cbd23ca3b8d6cfd04f50c5b6bd1aa4
d7509c72f4385fc24eec68675f0f62986d9ca037
27084 F20110320_AABVSV margolin_s_Page_61thm.jpg
c1cebc47bb64e2f1734436df779112d8
eec6b48dfb220a001b367614e132a10cf2cd4fa0
F20110320_AABVTK margolin_s_Page_14.tif
e82d93805269f3662e2d24726ee0cad8
1b30866c4ab5173618071f28fe34ee14aa18e0b6
2086 F20110320_AABVSW margolin_s_Page_57.txt
354bb8a0b5b4cc7b742dc48e7e41c248
ffe165b9eaf4e42939e5d8e5dbd0785c28a25296
228346 F20110320_AABVTL margolin_s_Page_17.jpg
e7a8c3e19945c673de93685140af7b61
09373777c21a75a089c72771550c39463404f113
50001 F20110320_AABVSX margolin_s_Page_66.QC.jpg
204ab2e375e8850cd8b24b8e70ef1ab3
7c6f347f856490e82c98b463cec85edbf307c074
231210 F20110320_AABVUA margolin_s_Page_36.jpg
32df50b85285b7a240f31b6e04c7e068
dce90842b7fba5161ea2258e40afa0d4f8c9db7d
593117 F20110320_AABVTM margolin_s_Page_47.jp2
6a7fb00e235da49adf53fa6b76351e3a
e9c58d563b51dbfa6a193fa0755acd9ddef82890
69775 F20110320_AABVSY margolin_s_Page_68.jp2
ffe85d33c8cce6af609156e6022b5386
b5b2fb2bdf1a7edd63ef4c4aa10b517930160039
25271604 F20110320_AABVUB margolin_s_Page_70.tif
35bf548a69124a810975ea66041baf1b
26720c9c2c416ae0b95add5afaa06b788b3415e2
F20110320_AABVTN margolin_s_Page_51.tif
1b71fd8bb245c6fb8489fcc82488e89f
8eeaafc02ebba821973012cbccf5697e2dd3af8d
54283 F20110320_AABVSZ margolin_s_Page_49.pro
200826e0b53140a28e8730c36760d05f
b20107ac05afecc530bea54149f9bce0802c8f7d
75787 F20110320_AABVUC margolin_s_Page_84.pro
cfe5e88c8ccfd8cff9a985878cc3ba62
38b3ac4166022f4d89f38ce2972f2a641463ef16
564759 F20110320_AABVTO margolin_s_Page_08.jp2
2f921bd1acc8a45ff6a17829e8263b8e
15d2ba184328ae15a6c87fe4ed86d9cf6c017d00
F20110320_AABVUD margolin_s_Page_42.tif
30a59550b5a33b99cf0c9e364954a04d
25a5a33d5bbf1a19fd165e53fe3a4e8b1e7caa47
F20110320_AABVTP margolin_s_Page_64.tif
ac9d01a0ddf2a3f83b476428480cd707
bea6b9618b6084dc861e41cdfec934d8ed63d375
F20110320_AABVUE margolin_s_Page_59.tif
67b760e48210e6841d354716b7817d06
8dfd60176caf3de0cfd4249c57e5e08b2b60e09d
131392 F20110320_AABVTQ margolin_s_Page_85.jp2
ea132394073706c4ccf43cfbe65cc3bd
d3cd58eea232d1e9b4e600b2868141fd9a8d9c22
10576 F20110320_AABVUF margolin_s_Page_03.QC.jpg
3841d708d4c26ed4adeeb24b43463e8b
02738e4df65f18e224f54a5b4150c77316735ad6
55858 F20110320_AABWAA margolin_s_Page_14.pro
402c2494d80f6807cdfc6bb3dc12c695
81f9b673a6d2eccb67c826cd2676ddfa5dbf6a71
2608 F20110320_AABVTR margolin_s_Page_06.txt
7ba709a76fd06f4ba4dc5b0f3c266397
3bfeec67006d176092513e180c50056d06bc3f99
54442 F20110320_AABWAB margolin_s_Page_17.pro
faea2dd8c2578acef92b94add3358254
0b4742932b766b96f43df6fa7a261c5c0cf8c9fc
115891 F20110320_AABVUG margolin_s_Page_80.jp2
e88352737e81ec903e0de92833583bef
5e2edd9fe04771528e949e786766a0f414eda627
87239 F20110320_AABVTS margolin_s_Page_40.QC.jpg
b0556fc54760836da0ecaf0c4ed978e3
8746d71dd9c0075dad063e0036ef7ca634f3c0c6
54566 F20110320_AABWAC margolin_s_Page_18.pro
411e854dde7fbed6e56c7e7eab14d7ae
f2359f69344d7bc11c64ea6919832a8b9ec3234b
14392 F20110320_AABVUH margolin_s_Page_04.pro
ee2d3f48c5ea1caddd73666488d7db95
8ad125bfc4d646d084655a1567fa6b7147ce3be6
27248 F20110320_AABVTT margolin_s_Page_20thm.jpg
bba26e909e2656275742938436df1d2f
9b7f91f63500d217a9c7968fe48b16402f3822b2
55285 F20110320_AABWAD margolin_s_Page_20.pro
74864edf20bc4f63fd4e3485858ed3e5
e2ea62450d05ca18eaa1b8b8a93e5dcd721571e3
27382 F20110320_AABVUI margolin_s_Page_28thm.jpg
8e1295e3b27b119074673a6474d347d3
2ccca6b99f50a36937a55f901952dcc05fd074e7
2029 F20110320_AABVTU margolin_s_Page_56.txt
b52ea31f646b46a026eaf3fd308493c6
09e0763900d8d188874d60aa8954f8eacc58078f
56941 F20110320_AABWAE margolin_s_Page_21.pro
6c59267b4aaed7e9dac940892d2212c2
c2d7fb4922c4bfd36d7c864950c7716608e290a7
236875 F20110320_AABVUJ margolin_s_Page_16.jpg
1452e6b08e111078f9d5d72f3d542beb
ed72cbceeb42af4cf6c1613ab98014f8afb2fd41
107515 F20110320_AABVTV margolin_s_Page_05.QC.jpg
0c5f1b62f5473f6fb344e4fdd3f85fb4
4d30bd8ec2728f8062665d859f7ae6ff8e6ad7d4
25607 F20110320_AABWAF margolin_s_Page_22.pro
b1fab5afcf0ef31f2ed73f0d85b13ba4
60804f3bca725697c99353fef0f3549a9abedeef
26588 F20110320_AABVUK margolin_s_Page_58thm.jpg
3978677ea6d8c3635322a22557d45416
b5f4e7fb74afb4501e1303c4a6029fcc82b98e8d
2165 F20110320_AABVTW margolin_s_Page_61.txt
df63a80bcba5ae47c51e5c121dc8eecd
7ceda186d407c1b1ee603c42a7875289cadafd4d
53889 F20110320_AABWAG margolin_s_Page_23.pro
e6e82ce0aef09207bbd305f92bb7bf96
6fa9d4d1e7167f46eaa3d0a2b0cd871411bcc349
F20110320_AABVVA margolin_s_Page_03.tif
88048651273de2f0b6a3fd2897b2d18c
fe4fa4bede6f415e0f36574d9fa571a7adf53dbe
56818 F20110320_AABVUL margolin_s_Page_40.pro
8706c72df50a5e9d95ce14238b17280d
fa97a5efb4bf9110fdc7f4a147a590f059616bbd
213484 F20110320_AABVTX margolin_s_Page_34.jpg
40b50d983028e352d65f0d287cc5e275
9354850cffdd7a50325687ca89b087a9a8f2d301
54155 F20110320_AABWAH margolin_s_Page_24.pro
a4050cc746f59b470dfc1f92e98e3f5f
4c2277633bed6fe3a144b77b4b8aaef9f94463ca
F20110320_AABVVB margolin_s_Page_04.tif
e61463776d5472c1f033bc93efa8231e
16dc7fb1b41ad6c3232941ab7a7d07665e4b55d2
36824 F20110320_AABVUM margolin_s_Page_11.jp2
46aca88edf2cdc87f4ae01c6f1acc34c
15c6e231c9fa854ad85cdcb17517354788c31960
1051979 F20110320_AABVTY margolin_s_Page_07.jp2
69d699b6d5c460403fde43dccc21858b
57b9c7b37ef196c2e0c9eb96a161dfef37a72b32
48458 F20110320_AABWAI margolin_s_Page_26.pro
54e21fa695573de52a031de7cfa8c220
b3d9b6040528a5e1fe97a7bdbe086e59d15391f1
F20110320_AABVVC margolin_s_Page_05.tif
d60a3bbced7c9abe369eb58206ca04dc
69a938f0fbb39e700898bc2d773d9d5ea6f6d51b
F20110320_AABVUN margolin_s_Page_36.tif
44e235cebbbba3f7df346d4f9bce9535
834202fa05a1e04fb8ed91dc1e93823d0aec7fb3
252525 F20110320_AABVTZ margolin_s_Page_70.jp2
c89a594cb59ac8d7e1f2345510bd0fc5
1af5c71ab27f80758c22476e2548fa019ea39c9f
56575 F20110320_AABWAJ margolin_s_Page_27.pro
25a2dd98c0a3ac0f08996eb905e752d9
cf6b2ab31339300c4e737edd02c087be0f57ac62
F20110320_AABVVD margolin_s_Page_06.tif
64c41658d56c93b138b1b1d9332734fe
9faf9d1da2bbc71413bc9d2b10fd7755be360565
225553 F20110320_AABVUO margolin_s_Page_49.jpg
87d6c9dd77b501a07013a0f75d21e545
0f3df8d9b4b579ba0170d324d51854d9359dfef6
56249 F20110320_AABWAK margolin_s_Page_28.pro
b97b63818a75c025deb7838af34bd31f
c856c9a430f2c9d63406a2c8525dc840dc3d4d8d
F20110320_AABVVE margolin_s_Page_08.tif
16e591ed81fd384c8d08dd8f2083b0a9
c878869ee4a917ac8b7430051367c4afeff8a7f6
F20110320_AABVUP margolin_s_Page_20.tif
bd5915d74dafb9c030e02e2322a42b9d
c7e3c1eded337ed41f6479b91ff9fc9c622607fb
58063 F20110320_AABWAL margolin_s_Page_29.pro
8d81ce8624bce82923bde7ff7d6c6a6c
5f14ebf58fcabbd05edf5b8c356447dfe93f092f
F20110320_AABVVF margolin_s_Page_09.tif
85e1b107a6bfd3333da85903b6535666
467d39f6347f51f019ea65e4cdc28a1e56501c72
4971 F20110320_AABVUQ margolin_s_Page_02.QC.jpg
8d00944087864f1f993e7a5aaf9fc7a7
a5ffe5d33970b49544300330ee09d1de69a11f7b
55346 F20110320_AABWAM margolin_s_Page_31.pro
04c45ebb9c4c7c1ab01e2f1a1357ce1f
fc20a018a5d3e2a8903f740c9853cd2bfebbd1cd
F20110320_AABVVG margolin_s_Page_10.tif
0dc6f829ecd1c917d63f41c2d82d4f4a
a5203376b519313a5a71bf7e9f9979fbd22187b4
36834 F20110320_AABVUR margolin_s_Page_69.QC.jpg
c37b54ad5cb893f8d947d9044d7d232f
4a6b93f7b8f03b5b7511594f60e43ba14c67f808
53012 F20110320_AABWBA margolin_s_Page_57.pro
9653891641452c1967dd70ae9347a11d
27981cc935510f36ff75b26f417c0df14a3a93d8
52493 F20110320_AABWAN margolin_s_Page_33.pro
38f13e5bfb06f9472d7787199b12921f
076d9a4a50aba6f7fce34339b4e0fc43238dc1c0
26724 F20110320_AABVUS margolin_s_Page_39thm.jpg
3298ee6de17c6b96695f71a4c9bca094
07035441807b5b443705ae110c3e9fda3b747e13
54253 F20110320_AABWBB margolin_s_Page_58.pro
f3df6ef6ed7dc16a969dd9b9d647299e
92ff33f547a7aee546a0ccab7570618f870e9b12
54959 F20110320_AABWAO margolin_s_Page_36.pro
0bbbe9edba94ec487c53d5331cfcddbf
9b51cfe1ec8a591f604a30fa0f6d5690b3bc15db
F20110320_AABVVH margolin_s_Page_11.tif
0bdfe77419fa37dbfa4fd05399d061e3
7b132fa1733e5d81bbf01bf2b95fc0cdb3c22e00
81240 F20110320_AABVUT margolin_s_Page_34.QC.jpg
f8fe30335290f3ad682829b43d1f22c1
81513912c63c93b59f2b83c96f5204f04552e919
56072 F20110320_AABWBC margolin_s_Page_59.pro
a430925b6277225072c86edcada6fd60
fa2a3212002b3d078cb0c1c00a256d08ad72f01f
F20110320_AABVVI margolin_s_Page_13.tif
cf8fbbe3da76810be01d75d38040e4d5
0407e09e879a8af2053ea87315ef7313f03271b2
2172 F20110320_AABVUU margolin_s_Page_20.txt
95d72de210e7823107abe675e4cfd698
22e821dbb0f34d347bd01dfad4f2afed51e2e784
55305 F20110320_AABWBD margolin_s_Page_61.pro
7d07d826045318cf79050cfea13133b5
f0239a8a51c1f97a8ba0379eb65cbe22c503e85b
55884 F20110320_AABWAP margolin_s_Page_37.pro
516bc1bf899ebc3a156a5beae99b52e9
ad5dd5a89bffe3e5cb8e583c145f1f9375b892f9
F20110320_AABVVJ margolin_s_Page_15.tif
165e6965a8fd422af331acbe0c57fd51
7625686948fec4786e46c1e3483f3a2660c69d8e
145612 F20110320_AABVUV UFE0009342_00001.xml FULL
cf19fcc460b973c9a8778c7fd88d3cc5
33539ac256ca112c3e493db425383b30cf177463
54048 F20110320_AABWBE margolin_s_Page_62.pro
ac3638fe3d05aba2e43253594a75fecb
8390ad527f03676ed31ddb09900bd69a4f9a174c
55693 F20110320_AABWAQ margolin_s_Page_39.pro
8ddc8e52dc7e6225b1a2765f402d1414
5aaa9555a5792b7e373a7d2831ee76a885806903
F20110320_AABVVK margolin_s_Page_16.tif
41bd4dacfe64cbd33cc66afab2e6b724
fcf96070aca6a69fc6714d6cfb1ca26e572862bb
54073 F20110320_AABWBF margolin_s_Page_63.pro
2c12b501e9f920d01aef8f97c9f53abe
b35e20afb400342e41554f507bf4df910e202127
F20110320_AABVVL margolin_s_Page_17.tif
4d0723c3f4a1fb2caf446ab63657502c
4ed7a10ddbba87d297bb807542e10c347936ad35
14715 F20110320_AABWBG margolin_s_Page_64.pro
0e27d517a4a5a6555678fb3dc14905ab
34bd529d470602b18c7ea2b41b0b75914fa23280
F20110320_AABVWA margolin_s_Page_39.tif
a68c7fb0e9d1cb14ea7dc13f0383688a
b34b1bb32538f3fb07f94c6666b4ea5fade6f821
54554 F20110320_AABWAR margolin_s_Page_41.pro
7502dabb94dc90fbf2e45a9a3eb6a2e3
b7ebf645d3a7e4f5f2ba50d365f81357a3a8183e
F20110320_AABVVM margolin_s_Page_18.tif
2d0196f4e0bccc842ca3ca6b8756a73f
749d4ac6224ba32996490cdbf45038d40bcadb0a
F20110320_AABVUY margolin_s_Page_01.tif
c196f1e01a8a93ed2c138cd84c811d33
7be7d0c8424af7e13972a208dde22c43c0bfc95e
38716 F20110320_AABWBH margolin_s_Page_65.pro
9d60fa8057d879127c52eceb13526c19
bf319b49957770d1d10244c734588b451157093f
F20110320_AABVWB margolin_s_Page_40.tif
960163eb6ab6118fb0fac34372474f16
97b89bae363a10338395987d63b04dd337d1a6a9
43485 F20110320_AABWAS margolin_s_Page_43.pro
04deb3dda432d112719a0423892cba65
334930f527c8cb9c7f6f155fef1ff10d49449aec
F20110320_AABVVN margolin_s_Page_22.tif
40ee1f903b5644998f0c6269d6b7079c
b867d9ab6bf30af0753b654a9bf7e9c1264827c7
F20110320_AABVUZ margolin_s_Page_02.tif
4b8b843b6306e156578be9391018cf73
a3b60ff3d1c5eccb8938bd7e69ee138e3981cea5
38313 F20110320_AABWBI margolin_s_Page_66.pro
d20b4db31c7a4041f9faa993300a44c1
09d399f6121cca4c2b1119a3f3bf6aa59f0e6a8a
F20110320_AABVWC margolin_s_Page_41.tif
6ced7b8eb325fc346835d4fd6ac1e8d8
9355d77e51386450d5ebe8e21f64317a1108e04b
47927 F20110320_AABWAT margolin_s_Page_44.pro
304102b863f7bfa1cccdb62ce05cd387
5830b739a647e6ea433f7659dd9f91949e3b20b7
F20110320_AABVVO margolin_s_Page_23.tif
c7fb526dd42c333779857b7a981dd7df
ae6f1a5b2715b5b4dba583c570899c8d34856c62
33721 F20110320_AABWBJ margolin_s_Page_68.pro
94a8e7bb62f3e266bf86ad5dce3f8578
bc04d22711ee2b81ae773132780defa94b9e2e19
F20110320_AABVWD margolin_s_Page_43.tif
2a73846ad41c631615b9181b2d3fa3bc
9fa576e030b9dca0e01f7f9d15c5a1b7878d600e
42700 F20110320_AABWAU margolin_s_Page_45.pro
bc555f411792a52489768be76f22136d
7c89c8112a5ee821babf311dae8e37f54b72d88c
F20110320_AABVVP margolin_s_Page_24.tif
649e676fb01d0dce0df39ecdf3e621ee
9442e9aeddf04f2ac26a1445fc6599f8af487771
31227 F20110320_AABWBK margolin_s_Page_69.pro
1e16fe5bb584d0e9fa3d7f4eaa3c9ed7
3d8ee92d8768edcbc040e8e8b9460ecc0e356fec
F20110320_AABVWE margolin_s_Page_46.tif
d2510dea12bd6d9cd88895c2e77b7ba2
44bbcb1b8025e944b7c0444738e799c1f5142271
43997 F20110320_AABWAV margolin_s_Page_46.pro
3c25ae34f304bab15c72d551ec71fab9
f86c2a3d7a84b24c9f2d475bf2356fd4238f146e
F20110320_AABVVQ margolin_s_Page_26.tif
50d050e6ca5582f6af6064d3d5a273af
4b272f1336c92e0c38e5bc145a75c029f0916ca3
55496 F20110320_AABWBL margolin_s_Page_73.pro
f8e6075ff4a6a790f6d21756893d8bb0
f35f8f880296bf80a867726bc2fda49cb8465300
F20110320_AABVWF margolin_s_Page_47.tif
c9ae632fb944caf3afcca72d80e0df81
567aaeb68b2156dd3a989638d3ca0558f7600070
53551 F20110320_AABWAW margolin_s_Page_48.pro
4c8af10a2d23f5b5b4537169a6094983
a441ccddbbd92ac4941abe95421d015590332969
F20110320_AABVVR margolin_s_Page_27.tif
54f7721ca0c094cf054a2f5ae29f7c33
5be1675a1ad23d56788d865c4c9629a758b115b6
58485 F20110320_AABWCA margolin_s_Page_01.jpg
47fa99f0676a6b990d5661e832a2d61e
4318dd00c252759b6575161e1e753240920214fe
56531 F20110320_AABWBM margolin_s_Page_74.pro
e15494e84483a15e23bd086da43dc9f6
f180aea0a3667ae9fbaa299ef43c22df5c59b42a
F20110320_AABVWG margolin_s_Page_48.tif
9ebee4cdbcc4fa1da465c1b4b5f60fdc
4f4b72fef5ffb625889a32be8ede707e6d98dc8e
52437 F20110320_AABWAX margolin_s_Page_53.pro
8c35ecca147d838fcef8ab4c4be83afa
e40c7b6fd908be2f713d40332eb182bbf4b6c130
F20110320_AABVVS margolin_s_Page_28.tif
b4fe13bf26a3e30a500253e326cac7e6
16d96a49e4d7de726ebdd88cf67476fa6068a14c
21700 F20110320_AABWCB margolin_s_Page_01.QC.jpg
22fee5ba34d6bf3402f53a811fd9747e
0279bad78f09a6cc67d69fe1f4b00aeb9cf46bdd
52918 F20110320_AABWBN margolin_s_Page_75.pro
3dfad20b6971987db9fa2af0ee9a8013
732e7d7a0cf6ee4a6da8d27898d3025a85208e46
F20110320_AABVWH margolin_s_Page_49.tif
6c6672351d750b8da6216c18bc70e921
b09404cd8e018f23286259797295cfe2bca12b54
52529 F20110320_AABWAY margolin_s_Page_55.pro
b2a3775d0565294b28f660a9f048d82f
c0796aa7579f2d71a3dcb2b4e47ffd2bceffface
F20110320_AABVVT margolin_s_Page_29.tif
c6eda78776c33141e3205b53ec0771b8
de8e35f55c82a683a1d8a881cf978b382627d489
13296 F20110320_AABWCC margolin_s_Page_02.jpg
70fdb047078797bb54b9f2182f6887c1
2c6c01555bfced9c492355db3f1c7ab644a31a13
55064 F20110320_AABWBO margolin_s_Page_76.pro
df4c24a7fa40deda608330e12b7f0962
8aedb1174b3526d67bc68b625cd2ae9c4d927ab8
51373 F20110320_AABWAZ margolin_s_Page_56.pro
f9a6ae4b0fa37527a981916ec100b809
f1510ab3a1f51c42c95d23114ccab38d14b60507
F20110320_AABVVU margolin_s_Page_30.tif
ab649b6be28edc672527e97fc9e45a3c
52b266ad3408f8e55f4909ce8a5c373c44a98257
25762 F20110320_AABWCD margolin_s_Page_04.QC.jpg
bc160286b91dbd9f167ff99fcb9cb6a8
fbd138d8e2414b6f05c2e0c60db160729ab5f0df
50925 F20110320_AABWBP margolin_s_Page_77.pro
00017228d9bf9a1d09d3b5150a0ac507
5be6fc31113dc41bd79b1d2df9dc5079a1dfd2fc
F20110320_AABVWI margolin_s_Page_50.tif
165ec30a0896b11d09de1d80eaf285e0
7e8f15df11721f728e0d44d0125a5138bd50f174
F20110320_AABVVV margolin_s_Page_32.tif
ac485b8402febc56eb14546d52cd0b88
4583b70413ad041760eeca198543c45ea5438acd
288822 F20110320_AABWCE margolin_s_Page_05.jpg
026a88f1789e8b52092781ed645ae15e
7cd6e707cb2f5a5efcb7c781ccbbf9e67b326905
52913 F20110320_AABWBQ margolin_s_Page_78.pro
394137d76b9d7501c8fcf33feca02bde
a01e8e178d18488343d87be6573eac5924997549
F20110320_AABVWJ margolin_s_Page_52.tif
6322ea5e18695a90829e7c7ff07f4462
454c257db495e97dae86e32af888220ef3e9dc3d
F20110320_AABVVW margolin_s_Page_33.tif
12c6d54071f4c874a5a288386bff1b01
148bd876edd0e4f6311dc400392e55ef9f0f71b3
322525 F20110320_AABWCF margolin_s_Page_07.jpg
72e74f096b753936365d0c87b041fdf6
3d6a05ce91cf5dc265c73bc5aa7b01a5b45eca97
53034 F20110320_AABWBR margolin_s_Page_79.pro
4578820751a42fa62678a95d4d533f55
d27b845cba1ebcce92b0748dc3b01d0d5e7fc983
F20110320_AABVWK margolin_s_Page_53.tif
5f993ece1e700a193f3305483799070a
22a987c2f92fbce6106cf5c696846c2f615e2a0d
F20110320_AABVVX margolin_s_Page_34.tif
d953a44c50e84cbd589399edb1096703
42a3e0480c59870ea24b0b1470bc98a1670694d8
127908 F20110320_AABWCG margolin_s_Page_07.QC.jpg
4818200009e9d9f26062a5203a3be977
1661130603cc42e12ad4236317aa6e805ba6a592
F20110320_AABVXA margolin_s_Page_81.tif
7f90cf40b0db67ec982f3060ba1ab304
5b4eb568fe853a749bbe20e468e4a042cdf05ec7
F20110320_AABVWL margolin_s_Page_54.tif
43decd753f19f7b5ef0a3f8787b617a9
9fe9ea284f5081aa0e3283c1f90fc83ed972c98b
F20110320_AABVVY margolin_s_Page_35.tif
0a66580f786a04632e0dc720022fdd37
71a23604d19d8887e22c7942d331019a8999559c
94919 F20110320_AABWCH margolin_s_Page_08.jpg
4adbdc0044ced118d03633ba33c6548f
cf69e70678bcb96be9497b6f2eddc21a304a691c
F20110320_AABVXB margolin_s_Page_82.tif
dd4dbc6f4ae40d6cffbf7844b7a27601
2cba23b092b0223448c6c2988f6fcb1b9bd5e038
53417 F20110320_AABWBS margolin_s_Page_80.pro
027aa86a6e94636910309f47995c33e3
4592b93c1f179636ae7d6c98a94b1713d0f2b230
F20110320_AABVWM margolin_s_Page_55.tif
7aaa92634247e19a3dd005988fee241e
ef1c28490e6bae2a93382abd99319e1f7efd25e3
F20110320_AABVVZ margolin_s_Page_38.tif
0c9df7a8e6b5ab3af7c2d209ea0e26b9
83c28ed16e1759be0ee5f3022ef416486dd1a3fc
83648 F20110320_AABWCI margolin_s_Page_09.jpg
f3b8599052e4d17cdb97e53cf8954ddf
f5e0b83874f735ddccde1fa56edd4844842d70c3
F20110320_AABVXC margolin_s_Page_84.tif
6b26a884e94b70c52a8bfb43ca0cdf6a
f329682b89f12e11fa945dcd9dca8ca01286814d
11294 F20110320_AABWBT margolin_s_Page_81.pro
60e666811a50c5b45897b1a3f8c20e66
7cb915d81258ed29ac72134078f71363ba89aa6f
F20110320_AABVWN margolin_s_Page_58.tif
475d5ab5d63488a150e00d1e3a53d7f2
a974e7cb3830e4c433dda8ae0cc6f9dfed6c5e9f
48170 F20110320_AABWCJ margolin_s_Page_09.QC.jpg
c65464c83f1350270e6cb3559dbfaadd
f9907a090f71e98285adbbb72093355c99f2feb1
F20110320_AABVXD margolin_s_Page_86.tif
991aed1244ab28e03b7aaa813761ad46
3ecb4cc921f2c34707b03f6bb6ec23827391fcb0
15850 F20110320_AABWBU margolin_s_Page_82.pro
fc0664faaad973c6b1c7c01f7d23a58f
4d614be4c7653cf7da9b70b3c889562c7531ff33
F20110320_AABVWO margolin_s_Page_60.tif
86a1d0ed19fdf2a5a7bd67e12fc547f5
465272c0ab14835f2b55be4cc3498c8c15acf89c
212549 F20110320_AABWCK margolin_s_Page_10.jpg
ec2b3c6620e76260fb0a54bbbb017edc
f0b725d710d7ce3dab4acee3573689b4bfcda4ea
F20110320_AABVXE margolin_s_Page_87.tif
d295d366037831d2a530f3c2b7abf4ec
596be7e83d6012256a8a345ad53aff1d12b6ae28
66508 F20110320_AABWBV margolin_s_Page_83.pro
200481259a91a9913ff1cc1209f6f1f4
b6f4bd9bce84bc7572e81769f5d60afff0475fea
F20110320_AABVWP margolin_s_Page_62.tif
88f7ae625278ae1fdefde012b2ba86da
c58d035a5783a16967c0956a4af64b1a770446c0
73176 F20110320_AABWCL margolin_s_Page_10.QC.jpg
e2aa4f5ef337ea7be9863267bc574e2e
a20abbcc3fddc017214b72016aa4e696b1a1d79e
F20110320_AABVXF margolin_s_Page_88.tif
4a0bfd1d021f275f9277791e4cfb8c80
eb941f1f47fe859054ad6b760a6f4b9f22af98e7
60089 F20110320_AABWBW margolin_s_Page_85.pro
faefd7c2b1153d2036a69674cc921796
8e93614237824cedc9b17954a7a3ba9b33acbb50
F20110320_AABVWQ margolin_s_Page_63.tif
e2c9abe8550e58599a691afd5e51f122
8f796705d42f4418863882e7b935aef0aed5f9eb
70819 F20110320_AABWCM margolin_s_Page_11.jpg
2f600629230166810d32013c4c5493bd
741c1c13e25f826d166c3ff7ba94c65c520790d8
F20110320_AABVXG margolin_s_Page_89.tif
9f216a4e52739a90efdb1130c1d125fb
938361e850280793d83aefe6f00fe64222af629c
61257 F20110320_AABWBX margolin_s_Page_86.pro
3b4183025849448b70389aa1a72dad10
783f5d3420658ebd24c0573929f8e0ef3d0516f1
F20110320_AABVWR margolin_s_Page_67.tif
9902bca6b311a8fdedb0d33e3b6b04d4
cb79107c94e926bc0f382efa55edc763bf6b9104
236794 F20110320_AABWDA margolin_s_Page_21.jpg
f33184bc0cfd8f405e85b531c87c511a
ed9d555c138ee3ea7252fdc9a11ddf7b96c6bfa9
28895 F20110320_AABWCN margolin_s_Page_11.QC.jpg
24f157bb151f706a61d024a9fee31169
bf1043b66678115b74dd2888922b4ce7669063bf
500 F20110320_AABVXH margolin_s_Page_01.txt
37d2fc6261c51b836eeb6908c8853c99
20bc7994d5ad48867f97e63d1776c14baf406eee
63991 F20110320_AABWBY margolin_s_Page_87.pro
378751c70a4e0782dd4eecdc079a4192
2617673424337bb2f07587611439feb93587acc7
F20110320_AABVWS margolin_s_Page_68.tif
a18d984348fb851c7a6c7f20d9ee3505
5a0496cc8aaf0b061841cb8439b0cab6489ed778
87642 F20110320_AABWDB margolin_s_Page_21.QC.jpg
4fa972fc7cdd79c0b3a5de430d10d589
7bc2100e6a08246181b3f8cca65d65e96c5a0265
220961 F20110320_AABWCO margolin_s_Page_12.jpg
7fcba2a92d2ecdbfbea147e5713bfa08
f08fe4c689018a27c75ff19d94fade8996329499
90 F20110320_AABVXI margolin_s_Page_02.txt
44dce644bbc3d06de7177eb3a6a86f4e
27a22ba75cb14efc9cbda824722bb6a7f4737d2c
47958 F20110320_AABWBZ margolin_s_Page_88.pro
4540239378fc73028141def2c21c1dc4
00927cdc1d57d96fb63f09d968983c7f47aa4e14
F20110320_AABVWT margolin_s_Page_69.tif
c12ff681bf54668fa0209c6bc22bb209
ad0fd0c8ecd3a436bf60f70f356ce256d99f619b
112235 F20110320_AABWDC margolin_s_Page_22.jpg
758dcaa88ac61f2904e26caa2aa1bde5
1bea7f9160aa141c96c8b4c14d98a138a463c6a8
232478 F20110320_AABWCP margolin_s_Page_13.jpg
e98906073cb528f28b4cdd5d9f76121e
56fc3236f29113478983d254b25f56ac998d8054
F20110320_AABVWU margolin_s_Page_71.tif
15725f1c9e4ae810a303383461dd9557
5f1e570e69334d4741e59b79fb9c2280f5eb5923
43532 F20110320_AABWDD margolin_s_Page_22.QC.jpg
173f23e76aded90359bf59452642b303
e393240bc26c42e1e46b605d87bd4e3e49db52c1
87746 F20110320_AABWCQ margolin_s_Page_13.QC.jpg
5385852f87e7ca5def40a4d9a81a8bd7
d41f68e9bc467813a34c0e9114195ff2c9f6b566
309 F20110320_AABVXJ margolin_s_Page_03.txt
9ba15ddd280da78a352e8a1073fa6cf0
7259a4bfb5ee38c8b1c8ca7f043e1d19a5415d84
F20110320_AABVWV margolin_s_Page_72.tif
1698e6e24cb870389c39e6fd165a642a
e0fbc4ba96f94e3fb5f64d31d60e39989287d817
224000 F20110320_AABWDE margolin_s_Page_23.jpg
bd8be6bd321a424994a67ea5b82add9a
2a2134f9ab3a1c9f1b766351e16d168e6c5c25cd
232177 F20110320_AABWCR margolin_s_Page_14.jpg
c9c296afd345c1be55386255deaa7efe
b2d783571c3c4fcaf96f5dbeb9ebe31384a098e5
2760 F20110320_AABVXK margolin_s_Page_07.txt
ad0f654876eac5f1028452614d669fef
58fc7382f0cf7f52446d0e4e179572272d8545c4
F20110320_AABVWW margolin_s_Page_74.tif
f41ae6a2a727354d23e92ff5645b0818
c35c8ea0ba580f594a6d688bcb31827892aab8b7
83184 F20110320_AABWDF margolin_s_Page_23.QC.jpg
3d86074e833e28e89449efeb353d1ab6
c125bac9e895e55ac1b92e9402119a7e8db71664
2209 F20110320_AABVYA margolin_s_Page_28.txt
af2cab86ae958e22037ab83750d76195
b9492dc07c6c1b911dec14c32c48d14805390d49
88732 F20110320_AABWCS margolin_s_Page_14.QC.jpg
9868d92e519befd8142e7c459a7b573c
d0da75170522dbd925cdc4767777a1e2c00c96a7
529 F20110320_AABVXL margolin_s_Page_09.txt
3db0d2d3596c6294756e0e9921774c11
7a0d78185915c129e0ff81f9a1eb3d9f52cd3060
F20110320_AABVWX margolin_s_Page_77.tif
fcfc653c77ce4fe1570858e59ee48bb7
62be2c100f9a5024d234cd532ad2806f22ed3ffa
230161 F20110320_AABWDG margolin_s_Page_25.jpg
09782a834bf946be48a582d6e38cdb55
8dc83d096a6b0c1caf8ac80919e03d1da99dfcdd
2272 F20110320_AABVYB margolin_s_Page_29.txt
1b68df82501033b564be42da81f9bb2a
8327c753e0116fd4682eb93dc93e529e099ebfd8
2081 F20110320_AABVXM margolin_s_Page_10.txt
0f596555c4a638f9d5e41926581490a7
9c912d47623b7194df95a660af0ba2e816eb40c6
F20110320_AABVWY margolin_s_Page_79.tif
d9877a7657ef4c12153ce9250cd3d25c
1c81c16ca37c27c41d7230a53129daff2b73e2a7
87003 F20110320_AABWDH margolin_s_Page_25.QC.jpg
e9a28262bfaa15af681ca90d7bd2ed7a
204367fdeaf4da1360035ae6454faf911e44862c
F20110320_AABVYC margolin_s_Page_30.txt
5dfeb43ff873f8b27531327fa2a6f5d9
ac2392da2e36af3d36a9d8a1c4d220451a70ee7c
228478 F20110320_AABWCT margolin_s_Page_15.jpg
dea1be57b43f7a999b17a33657829510
afcd2613bbe8861240f2a837b73b5b84a5d3d8ce
618 F20110320_AABVXN margolin_s_Page_11.txt
40feaaa828c6eddc0fde26545b8fc4fa
3b8f63c0c5b254c0b4b647f7651d79630845d072
F20110320_AABVWZ margolin_s_Page_80.tif
58cf6ef1a35270d0f6a44e1a271ed31a
4beedf860174b4a64754c5523e4e19241ed256fe
201820 F20110320_AABWDI margolin_s_Page_26.jpg
8cf2b1c6993d03d8d89904ae95fcd2b1
c1e37499295429feb9aec8e357cc53480839051b
2167 F20110320_AABVYD margolin_s_Page_32.txt
170205a4c5360315389fcadceaa00068
c031f889bb933e8751c25ecbc2732e45c018618d
88761 F20110320_AABWCU margolin_s_Page_15.QC.jpg
e06a090d525baf5e84d15680871c02e5
f81c83f1adf490863af7eaae12b232373d2345bd
2111 F20110320_AABVXO margolin_s_Page_12.txt
b3cbb554f6fe1521b965b6aab62fb3d0
b6475cfb80bd87a82b1ef01265f6b3b111301374
237183 F20110320_AABWDJ margolin_s_Page_27.jpg
d4cbab55338ccf9a7fdee3c7f00809e7
83ec87648b6cec6a17f31abd5d4969cdd512e720
1996 F20110320_AABVYE margolin_s_Page_35.txt
565a6b884a574b704ff87dbfc84e20a2
c75355e285d5c0f6104187b77308e703cb7ea9fd
86769 F20110320_AABWCV margolin_s_Page_17.QC.jpg
1811aca6eac77afff96bfd484ab2791e
d787a89483251fa331bab975dd0ce7ad07f41a50
2192 F20110320_AABVXP margolin_s_Page_14.txt
855ac55fce5c7f7eca1b387d507d0d09
88c5fec5fb80bc4cb7106eb4caf565ffe66f396b
90466 F20110320_AABWDK margolin_s_Page_27.QC.jpg
a5f7f0e0299cf165e1cea66b1ab6b15c
925bde54e38d2f0678a533a3652cd04a2b41e249
F20110320_AABVYF margolin_s_Page_36.txt
310fd480e4a8ad93ef9b1c75521c01f0
810d5166d3f33e570cd298154b4dd5e59b312f97
227385 F20110320_AABWCW margolin_s_Page_18.jpg
52b7f881675215d50955b44b61e5f565
6a5aafb44529ff5ac1977ccffd41491fb574421d
2169 F20110320_AABVXQ margolin_s_Page_16.txt
ab136530bb5476f3cf3d453409b65856
dbf76eb99d8a01896f07f851ae7369caf9ca69ce
237903 F20110320_AABWDL margolin_s_Page_28.jpg
da26077423d539ebba49099a1033e940
ec8f37a7bbf262aa1e63973cc8ef8cf25734ab2c
F20110320_AABVYG margolin_s_Page_37.txt
dafb21dc702afa2b929962aece8e592c
e2338ad02af21990976a5b6d72862c9e9fd633d4
86613 F20110320_AABWCX margolin_s_Page_18.QC.jpg
9dc2da0cb80d5a91fd42ca8f60fbc127
a56fb60e50bcdbf39f7efbde6a36a74066d50cf4
2142 F20110320_AABVXR margolin_s_Page_18.txt
ceedfc98ae828afeafe87bf78feb2ed2
5bc44c4fb4c293f1617ba8ac84a9df32e846d790
231190 F20110320_AABWEA margolin_s_Page_41.jpg
b60409315dc0967ac0e7c095497626ef
6814f16fa0a50817e27a8ff148f669b79524b6a6
89490 F20110320_AABWDM margolin_s_Page_28.QC.jpg
c713099a3ca9d04647d794a3e5880a74
e32f35b88238c4ccdb1c1613471e904559c98581
2269 F20110320_AABVYH margolin_s_Page_38.txt
e54a01a6c62186033e9914172a167533
3bbb3eca76ff825105efa5266fabcc8e2b0c3539
230567 F20110320_AABWCY margolin_s_Page_20.jpg
d07217994f1e85e9a22c15a832cc00d1
88f95aec98e848737bf852e58be6d7d3590978fa
2228 F20110320_AABVXS margolin_s_Page_19.txt
9b6e1fbdfb7c7d0b1dac632cdf7b502d
b282cce182922d7783587ffe7ca5a4ce932b0ca7
86039 F20110320_AABWEB margolin_s_Page_41.QC.jpg
b82942338b01482568d42912973e647f
f05473c2cfd50b5a5590bf1d3d2e555344dca9d4
238817 F20110320_AABWDN margolin_s_Page_29.jpg
3959552bd561a154225d6b8bb129f280
41205de5583fc75e54865f5a9fe8db3a72044984
2221 F20110320_AABVYI margolin_s_Page_39.txt
f41cf062ffad79aa75141522f0f4ce9f
e1fd9473162e857113577ff7afd2a51df56d9482
86744 F20110320_AABWCZ margolin_s_Page_20.QC.jpg
50d32578f2e41b4be5518253d7f4200e
9805e96a57fc94b74bd1138ff18be98e8b18a73c
2262 F20110320_AABVXT margolin_s_Page_21.txt
89c4c109a92fbe3b202c5c7b039471da
bd5f76cdac436b0196b5418ab2bdd5efd6b2a114
236221 F20110320_AABWEC margolin_s_Page_42.jpg
689b9e27ff937d775f6728c714effbc5
ba382858413ab44d423afa74f907f5cb6427759d
87515 F20110320_AABWDO margolin_s_Page_29.QC.jpg
c261ae3fa783813bbf8d0aa206859e81
61c2b3e523f8e0288499077ff33b2f1147eddb33
2222 F20110320_AABVYJ margolin_s_Page_40.txt
f4523bcdb71dfa573d8dc83ff519f220
75269dbb0614dff4e8e286ac3a5d75926414563f
1021 F20110320_AABVXU margolin_s_Page_22.txt
c1b9aefb793dfbf1c8a30d4ed9fe35ab
042dab3783432f18f7527b6c8d6c6aa5deadce2e
88565 F20110320_AABWED margolin_s_Page_42.QC.jpg
9f84189432ec8a68611a1990c476ff92
b312fe8011216a5603e60c80250a7dca846a6731
90257 F20110320_AABWDP margolin_s_Page_30.QC.jpg
6f58723eee3a65745bd198a72750dc51
58796a9e4b25c4da7a3b245f132f2ff5cbf415dc
2230 F20110320_AABVXV margolin_s_Page_23.txt
f4b43058e8e6b5c3f7f7a12412714f4d
23c662a90e961dd978c535b93aac9cc42343707a
180641 F20110320_AABWEE margolin_s_Page_43.jpg
bfb4d94024f4610e581e914d984196f3
90314d022862b0b15d673d425cc0bdf4c1d55a5f
228408 F20110320_AABWDQ margolin_s_Page_31.jpg
6af73bad106002bccec5b97461c37582
59325a1711c8c064112f38edf2b67baa8d4f5520
2140 F20110320_AABVYK margolin_s_Page_41.txt
28a31e77177e108eb9ed7647c3056cfd
c090f26a382b27df24ba805a0774d23851f55400
F20110320_AABVXW margolin_s_Page_24.txt
251b89c18f054d19c932e6c2fe1b97ad
7c0cef661b8be49129cab2c615c01d182bf34578
68777 F20110320_AABWEF margolin_s_Page_43.QC.jpg
c0e7114d0e483883b90443232771719b
42e80f8f3d63514e2e82ec22b880e013b330706e
87695 F20110320_AABWDR margolin_s_Page_31.QC.jpg
14d54deec5ecbc5ba8ad4c1e45c0586a
99fd32e964568d7c7408b4c03f08891318d3737c
2229 F20110320_AABVYL margolin_s_Page_42.txt
ade8aee0da8ebb1037709cdfede2f042
cbd5f8d24d2c1ec277f166d2ed0c8b2cc4ac0f8c
2226 F20110320_AABVXX margolin_s_Page_25.txt
71a350e8578c4e957eb6d67d0c76206c
eda036709033fd612179e27a27284ca50e43d359
168504 F20110320_AABWEG margolin_s_Page_44.jpg
2ae638a8e8d47720e78884bbaea81946
9bb6d1b87333ef8dba63f2b80156c7a74e10c3cf
2124 F20110320_AABVZA margolin_s_Page_63.txt
a06a65d594aa968d36c370e47524c2e5
9673d43b948af10ae259791b78da240d08843cb3
86758 F20110320_AABWDS margolin_s_Page_32.QC.jpg
8ce372017315f00975484714eed6a948
67636bcdc13b201735953ff7d586abf7743337b5
1728 F20110320_AABVYM margolin_s_Page_43.txt
f50b6d16c7de1f4d960210058f2e8ad7
ea45e62e308ee0061ddfb7acf22f3bee38c058f0
1959 F20110320_AABVXY margolin_s_Page_26.txt
7cefc79c535e4557424cef0c087d1a15
bd86270eab10e9002db145050561c2d1a2835238
50944 F20110320_AABWEH margolin_s_Page_45.QC.jpg
09066f7929b74663c7a6236b7d6a28f3
08f74bacfcd2158f7c082b581baf2ab688dac469
584 F20110320_AABVZB margolin_s_Page_64.txt
e4a276eb3caa4e2349a9c9c41a765bab
896f6e6bc43045360f0502b5c9390f06dfd027c8
223323 F20110320_AABWDT margolin_s_Page_33.jpg
322b7b99bb972c0a62b227f4832ced86
0027cb27ddeddecb77f2fd4ef523c4e7d616655a
2506 F20110320_AABVYN margolin_s_Page_44.txt
67e1fcdcf553ee400f67d2369b99ae61
c52c773b410f12e0a08c4114d35eda6d20e5864a
F20110320_AABVXZ margolin_s_Page_27.txt
649153c327fbd0c30d13baba36285945
242815501acf11c603fb0f7b6bb398c3b62329fb
153646 F20110320_AABWEI margolin_s_Page_46.jpg
26406b7727a78c6bb92a79afb01e42a6
de7302fe10a5cc6a3d8ca44ed7a1fdf525d2e58f
2130 F20110320_AABVZC margolin_s_Page_65.txt
e9dd5268bd8208594c6217f652a2554b
a32c9b5ba4d30a2970ddb86a1f7d2a43e6ad8f0d
53117 F20110320_AABWEJ margolin_s_Page_46.QC.jpg
e4bef9792839f371a36006925803e61a
1fe501ff5c02749d118458556cb08c5dcb18c15f
1931 F20110320_AABVZD margolin_s_Page_67.txt
5f2dfbe44b86da6ce2bb86d1e35835af
9932dd8aa5e981337e6d21a57039f4b3ab2af74f
85162 F20110320_AABWDU margolin_s_Page_33.QC.jpg
8e7fbc6d753efd02da06e10c5ad3beb3
8f64926c747eafc81ce2fa7c729b6adbdc3365ea
2516 F20110320_AABVYO margolin_s_Page_45.txt
8ec8343db0dff7cc6c6b280fcac285f0
d4a0c835015b91408d7ccf395082f2a246f48391
66822 F20110320_AABWEK margolin_s_Page_47.QC.jpg
47fbc56dcb5bc1a0170906feb3ab7186
031e1bec6f18d16f137e14a47fc9e2f5ce2b2ad8
1652 F20110320_AABVZE margolin_s_Page_68.txt
078689fc87f4cdae0f891011129e5518
ef46285b1839fa07485cbe41d32222ac4a31de8e
77903 F20110320_AABWDV margolin_s_Page_35.QC.jpg
ea26849af917df798d8f046006642afc
18b231bcfa85a403d0d2bcc1cd7a9585461c2c41
2612 F20110320_AABVYP margolin_s_Page_46.txt
d6c00d55d620cbca2d5578fe2a8a0e48
8cf14ea1bdd245fc2890c2339466d9d4af5766c8
228224 F20110320_AABWEL margolin_s_Page_48.jpg
2aa0aaaff321aaedc21fb0c10d8865d7
151271d26e199ff94337abb40e16c063f1383225
1676 F20110320_AABVZF margolin_s_Page_69.txt
7f256e94a132f9e6051684d51f21ed67
ca6cd88733e6becb65250630a1fd7a6987f46d17
237443 F20110320_AABWDW margolin_s_Page_37.jpg
0eda7a1c602f591865f2b07b76abf6ed
2873911ac680c91149d9a9e6403192ec80a5c3a2
839 F20110320_AABVYQ margolin_s_Page_47.txt
c87e82f7a64dbd8d4aed32e8d300f3f5
1f8b1bda52f5de56b385b02380be30afdb4eb99b
85056 F20110320_AABWFA margolin_s_Page_57.QC.jpg
b2134625bf00e1f38d62fec69cb05b71
d5b7d95667bc2020df3dafbc75eaa0e962c61285
86411 F20110320_AABWEM margolin_s_Page_48.QC.jpg
e270a5917ce5c9498f2e25f9d7c86a33
0545f7c1440900dd4f709dfaac6ac83949284775
443 F20110320_AABVZG margolin_s_Page_70.txt
9852c8b2850a7893e30abc7c299bc282
ea2c376ba46f2483cb54cc0f2af8f1ed34c1600d
237301 F20110320_AABWDX margolin_s_Page_38.jpg
b6e3df9aa7a21d5a2d781ac86fdf6c9f
25d7fd91131d3b575e59212359d372df97b54212
F20110320_AABVYR margolin_s_Page_48.txt
f745dd7ac2f2062c10d212fc02a5cf85
6c245f38dd8cb737e4eefa479aebc93c38b8ae63
229193 F20110320_AABWFB margolin_s_Page_58.jpg
8b20ea3bb8c4a107b86d0a6816db0437
68ea26ac8e90a880cefa9bb1362bf38c2d260eff
87481 F20110320_AABWEN margolin_s_Page_49.QC.jpg
c8710b627c755f7f9b6d6f8ed8b27b38
a3c02993e0a8d46821adbca7fc49fe1514f7d2e3
F20110320_AABVZH margolin_s_Page_74.txt
26065dc6b857860122cc0945bdb995ac
dd2a2bdbcbcf121873d7cac6315316906770ecbf
91516 F20110320_AABWDY margolin_s_Page_38.QC.jpg
aca6e9ee04aed8b5344ec737d819c111
60cc1df3cdc04c0de1de4a9fef8aa1960c16b52d
F20110320_AABVYS margolin_s_Page_50.txt
f9d337f4f91a5c0b9b0ac21589d4463c
4be7cdaf23819d7f8be3c5fa8be8b68e3f1c8c43
84783 F20110320_AABWFC margolin_s_Page_58.QC.jpg
c8152577791c48850f352365226c03c4
5ff4da1c4d2e9c2cc78a8e99921a95d4c746b0b9
232274 F20110320_AABWEO margolin_s_Page_50.jpg
41496657f3f5f81f54f2986da662fe61
eb7b7b319233a6b1d7e76a17f5dd27884af4ebd0
2197 F20110320_AABVZI margolin_s_Page_76.txt
d5100cfe2c8c62336767ae1e8cb56d18
27b09f364ec501f171c140f4a43de2078f83a6f4
233084 F20110320_AABWDZ margolin_s_Page_39.jpg
1b823dc334a099871e1c34205e6ad63b
a7703e0f55e79b39d29b033a8a474f95266011d7
1973 F20110320_AABVYT margolin_s_Page_51.txt
abfb341773835a5dd62a0d3109a85cfa
38a02790da47d8da0f66e3e8aab3e88310585e64
84559 F20110320_AABWFD margolin_s_Page_59.QC.jpg
61261067a148e793778b47c692f2e53a
5c268670138755049cf3f7514c9e77fd82e9e309
86808 F20110320_AABWEP margolin_s_Page_50.QC.jpg
14021654efed1e6cd30678bb1db3e6bd
6aa78a3f2e7cfae3b19263a87633a5c072b72d91
2006 F20110320_AABVZJ margolin_s_Page_77.txt
aa40cabb5e30373951fcc1d95de24412
312d4a1394a0f37ce147a463ec4ebb0dbfa9e194
1984 F20110320_AABVYU margolin_s_Page_52.txt
63a1638663e097deeb043a0730f166a4
ed44f9846611fcb93868d480b0652f06ed9ef22e
89251 F20110320_AABWFE margolin_s_Page_60.QC.jpg
b74dfc6ef92e2872ddede8c0927b0171
2c05e5e2083e8bd6abe05a9c4fc86b3115805e64
206784 F20110320_AABWEQ margolin_s_Page_51.jpg
331a0c9548fdbf75fd738339f158a823
3f0e5ac8ba18d80b1547b84f056a475e8f6f1328
F20110320_AABVZK margolin_s_Page_78.txt
fa4f988bc4561b0329b1053a24267773
c753ec476dfcfb68f9cf6cc9c1da47bea25b9ee9
2108 F20110320_AABVYV margolin_s_Page_53.txt
380b98815b244dec8141ec3381295bff
e81710a4e68c39ccf6ff8d510491e44e45c73d26
231471 F20110320_AABWFF margolin_s_Page_61.jpg
ecf5fbf6a03848a144d374f2608ee06f
f16bb3f41e5e072caad513b0051c496ebb241788
221043 F20110320_AABWER margolin_s_Page_53.jpg
4225bd4c0ebe9aeda3da2df018f71d5e
07133e2715aa7eec7c29d24d14e1f35ac6c47d3a
F20110320_AABVYW margolin_s_Page_58.txt
8c2303af9ef2e56944dfd50212eecd08
2a25f9382e311671b550917cc8707861755e20ad
87368 F20110320_AABWFG margolin_s_Page_61.QC.jpg
d75213e725625436cc541408c2d66306
9f84c862e02214ef6cd847fd5b32550be1651551
80850 F20110320_AABWES margolin_s_Page_53.QC.jpg
6a50f1567d1625030702b70008ee4849
fdf3472c81bc2a4d2c7292393a67e34ec710a741
2127 F20110320_AABVZL margolin_s_Page_79.txt
09b4cf694fc5ec360a294d18988b5cb6
4d49d939cc904ab58b4eafc60ca8db29612891a6
2235 F20110320_AABVYX margolin_s_Page_59.txt
2d4d34c5da8298375d0093da8644b878
4b9b7f42a555691e33fcf0324f9f83ab380241e8
226905 F20110320_AABWFH margolin_s_Page_62.jpg
798069569a7c9111b546d345ac868e5f
59fb00b8b6fd076f5dc7a600ac49c38738e719bc
222823 F20110320_AABWET margolin_s_Page_54.jpg
7687c14547522685851f3cf7c19b9312
3a01a20541f3981549bb5ccc5bc1dbfb46fb7157
2103 F20110320_AABVZM margolin_s_Page_80.txt
59675198dcb6bc07d71efebcac4a86c7
3af9c9235be7d67bd88beccf42007e9b396cfeb6
F20110320_AABVYY margolin_s_Page_60.txt
74ffd94f2cd4a355b6e08f63fd7778c5
70e942a114daa2df2606e5b5d7655cacba3dfa66
85714 F20110320_AABWFI margolin_s_Page_62.QC.jpg
cdf3c2ee746245fdd08f8924d1567a21
fb87e4fc67eef1e01ebb40bd0e3b2a0f2cc9ce14
83676 F20110320_AABWEU margolin_s_Page_54.QC.jpg
e098fc914161a671a388fda207e3bec9
a8448d27297af5f07e2005a13da03b1f0efae4a4
450 F20110320_AABVZN margolin_s_Page_81.txt
894f11356d2a8f8d7f229b3d52e79058
956a782ea0e33fc4c8b7e0449384de032dd92ae7
2117 F20110320_AABVYZ margolin_s_Page_62.txt
fb88b51bc05ee4e6010e4000c7b3f582
0165234bf76757d63a9408cd0157b4f8b1f447a5
84725 F20110320_AABWFJ margolin_s_Page_63.QC.jpg
702ae355e1d3694a9d857e68a918a61f
0d240dc9713b8f88c2da6e388d8cebded3b769bf
2987 F20110320_AABVZO margolin_s_Page_84.txt
308e1f4846ced88a0eaac1bfcf00b3cb
d8f15d797b0ec171b5031e0db17dd8a9321eb70b
68103 F20110320_AABWFK margolin_s_Page_64.jpg
7d5c0a3fb3ccda5b50cc96d852c16bb8
2d0d495962e1c867834efeaef20e27076cc573c7
224513 F20110320_AABWEV margolin_s_Page_55.jpg
f44101e0fc7b14305b913c0aa5ea46ba
a47b929d4f3325d3483f0c5dad3abe285ba93404
2424 F20110320_AABVZP margolin_s_Page_85.txt
66a4565ea96826fa0dd02db27b0d5c53
4a990d1118d9e0550ea97853f7983966fd2cd06f
27334 F20110320_AABWFL margolin_s_Page_64.QC.jpg
268e7e156e81dfcd730234016141a02d
18cf9c9749e4a6aa7d9caa3f4a636ff7f7ef4983
85323 F20110320_AABWEW margolin_s_Page_55.QC.jpg
a8ecdd65c1f48f7d71cf296072276cda
6fe8584faccba88578e2b7ec5bf864458a2c2ba6
1953 F20110320_AABVZQ margolin_s_Page_88.txt
c56b4262760a18aad783608a2dee098e
240a1e0d08963e4f0258ced61005e2a4cfeb1443
224792 F20110320_AABWEX margolin_s_Page_56.jpg
8741bb3ef95df627751c5068b7073e62
36e0e26d4345ba336944726542fbb28707d6235e
1256 F20110320_AABVZR margolin_s_Page_89.txt
e873fe54c2a027c388fcf18135478b00
4137edfd015601527d17aec73a796a014785200f
220316 F20110320_AABWGA margolin_s_Page_75.jpg
788a41d2c9c508120f51a1cf2710a6bc
a1ad01d08eb621355b22ac91d83c8aab61770d22
156764 F20110320_AABWFM margolin_s_Page_65.jpg
d1f96e91d20d06cda664513a50023c34
0b70001877cbe2a8cd5d51f9023fe16b8ee7f44e
83039 F20110320_AABWEY margolin_s_Page_56.QC.jpg
44c0378cf9906910a1a5065ada393f86
01c3b4a153286e70a2d82eed453d0acfb0cbcb18
8557 F20110320_AABVZS margolin_s_Page_01.pro
26804b1565baf97b0fb796448335a47b
c143fbe76b404282cea4021b367cb85e7e684c0a
82451 F20110320_AABWGB margolin_s_Page_75.QC.jpg
8fcb76f2544a8067559315d2d8e91329
95dedaacf9879719741c769a646709cbbb5b0c5e
57183 F20110320_AABWFN margolin_s_Page_65.QC.jpg
7c5b54ce2b3225c04a8ee3b502fd3704
09a4096ce9f1b40a378605183ca68a5a246a1744
223484 F20110320_AABWEZ margolin_s_Page_57.jpg
bce7830e5cfe7824502da6c3e4a61ece
7bb7013d33fb7de7184a3c03bd799e52fafda27c
888 F20110320_AABVZT margolin_s_Page_02.pro
aa0ac7a9210f38debdcb948d6865fd4f
d748a57b3c846effacd17c2d11f1697cc663504e
224094 F20110320_AABWGC margolin_s_Page_76.jpg
82a27af8d31862713b70d4c875852434
f51ffcb1bf74d2b53bb70f1a24a413b2b80482c8
155872 F20110320_AABWFO margolin_s_Page_67.jpg
1d3180f21f7c24f49775f7e2837b70d4
296a08a7d57580002dc30889a9257978e9b0455c
94795 F20110320_AABVZU margolin_s_Page_05.pro
77893dabb65fcdd4ffb8e2c41e23c1d8
0f4ab0f22d1c51197786ca79e691363216a554ee
83750 F20110320_AABWGD margolin_s_Page_76.QC.jpg
adb90b2903a859348de4c556dc94ee2a
74ecf7e36d04c06b33c47bd18e6983cb5eba7fd8
52246 F20110320_AABWFP margolin_s_Page_67.QC.jpg
9f2d0a0ef9b6c726625419aeb63a3497
611cbab05852df366fdc0a88e4ed62a7538b7fbb
62996 F20110320_AABVZV margolin_s_Page_06.pro
c1f7dbaf579af1845407819f8371e272
fd60a1ae3587a7cad6d017525a55fe78f27b7790
214834 F20110320_AABWGE margolin_s_Page_77.jpg
ef65ecb5b3dd79cab180a2efc7b135ac
7918a2727b46a559a4faf035460d6ec4a15f40e9
138989 F20110320_AABWFQ margolin_s_Page_68.jpg
23425208b123c5e81c62e3e3ef0366cc
0eb0b703e1ac43249d9fe4d178dfead1ef17ea55
12228 F20110320_AABVZW margolin_s_Page_09.pro
f5c62c92b02f580c9bcce28ff3b866c8
a009e14767a041946b927fa59bce3e4254872ab1
80568 F20110320_AABWGF margolin_s_Page_77.QC.jpg
bc16d32e604416ef36e3d317c58a54b9
7d300432754fbd0bbd867c367fd9de2a6fbc6d08
46823 F20110320_AABWFR margolin_s_Page_68.QC.jpg
3089e82ab9f6701beecf99c397830661
1a088599132ee2d5ca0a4edeb0ad4e4fffbf3d2d
15536 F20110320_AABVZX margolin_s_Page_11.pro
6f095a45afc76e30a1b3859c960f84b9
c9c8a7a0170278cbe22fcf4307d46d066114fd6a
220287 F20110320_AABWGG margolin_s_Page_78.jpg
96cce82d5c7d11f47cd3e21fd7db7c25
833abc12940843be7cd8d32c9d12eab4296ed315
103794 F20110320_AABWFS margolin_s_Page_69.jpg
948df7abb09ddbcf32271ba1bddcf2d7
be58c03add596ca102a35f16102ecc8bdd8d35dd
51743 F20110320_AABVZY margolin_s_Page_12.pro
38f7d34fe7629d1490da7b48453fe00b
4178cf8a893ac33d1c0f5576ff30532a82d987f6
82701 F20110320_AABWGH margolin_s_Page_78.QC.jpg
3ddb8aab240a1d9c2a3a161e716af22c
c329935992403bd375d642b6a43311f8b696ce71
77979 F20110320_AABWFT margolin_s_Page_70.jpg
0cfdf650d14776b1e98294b639ac8f75
77962efec091232ffcf7b742bd8183fd1ae2c77c
55673 F20110320_AABVZZ margolin_s_Page_13.pro
94a4e67e985c9ddabf61599870108e8e
6d0acf64f1703a2524936a46e3bd0a4adbb3cc5a
221309 F20110320_AABWGI margolin_s_Page_80.jpg
1c12b33d49b270f7e5f0b812651123b1
6b6081264d7d00a2a08407810f0f34bd35b09f80
46198 F20110320_AABWFU margolin_s_Page_70.QC.jpg
dc18de5337d7d2863ffe2c612d3e8b7c
e186e80cf4eacf0c4334b52294a4aa242042b748
84760 F20110320_AABWGJ margolin_s_Page_80.QC.jpg
86372065d7a4a18d5f29965eefef2f43
ab39bc6952fdcbf60a80b3f51b1e77a813db2a9d
224244 F20110320_AABWFV margolin_s_Page_71.jpg
f6e3e168363fc50a3b53507d5dd234a6
67a6d58d3b48056213a599b1409d13b3d92baf09
53626 F20110320_AABWGK margolin_s_Page_81.jpg
08a87cce8c3c8c8e05261f15012f63e8
91508436fa69e27a22900b7bab16a893ac5c8576
21122 F20110320_AABWGL margolin_s_Page_81.QC.jpg
8b56dbd0a2f78b6276d69f137e051716
f508c3ecef40dc70542d9137281dc69a6ec5bd38
226780 F20110320_AABWFW margolin_s_Page_72.jpg
c393ae8d0352d17b2858a1d59324d02b
12803eb8a7eca923eab534084bca0af63520a7ec
51605 F20110320_AABWHA margolin_s_Page_89.QC.jpg
2e5bb4b069b9e4855c461ac4940ec206
3ea343f17b74a59ef7f69c57590cf6bde8605900
74337 F20110320_AABWGM margolin_s_Page_82.jpg
34692f08c48633eca3002ad3db3d61a2
8e748d4eec3a1a870811b56c6e6fe832ee8d2de9
228551 F20110320_AABWFX margolin_s_Page_73.jpg
3bcb9406bb791ac5d4e3c1295c69fdf6
456f4d62e90a2776a10c9fdb517c878311ae5419
14955 F20110320_AABWHB margolin_s_Page_03.jp2
579d0257024cf56b26aac9f0ebb45efe
a1f3328c4ed2175baa91a9fbd5d4dd44cabc42be
27925 F20110320_AABWGN margolin_s_Page_82.QC.jpg
8969262b208a73e7e7a7dec2d8c08bf0
0946c1df2b88b61c743327adc4e17e73b7c9d306
234495 F20110320_AABWFY margolin_s_Page_74.jpg
c984c5561d3f21106f039d8c4c3780f7
120bce8de29369786730b3c296794a93bf4a1431
1051984 F20110320_AABWHC margolin_s_Page_05.jp2
9fc54e6570e7cd3818bd99e36affc692
ae4a3e9aa19339167c266341e423ade2ffed3e9a
275065 F20110320_AABWGO margolin_s_Page_83.jpg
85d82fbf7302e5869b556b7b29647b10
1da4dfa3fa59e862a4edec5effb5199764b7751d
89203 F20110320_AABWFZ margolin_s_Page_74.QC.jpg
f3b469d51b49fc616ffc46dd92288507
f813eaac818597cd33ae206b6d3938847688047e
1051974 F20110320_AABWHD margolin_s_Page_06.jp2
048461cfe2bee248cffc5ba6476956ee
aef436e390c37e63a442b3f76f2d3bf286f12fab
309550 F20110320_AABWGP margolin_s_Page_84.jpg
1cb0e2af68db57bd36af50e70a5a1275
cfa214d1dc0bb9779c09af24453da843f5d99fe7
425044 F20110320_AABWHE margolin_s_Page_09.jp2
5849265d8c3cd89cb54694fe5a156509
81a92da0a1bbb33724db336a450e33ade96807c6
89164 F20110320_AABWGQ margolin_s_Page_84.QC.jpg
553af027cdad50333a39c5a5909003c1
f3b31d4f7f56cd78c06393556f355ee1cbf5b014
107079 F20110320_AABWHF margolin_s_Page_10.jp2
95b16ab3c274199a0008fb000fbb6ab0
9a7055f7e5690b157eb209f153d51b5d15edd935
235653 F20110320_AABWGR margolin_s_Page_85.jpg
93ff5c75a944f951f0a140dac1615863
f7678350189f549eabb03e57324a998489d9a2e2
112400 F20110320_AABWHG margolin_s_Page_12.jp2
b0828ddd871b44c5779fd0687db335e9
c5af0418b08827042697f845be247b21c2a5e3af
80658 F20110320_AABWGS margolin_s_Page_85.QC.jpg
22369c705da735faeedb7575120826cc
adf4548d6e606b243a931eeb020dd8393cec5d1d
118876 F20110320_AABWHH margolin_s_Page_13.jp2
5558fa5966d706c99389574d6f846813
0edb146bc15631a4fb264b210f0f110d615710db
260986 F20110320_AABWGT margolin_s_Page_86.jpg
167dbdb4bf7d2b6f3f51bcbb59aa9cb0
415cc1dcc70306f171e9b2a4ac996fe8ac2033de
119869 F20110320_AABWHI margolin_s_Page_15.jp2
5e856cc9991ad97bda44c30ef0f9e820
0821dbb853d3171d0590f1e76174047e7faa3bc7
86688 F20110320_AABWGU margolin_s_Page_86.QC.jpg
694a7174180d82b14168fd055c80de24
52c6bcb6e6def5ff5277aee16d3c3cdfd586b0c5
120796 F20110320_AABWHJ margolin_s_Page_16.jp2
a66e3f9e85ac61caf2cc02421ebfb9c2
6fc35247f07748072cb6578ffd98279f1e7c5905
283021 F20110320_AABWGV margolin_s_Page_87.jpg
93402996290032098ca717445342a70a
7d72a6baf4505f69b8354f325fe052a78bbe64e9
117108 F20110320_AABWHK margolin_s_Page_17.jp2
600899afe338e26898cbfe6c1f3c5cb3
f28c256a9c55af22c3f141f59fd20411759c957d
92192 F20110320_AABWGW margolin_s_Page_87.QC.jpg
4b14018f86068b0dfcf6fb06cf44ca26
75473f1245ce3340458b92ce9a11b8f1eac047d9
120181 F20110320_AABWHL margolin_s_Page_18.jp2
d70093d6e9931486193413b07d491ff4
93f89e572f44e7ba95c3cfe514393363293789e9
119709 F20110320_AABWHM margolin_s_Page_19.jp2
c6b3428b4babdee7e29d492870a27abf
2723ab7aed6ef8d49665df92010355979365c7da
205994 F20110320_AABWGX margolin_s_Page_88.jpg
a29ad7a369bcf77a9fbc6eec41e74431
32b94ab57bb054269849931c383d316e8cea6fd3
122330 F20110320_AABWIA margolin_s_Page_42.jp2
c28ece17892893ebe03e3915363c81ff
a4dbf550863a0039c66800c74b44e312778e0aad
114538 F20110320_AABWHN margolin_s_Page_24.jp2
c987b73f2f8d999dbd55da496ce7c4c7
60220b8adb475faf23af900ff5b5244b9733d514
69560 F20110320_AABWGY margolin_s_Page_88.QC.jpg
29e97572587346a0089722734f872ce7
3b8875e0e992e0cf79fe6e111b35f2b51f623afb
95890 F20110320_AABWIB margolin_s_Page_43.jp2
3e8445ec8214e0f9ef392927d8b86655
3b6a5fecf930eb087dd02dcb28d3ac6b5b9128f8
118822 F20110320_AABWHO margolin_s_Page_25.jp2
75382749491c24fb8f67348347e7a8fe
07a1b80d77d6eda26aa4a091f4074a3528709e54
138008 F20110320_AABWGZ margolin_s_Page_89.jpg
56bbe84049e6f8facab0761992adcfd1
f949a31d9d4477fe16401f8be622a7753f204277
80129 F20110320_AABWIC margolin_s_Page_45.jp2
1be159f6081a13e6912a778161654357
bac7c3473f8f94f7a3a04f1a762f56876d4b174b
106436 F20110320_AABWHP margolin_s_Page_26.jp2
f4582a5585b031b1be1b98bfa545dda2
7685738997f842707fa30f924b4feffd069033ea
75040 F20110320_AABWID margolin_s_Page_46.jp2
099dbef5c4ad640c8c35ed913b3c7ecf
71c9fce11b02cf5ec93dd51f9b7ff5de6af8cf9c
123503 F20110320_AABWHQ margolin_s_Page_27.jp2
8706cfcf1e82baebfe3c10e77d06fdd5
0370c3ffc3d8eec8f92e1589981737b577bec75b
117837 F20110320_AABWIE margolin_s_Page_48.jp2
5853ed88c1cb3f9da1385de71319e121
bae9c2bea9becc37703b80adc4ab5efc3f46d0be
122402 F20110320_AABWHR margolin_s_Page_28.jp2
0f682e034bf55e666c630a5cf3eb5401
f4717bc46769c77f7456611336508e10640cdd09
117014 F20110320_AABWIF margolin_s_Page_49.jp2
16b1c581120508661e2eb1f7e38c3ebf
abe0946dcd46742401113fba8b64889bd1c7d7bf
110642 F20110320_AABWHS margolin_s_Page_34.jp2
d8bff10281deea6ba5769dca336e72c0
d35bcbb11ba50547a24c3b004eff590b226e0e9f
120449 F20110320_AABWIG margolin_s_Page_50.jp2
1f78ab7708adc6f259caa2101e14b5ed
71c54e7132a3781d839ecd7b7dc4c595856fbd84
111675 F20110320_AABWHT margolin_s_Page_35.jp2
c26273c77940d40730bd087fb2c69d5f
e2bd89bf2b9ca0950f04023dade098b0d20a6a89
107382 F20110320_AABWIH margolin_s_Page_51.jp2
bae98c89f921c33dfeae94cb4e309d49
a43074609f9cdbaeea7c3f554ea378f50147b03a
120340 F20110320_AABWHU margolin_s_Page_36.jp2
7f5e493ad7fb6dd0a522bbba4ff06d2e
9ceac5a05621ff4643d4c3dbea70f97e09de7a44
109416 F20110320_AABWII margolin_s_Page_52.jp2
aa00899a00cc3b2113b94ab45bad65fa
24f59da07cebd7e3da3c6934b603af0765544c81
122992 F20110320_AABWHV margolin_s_Page_37.jp2
473f484a696aee039b0901bd19c843a5
a12c1a4ef1f350b517d32794666c3648690dbf61
117130 F20110320_AABWIJ margolin_s_Page_54.jp2
4c4bf5fbf386a083d3950b79fbddd4f4
a8d408c10ad5aadb261e73cf1fa7829d34ca3503
125355 F20110320_AABWHW margolin_s_Page_38.jp2
ce8559996f4170dd352b60ac57e787b7
cb371beb76db1705aff2756cf9d0593dd103a58d
115108 F20110320_AABWIK margolin_s_Page_55.jp2
ba27eb18a0dd476726b379438868389f
170c67f26a916325bda99be642c1bc2d512c5897
121633 F20110320_AABWHX margolin_s_Page_39.jp2
6634c36d1f2ac4c2945b4163d693e06b
cf7f37d7bf26e011f7f9c716d8f6fa7faa010530
116516 F20110320_AABWIL margolin_s_Page_57.jp2
a584e500f16ef60f76307bc8980af9f8
65ee9ba335cea94adb77d14543bd0dd016623dfc
110060 F20110320_AABWJA margolin_s_Page_77.jp2
8921628c3ee8c4e3feeb82281c5b4077
191c728d7bca7fa7b7eda8aa406f832a8bc48f35
117571 F20110320_AABWIM margolin_s_Page_58.jp2
048df33ed09f7c4d3399133e449a2ad7
8bf6e90173e6922a07e42f35d0c4eac3aa7e3829
120624 F20110320_AABWHY margolin_s_Page_40.jp2
74f22534d98b0a8d36fefe33b3de1b4b
0f4564ae4a21358f6eac2626d772a3907abfec21
27788 F20110320_AABWJB margolin_s_Page_81.jp2
e76667bde0404925046e557a7ccb5002
339abfe12a1e0941a8d8764f7ede0f39a395262b
122340 F20110320_AABWIN margolin_s_Page_60.jp2
205db349e8d9809ee590969902725994
1c2e41dcafa7f42c06510ec2c5c4fb56c3a77e32
119702 F20110320_AABWHZ margolin_s_Page_41.jp2
46058e652db7684ec0284ea95f63ff88
7bded59b5e154d6bedae2569e19ac4dd9c958464
38521 F20110320_AABWJC margolin_s_Page_82.jp2
85b3a6c925427f16e81591b5fb64491b
9262d4b294e8c008ba387841844622d5f3ef26d5
120927 F20110320_AABWIO margolin_s_Page_61.jp2
070e4ffd7bdb31fa6957059bd7d1b00f
71834c3bf2559eb502ae427a86a308ae14816c4a
152212 F20110320_AABWJD margolin_s_Page_84.jp2
5fe16fc715aca5228e381921b7adb273
8e5d29fe16a9ab732f8a1efebace75c2bf962238
118030 F20110320_AABWIP margolin_s_Page_62.jp2
7a91ff8a1f9101e8474e9b4968dadb53
95b97a39dbff14929559cee5a63bcfdea75accc9
108136 F20110320_AABWJE margolin_s_Page_88.jp2
3e12982896b01e5b0b123117f9e35c5f
14e863db24f66bf28a3988e5cca817af4e345eae
115074 F20110320_AABWIQ margolin_s_Page_63.jp2
fc7b6b28fa65163914dbe35200a87672
fb2afff12e289ec545d1d80f74707f459a620f1e
2869 F20110320_AABWJF margolin_s_Page_02thm.jpg
fa85efa8e9776b11706ff3039b2cf32f
fbf89a45c67fc64bada3aee7a2461a2c7ce384da
35809 F20110320_AABWIR margolin_s_Page_64.jp2
4c932f9437b10f69dacaaee35f9587a3
699bd6e3c8444ef47d10e4a6be472f7b7a4cd742
4914 F20110320_AABWJG margolin_s_Page_03thm.jpg
e3fd2397acb634e6efb79ca49c777c94
c16053793c3a7e8bd49eddabb7005ee3c77058ef
80340 F20110320_AABWIS margolin_s_Page_65.jp2
ce885171f5d52a8fe6eea5f3f541beac
f78e510208e0acffb55f9df026a5cb1660cd8384
8768 F20110320_AABWJH margolin_s_Page_04thm.jpg
7dad61bd795818245e54fc977219487b
8877465e6d55cc032a07c5df2c376cac4b842674
78976 F20110320_AABWIT margolin_s_Page_67.jp2
0211f95aa96f837ab302e20a898fb207
f1c514c3eb4766e607070456730ed60ecf0169fb
54340 F20110320_AABWJI margolin_s_Page_07thm.jpg
0f54a306f02cf97851d2823da2fd311a
d34a6f0409ebc2903edb5ca5d19e74f3e03322a7
51639 F20110320_AABWIU margolin_s_Page_69.jp2
f832b1edbc7b109711d8c1a792507a55
9498ea16715a94a6381a1b7af3622888af662b4f
32193 F20110320_AABWJJ margolin_s_Page_09thm.jpg
a7e8a0f532749add0bd91d86d09bb652
04dc2abf4cd115ed32d7f7e698ca564a3e9c7d66
116212 F20110320_AABWIV margolin_s_Page_71.jp2
eba775d790d8480e61ac44688261a88e
5b4ec36e5abe3ff3f14b725d1451493be48ae5f8
23548 F20110320_AABWJK margolin_s_Page_10thm.jpg
8d6b3f425a738c94a1ae6602fc08d497
bb7c4952b1f742718510666a7d566bb7ed9d7ebf
114756 F20110320_AABWIW margolin_s_Page_72.jp2
34cce84f064ee3f3c574d3599a745f2d
daf304bb05dc924d70e57dac66a165d520d7f566
26705 F20110320_AABWJL margolin_s_Page_13thm.jpg
abba5e0719d850d8ed4c26863b89970b
85bf0d97c0d15e29edc218963ba6649f68a2eb95
118392 F20110320_AABWIX margolin_s_Page_73.jp2
bb2bd4fba83da8eb18cb72289a3b058e
5f05b131cdb48b1722f921405406a58f5dec2789
25194 F20110320_AABWKA margolin_s_Page_35thm.jpg
b44f303440c34a93307ee2940057ad36
eae23348b6c319866c71cba9bdf618f6fc052654
26983 F20110320_AABWJM margolin_s_Page_14thm.jpg
8aa024cae907e72379f4c6dc1d9f6695
9bcc6e0b0977a3df2be4dfb1bd8b2134cd0abbc3
113650 F20110320_AABWIY margolin_s_Page_75.jp2
82f51332e88129d0a904abd7444e43f9
f03655bc7af5db7d770a05d5764274c4341d76e8
27626 F20110320_AABWKB margolin_s_Page_37thm.jpg
8eceb9d94287eba061ccb21bfde0435d
5e87ecb5b8b21bd7636dbe42b5789cbabdd52155
25984 F20110320_AABWJN margolin_s_Page_17thm.jpg
11f81af5732aadfc6072115a2bdec8db
3080f635cdd81175a2d3e6ef1af02a060f314de0
27960 F20110320_AABWKC margolin_s_Page_38thm.jpg
550dae88c0d4ba477699ac792fdd0a20
1693ab1ff0441474663fed8852bba9f21c286a90
26974 F20110320_AABWJO margolin_s_Page_18thm.jpg
2601b196142a699c377c5ef454501f75
1aaff9f8f22988ba39bfa6489c42ec5f704bbaea
116925 F20110320_AABWIZ margolin_s_Page_76.jp2
70b59919a60d558097bf0559518bdcf1
0979782af604eb1f00d39b1e6fd29c1ab83765e6
26678 F20110320_AABWKD margolin_s_Page_41thm.jpg
8fd43e28d61403745fdf611a1606cb58
acea56ba0c08d9073f69ef364b8bbd9686b3e021
27620 F20110320_AABWJP margolin_s_Page_19thm.jpg
6817cadff9441e56e783a960c74d1497
9b5d24d71edaa9c2bee6727a06e2ef74c89ed655
27319 F20110320_AABWKE margolin_s_Page_42thm.jpg
014e5c83e74b34ed3637b463c4c804f7
073ccbb13094a9364f6983862ccad24c909113c6
27324 F20110320_AABWJQ margolin_s_Page_21thm.jpg
6c3942e97487e559ef36c201a21d213d
b5cbf8215057a49d6029e396c5f90f97662ed24c
18791 F20110320_AABWKF margolin_s_Page_44thm.jpg
10e93431310400f12f2228e9cf845999
5225332a698da9d1c9cf741b1b4c67695c8c2a36
13659 F20110320_AABWJR margolin_s_Page_22thm.jpg
b43797872ddf6bd28b9c85d0168e0237
9bdc9c444b31bedb94d8dc303c071fdcfb4a1269
17072 F20110320_AABWKG margolin_s_Page_45thm.jpg
ad93c6bed42be3ca05ca1f3fee28f3ae
7068e87b5cb65e280b701bc4a0a7792c549adae9
26760 F20110320_AABWJS margolin_s_Page_23thm.jpg
8729cd8c1c321687b4e31cba1ed7169a
e5711f6670550b153a6fc5f4c84113eb2bfc8638
17878 F20110320_AABWKH margolin_s_Page_46thm.jpg
444eccc7ce99f01bf2866e9de76407e2
ddb21a85f2fc5a0b0cfb9edec9391eafc2569426
25184 F20110320_AABWJT margolin_s_Page_24thm.jpg
f8eef25bc6d7a8f3e146874e15b37d6e
c4fd7785bcdbde21513e6d118292d0eee28c2cd0
26072 F20110320_AABWKI margolin_s_Page_50thm.jpg
ce8d1b01235f3b3361db777a75fdf389
a919f0413b54c4a61d232b219af188173d609185
26804 F20110320_AABWJU margolin_s_Page_25thm.jpg
a03167283037a97b721c7e0704098c61
53a2e3af15253863e4fb9ce664dd4ef1c1e37318
27431 F20110320_AABWJV margolin_s_Page_27thm.jpg
28ef334f85d0a1a1f56110f1ca003366
78d010556491764da9ece84bcf91370c69faee0f
23978 F20110320_AABWKJ margolin_s_Page_51thm.jpg
62b2b0896d337a0e8312701df76246bf
a565fb70e399a62680b0b6badd5a81b898405341
26513 F20110320_AABWJW margolin_s_Page_29thm.jpg
e3b5952bcf206ed7e9dedd55d263e575
d8118d42a905cfa21b7e41d271d77f9df3bea8fd
24572 F20110320_AABWKK margolin_s_Page_52thm.jpg
bb8f8391374fef3923fd850fee8164aa
a59b1a80a458f61d7f8e147ef0510b11f4f44108
27796 F20110320_AABWJX margolin_s_Page_30thm.jpg
b072a68d80c2b40367840d988b91a13b
0d0d078c691f64247beb477196ef500607651923
25199 F20110320_AABWKL margolin_s_Page_53thm.jpg
5ce1bf6cb5bcc51ded740bcec76dd508
db158cc9a06cc025c288dccc4907846fe2899f37
25688 F20110320_AABWJY margolin_s_Page_31thm.jpg
dea5a1180c82aaf3d28a5e9002b32d0e
a1c89e4a2288ff1f57af6a24d74f702ade7de27e
27940 F20110320_AABWLA margolin_s_Page_74thm.jpg
ce99f9b117e2e9421fe1745d17d4dff8
45b859665df07472b057249ef4b8f9c03e3de8f1
26370 F20110320_AABWKM margolin_s_Page_54thm.jpg
c93d3339db1b8288aefb1ececfb69812
c3102b378d3b9678cc9426815d3788fd02138ba0
23479 F20110320_AABWJZ margolin_s_Page_34thm.jpg
54f0be77ca0b92599c00c5ad7902dca1
af7e77de7c5cd0bbe65d877d5a506522ab01bb68
25629 F20110320_AABWLB margolin_s_Page_75thm.jpg
4a797ef88c752746876a6384420bf522
cea214350a9b2510ac2f64ebcc541e8bd6b3cb8d
25476 F20110320_AABWKN margolin_s_Page_55thm.jpg
94b99b9b7ceef1aeaa54ea80981b8ac8
65fbadf49090169a4f075f713b59345302e614ce
25699 F20110320_AABWLC margolin_s_Page_76thm.jpg
7c67d9bfacb5c84077a881eed51433b8
6d809f0bbbd8d4e7ac418cc13af06ab2eb2eb890
25919 F20110320_AABWKO margolin_s_Page_56thm.jpg
1436b718ba580d9cac8c17580ff390df
bc598965ea2f908b7388a55bc09363981ca0bd38
25419 F20110320_AABWLD margolin_s_Page_77thm.jpg
4791e025e8216af8561dd5dd86926cbd
aaefa8a1bee84e4672edc041ec236768e2ed6b72
26354 F20110320_AABWKP margolin_s_Page_57thm.jpg
d0b3528fdcc871ef20da8adbff3108c9
d51e4ccc0ad74f41549754d07420be8d0a7bf13d
24072 F20110320_AABWLE margolin_s_Page_79thm.jpg
0fd4b2b4071f4102e43810ba48052129
4824b8364563b1a310a2a50d6d30da6aa1a96cde
25717 F20110320_AABWKQ margolin_s_Page_59thm.jpg
a2006101e25d0f5fd03a47e8643e9fdf
769d0a5468e6ea135880f3804f7b190599bfbbfb
7464 F20110320_AABWLF margolin_s_Page_81thm.jpg
d7b8e037cae577214f4bbcd3db96f3a0
3dd8ea7383a4e8f9b236412b1c6b4593aa25ce00
27718 F20110320_AABWKR margolin_s_Page_60thm.jpg
3b7e6d33acb105cc9bdd4a9077cb77f3
ab6af4e0498a8a3af69df38679ab19da634a7ab7
9872 F20110320_AABWLG margolin_s_Page_82thm.jpg
014edb53feda60f0b3850599c2c4cbe8
ea50ca5eda20786dc735ca8d4f26bd913cf6588d
26539 F20110320_AABWKS margolin_s_Page_62thm.jpg
a5c6b2fb0c27b3d0d8935e8486e74f6e
7927733a81166ef2a63ffb6dcfc07465585fc0d2
24378 F20110320_AABWLH margolin_s_Page_83thm.jpg
e667a56e22134c744543ee0b0f1e0e24
e93417979ff583773ac7be6a41bc0e853ff3e986
8806 F20110320_AABWKT margolin_s_Page_64thm.jpg
6e0c1afcee5a6182e9725053fb3a2c60
24c05cf95e173ffa02a8248b731f1c10c8993f14
27101 F20110320_AABWLI margolin_s_Page_84thm.jpg
eabf60d66559d2f59bc5e205c8dade3b
905ad42eac4664b8b59602116fa91b64811694b7
18081 F20110320_AABWKU margolin_s_Page_65thm.jpg
329dff7ce2342c3f7cddd94522b71a31
cf567fc6ba2be93d53df737deb3569db868d7dd7
24743 F20110320_AABWLJ margolin_s_Page_85thm.jpg
d8c5a482dd9901e4e4592ec3033a02f7
f2cba8f7008b7d0b100a3f89538243ce41e5aa57
16696 F20110320_AABWKV margolin_s_Page_66thm.jpg
e767cc331e1ccf4397ac62f3695a3e80
2fd8f5f328a1d42a5803f03e19792efdf23aa227
27358 F20110320_AABWLK margolin_s_Page_86thm.jpg
65184dfd6c752d2a639f6fe0c0ef0479
aa83c4a2d47e8c1b2c3cdb750de20b712f5ef77e
16897 F20110320_AABWKW margolin_s_Page_67thm.jpg
78c6fd7c8924327629c2c7b04486c339
8487337ba5a624e68d5435f55a878acd59a31c58
26507 F20110320_AABWLL margolin_s_Page_87thm.jpg
6b763d6c90aaf1258b7577478f502fd5
354fb9faf221653e783db450757d8ca7460ca48c
15608 F20110320_AABWKX margolin_s_Page_68thm.jpg
c7bb7f79547a5a444115bcedc87b1f66
d4969b30addf4e21bfca5fd30496ce702ce6cc5f
21380 F20110320_AABWLM margolin_s_Page_88thm.jpg
759d28d042d7f4a8683ad7ba1334140d
76d0a4e2ae61a79cbedc9f43185a291fdd2cd327
12481 F20110320_AABWKY margolin_s_Page_69thm.jpg
a2bac1ace5be0bdfb6818e4912f9113e
48c387a4405c92143c56c227fb680b9d2871ad7a
104758 F20110320_AABWLN UFE0009342_00001.mets
3136756c2e4f776bea16d30a9881b4fc
8df8af87be02c96959aab5d648b56f489b7bed69
33740 F20110320_AABWKZ margolin_s_Page_70thm.jpg
250090091703c02fd6727f9e269358f1
eb99416c4378d9628d792ffc2ff6dd7cafb821a3
116627 F20110320_AABVNK margolin_s_Page_33.jp2
79e3d23565fed3cc0cb89896132d48c1
bb6ee65055f671a64d29d962384b35ccf972111e
50327 F20110320_AABVNL margolin_s_Page_05thm.jpg
f67c11306498b1688c30c6db3ff2d944
57bdd385e1f8136df3d3f997409fea8256d29c38
224039 F20110320_AABVOA margolin_s_Page_79.jpg
9f6a4a99cdfe569541ec6cabc1f8da85
4027cdc5302b6d72de5e387758f51e3f71800cb0
25274 F20110320_AABVNM margolin_s_Page_78thm.jpg
dccbc35e759dc9a179525165a8ba6419
2255adccae84d7e876c4b0f6002d4f5f0ef72e89
2166 F20110320_AABVOB margolin_s_Page_49.txt
1d4966bbdcee0183dd3c40185b6baaea
90e2aaa694d3094fb36c63cc7327e7956f067ba3
231564 F20110320_AABVNN margolin_s_Page_19.jpg
ae0f83765ca519814357564595e8f2c1
9401d04d1dcdc9eccd9deaa4d495a0d2f232e756
82861 F20110320_AABVOC margolin_s_Page_79.QC.jpg
fd50a3eed89e7602c5884a7a9b89da1a
677368492567b3ff9943f323d99ecc7d6a1e7bc3
F20110320_AABVNO margolin_s_Page_31.tif
d202db6a30db1a3fec80ba64b763e760
553dbc6bee30eabf4bc8390699505c347bfef406
139712 F20110320_AABVOD margolin_s_Page_87.jp2
98d03f062e84a85391a77d585fecdbfb
408570a56bcb2980b38c363b92cc117e8d5f8dbd
26232 F20110320_AABVNP margolin_s_Page_32thm.jpg
2c55b55634b6c9d3130761e4a65a3606
b1b7cfd7681482f53fd130e4df1e1d2797076410
120483 F20110320_AABVOE margolin_s_Page_29.jp2
0324600418c1d2caa41e9b57f1ac55cf
41538b893b3ba97f0ae66800c23744ac8d84c067
1980 F20110320_AABVNQ margolin_s_Page_34.txt
073143c1b4640f167708a8a3c5c6b5ac
9585c6218fe748edaf4b2f3cdcf19ddbafbb1b1c
119994 F20110320_AABVOF margolin_s_Page_31.jp2
9ae2f0494bc3d772fc1599e7c090faec
0a8ef8fa846fec1582837c8e6a3fb63c26fb7533
52999 F20110320_AABVNR margolin_s_Page_54.pro
4d827c9c5b2090acbc06b2dce6d4107c
08c9c6afb307d61d0f35fc767a758726a6ca184c
30503 F20110320_AABVOG margolin_s_Page_89.pro
bb56c062d7524bec801e5f2136b4ba64
dc7b7f12c1de251b5c313fcb46cd505feeb82495
21088 F20110320_AABVNS margolin_s_Page_43thm.jpg
7541842251971903fc145b5c39fa5017
424e62a8b7a20949b14e290141e5312963eaab93
26034 F20110320_AABVOH margolin_s_Page_49thm.jpg
1d8ac1baf9de6a6beb7aa78a2a91b195
f5843b6076367c03e90cdb686fa0eb93e662da2a
F20110320_AABVNT margolin_s_Page_83.tif
386faacd80f7014bc72f7ae39eae7e1f
fbb5a6006f1690990f715cfab54bffac02e801fe
43449 F20110320_AABVOI margolin_s_Page_06thm.jpg
2b191de3bd85a8f4fbed18b65a1e67c2
d12c05bbfdf53761209f02bdf116ff3c2e058e1c
115181 F20110320_AABVNU margolin_s_Page_56.jp2
7f89f398fd2342cc3347f679ab958fd3
bc9a1ab5d3aa1f75e68b074514f508078be47c2c
15936 F20110320_AABVOJ margolin_s_Page_47.pro
bd2889b259b7732959c371342e956c2c
06fb593252f552140e0061f47e807dccf49a738c
120779 F20110320_AABVNV margolin_s_Page_21.jp2
5732835e06b32986b217046b9889dd63
82618f3f5b3bd057353d5d1c5a86caed8c7aab18
3881 F20110320_AABVOK margolin_s_Page_05.txt
bd6ccac5e6fc7443b08bde8773a58cdb
118f50f5522a573786e6a723086b16e07cf56097
715 F20110320_AABVNW margolin_s_Page_82.txt
d8b0e9bd6866be135b682b208547d963
73a875eb78676f197dacd1ace87f7352d9bfd356
119011 F20110320_AABVOL margolin_s_Page_20.jp2
98b83085f3466bdad76fb04fa55a1615
b96a98c1dab4dc0ffc3c1a0cd11f2178f9ba1d8e
F20110320_AABVNX margolin_s_Page_07.tif
84928f5044e0d2fd927e11471776a24b
9d647a3d8fab80841332eef79209913fb604914f
242029 F20110320_AABVPA margolin_s_Page_30.jpg
9b5d2a98d6c7a84244879c43094f9b7f
6a6892676386a8db589ec2370b01f1e0ff7368c2
F20110320_AABVOM margolin_s_Page_37.tif
d7cdeff598f8a8c30eb964e140ef633c
f0f77d29dba86f78de0a9cce4314e4fe88ac8dc4
16937 F20110320_AABVNY margolin_s_Page_08.pro
e36003b2ee68b2cae8a93487d1039ebb
0c8332cd0cb4549717d211df4b058085a76379fa
15953 F20110320_AABVON margolin_s_Page_89thm.jpg
d7b26144f33c9b480059b561707d3d3f
abacc01da2eeb4dec71ee29424cf08579c41b170
317790 F20110320_AABVNZ margolin_s.pdf
22707b33785613dc362d2991b579fe91
b95b866639810ad458863f53591d488d431a73a6
148227 F20110320_AABVPB margolin_s_Page_66.jpg
6b5bc27b6b7d6cc86aa4f1362e6c3df3
3ff9312cc6026e6cadd94b52573887793c4ca4de
F20110320_AABVOO margolin_s_Page_85.tif
c4084c02758ffd1ce803bfe22a05d983
89b747fdf339ff40fa9250bad5c36b5bc47d9a97
134680 F20110320_AABVPC margolin_s_Page_86.jp2
21dd99e0f2367c802e3196df8a50c0ac
6233cf0ee909f9c149741566952292527792e013
78323 F20110320_AABVOP margolin_s_Page_66.jp2
9563313000200c3ef3ad7804a41c5e2e
fda25df0530362aca8de1bbd0e392ebff86581d6
86889 F20110320_AABVPD margolin_s_Page_44.jp2
06cb6df0bbcacc725dfe85210693bb84
87d9a23feaf387179b40fc8ba3aff5e0765deb2a
135193 F20110320_AABVOQ margolin_s_Page_83.jp2
5c5c4b64b3359252aee089fadf90a0d6
6096257f36d08de41b883e93893a58f396477bf2
F20110320_AABVPE margolin_s_Page_61.tif
dfc153546648aa4c82fdf9505a0821c8
fc318be7e58f086483ec834ed176cbe55ee2a390
2653 F20110320_AABVOR margolin_s_Page_83.txt
c32bfd0861b586f491d62fe8ed3c6ba6
d46ebb72760397af5f8b21fa51748aac02ede09b
F20110320_AABVPF margolin_s_Page_45.tif
31521d9a15608fd6a2499ddc97c347d8
87f070e1df040616105c24f2b7d6193bb0ec3334
2178 F20110320_AABVOS margolin_s_Page_73.txt
59c1c73322f1cb56ec738fcec5fe9c93
c29b51df75414f4b9b0722594efc399e44efe3d2
68727 F20110320_AABVPG margolin_s_Page_07.pro
2a2b31ae5257c037851d4c7f90ce39d4
66d26762cb464d63b7e11ac15602ec214a3138b7
F20110320_AABVOT margolin_s_Page_78.tif
42312828127949e8a98b9718a02ea4a4
61361fe2396cb6c6c042022dc5981de7b162c1b9
114250 F20110320_AABVPH margolin_s_Page_79.jp2
9d414fc4119b1299b36737da30985a94
4374c3d3a824e1bbba60fa10c1b061b5a1b00c74
55376 F20110320_AABVPI margolin_s_Page_32.pro
287c564662915d317be85f9982b6ad8a
4ccc09cf5ad1f4fdfb63f43c3a677efeab196061
48168 F20110320_AABVOU margolin_s_Page_10.pro
bc221bdd84c99911508aa2522409f237
77ef18fe0a4db8590a4e25d0fe147b5fe9083802
F20110320_AABVPJ margolin_s_Page_73.tif
929eeb9d97f9644156682aed6be4d2ef
5120d173e463e7b26517a190eb6144d9444e3f27
25379 F20110320_AABVOV margolin_s_Page_71thm.jpg
76838845d55782a5e8e578ff4aec52d7
6a47cef853e37388958ee2b20ad1b59c3b3048e3
5674 F20110320_AABVPK margolin_s_Page_03.pro
9659e03e542a79f9e69e6f60ae07b97b
d71ef80c52c126b853dd3ab4a0c247e4332e34b5
204246 F20110320_AABVOW margolin_s_Page_06.jpg
79c96189db6de8dbfbc8fe1a91eacb80
7539a76e65d9487c3544070f20cf78b924344d6e
83129 F20110320_AABVQA margolin_s_Page_12.QC.jpg
c3765b989e74aff751534110e0d57a99
ae7399857610a1b0e361243717f29ff902ede5e8
50301 F20110320_AABVPL margolin_s_Page_34.pro
3dc8282e06931f52020f0715fb115ee6
b6c7b427c7f742fe60fb655aec620b4cbd1214e5
111019 F20110320_AABVOX margolin_s_Page_53.jp2
b323e1b1443b8fbc5c2880e15c4d17c4
6f0be7c2e399549093d15116d587dc8e1ed9fe51
25174 F20110320_AABVQB margolin_s_Page_12thm.jpg
49c6c4a5b8fcee77e44dc8ec5ba3a0e9
b8418327c6a8ce06b22705e77a3d61e0c405a096
85777 F20110320_AABVPM margolin_s_Page_24.QC.jpg
731d95187be72b4139dc150b64bdcadb
f69bdf594f969d9087bccf5aba933005ce590e29
231259 F20110320_AABVOY margolin_s_Page_40.jpg
ef83a379fe6ed2e7d88a49c407e55e75
6a39a1623e7fa4078bf2e18243ef24c2fab5327f
F20110320_AABVPN margolin_s_Page_19.tif
9823cf66a82619d4ffc11e408178b7cc
4bf8e9c66a846a23fe351a1b0732890025f7c891
F20110320_AABVOZ margolin_s_Page_44.tif
99b6c3295cbaa5417c23b8cdeff62084
a812274aa81c896a4874feecbd8144d74fa9e2a9
116349 F20110320_AABVQC margolin_s_Page_32.jp2
d2fdee7bf8c65a8098ad3b6eeb4f0c69
c8ca97cba4b2dcf29f16000aa528f658c3968160
115595 F20110320_AABVPO margolin_s_Page_23.jp2
bf820160e503f957628d0080807b9516
104cd095fef088e73c97ab98b37a9c81efe8e112
89559 F20110320_AABVQD margolin_s_Page_37.QC.jpg
c6cefaff8a8a4a5ed58d81c3c33a0c81
9081b94ad054891718adba7b965c4c1b94c66a39
2077 F20110320_AABVPP margolin_s_Page_54.txt
f4ebfe02f03258b845497e9efda9c16c
31e5fccfcfc07ecf5f5ff837f4549015e0f79ac3
55881 F20110320_AABVQE margolin_s_Page_50.pro
9315ededee23c2b8d7b874e2c0a43413
b5c76058540935bfeb5ef3d53ead9a20c0a299a3
239547 F20110320_AABVPQ margolin_s_Page_60.jpg
78194eb5f16bb2988a8845edc64b5d0d
68976b65b56cc0168efa3d88e8b797df69e552fe
9114 F20110320_AABVQF margolin_s_Page_11thm.jpg
ff75c38899f10c6f2337aa8eaf4ebc4b
a1dfb081cbdd652849e869d8a393630a1eb423f9
138724 F20110320_AABVPR margolin_s_Page_47.jpg
72fb885a3f98e830b9f77eaf65836acb
e8e3ced1e4a70768a11a2645dbe857126bfa0695
2066 F20110320_AABVQG margolin_s_Page_55.txt
6b1a894789cac880f0d8246e9fcf8534
91213635c0044bb650421ac49a3f9f911b383c46
216929 F20110320_AABVPS margolin_s_Page_35.jpg
df54d0c91282b8ec0822388731c62772
a957e88fe8a0279402aceec7173eaf0f3ea9042a
56762 F20110320_AABVQH margolin_s_Page_25.pro
7952f2d58a8c1674ac0bcb3128d25624
2066df8c58b4a3b80c7c2da4facbf991ed962331
7372 F20110320_AABVPT margolin_s_Page_70.pro
16a53840b863f9ca626c6d773b1d4ff9
6fb3f6ffe9aef38d741b08a14dc51e865c463daa
611 F20110320_AABVQI margolin_s_Page_04.txt
a89521703c3215523908f833c7fcd92d
5a89c7b793f9e786c4baf7bc420e5501eaccc669
2149 F20110320_AABVPU margolin_s_Page_71.txt
236ac92caf2779033b0220ae5799e82f
927db946e4dea9f67e0ec8d69d8b9b993233a634
F20110320_AABVQJ margolin_s_Page_75.txt
4629f5db244375b83bb2d31f86de23d4
fe5ee622057d586900ed56e3cde36c3e52da8446
35466 F20110320_AABVPV margolin_s_Page_04.jp2
2341f3bf715c95693ba9ba5516d0590c
931f4dec436f10a476a7feaa84f66d2c33cdff7c
119740 F20110320_AABVQK margolin_s_Page_14.jp2
f076d546866d3cdb9c23780cd5a362d4
51a85ee12a310a1b320e0ad6be9d1fe1f72b65c3
26538 F20110320_AABVPW margolin_s_Page_73thm.jpg
d8ae8d04f2d270d6ebaf512225ba99ce
379ee1f8792ae819d0af4cb7f026e6404f169e72
76912 F20110320_AABVQL margolin_s_Page_51.QC.jpg
98efdcb8492fc0c7dc19ae1d79e14be7
5fd2a37b01a8ef15a529143e2b42ec9b26bcd965
78289 F20110320_AABVPX margolin_s_Page_52.QC.jpg
59918a1b5833d21dec3c6cfc29f019e8
3d21e822f0df33192f173c016d26f7b22ba493ac
F20110320_AABVRA margolin_s_Page_57.tif
ba9e49f7ce8bb8dd6216e15c53414c8a
7a942682d68ceb57566b05794aaeee9eec0b5fdb







NEGATION COMPREHENSION AND AGING: THE ROLE OF INHIBITION AND THE
IMPACT OF COMPENSATION DURING READING




















By

SARA J. MARGOLIN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007

































2007 Sara J. Margolin

































To those in my life who have encouraged, supported, inspired, and helped me achieve my
goals-It is with gratitude that I complete this work, and dedicate it to my husband, my family,
and my dearest friends.









ACKNOWLEDGMENTS

I thank the young and older adults who participated in my experiments; their help and

generosity are deeply appreciated. A special thank you goes to Jennifer Eastman, Max Thomas,

Giovanna Morini, and Lindsey Bartlett for their assistance in stimuli development and data

collection. I gratefully thank my committee members for their thoughtful insights into guiding

me through this project. Above all, I thank Lise Abrams, whose attention, comments, and

guidance made this project possible and the best it could possibly be.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F TA B L E S ....................... ....... ............................................................... 7

LIST OF FIGURES .................................. .. ..... ..... ................. .9

A B S T R A C T ......... ....................... ............................................................ 10

CHAPTER

1 INTRODUCTION ............... .............................. ............................. 12

The Impact of Negation on Reading Comprehension .................................... ...............15
Theories of Language Comprehension and Aging ............................................................17
S p e c ific A im s .................................................................................................................... 2 1

2 EXPERIMENT 1 ............................... ...... ... ..................23

H y p o th e sis 1 ........................................................................2 3
H y p o th e sis 2 ...........................................................................................................................2 4
Methods ........................................ 26
P articip an ts ................................................................2 6
D e sig n ................... ...................2...................6..........
M a te ria ls ..........................................................................................................2 6
P ro c e d u re ......................................................................................................................... 3 0
R results ..... ......... .... ........................... .................................. 3 1
Demographics, Operation Span, and Stroop Test ........................................ ....31
R e a d in g T im e s ........................................................................................................... 3 2
O v erall read in g tim es ...............................................................................................3 3
Target w ord reading tim es............................................ ......... 33
Post-target word reading tim es ................ ....................................... ... 34
Modifier versus target reading times for negative sentences only ........................34
Probe N am ing Tim es ..............................................................................35
Com prehension A accuracy .................................. ... .....................36
The Relationship between Reading Times, Naming Times, and Comprehension
A c cu ra cy ................................................................ 3 8
D iscu ssio n ................... ...................3...................9..........

3 EXPERIMENT 2 ............................................................................... 48

H y p o th e sis 1 ........................................................................4 9
H y p o th e sis 2 ...........................................................................................................................5 0
Methods ........................................ 51
P artic ip an ts ................................................................ 5 1










D e sig n ................... ...................5...................1..........
M a te ria ls .......................................................................................................... 5 1
P ro c e d u re .........................................................................................................................5 2
R results ...... ....... ........... .................. ... .................... ........ 53
Demographics, Operation Span, and Stroop Test ........................................ ....53
R e a d in g T im e s ........................................................................................................... 5 3
O v erall read in g tim es ...............................................................................................54
Target word reading times............................................ .........54
Post-target word reading times ............................................... .............. ....... 55
Modifier versus target reading times for negative sentences only ........................55
P rob e N am ing T im es ............................................................................... 56
C om prehen sion A accuracy ......................................................................................... 56
D iscu ssio n ................... ...................5...................9..........

4 GENERAL DISCUSSION ................................................. ........ ..................71

L im itatio n s ................... ...................7...................6..........
F utu re D direction s ................................................................79

APPENDIX

A S A M P L E S T IM U L I ............................................................................................................... 82

B EXPERIMENT 1 INSTRUCTIONS .............................................. ...............83

C EXPERIMENT 2 INSTRUCTIONS .............................................. ...............84

L IST O F R E F E R E N C E S ............. ................. ...............................................................85

B IO G R A PH IC A L SK E T C H ................................................................................................... 89























6









LIST OF TABLES


Table page

2-1 Length and frequency averages for each type of probe word........................................44

2-2 Means and standard deviations (st. dev.) of demographic measures ..............................44

2-3 Means and standard deviations for overall reading times (msec)................................44

2-4 Means and standard deviations for target word reading times (msec)............................45

2-5 Means and standard deviations for post-target reading times (msec)..............................45

2-6 Means and standard deviations for modifier and target word reading times for
negative sentences only (m sec)............................................................... .....................45

2-7 Means and standard deviations of probe word naming times (msec)..............................45

2-8 Means and standard deviations for comprehension accuracy (%) ................ .................46

2-9 Means and standard deviations for comprehension accuracy with both
sentence type and probe type (% ) ..................................... ................................. 46

2-10 Results of regression analysis on negative sentence comprehension accuracy .................46

2-11 Results of regression analysis on non-negative sentence comprehension
accu racy ...................................... ...................................................... 4 6

3-1 Means and standard deviations (st. dev.) of demographic measures ..............................65

3-2 Means and standard deviations for overall reading times (msec)................................65

3-3 Means and standard deviations for target word reading times (msec)............................. 66

3-4 Means and standard deviations for post-target word reading times (msec)....................66

3-5 Means and standard deviations for modifier and target word reading times for
negative sentences only (m sec)............................................................... .....................66

3-6 Means and standard deviations of probe word naming times (msec).............................67

3-7 Means and standard deviations for comprehension accuracy (%) ..................................... 67

3-8 Means and standard deviations for comprehension accuracy by probe type (%)..............68

3-9 Results of regression analysis on negative sentence presented with symbol
com prehension accuracy .............. ............................................ ... ............... 68









3-10 Results of regression analysis on negative sentence presented without a symbol
com prehension accuracy .......................................................................... .. ..................69

3-11 Results of regression analysis on non-negative sentence presented with symbol
com prehension accuracy ......................................................................... .................... ..... 69

3-12 Results of regression analysis on non-negative sentence presented without symbol
com prehension accuracy ......................................................................... .................... ..... 69









LIST OF FIGURES


Figure page

2-1 Target word reading times as a function of age group and sentence type .......................47

2-2 Comprehension accuracy for questions after each probe type as a function of
sentence type. .............................................................................47

3-1 Post-target reading times by sentence type and presence of symbol...............................70









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

NEGATION COMPREHENSION AND AGING: THE ROLE OF INHIBITION AND THE
IMPACT OF COMPENSATION DURING READING

By

Sara J. Margolin

August 2007

Chair: Lise Abrams
Major: Psychology

The present research investigated the influence of several factors on young and older

adults' reading comprehension of real-world texts. The first experiment determined the effects of

negation on comprehension of written sentences and examined the level of text representation at

which negation most severely impacts comprehension. The second experiment assessed the

effectiveness of a specific compensatory strategy, the use of a negation symbol, in

comprehending negation. Using online, activation, and content measures of comprehension,

participants read sentences modified from fiction novels. Participants read sentences one word at

a time at their own pace, then immediately named a probe word that was verbatim from the

sentence, related to the sentence, or unrelated. Participants then answered a comprehension

question about the sentence they had read. The results showed that negation affected

comprehension in both age groups, such that sentence reading times were faster for negative

sentences, probe naming times were slower following negative sentences, and comprehension

accuracy was lower for negative sentences, relative to non-negative sentences. Negation

equivalently slowed verbatim and related probes' naming times, suggesting that negation had

similar effects on the situation model and surface levels of representation. However, these effects

of negation were not exacerbated with age, although older adults' comprehension overall was









poorer than young adults. With respect to compensation, providing the negation symbol during

reading did not facilitate the processing of negation specifically but instead encouraged readers

to focus more attention on all of the sentences. These findings suggest that age and inhibition are

less important in influencing negation comprehension than working memory. Furthermore, the

lack of age deficits in negation comprehension has positive implications for older adults'

functioning in a world where negation is present in many contexts, such as prescription labels

and road signs.









CHAPTER 1
INTRODUCTION

The present research explores a topic that concerns an activity people engage in every day

through their interactions with other people and the world: language comprehension. In tasks

such as reading or conversing with friends and co-workers, processing and understanding the

language with which we are presented becomes critical for functioning. However, there are some

significant differences between comprehension of spoken versus written language.

Comprehension of spoken language is aided by various cues, such as feedback provided from a

conversational partner, or prosodic cues, such as intonation. In contrast, reading is dependent on

the reader and his/her cognitive capacities. The process of reading is also complicated because it

involves various levels of comprehension (e.g., surface level, text base level, situation level) to

reach full understanding of the text (e.g., Kintsch, 1998). Given that reading is important for

functioning under many real-world circumstances (e.g., menus, road signs, voting, medication

labels, and instruction manuals), language comprehension through reading becomes important

for study, and the variables that may impact the ability to perform such a task need to be

examined.

Research investigating language comprehension through reading has used various

experimental tasks to study the processes that take place during reading, the information from the

text that is stored in memory, and the kinds of representations that are formed from the

information presented in the text. Zwaan and Singer (2003) reviewed various techniques in the

literature, including "online" measures like self-paced reading and the moving-window

technique, "activation" measures like lexical decision, recognition, and naming, and "content"

measures like the think-aloud protocol, question answering, and sentence completion, to describe

their use in examining reading through different means.









Online measures of reading comprehension (e.g., Kemtes & Kemper, 1997; Stine, 1990;

Stine-Morrow, Loveless, & Soederberg, 1996) test reading as the process is occurring. Self-

paced reading presents segments of text (e.g., word, phrase, or sentence) on a computer screen at

a rate controlled by the reader. The reader can move from one segment of text to the next by

pressing a key on the keyboard. For the self-paced reading technique, only the segments of text

presented for reading are present on the screen, while no other information about the length of

the upcoming words or length of the passage is given. This presentation is different from the

moving-window technique, another type of self-paced reading, where an entire passage is

presented on the screen with all of the letters replaced with dashes, which give clues about the

length of the words to be read. The dependent measure of both self-paced reading and the

moving-window technique has traditionally been reading times between key presses. Longer

response times provide evidence for increased difficulty in reading that piece of text.

In activation measures (e.g., Hamm & Hasher, 1992; Radvansky & Curiel, 1998;

Radvansky, Zacks, & Hasher, 2005), the accessibility of the information that is learned from the

text is tested. The term "activation" describes how accessible the text's meaning, or occasionally

verbatim text, is in memory. If the meaning or verbatim text is readily available (i.e., still

activated), then that piece of information is considered an important part of the mental

representation of the text. One activation measure often used is the naming measure. In the

naming measure, a word, either from the text or semantically related to a word from the text, is

presented after reading. The participant is required to say aloud the word shown. The dependent

measure for activation tasks is the time required to say the presented word. The faster the

response times, the more available the word is in the mental representation of the text, indicating

its importance to the situation described in the text.









Content techniques (e.g., Radvansky, Copeland, Berish, & Dijkstra, 2003; Radvansky,

Curiel, Zwaan, & Copeland, 2001) measure the kind of representations that are being formed or

have been formed about the text (depending on when the test is given), as well as the information

that is contained in that representation. A common content measure involves memory and

includes free or cued recall. Recall tests require participants to either reconstruct the text that

they read from memory (i.e., free recall) or to fill in a missing part of the text given a portion of

the text as a cue (i.e., cued recall). Unlike the other measures, content techniques focus on

accuracy, where inaccurate responses are evidence that the text may not be stored clearly or

correctly in memory.

In the aging literature, older adults' abilities have been examined through a variety of these

methodologies, often combining methodologies to expand the scope of exploration. General

findings from online measures suggest that older and young adults allocate their reading time

differently (e.g., Connelly, Hasher, & Zacks, 1991; Kemtes & Kemper, 1997; Stine, 1990; Stine-

Morrow et al., 1996; Waters & Caplan, 2001). Older adults are more likely to pause at both

clause boundaries (i.e., the end of a single idea in a sentence, often indicated by the presence of a

comma) and sentence boundaries (indicated by a period), where young adults will typically

pause only at sentence boundaries. The different strategies for older adults' allocation of reading

time do not suggest that older adults were less able to read and understand text, but rather that

their approach in doing so was different. In activation techniques (e.g., Hamm, & Hasher, 1992;

May, Zacks, Hasher, & Multhaup, 1999; Radvansky & Curiel, 1998; Radvansky et al., 2003;

Radvansky et al., 2001; Radvansky et al. 2005;), older adults have shown little, if any, deficits in

comprehension (for a review, see MacKay & Abrams, 1996). While older adults may be slower

to access information from the text, their accuracy in accessing that information is similar to









young adults. In content measures (e.g., Kemtes & Kemper, 1997; Stine, 1990; Stine-Morrow et

al., 1996), older adults show some deficits relative to young adults in that they often recall less

specifics from the text, although they usually can retrieve the gist of the text processed during

reading. However, these techniques rely on retrieval from memory, so age differences in content

measures may actually result from age-related declines in memory, rather than difficulties in

comprehension (e.g., Stine-Morrow et al., 1996).

Using these techniques, research on language comprehension has examined many variables

that influence a reader's comprehension of text (e.g., syntactic complexity, negation, lexical

ambiguity, morphological complexity) and provides evidence of how and when difficulties occur

in the reading process. Studies of the above variables give insight into the mechanics involved in

the reading process through an examination of constructions that naturally occur in language and

that may increase difficulty in understanding. One specific focus of past research has been on

young adults' comprehension of negation in text; however, older adults' comprehension of

negation has not been explored. The examination of negation as a variable in language

comprehension is important for older adults because of the impact it may have on reading

prescription medication labels, road signs, and even voting forms, all of which may contain

negation.

The Impact of Negation on Reading Comprehension

The general findings in the negation literature are that sentences that include negation take

longer to process during reading and that young adult readers show poorer comprehension when

sentences contain negation compared to sentences that do not contain negation, i.e., non-negative

or affirmative sentences (e.g., Cornish, 1971, Experiment 1; Cornish & Wason, 1970; Hoosain,

1973; Just & Carpenter, 1971; Kaup, 2001; Kaup, Dijkstra, & Ludtke, 2004; Kaup, Ludtke, &

Zwaan, 2005; Kaup & Zwaan, 2003; Kaup, Zwaan, & Ludtke, in press; MacDonald & Just,









1989; Sherman, 1973). Both activation and content measures have shown that the presence of

negation slows responses times to and decreases accuracy of comprehending negated concepts

(Cornish, 1971; Cornish & Wason, 1970; Just & Carpenter, 1971, Experiment 1; Kaup, 2001;

Kaup et al., 2004; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup et al., in press; MacDonald &

Just, 1989, Sherman, 1973). For example, MacDonald and Just (1989) examined the impact of

negation on comprehension of a sentence via two activation measures, probe recognition (where

participants indicated whether they recognized a word as having been in the sentence they had

just read) and probe naming (where participants said the probed word out loud). Results showed

that participants' responses to both recognizing and naming a probe were slower if the probe

word had been negated in the sentence they had read. These results suggest that negation may

reduce the availability of negated concepts in the mental representation by deactivating the

negated concept.

Further examination of the availability of negated text has been examined by Kaup (2001).

In two experiments, participants completed a probe recognition test, where the probe word was

part of a definite/indefinite noun phrase from the sentence (Experiment 1; e.g., "the table" or "a

table") or was a creation/destruction verb used in the sentence (Experiment 2; e.g., "bake" or

"burn"). In the first experiment, the meaning of the probe word (e.g., "table") was the same, but

the specificity of the word before it (e.g., "the" vs. "a") differed between the two examples. In the

second experiment, the meaning of the probe words was different. The results for Experiment 1

showed longer response times and lower accuracy for recognizing probe words that had been

negated relative to non-negated probe words, but the type of noun phrase (definite vs. indefinite)

had no effect on the speed or accuracy of responses. The results of Experiment 2 showed slower

recognition times for negated probes than non-negated probes, with a greater difference for









creation verbs. Together, these findings suggest that negation has greater effects on

comprehension of text at the level of meaning rather than the specific word level.

Similar effects of negation have been found with content measures. For example, Cornish

and Wason (1970) examined negation's effect on recall, i.e., whether negative sentences would

be recalled less than positive sentences. Results showed that affirmative sentences (e.g., "It is

bright") were recalled more often than the negative sentences (e.g., "It is not bright"), and more

errors were made on the negative sentences. These errors often included changes in syntax but

not meaning (e.g., recalling "It is not bright" as "It is dull"); the negation word was deleted,

changing the sentence structure, and was instead included in the meaning of another word (i.e.,

"not bright" is recalled as "dull"). The implications of these results are that negation not only

makes ideas less accessible during reading, but it also makes them difficult to store and retrieve

from memory.

Theories of Language Comprehension and Aging

Previous research on reading comprehension has been framed in terms of several theories.

The general theories of comprehension give an overview of how information from reading can

be represented in memory and how comprehension may progress independent of age (e.g., Just

& Carpenter, 1992; Kintsch, 1998). The aging theories then focus on how language

comprehension can be influenced by variables such as working memory, which may

fundamentally change across the lifespan (e.g., Hasher & Zacks, 1988; Light, 1988).

One theory of comprehension, a capacity theory of language comprehension (e.g., Just &

Carpenter, 1992), explains how working memory resources are critical for language

comprehension during reading. The principal element of this theory is that the amount of

information that can be held in working memory, i.e., capacity, influences the degree of

comprehension during reading. However, there is an inherent limit to how much working









memory can hold. Difficulties arise when the capacity needed to comprehend a sentence exceeds

an individual's working memory capacity. Context may relieve the burden on working memory

to some degree. When sentences are read in context, i.e., a framework of text involving related

ideas, there is some amount of pre-activation of ideas (i.e., preparation for what ideas are to

come next), which lessens the amount of capacity needed to process subsequent ideas. This

theory's basis in working memory capacity has implications in the aging literature such that older

adults are believed to have smaller working memory capacities (e.g., Light, 1988; Waters &

Caplan, 2001), and thus may show decrements in processing sentences that tax their working

memory limitations.

Given that negation seems a pervasive construct that is likely to cause difficulty in

comprehension, it is important to understand how it is processed. Negation may tax working

memory, as processing negation may involve more steps than reading non-negated text (e.g.,

activating and then deactivating a particular concept). MacDonald and Just (1989) proposed that

when we read negation, all ideas that are negated become deactivated. For example, after reading

the statement "Elizabeth baked some bread but no cookies," "cookies" becomes deactivated and

therefore is removed from working memory. Additional time and processing resources are

needed to deactivate the negated concept, which makes processing of negation more difficult.

Various types of negation may differentially burden working memory. The type of

negation that shows the most detrimental effect on reading comprehension is the negation that

involves changes in syntax (i.e., adding "not" into the sentence; Just & Carpenter, 1971,

Experiment 1; Sherman, 1973). For example, Just and Carpenter (1971) asked participants to

verify sentences in comparison with a presented picture and examined three types of negation:

explicit negatives, implicit syntactic negatives, and semantic negatives. Explicit negatives were









syntactically negative because they overtly included the negative by use of a negative modifier

(e.g., "The dots are not red"). Implicit syntactic negatives were negatives that did not include a

negative modifier but described a subset of items that were negated (e.g., "Few of the dots are

red") and included a word that can be used in different syntactic structures; for example, in the

sentence "Few of the dots are red," "few" can either indicate that a few dots are red or that some

other amount of dots are not red. Semantic negatives also refer to a subset of items but do not

contain a syntactic property for marking the negation (e.g., "A minority of the dots are red"). The

results showed that response times were longer following explicit and implicit syntactic

negatives relative to semantic negatives, suggesting that syntactic negatives are more difficult for

comprehension than other types of negatives and that negation can be viewed as a syntactically

complex structure.

Another theory of comprehension is the Construction Integration (CI) Model (Kintsch,

1998). In this model, a network is formed during reading, containing a "text representation" of

the newly read information. For the CI Model, there are three levels of representation that a text

can take in memory. The first of these representations is the "surface level," where the

representation is only the verbatim words themselves. The second level is the "textbase," which

represents the words as well as their meanings. The highest level of representation is the

"situation model," including the meaning of the text as well as conclusions and inferences drawn

from the text. While all readers use the surface level, textbase level, and situation model level for

processing and storage of text information, there are some age differences in the representations

on which readers tend to rely. There is evidence that older adults are better able to understand

and store text at the situation level than at the surface and textbase level, while young adults are

better able to store text at the surface and textbase levels (e.g., Radvansky et al., 2001). Given a









limited working memory capacity, older adults may continue only store the meaning/situation

model instead of the lower levels of representation. Young adults may tend to store exact

wording as well as the situation model representation because they have additional working

memory resources, which leaves more room to store additional information which may increase

the specificity of the memory representation of the text.

This theory involving different levels of representation of text (e.g., Kintsch, 1998; Zwaan

& Radvansky, 1998) can be applied to the comprehension of negation. In the propositional nodal

network, the ideas in the sentence that were activated during reading are subsequently

constrained to deactivate unneeded information. In the case of negation, the negated word is not

needed in the situation that is described. For example in the sentence, "Elizabeth baked some

bread but no cookies," "cookies" are not necessary for representing what happened in the

situation described in the text. In other words, not all of the concepts that are presented in the text

are necessary for the situation model and therefore do not need to remain as highly activated,

allowing a more accurate representation of the situation to be created. The CI Model as applied

to negation can explain some of the previously-reviewed studies. For example, the results of

Kaup (2001), where negation had an impact on comprehension only when the meaning of the

text was changed, suggest that negation affects the situation model representation of the text

rather than the surface level. The results from Cornish and Wason (1970) also suggest that the

meaning of the negation is more important than specific wording, and that readers remember the

gist or the situation model representation instead of the exact words.

A final theory, related to Just and Carpenter's capacity theory of working memory that

relates working memory difficulties to aging, is the Inhibition Deficit (ID) Theory (Hasher &

Zacks, 1988). This theory proposes that inhibition from working memory is the focus of









problems in comprehension in aging. When inappropriate information is held in working

memory, unnecessary amounts of working memory capacity are taken up and are not available

for appropriate information processing. If inappropriate information gets into working memory,

it is the duty of the inhibitory mechanism to deactivate that information (i.e., get it out of

working memory). However, if the inhibitory mechanism is not working properly, inappropriate

information will not get de-activated as quickly, and the resulting capacity remaining to process

the appropriate information from the sentence is diminished. Hasher and Zacks (1988)

specifically proposed that older adults have inefficient inhibitory mechanisms. Therefore, older

adults may not be able to efficiently or appropriately deactivate a negated concept that is

unimportant to the situation model representation of the text.

Specific Aims

In light of the potential for difficulty in older adults' comprehension of negation, the aims

of the present studies are to: (1) determine the effect of negation on older adults' comprehension,

(2) assess the function of inhibition in negation processing, (3) demonstrate the impact of

negation on multiple levels of text representation, and (4) test the effectiveness of a

compensatory strategy in facilitating young and older adults' comprehension of negation. All of

these aims were tested within the context of texts that readers may encounter in real life, as

opposed to texts that have been created by an experimenter, a technique not previously used in

negation research. Many studies of negation assume that the simple laboratory experiments

extend to real-life situations. For example, Cornish (1971, Experiment 1) had participants

compare negated sentences to a multi-colored circle (e.g., participants verified the sentence "The

dot is not red" in reference to a picture they saw). This task seems very different from the way

that negation is encountered in everyday life, such as through reading a novel, newspaper, or

even a prescription label. These experiments may actually be testing logic in that comparing









sentences to pictures may be more reminiscent of "if-then" reasoning than reading

comprehension. As Neisser (1978) suggested, a more ecological approach should be taken

toward laboratory research because no practical impact can be made from research that is not at

least partly based in reality. By using logic sentences (e.g., "The dot is not red"), previous

experiments have not sufficiently asked or answered questions of comprehension, such as: How

does negation affect our reading? How does negation affect our representation of that text? How

does negation affect our retention and retrieval of the information we have read?

The present research will demonstrate whether young and older adults are affected by

negation when reading more real-world types of texts. The next chapter will describe the first in

a series of two experiments, which use online, activation, and content techniques to more

thoroughly assess negation's influence on reading comprehension.









CHAPTER 2
EXPERIMENT 1

The purposes of Experiment 1 were to determine the effects of negation on young and

older adults' comprehension of negation and to examine the level of text representation at which

negation most severely impacts comprehension. Specific hypotheses are presented below.

Hypothesis 1

Older adults will have more difficulty processing negation than young adults because of an

increased difficulty with inhibition. As proposed by the IDH, older adults will have trouble

inhibiting the concept that is modified by the negation, resulting in the concept's inclusion in

their memory representation of the text. As an older reader takes in the information from the text,

he/she will be less likely to disregard the negated concept and will include the concept in their

memory representation of the text as if it were never negated. Representing the negated concept

(which should not be represented because it is not relevant in the situation described by the text)

takes up some working memory resources. According to Just and Carpenter's capacity theory of

working memory, using working memory resources for processing negation would cause

problems storing the rest of the appropriate/relevant content of the text in working memory,

especially when that text is difficult to process and would require a lot of resources. The result is

that for older adults, whose working memory resources may already be limited (e.g., Light,

1988), an inhibition deficit will result in incorrectly representing the negated concept in memory

and prevent appropriate processing of the text as a whole. The alternative hypothesis is that older

adults will not show a difficulty with inhibition as it relates to negation processing, contrary to

the IDH. Burke (1997) suggested that older adults can demonstrate inhibition to the same degree

as young adults, such as in completing sentences (e.g., Burke & Harrold, 1988), metaphor

comprehension (e.g., Newsome & Glucksberg, 2002), or picture descriptions (e.g., James, Burke,









Austin, & Hulme, 1998). Therefore, older adults may represent negation appropriately and be

able to prevent the negative concept from being included in their mental representation of the

text.

The potential for increased difficulty in negation processing for older adults compared to

young adults may be caused by factors other than inhibition, even if working memory limitations

are still relevant. Both Gough (1965) and Mehler (1963) suggested that negation, as well as any

type of complex sentence, can be processed in terms of its simpler form (called its "kernel") and

then transformed back to the original, more complex, form later for retrieval or verification. In

terms of negation, this idea is that a negative sentence would be processed in terms of its

affirmative counterpart and then marked with a "footnote" that indicates that the sentence was

negative. Later, both the sentence and its footnote are retrieved, and the sentence is reversed into

its negative interpretation once again. Older adults would have difficulty in this instance because

they would need to hold both the affirmative form of the sentence and the note about its

transformation in their working memory. This additional information would take up working

memory resources and cause difficulty, not because the reader was unable to inhibit the

necessary information to make an appropriate representation, but rather because older adults'

working memory capacities are limited. Regardless of the mechanism for making negation

difficult to process, limited working memory resources for older adults would be

counterproductive in processing negation.

Hypothesis 2

Older adults will show the most difficulty in representing negation at the situation model

level of representation. The situation model, i.e., the highest level and most meaningful level of

text representation, is the level of text where information is stored and used for many tasks,

including learning. In relation to aging, there is evidence that older adults more often understand









and store text at the situation level than at the surface and textbase level, while young adults store

text at the surface and textbase levels (e.g., Radvansky et al., 2001). Given that older adults

represent information from text at the situation model level, this level may be the most affected

by any difficulty in representing and remembering the information presented in the text.

The research on young adults' negation comprehension suggests that negation most

adversely affects the situation model representation of text (Kaup, 2001). In addition, older

adults most easily remember text from the situation model because it is the most meaningful

representation, and older adults' limited working memory capacity prevents them from

necessarily retaining multiple representations of the text (e.g., Radvansky et al., 2001).

Combining these results with the prediction that older adults may have more difficulty correctly

understanding negation than young adults, older adults' difficulty in processing negation is

expected to be most evident at the situation model level. To measure the availability of that level

of representation in a reader's memory for that text, activation measures will use words

representing the situation model level (i.e., words that show the gist meaning of the text) as well

as the surface level (i.e., words verbatim from the text).

An alternative hypothesis can be made based on older adults' expertise in using situation

model representations and their more extensive practice at processing at this higher and usually

important level (Radvansky et al., 2001). Given that older adults primarily store and retrieve text

at the situation model level, they may be more experienced at using this level of representation.

Thus, older adults may be able to offset, i.e., compensate for, a difficulty that would occur at the

situation model level with their extensive experience. The result would be that older adults

would be relatively unimpaired by the negation at the situation model level, compared to

processing at the surface and textbase levels. While this prediction also supports the idea that









older adults tend to process and store text at the situation model level, it differs in that older

adults will be able to use their experience to offset their difficulties with negation at levels in

which they best process information from text.

Methods

Participants

Fifty-four young (40 female and 14 male) and 54 older adults (35 female and 19 male)

were tested. The young adults ranged from 18 to 23 (M= 19.09, SD = 1.22) years of age and

were recruited from introductory psychology classes at the university. The older adults ranged

from 64 to 87 (M= 74.59, SD = 5.98) years of age. The older adults were community-dwelling

older adults, recruited from churches, clubs, and libraries surrounding the university. Participants

were fluent speakers of American English and reported that they not previously read the "Harry

Potter" series by J.K. Rowling or seen the corresponding movies.

Design

The present experiment used a 2 x 2 x 3 factorial design with age group (young and older),

sentence type (negative and non-negative), and probe type (verbatim, related, and new) as

factors. Sentence type and probe type were within-subjects factors, and age group was a

between-subjects factor. The dependent variables were the reading times for words in the

sentence, the naming time for the probe words, and comprehension question accuracy. Reading

times measured online processing (i.e., processing during reading), naming times measured the

activation levels of the probed words, and comprehension question accuracy indicated how much

and how accurately a reader retained the content of the text specifically relating to the negation.

Materials

One hundred and two experimental sentences and 30 filler sentences were used in the

present experiment. Experimental sentences came from the "Harry Potter" series by J.K.









Rowling. As reported by Laurie May of the Alachua County Library System, this series was

selected as text that both young and older adults would like to read, based on observations from

library employees and rates of check-out (L. May, personal communication, June 1, 2006). In

addition, an article in the Gainesville Sun newspaper reported "Harry Potter" as one of the young

adult books that other age groups are beginning to enjoy reading ("Grown-ups are turning to teen

books," 2006). Sentences were modified to include or exclude negative modifiers, and changes

in grammar were made as necessary. Names of characters were also changed to obscure the

source of the text and to avoid participants bringing preexisting ideas about what they would be

reading. Each sentence (e.g., "Trying very hard to not imagine what Katie might look like if they

found her, Adrian led the way forward. ") contained a target word, e.g., imagine. Each target was

preceded by either a negative modifier (e.g., not imagine) or no modifier (e.g., imagine). A

negative modifier was defined as a word that denotes the absence of the word that it modifies.

Target words were of varying parts of speech, including verbs in 63.7% of sentences, nouns in

10.8% of sentences, adverbs used in 1% of sentences, and adjectives in 24.5% of sentences.

Modifiers were "no," "not," and "never," with "no" used in 11.7% of sentences (7.7 %

modifying verbs, 84.6% modifying nouns, and 7.7% modifying adjectives), "not" in 87.3% of

sentences (71.6 % modifying verbs, 1.1% modifying adverbs, and 27.3% modifying adjectives),

and "never" in 1.0% of sentences (100% modifying verbs).

Each sentence was associated with three possible probes for the naming task. The verbatim

probes were identical to the targets, e.g., "imagine." The related probes were semantically related

to the target, e.g., "dream," specifically synonyms of the target word, as defined by a thesaurus,

which could replace the target in the text without a change in meaning. The new probes were

words that were not present in the sentence, were unrelated to the situation described by the text









(e.g., "adjust"), and could not replace the target word in the sentence without a significant

change in meaning. All probe types were relatively high in frequency (e.g., Francis & Kucera,

1982), and were similar in length (Table 2-1). The presentation of each type of probe word was

counterbalanced across participants, along with negative modifiers and no modifiers, creating 6

versions of stimuli with 17 sentences in each condition. Each sentence was also associated with

one comprehension question, specifically focusing on the target and its representation in the

situation model of the text. Appendix A displays sample stimuli, including sentences in both the

negative and non-negative conditions, their corresponding probes (verbatim, related, and new),

and comprehension questions. In addition, 30 easy, non-negative filler sentences, with probes of

all three types and comprehension questions, were presented to build confidence in less skilled

readers who might have had more difficulty with the experimental comprehension questions and

also to hide the pattern of questions always referring to the negative portion of the sentence.

Sentences, probes, and comprehension questions were presented on a computer screen via

a Gateway E-series Pentium 4, 1.8 GHz PC-compatible computer using a program written in

Visual Basic 5.0. Naming times were measured using a Microsoft Multimedia Control sound

recorder written into Visual Basic 5.0. The sound recorder began recording at the onset of the

probe word and stopped recording when participant finished saying the probe and pressed the

"enter" key. Each sound file was manually examined using Audacity, a sound recording and

analysis program that displays the waveform of the sounds recorded, and the time between probe

presentation and naming onset for each probe word was extracted.

Young and older participants were tested on several measures of cognitive ability,

including a 25-item multiple-choice vocabulary test, forward and backward digit span tests, a

Stroop interference task (e.g., MacLeod, 1991; Stroop, 1935) and an Operation Span task (Turner









& Engle, 1989) to provide measures of semantic knowledge, inhibition, and working memory

capacity. Participants also completed questionnaires on other background demographics,

including questions about their education, reading, and writing habits. In addition, older

participants completed a test of mental status known as the Mini Mental State Examination

(Folstein, Folstein, & McHugh, 1972), a 30-point questionnaire surveying older adults'

orientation, attention and calculation, recall, and language processing abilities, where all older

participants scored at least 27 out of 30, M= 28.83, SD = 1.24.

For the Stroop interference task, participants saw a color word written in a particular color

on a computer screen. Underneath the word, there were four boxes: a red box, a blue box, a

yellow box, and a green box. The task was for participants to click the mouse on the box that

matched the color in which the word was written. A motor response was used instead of verbal

responding, similar to Keele (1972), in order to fully computerize the task and more precisely

time the sessions. The first 20 trials were congruent, such that the color and the word matched

(e.g., the participant saw the word "red" written in red). The next 20 trials were incongruent,

where the color and the word did not match (e.g., the participant saw the word "red" written in

yellow). Trials were blocked by type to reduce confusion from transitioning from one type of

stimulus to another, and congruent trials were presented first to simulate the "control first"

method originally used by Stroop (1935). The time to complete each block1 was recorded, and

the difference between the incongruent and congruent blocks was calculated as a measure of the

Stroop effect. A larger Stroop Effect indicates more difficulty with inhibition.

For the Operation Span task, participants were presented with a math equation to verify

(e.g., Is (9/3)-2=3?), followed by a word to remember. These two types of materials were

1 Individual word timing has been investigated in previous studies and found to yield similar amounts of Stroop
interference as the timing of blocks (e.g., Sichel & Chandler, 1969).









presented several times (i.e., a math equation, then a word, then a math equation, then a word)

until participants were presented with a number of words to remember, ranging from two to five,

with set size presented in the same random order for all participants (stimuli was the same as was

used in Turner & Engle, 1989). Participants were then prompted to recall out loud the words

from that set, i.e., the group of words presented for recall. This process was repeated for 12 sets

(three sets of each size from two to five words), the proportion of words recalled from each set

was calculated, and the average of these proportions was calculated to make each participant's

final operation span score. Although there are other scoring methods, this method was chosen

because it was a straightforward measure of how many items participants could recall (Conway,

Kane, Bunting, Hambrick, Wilhelm, & Engle, 2005 present a summary of scoring techniques).

Procedure

Upon beginning the experiment, participants were asked to sign an informed consent form,

giving their approval to continue with the study. Older adults also completed the MMSE

screening at this time. Then, participants received instructions on the experimental tasks (to be

described below) and completed three practice trials. Appendix B presents the instructions that

were seen by the participants. Following the practice trials, the experimenter answered any

questions the participant may have had, and data collection began for each of 132 trials (102

experimental and 30 filler trials randomized for presentation). Each trial consisted of the reading

task, the naming task, and one comprehension question. The tasks were completed as follows.

The first task was the self-paced reading task, where participants read sentences one word

at a time by pressing the "space bar" with their dominant hand to move from one word to the

next. Sentences appeared on the screen with letters replaced by dashes while preserving spaces

and punctuation. With each key press, one word was revealed while the previous word returned

to dashes, and the reading time for each word was recorded by the computer. After reading the









final word in each sentence, a 500 msec pause was presented, and participants were presented

with one of three possible probe words (verbatim, related, and new) for the naming task. The 500

msec delay was determined by pilot testing young adults on various delay presentations (i.e.,

250, 500, 750, 1000, and 1250 msec) to determine the longest amount of time possible to show

the probe while still getting an effect of negation (i.e., a longer naming time for negatively

modified probes compared to non-modified probes). The delay was presented to make a negation

effect able to be detected in both age groups, since older adults tend to be slower in processing

than young adults in many domains (e.g., Salthouse, 1996). Upon presentation, participants said

aloud the word shown. The multimedia sound recorder recorded the sound, and the naming

latency, i.e., the time between probe presentation and the onset of the participants' voice, was

extracted. After naming, participants pressed the "enter" key to immediately show a

comprehension question. This question was answered verbally, using one or more words, and the

experimenter recorded the answers. After answering this question and pressing the "enter" key,

the next sentence appeared on the screen. The process was repeated for all 132 sentences.

When the reading, naming, and comprehension trials were completed, participants

completed the background questionnaires and cognitive tests, as described earlier. After these

cognitive tests, participants were debriefed and thanked for their participation. Older adults were

given monetary compensation, and young adults were given course credit.

Results

Demographics, Operation Span, and Stroop Test

Independent samples t-tests on the demographic measures (Table 2-2 shows means and

standard deviations) indicated that older adults had more years of education, t(106) = 9.21, p <

.01, had greater vocabulary scores, t(104) = 15.20, p < .01, and reported spending marginally

more time watching television, t(105) = 1.94, p < .06, and doing crossword puzzles, t(105) =









1.74, p < .09, than young adults. Young adults had marginally larger forward digit spans than

older adults, t(105) = 1.92, p < .06, and reported spending more time writing, t(104) = 5.79, p <

.01, than older adults. Both young and older adults rated themselves similarly on health, p > .59

their spelling training while they were in school, p > .14, and their spelling ability now, ps > .27.

In addition, both age groups reported spending similar amounts of time reading, p > .35, and had

similar backward digit spans, p > .20.

For the Operation Span test, young adults recalled a higher proportion of words than older

adults, t(105) = 4.10, p < .01, suggesting that older adults had smaller working memory

capacities. For the Stroop task, an Age Group x Trial Type (congruent and incongruent) ANOVA

was conducted on mean response time for each block of trials, resulting in a main effect of age

group, F(1, 106) = 125.13, MSE = 105.93, p < .01, a main effect of trial type, F(1, 106) = 34.66,

MSE= 41.20, p < .01, and an interaction, F(1, 106) = 24.95, MSE = 41.20, p <.01. Follow-up

tests on the interaction indicated that young adults had similar response times for congruent and

incongruent trials, p > .53, but older adults had longer response times on incongruent trials

relative to congruent trials, p < .01. The lack of Stroop interference for young adults may be due

to the use of a motor task, which tends to have smaller Stroop interference effects (e.g., Keele,

1972).

Reading Times

For all participants' reading times, reading times exceeding plus or minus three standard

deviations from the mean of each age group were excluded, so as not to include instances where

participants skipped over words, stopped during reading for a break, or paused to ask a question.

For young adults, this technique eliminated 1.9% of the data, and for older adults, this technique

eliminated 1.7% of the data. Tests of reading times measure the difficulty that older adults may

have in comparison to young adults in reading on-line. Longer reading times indicate difficulty









with that portion of the sentence, and provide a specific time frame during reading, where young

and older adults may differ. These tests will provide a test of Hypothesis 1, investigating whether

additional time is necessary for processing negation because it is a difficult sentence

construction.

Overall reading times

To examine the speed with which readers in each age group were reading the words in the

presented sentences, a 2 (Age Group) x 2 (Sentence Type) ANOVA was conducted (Table 2-3

shows means and standard deviations) on the mean reading time per word, averaged across all

words in a sentence. Results showed a main effect of age group, F(1, 106) = 35.92, MSE =

31385.87, p < .01, where older adults (M= 603.02) spent more time reading each word than

young adults (M= 458.54). In addition, this analysis revealed a main effect of sentence type,

F(1, 106) = 5.33, MSE = 115.32, p < .02, where participants spent less time reading words in

negative sentences (M= 529.09) than non-negative sentences (M= 532.46). The interaction

between age group and sentence type was not significant, p > .22.

Target word reading times

To distinguish specific locations within each sentence that participants may have increased

or decreased their reading speed, analyses of the different types of words in each sentence were

conducted. First, a 2 (Age Group) x 2 (Sentence Type) ANOVA was conducted on the time spent

reading the target word, i.e., the word that was modified by the negation in negative sentences or

was not modified in the non-negative sentences (Table 2-4 shows means and standard

deviations). A main effect of age group was revealed, F(1, 106) = 30.79, MSE = 44719.88, p <

.01, such that older adults (M= 609.53) spent more time reading target words than young adults

(M= 449.85). While the main effect of sentence type was not significant, p > .13, a marginal

Age Group x Sentence Type interaction (Figure 2-1) was shown, F(1, 106) = 3.02, MSE =









839.01, p < .09. Further analyses on this interaction revealed that young adults spent less time

reading the target word when it was in a negative sentence than when it was in a non-negative

sentence, p < .02, whereas older adults had equivalent target reading times regardless of sentence

type, p > .88. For both negative and non-negative sentences, older adults read the target word

slower than did young adults, ps < .01, but the age group difference was slightly larger for

negative sentences.

Post-target word reading times

A 2 (Age Group) x 2 (Sentence Type) ANOVA was conducted on post-target reading

times, the word immediately following the target (Table 2-5 shows means and standard

deviations). Results showed only a main effect of age group, F(1, 106) = 41.52, MSE =

28325.21, p < .01, where older adults (M= 604.18) read after-target words slower than young

adults (M = 456.60). The main effect of sentence type and the interaction were not significant, ps

> .68.

Modifier versus target reading times for negative sentences only

To compare the reading times of modifier words with the reading times of the words they

modify (i.e., the target words), a 2 (Age Group) x 2 (Word Type) ANOVA was performed on the

reading times of the modifiers and the target words in negative sentences only (Table 2-6 shows

means and standard deviations). Only negative sentences were evaluated because there were no

modifier words in non-negative sentences. The analysis revealed a main effect of age group, F(1,

106) = 31.87, MSE = 40046.29, p < .01, and a main effect of word type, F(1, 106) = 96.19, MSE

= 1715.21,p < .01. These main effects were moderated by an Age Group x Word Type

interaction, F(1, 106) = 5.16, MSE = 1715.21, p < .03, such that reading times for targets were

slower than for modifiers for both age groups, but this difference was larger for older adults, ps <

.01.









Probe Naming Times

Any inaccurate responses, including those where a participant laughed, stuttered, coughed,

cleared their throat, or pressed the "enter" key before they said the probe word, were not

included, eliminating 17.3% of potential responses from young adults and 16.0% of potential

responses from older adults. Similar to reading times, outliers in participants' naming times (i.e.,

the latency between the onset of the probe word and the time the participant began saying the

word) were accounted for by excluding naming times plus or minus three standard deviations

from the mean of each age group. This method excluded 1.8% of the data for each age group.

Probe naming times test the activation level of probe words represented in different levels of

representation of the text, which provide a test of Hypothesis 2.

A 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) ANOVA was conducted on mean

naming times to determine the availability of the probe words in memory after having just read a

negative or non-negative sentence (Table 2-7 shows means and standard deviations). This

analysis revealed a main effect of age group, F(1, 106) = 4.31, MSE = 97162.48, p < .04, where

older adults had longer naming times (M= 789.10) overall than did young adults (M= 738.26).

In addition, a main effect of probe type, F(1, 106) = 36.02, MSE = 2603.96, p < .01, showed that

both verbatim (M= 740.39) and related probes (M= 770.07) had faster naming times than new

probes (M= 780.57), ps < .01. Verbatim probes were also named more quickly than related

probes, p < .04. Finally, there was a marginally significant main effect of sentence type, F(1,

106) = 2.72, MSE = 2339.81, p < .10, where probes following negative sentences (M= 766.81)

had slower naming times than probes presented after non-negative sentences (M= 760.54). No

interactions among any of the variables were significant, ps > .19.









Comprehension Accuracy

Comprehension accuracy was coded to indicate whether an answer was either correct or

incorrect. Correct responses included answers that used either the verbatim words from the

sentence or synonyms of the words in sentence, just as long as the answer captured the meaning

or the gist of what happened in the sentence. Answers that were the opposite or different from

what happened, off the topic from what happened, or indicated that the participant did not know

the answer, were marked as incorrect. On rare occasions when there were uncertainties,

additional coders were consulted, and a 2/3 majority ruled the answer as correct or incorrect.

Tests of comprehension accuracy provide an analysis of the content measure, testing Hypothesis

1 and examining the storage of the information read by young and older adults.

To examine participants' comprehension, a 2 (Age Group) x 2 (Sentence Type) ANOVA

was conducted on mean accuracy in answering the comprehension questions, which focused on

the portion of the sentence containing the target (Table 2-8 shows means and standard

deviations). This analysis revealed a main effect of age group, F(1, 106) = 7.31, MSE= .023, p <

.01, such that older adults had poorer accuracy overall (M= 68.0%) on the comprehension

questions than did the young adults (M = 73.5%). In addition, a main effect of sentence type,

F(1, 106) = 19.29, MSE = .007, p < .01, showed that participants' accuracy was poorer for

negative sentences (M= 68.3%) than for non-negative sentences (M= 73.3%). The interaction of

the two variables was not significant, p > .53.

To determine the extent of the impact of working memory capacity and inhibition on

comprehension accuracy in contributing to the main effect of age group shown above, two

ANCOVA analyses were conducted with age group and sentence type as variables, as in the

above ANOVA. The first of these analyses included Operation Span score as a covariate. No

significant differences based on age group (Molder adjusted = 69.2%, Myoung adjusted = 72.1%), F(1,









104) = 1.90, MSE= .02, p > .17, nor an interaction of age group and sentence type, F < 1, were

revealed, showing that the age difference in comprehension accuracy disappeared when working

memory between the age groups was equated. The second ANCOVA analysis included Stroop

Effect score (the measure of inhibition) as a covariate. Results for this analysis showed a

marginal main effect of age group, F(1, 105) = 3.74, MSE = .02, p < .06, such that older adults

still had lower comprehension accuracy (Madjusted = 68.6%) than young adults (Madjusted = 73.0%).

Unlike the ANCOVA with working memory as a covariate, this analysis showed that age group

differences remained in comprehension when the measure of inhibition was equated across age

groups. The interaction of age group and sentence type was not significant, ps > .92.

To determine whether the presentation of a probe word for naming (between reading and

answering the comprehension question) could influence comprehension accuracy, a 2 (Age

Group) x 2 (Sentence Type) x 3 (Probe Type) ANOVA was conducted on mean percent accuracy

in answering the comprehension questions (Table 2-9 shows means and standard deviations). As

probe type is the only new variable in this analysis, only effects involving probe type were

presented. A main effect of probe type was revealed, F(2, 212) = 16.72, MSE = .015, p < .01,

such that accuracy was lower for comprehension questions that came after new probes (M=

66.8%) relative to verbatim probes (M= 73.5%) and related probes (M= 71.5%), ps < .01.

Comprehension accuracy for questions that came after verbatim probes was marginally higher

than for questions that came after related probes, p < .093. Finally, a Sentence Type x Probe

Type interaction (Figure 2-2) was shown, F(2, 212) = 4.52, MSE = .015, p < .01. Further

analyses on this interaction demonstrated that comprehension accuracy was poorer for negative

sentences than for non-negative sentences when the questions came after verbatim probes, p <

.01, or after related probes, p < .01, but there was no difference in comprehension of negative









and non-negative sentences following new probes, p > .34. Within sentence type, there was an

effect of probe type for non-negative sentences, p < .01, such that accuracy after saying verbatim

probes and related probes was similar, p > .17, but accuracy for these questions was significantly

greater than accuracy for questions presented after new probes, ps < .01. For negative sentences,

comprehension accuracy was similar following the three different probe types, p > .15.

A series of regression analyses were conducted to further examine the influence of

working memory on comprehension accuracy, and to determine if any other variables may also

have an impact on comprehension. Stepwise regressions were conducted using either operation

span, Stroop effect, or age as the first variable entered because of the theoretical implications of

each of these variables on comprehension. The remaining variables entered were vocabulary

score, years of education, and time spent reading each day. The model that predicted the most

variance for negative and non-negative sentences entered operation span first; for both types of

sentences, only operation span score accounted for a significant amount of variance (13.1% for

negative sentences and 16.5% for non-negative sentences). After accounting for this variable, no

other variables entered into the regression model. Results from the regression for negative

sentences are presented in Table 2-9, and results from the regression for non-negative sentences

are presented in Table 2-10.

The Relationship between Reading Times, Naming Times, and Comprehension Accuracy

Pearson correlations were conducted to determine the relationships between modifier

reading times, target reading times, probe naming times, and comprehension accuracy, separately

for non-negative and negative sentences. For young adults, a significant positive correlation was

demonstrated for target word reading times and probe naming times for both negative sentences,

r(54) = .27, p < .05, and non-negative sentences, r(54) = .36, p < .01. Young adults also

demonstrated a marginally significant correlation for modifier reading times and probe naming









times, r(54) = .26, p < .06, in negative sentences. However, no significant correlations occurred

with young adults' comprehension accuracy, ps > .70. In contrast to young adults, older adults

showed no significant correlations between any of the measures, ps > .17.

Discussion

The results of this experiment showed that negation detrimentally impacted reading

comprehension via two of the three measures used (probe naming times, comprehension

accuracy), consistent with previous work in negation comprehension (e.g., Cornish, 1971,

Experiment 1; Cornish & Wason, 1970; Hoosain, 1973; Just & Carpenter, 1971; Kaup, 2001;

Kaup et al., 2004; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup et al., in press; MacDonald &

Just, 1989; Sherman, 1973). The presence of negation within sentences decreased comprehension

accuracy and marginally increased naming times of a probe word, independent of whether that

word had actually appeared in the preceding sentence. However, both young and older adults

were adversely affected by negation to the same degree, and negation did not differentially affect

one level of representation over another, for either age group.

The first hypothesis, which predicted that older adults would have more difficulty

processing negation than young adults, was not supported. There were several indicators that

older adults did not have an increased difficulty with inhibition when processing negated

sentences, as proposed by the IDH (e.g., Hasher & Zacks, 1988). First, the covariance analyses

showed that controlling for inhibition (using the Stroop interference task) did not eliminate age

differences in comprehension accuracy. Second, the regression analyses showed that inhibition

did not significantly explain any of the variance in comprehending negative sentences. Third,

although probe naming times were marginally slowed by negation, the negation effect did not

interact with age group. Finally, overall reading times showed no interaction between age group

and sentence type, and in fact, negation speeded rather than slowed overall reading times for









both groups relative to non-negative sentences. The lack of increased inhibition problems during

reading for older adults occurred despite older adults having more difficulty with inhibition in

general, as measured by the Stroop task, in support of the alternative to the first hypothesis.

Therefore, it can be concluded that the older adults did not demonstrate a difficulty with

inhibition in reading negation in this experiment and may not have a difficulty as it relates to

negation processing, contrary to the IDH.

Although negation did not differentially affect older adults in the majority of tasks

administered here, several measures did reveal a differential impact on young and older adults'

reading times. With respect to target reading times, young adults increased their reading speed of

the target word when it was modified by a negative modifier relative to no modifier, whereas

older adults' target reading times were equivalent in both types of sentences. Young adults may

have been sensitive to the fact that negation is more difficult to process and therefore slowed

their reading times after encountering a negation word. Young adults' target reading times also

correlated with their probe naming times for both negative and non-negative sentences, whereas

older adults did not show this relationship. These results suggest that target words that were

easier to read were also easier to represent in the memory representation, allowing young adults

faster access to the representation that they had stored. In contrast, older adults' reading times

were less indicative of the ease with which they accessed the representation they created,

possibly because they did not adapt their reading times when more difficult circumstances (i.e.,

negation) arose. Another age difference occurred when focusing on negative sentences only,

where young adults and older adults read the modifier word more quickly than the target word,

but the difference between the two words' reading times was greater for older adults. Similar to

the way a person may read the words "the," "it," or "and," a reader may also skim over the word









"not" based on its length, as these shorter words tend to be less meaningful than other words in

the sentence. Older adults may be especially likely to read these words quickly because of their

limited working memory capacity; they are aware of their limits and instead focus on words that

they think are more meaningful and can help facilitate their comprehension, such as the target.

There was also evidence of deficits in older adults' reading comprehension independent of

negation, consistent with other research on aging (e.g., Connelly et al., 1991; Kemtes & Kemper,

1997; Stine, 1990; Stine-Morrow et al., 1996). Older adults spent more time than young adults

reading both negative and non-negative sentences, suggesting that they require more processing

time than do young adults (e.g., Salthouse, 1996). The results on comprehension accuracy also

showed declines for older adults, who had poorer overall accuracy than young adults, despite

spending more time reading the sentences. In fact, neither young nor older adults showed a

correlation between reading times and comprehension accuracy. Content measures require

retrieving information from memory after reading the sentences has been completed, which may

put older adults at a disadvantage because of their smaller working memory capacities (e.g.,

Light, 1988; Waters & Caplan, 2001). Consistent with this view, operation span as a covariate

removed the effect of age on comprehension accuracy. Furthermore, the regression analyses

demonstrated that working memory accounted for the most variance in comprehension accuracy.

These findings suggest that working memory, rather than age per se, was more critical for

comprehension accuracy in the present experiment, although one would have expected it to be

especially relevant in comprehending negative sentences, which did not occur. While working

memory is a primary determinant of comprehension accuracy (e.g., Just & Carpenter, 1992), the

regression analyses indicate that there is considerable variance unaccounted for, suggesting that









many other factors may be useful in understanding negation, such as attention, experience with

negation, and the appropriateness of the negation used in the text.

The second hypothesis predicted that older adults would show the most difficulty in

representing negation at the situation model level of representation. However, given that older

adults did not show more difficulty with negation than did young adults, the question becomes

whether one level of representation (e.g., Kintsch, 1998) was affected more than another by

negation for either age group. This hypothesis was not supported, as naming times for related

probes (the ones representative of the situation model) were faster than new probes for both

negative and non-negative sentences. If negation had increased the difficulty in representing the

situation model, naming times would have been slowest for probes representing this level of

processing. Therefore, negation had equivalent effects at surface and situation model levels of

representation, for both age groups, which also contradicts the alternative hypothesis that older

adults would be unaffected by negation at the situation model level. However, independent of

negation, the probe naming time analysis revealed that readers did have differential access to the

levels of representation, evidenced by longer naming times for new and related probes, compared

to verbatim probes. This result suggests that for all sentences, both the gist and exact wording

information were retained in memory representations, but the exact wording remained more

highly activated. Two possibilities could explain why exact wording was more highly activated

than the gist. The first is that the reader did not form the situation model representation (e.g.,

Kintsch, 1998) because there was not enough time to do so between reading and probe word

naming. This possibility seems unlikely given that reading occurs at a fairly fast rate, and a

reader is used to quickly consolidating information between consecutive sentences of a text. The

second possibility is that the constraining of activation that occurs after reading may have









included only the exact wording from the text because these words were the most effective in

communicating the ideas in the sentence. Given the richness and complexity of the text used in

the present experiment's sentences, the exact wording may have been preferable in order to retain

the meaning of the message, as opposed to simpler sentences used in previous research where a

gist representation involving a character (e.g., Elizabeth) and the items he/she created (e.g.,

bread) would be sufficient.

Unexpectedly, the presence of the probe words for naming also influenced comprehension

accuracy. Readers' accuracy was greater following verbatim and related probes than new probes

for non-negative sentences, whereas comprehension accuracy was equivalent for negative

sentences independent of the probe. For non-negative sentences, the probe word was more easily

incorporated into the situation model of the text as additional information when it was relevant to

the text (i.e., verbatim or related) than when it was not (e.g., new). However, correctly

representing negative sentences meant excluding it from the situation model; therefore, none of

the probes would be relevant to the text and should not be incorporated into the situation model.

If the probe word was incorporated into the situation model of a negative sentence, then the

representation of the sentence would now be incorrect. Thus, readers were sensitive to the

presence of the probe word, potentially because of the time at which it was presented (i.e.,

between the sentence and the comprehension question) and let it influence the representation of

the text that they created.









Table 2-1. Length and frequency averages for each type of probe word.
Probe type Average frequency Average length
Verbatim 204.39 6.53
Related 154.43 6.90
New 131.51 6.56


Table 2-2. Means and standard deviations (st. dev.) of demographic measures.
Age group
Young Older
Mean St. dev Mean St. dev.
Years of education* 13.07 1.27 16.78 2.67
Health rating (out of 10) 7.74 1.51 7.90 1.50
Vocabulary (out of 25)* 13.3 3.22 21.32 2.09
Forward digit span*** 7.56 1.41 7.06 1.28
Backward digit span 4.98 1.51 5.32 1.19
Hours spent writing* 2.57 1.53 1.20 .79
Hours spent reading 3.11 1.89 2.81 1.38
Hours watching TV*** 1.96 1.47 2.58 1.84
Hours doing crosswords*** .23 .50 .49 .97
Spelling training (out of 10) 6.42 1.72 7.08 2.68
Spelling ability now 6.54 1.36 6.15 2.17
(out of 10)
Operation span* .90 .08 .82 .12
Congruent Stroop* 22.33 5.26 33.63 8.29
Incongruent Stroop* 23.11 6.10 43.14 12.68
Note: indicates that the age differences were significant at the p < .01 level; ** indicates that
the age differences were significant at the p < .05 level; *** indicates that the age differences
were marginally significant at the p < .10 level.


Table 2-3. Means and standard deviations for overall reading times (msec)
Age group Sentence type Mean reading time Std. deviation
Young Negative 457.75 106.53
Non-negative 459.31 103.16
Older Negative 600.42 143.82
Non-negative 605.61 142.58









Table 2-4. Means and standard deviations for target word reading times (msec)
Age group Sentence type Mean reading time Std. deviation
Young Negative 443.42 120.95
Non-negative 456.27 114.32
Older Negative 609.96 187.24
Non-negative 609.10 168.42


Table 2-5. Means and standard deviations for post-target reading times (msec)
Age group Sentence type Mean reading time Std. deviation
Young Negative 456.64 100.17
Non-negative 456.57 91.21
Older Negative 602.23 139.56
Non-negative 606.14 142.63


Table 2-6. Means and standard deviations for modifier and target word reading times for
negative sentences only (msec)
Age group Word type Mean Std. deviation
Young Modifier word 400.95 107.78
Target word 443.42 120.95
Older Modifier word 541.88 149.07
Target word 609.96 187.24


Table 2-7. Means and standard deviations of probe word naming times (msec)
Age group Sentence type Probe type Mean Std. deviation
Young Negative Verbatim 714.30 105.65
Related 746.09 111.98
New 761.83 106.54
Non-negative Verbatim 709.24 110.77
Related 739.40 117.44
New 758.69 120.55
Older Negative Verbatim 775.47 161.19
Related 798.65 159.83
New 804.52 152.50
Non-negative Verbatim 762.56 146.63
Related 796.13 156.45
New 797.25 144.80










Table 2-8. Means and standard deviations for comprehension accuracy (%)
Age Group Sentence type Mean Std. deviation
Young Negative 70.68 12.20
Non-negative 76.40 9.52
Older Negative 65.87 15.24
Non-negative 70.14 10.95


Table 2-9. Means and standard deviations for comprehension accuracy with both
sentence type and probe type (%)
Age group Sentence type Probe type Mean Std. deviation
Young Negative Verbatim 72.02 18.11
Related 69.24 17.71
New 68.32 13.51
Non-negative Verbatim 81.58 13.64
Related 77.11 13.52
New 70.95 13.33
Older Negative Verbatim 66.95 17.45
Related 66.39 18.60
New 63.80 19.17
Non-negative Verbatim 73.56 17.13
Related 73.44 12.68
New 64.28 14.44


Table 2-10. Results of regression analysis on negative sentence comprehension accuracy
Variable 13 p
Operation span .34 < .01
Stroop effect .04 .68
Age .01 .93
Vocabulary .02 .81
Years of education -.14 .13
Hours spent reading .06 .50


Table 2-11. Results of regression analysis on non-negative sentence comprehension
accuracy
Variable 13 p
Operation span .42 < .01
Stroop effect -.13 .18
Age -.16 .11
Vocabulary -.14 .14
Years of education -.11 .23
Hours spent reading .02 .80











Marginal Age x Sentence Type Interaction for Target
Reading Times
Young Adults
0 Older Adults


S600

E 550

E 500
I,
450 -

400

350
Negative Sentences Non-Negative Sentences
Sentence Type


Figure 2-1. Target word reading times as a function of age group and sentence type.


Sentence Type x Probe Type Interaction for
Comprehension Accuracy


* Negative
1 Non-negative


Related
Probe Type


New


Figure 2-2. Comprehension accuracy for questions
sentence type.


after each probe type as a function of


85

80

75

70

I 65

60

55


50


Verbatim









CHAPTER 3
EXPERIMENT 2

Experiment 1 demonstrated that negation detrimentally impacted reading at multiple points

in the reading process, for both young and older adults, similar to other complex structures such

as syntactic complexity (e.g., Kemper, 1987; Norman et al., 1992; Norman, Kemper, Kynette,

Cheung, & Anagnopoulos, 1991; Stine-Morrow, Ryan, & Leonard, 2000; Waters & Caplan,

1996). However, there is evidence that compensatory strategies, such as slowing reading times at

specific points in the reading process, can increase recall by encouraging wrap-up and integration

of the previously-read clause or portion of text (e.g., Stine, 1990; Noh, Shake, Joncich, Hindin,

& Stine-Morrow, 2006; Stine-Morrow et al., 2000). For example, Noh et al. (2006) instructed

older adult readers to slow down when they reached a comma in the text (indicating the end of a

clause), and this change in their reading strategy led to improved recall of the text. Compensation

can also occur spontaneously without instruction; Stine (1990) measured reading times and recall

and found that older adult readers, who slowed their reading at clause boundaries instead of only

sentence boundaries, had perfect recall. These studies suggest that compensatory strategies help

to reduce the burden on working memory during reading; therefore, it seems likely that

comprehension of negation may also benefit from a compensatory strategy. A compensatory

strategy may make available some additional working memory resources for processing

negation, resulting in appropriate activation and de-activation levels of negated and non-negated

concepts from the text.

One problem with the above compensatory strategies is that the reader must actively

commit to using the strategy, which may or may not occur in real-life reading situations. The

present experiment explored the possibility of compensation during reading that is offered within

the texts themselves. For example, a compensatory strategy may be seen in the prescription









medication industry. Sometimes appearing on prescription labels are instructions that include

negation (e.g., "Do not take this medication with aspirin") accompanied by a symbol (e.g., a

circle with a line through it). These symbols may serve the purpose of helping readers,

particularly people with increased difficulties in comprehension, to understand the instructions

that pertain to the medication they are taking. However, if the symbol is difficult to interpret or

inapplicable to the instructions they are given, more working memory resources may be used,

resulting in an increased difficulty in comprehension. Research involving the use of pictures in

the interpretation of medication instructions has demonstrated that people preferred instructions

given with pictures to demonstrate the instruction compared to text-only instructions (e.g.,

Sojourner & Wogalter, 1997). However, Dowse and Ehlers (2002) showed that pictures should

accompany, but not replace, textual instruction, as misinterpretation of pictorial descriptions can

occur.

The purpose of Experiment 2 was to assess the effectiveness of a specific compensatory

strategy, the use of a negation symbol, in comprehension of negation. Both age groups had

greater difficulty understanding negation relative to non-negated sentences in Experiment 1,

suggesting that there is a need to facilitate the processing of negation.

Hypothesis 1

Symbolic compensation will release working memory resources and assist in

comprehension of negation. Given that working memory capacity is particularly vulnerable to

the changes that accompany aging, compensation is expected to have a greater impact on how

older adults process negation. Just and Carpenter (1992) more specifically defined the

mechanism underlying compensation as activation of relevant ideas. They proposed that when

sentences are read within a broader context (e.g., a whole paragraph focused on a particular

topic), that context can serve as a framework of ideas surrounding the information presented in









the sentence, creating pre-activation of the ideas to come next. This pre-activation prepares the

reader for the concepts that will be presented in the sentence, leaving more working memory

resources available for processing and facilitating comprehension.

In the present experiment, a symbol is used to provide a context or framework, prior to

reading each sentence. If the symbol provides a context for negation, then the pre-activation

provided by the symbol could release some working memory resources to assist young and older

adults' processing of the negated text. Therefore, compensation may assist in the process of

reading and the representation created by the text. With respect to probe naming times, a symbol

is expected to facilitate negation processing such that naming times for negatively-modified

probe words will not differ from naming times of probes not modified, relative to sentences

without a symbol. For comprehension accuracy, the presence of a symbol is expected to improve

comprehension for negative sentences, resulting in no difference between negative and non-

negative sentences, contrary to sentences presented without a symbol.

Hypothesis 2

Benefits from the presence of a symbol are expected be more dramatic at the situation

model level. Although Experiment 1 did not demonstrate differences among the levels of

representation at which negation impacted reading comprehension in probe naming, the symbol

in Experiment 2 provides additional information that can function as pre-existing knowledge

about the text, creating the possibility for compensation to impact comprehension at higher levels

(the situation model level) than others. According to research on older adults' reading

comprehension, text representations at the situation model level tend to be the most easily

remembered for readers of this age group (e.g., Radvansky et al., 2001). Given that older adults

process information at this high level, which also includes information inferred from or

previously learned about the topic in the text (e.g., Kintsch, 1998), older adults may be more









experienced at incorporating outside information into the information that they are reading, thus

creating a better situation model representation when a symbol is presented than when it is not.

Specifically, the predicted lack of an effect of negation on probe naming times described in

Hypothesis 1 is expected for verbatim as well as related probes.

Methods

Participants

Sixty young (45 female and 15 male) and 60 older adults (33 female and 27 male) were

tested. Young and older adult participants were recruited from similar sources as in Experiment

1, although none had previously participated in Experiment 1. Young adults ranged in age from

18 to 24 years (M= 19.32, SD = 1.33), while older adults ranged in age from 63 to 88 years (M=

75.46, SD = 5.81). Participants of both age groups were fluent speakers of American English and

had not read the books nor watched the movies from the Harry Potter series.

Design

The present experiment used a 2 x 2 x 3 x 2 factorial design with age group (young and

older), sentence type (negative and non-negative), probe type (verbatim, related, and new), and

symbol (symbol and no symbol) as factors. Sentence type, probe type, and symbol were within-

subjects factors, whereas age group was a between-subjects factor. As in Experiment 1, reading

time served as an online measure of processing, naming time measured activation, and

comprehension accuracy measured retention of the content of the text.

Materials

The same 132 sentences (both negative and non-negative, filler and experimental), probe

words (verbatim, related, and new), and comprehension questions from Experiment 1 were used

in Experiment 2, with one modification. Half of the experimental sentences were presented with

a symbol, which differed as a function of modifier. For sentences with a negative modifier, the









symbol was a circle with a line through it, denoting absence. For sentences without a negative

modifier, the symbol was an open circle. Unlike negative sentences, the symbol for non-negative

sentences was neutral and was used so that symbols did not only appear for negative sentences.

Both symbols were approximately 1.5 inches in diameter and were white with a thin, black

outline. As in Experiment 1, the Operation Span test, Stroop test, and other demographic

measures obtained were also administered.

Procedure

The procedure for Experiment 2 was identical to Experiment 1, except the reading task was

modified to allow the symbol to be presented. Participants were instructed that they may see

some symbols presented before the text, that the symbol may or may not seem important to them,

but to make sure they looked at the symbol before beginning to read the sentence (Appendix C

shows exact instructions). In the symbol condition, the symbol appeared directly to the left of the

sentence, with dashes that represented the words of the sentence to the right. After looking at the

symbol, participants pressed the "space bar" to read the first word of the sentence. Participants

continued to read each sentence at a self-paced rate by pressing the "space bar." The symbol

remained on the screen for the entire duration of reading the sentence and disappeared with the

dashes and the last word when the participant finished reading the sentence.

As in Experiment 1, after reading each sentence, participants were presented with a probe

word for naming after a delay of 500 msec. The probe word was verbatim from the passage,

related to the passage, or not presented at all. After naming the probe word, participants pressed

the "enter" key, and a comprehension question was presented. Participants answered each

question aloud and pressed the "enter" key to move on to the next sentence.









Results

Demographics, Operation Span, and Stroop Test

One older participant's data was excluded because of improper data recording. Independent

samples t-tests on the demographics measures (Table 3-1 shows means and standard deviations)

revealed that older adults reported more years of education, t(116) = 7.13, p < .01, had higher

vocabulary test scores, t(117) = 13.48, p < .01, spent more time doing crossword puzzles, t(117)

= 2.14, p < .04, and had more rigorous spelling training, t(117) = 2.82,p < .01, than young

adults. Young adults reported spending more time writing each day than did older adults, t(117)

= 4.27, p < .01. Young and older adults did not differ on their reports of time spent reading, p >

.36, time spent watching television, p > .61, ratings of their spelling ability, p > .97, or ratings of

health, p > .41. Young and older adults also did not differ on either forward digit span, p > .47,

or backward digit span, p > .23.

For the Operation Span test, young adults recalled a higher proportion of words than did

older adults, t(117) = 2.85, p < .01, demonstrating young adults' larger working memory

capacities. For the Stroop task, an Age Group x Trial Type (congruent or incongruent) ANOVA

was conducted on the mean response time to complete each block of trials. A main effect of age

group, F(1, 117) = 182.14, MSE = 132.27, p < .01, and a main effect of trial type, F(1, 117)=

14.22, MSE = 53.54, p < .01, were revealed. These main effects were moderated by an Age

Group x Trial Type interaction was revealed, F(1, 117) = 27.15, MSE = 53.54, p < .01, where

older adults showed a Stroop effect with slower times for incongruent than congruent trials, p <

.01, but young adults had equivalent response times for the two blocks, p > .31.

Reading Times

For all participants' reading times, reading times exceeding plus or minus three standard

deviations from the mean of each age group were excluded, so as not to include instances where









participants skipped over words, stopped during reading for a break, or paused to ask a question.

This method removed 1.5% of young adults' reading time data and 1.5% of older adults' reading

time data. Reading time analyses were conducted to test part of Hypothesis 1, where the impact

of a symbol could be more important in how older adults process text during reading, compared

to young adults.

Overall reading times

To examine the average speed of reading the sentences, a 2 (Age Group) x 2 (Sentence

Type) x 2 (Symbol) ANOVA was conducted on the reading times averaged over all types of

words in the sentences (Table 3-2 shows means and standard deviations). This analysis revealed

a main effect of age group, F(1, 117) = 31.23, MSE = 88807.42, p < .01, where older adults (M=

589.70) read more slowly overall than young adults (M= 437.04). In addition, a marginal main

effect of sentence type, F(1, 117) = 3.40, MSE = 357.26, p < .08, was revealed such that readers

read negative sentences (M= 511.86) more quickly than non-negative sentences (M= 514.88).

No other main effects or interactions were significant, ps > .19.

Target word reading times

To further break down where readers spent more time and less time during reading,

reading times of different types of words were analyzed. A 2 (Age Group) x 2 (Sentence Type) x

2 (Symbol) ANOVA was conducted on target word reading times (Table 3-3 shows means and

standard deviations). Results from this analysis showed a main effect of age group, F(1, 117)=

31.77, MSE = 124218.38,p < .01. In addition, a marginal interaction of sentence type and age

group was revealed, F(1, 117) = 2.73, MSE = 2261.77, p < .10, such that older adults read target

words slower than young adults for both types of sentences, but the age difference was larger for

negative sentences, ps < .01. The interaction did not emerge when examined within each age

group, as targets in negative sentences were read equivalently to targets in non-negative









sentences for both young adults, p > .28, and older adults, p > .21. No other main effects, p > .60,

or interactions, p > .46, were revealed.

Post-target word reading times

Next, a 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol) ANOVA was conducted on post-

target reading times (Table 3-4 shows means and standard deviations), the word that came

immediately after the target word. A main effect of age group was revealed, F(1, 117) = 23.77,

MSE = 129994.86, p < .01, such that older adults read post-target words slower than young

adults. In addition, sentence type and symbol showed a significant interaction (Figure 3-1), F(1,

117) = 3.80, MSE = 4837.91, p < .05, where readers took longer to read post-target words in

negative sentences in the presence of a symbol than without a symbol, p < .04, but when reading

non-negative sentences, post-target reading times were equivalent whether a symbol was present

or not, p > .74. No other effects were significant, ps > .19.

Modifier versus target reading times for negative sentences only

To examine how much processing time readers used for modifier words compared to target

words, a 2 (Age Group) x 2 (Word Type) x 2 (Symbol) ANOVA was conducted on negative

sentences only (Table 3-5 shows means and standard deviations). This analysis revealed a main

effect of age group, F(1, 117) = 35.15, MSE = 103718.99, p < .01, where older adults (M=

581.23) read more slowly than young adults (M = 406.20), and a main effect of word type, F(1,

117) = 122.47, MSE = 3127.90, p < .01, such that target words (M = 522.08) were read more

slowly than modifier words (M= 465.34). These main effects were moderated by an Age Group

x Word Type interaction, F(1, 117) = 7.77, MSE = 3127.90, p < .01. Follow-up tests on this

interaction revealed that both young and older adults read the target words more slowly than

modifier words, but the difference was larger for older adults, ps < .01. No other effects were

significant, ps > .52.









Probe Naming Times

Recordings from probe naming where a participant coughed, cleared their throat, laughed,

made a comment, or pressed the "enter" key before they said the probe word were discarded.

This process removed 29.5% of potential responses from young adults, but only 11.4% of older

adults' potential responses. Similar to reading time data, participants' naming times outside of

three standard deviations above the mean for each age group were discarded, which eliminated

2.0% of young adults' naming time data and 1.6% of older adults' naming time data. Probe word

naming time analyses were conducted to examine the impact of symbol on different levels of text

representation, as predicted in Hypothesis 2.

A 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) x 2 (Symbol) ANOVA was

conducted on mean naming times (Table 3-6 shows means and standard deviations). This

analysis revealed a main effect of age group, F(1, 117) = 4.69, MSE = 246601.37, p < .03, a main

effect of sentence type, F(1, 117) = 6.61, MSE = 5213.40, p < .01, and a main effect of probe

type, F(1, 117) = 41.46, MSE = 7823.85, p < .01. Young adults (M= 777.64) named probes more

quickly than older adults (M= 835.07), and probes following negative sentences (M= 811.31)

were named more slowly than probes following non-negative sentences (M= 801.40). New

probes (M= 830.90) were named more slowly than related probes (M= 809.61), which were

named more slowly than verbatim probes (M= 778.55), with all differences significant, p < .01.

The main effect of symbol was not significant, p > .22, and no interactions among any of the

variables were significant, p > .12.

Comprehension Accuracy

To examine comprehension accuracy, a 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol)

ANOVA was conducted (Table 3-7 shows means and standard deviations). This analysis

revealed only a main effect of sentence type, F(1, 117) = 35.83, MSE= .01, p < .01, where









comprehension accuracy for negative sentences (M= 68.8%) was poorer than for non-negative

sentences (M= 74.4%). No main effect of age group, p > .31, or symbol, p > .53, was revealed,

and no interactions were significant, p > .75. Comprehension accuracy analyses tested the

content of the memory representation, and tested Hypothesis 1, where assistance of a symbol is

predicted to help the storage of information.

Given that the probe words were presented between reading and presentation of the

comprehension question, a 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) x 2 (Symbol)

ANOVA was conducted on mean comprehension accuracy to examine the impact of probe

words on comprehension of the information read in the sentence (Table 3-8 shows means and

standard deviations). As probe type was the only new variable in this analysis, only effects

involving probe type are reported. A main effect of probe type was demonstrated, F(1, 115)=

17.06, MSE= .03, p < .01. However, this main effect was moderated by its interaction with

sentence type, F(2, 230) = 2.46, MSE= .03, p < .03. Follow-up analyses on this interaction

revealed that for non-negative sentences, accuracy was greater following verbatim probes than

related probes, which was greater than new probes, with all differences significant, ps < .01. For

negative sentences, the only significant effect was that accuracy following verbatim probes was

better than accuracy following new probes, p < .05; accuracy following related probes was not

different from either verbatim probes, p > .58, or new probes, p > .11. The main effect of age

group was not significant, p > 59, nor was the main effect of symbol, p > .25, or other

interactions, p> .31.

To determine whether readers were better able to use the symbol and integrate it with their

memory representation of the text, data were split into thirds, and a 2 (Age Group) x 2 (Sentence

Type) x 2 (Symbol) x 3 (Group Order) ANOVA was conducted on comprehension accuracy. It is









important to note that because fillers were randomly distributed throughout the stimuli

presentation, these groups have different numbers of experimental stimuli. With respect to

differences in group order, a main effect was revealed, F (2, 234) = 4.35, MSE = .04, p < .01,

such that comprehension accuracy was greater for the last third of the comprehension questions

compared to the first third of comprehension questions, p < .01, and comprehension accuracy

was marginally greater for the middle section of comprehension questions than the first third of

comprehension questions, p < .07. However, there was no difference between the second and

third sections of questions, p > .32. This effect of group order did not significantly interact with

age group, p > .10, symbol, p > .97, or sentence type, p > .34.

A series of stepwise regression analyses were conducted to examine the influence of

operation span, Stroop effect, age, vocabulary score, years of education, and time spent reading

each day on comprehension accuracy of each type of sentence: (1) negative with a symbol, (2)

negative without a symbol, (3) non-negative with a symbol, and (4) non-negative without a

symbol. Variables entered in this order present the best fit model, although operation span,

Stroop effect, and age were each entered first in every combination with the remaining variables,

similar to Experiment 1. Results are presented in Table 3-9, Table 3-10, Table 3-11, and Table 3-

12. Results showed that for negative sentences presented with a symbol, a model containing

operation span, vocabulary score, and age accounted for 20.2% of the variance in comprehension

accuracy, with operation span accounting for 10.2% of the variance, vocabulary accounting for

an additional 4.2%, and then age accounting for an additional 5.8%. For negative sentences

without a symbol, non-negative sentences with a symbol, and non-negative sentences without a

symbol, only operation span accounted for a significant amount of variance: 8.7%, 2.6%, and

6.8%, respectively.









Discussion

The present experiment attempted to offset the difficulties in comprehension caused by

negation by exploring the impact of presentation of a symbol as a compensatory strategy. The

results showed that the presence of a symbol did not increase comprehension for negated

sentences specifically, for either age group. Effects of negation were minimal for reading times

but emerged for probe naming times and comprehension accuracy. Within comprehension

accuracy, readers did improve significantly over the course of the experimental session.

However, relative to Experiment 1, some changes in negation processing did occur, specifically

readers' post-target reading times in negative sentences were slowed by the presence of a

symbol, and young adults' target reading times did not slightly decrease in negative sentences

compared to non-negative sentences. Furthermore, older adults' comprehension accuracy

increased (i.e., was not poorer than young adults' comprehension accuracy) in this experiment,

suggesting that having a symbol present for some sentences may have been beneficial to reading

by generally increasing attention during reading.

Although the first hypothesis, that symbolic compensation will release working memory

resources and assist in comprehension of negation, was not supported, there was evidence to

suggest that the presentation of the symbol may have had some effects on the reading process.

First, for sentences with a negative modifier, reading times of the post-target word were slowed

by presentation of a symbol relative to sentences without a symbol, an effect that did not occur

for non-negative sentences. This result suggests that the readers may have been sensitive to the

fact that spending more time processing the meaning of the negation symbol with the

information presented in the text might help them integrate that information. It is worth noting

that the location of this additional reading time was specific to the post-target word; no effects of

symbol were seen on target reading times. These results indicate that readers waited until after









the negated phrase to incorporate the meaning of the symbol into what they had read, resulting in

longer reading times immediately after the target word, where they were able to pause for wrap-

up and integration (e.g., Stine, 1990; Stine-Morrow, 1996). Another piece of evidence supporting

an influence of the symbol is the young adults' equivalence in reading times for negated and non-

negated targets, unlike Experiment 1 where young adults were slightly faster on negated targets.

The symbol may have made young adults more aware of the nature of the sentences and realize

that they needed to slow down to process the negation more thoroughly.

This additional effort during reading may have been especially beneficial for older adults'

comprehension accuracy, as they had equivalent accuracy to young adults, unlike in Experiment

1 where older adults' comprehension was poorer than young adults. While slower reading times

did not directly lead to better comprehension accuracy, e.g., older adults always had slower

reading times than young adults in Experiment 1 and still had poorer accuracy, the symbol

presented in the present experiment may have served as an "attention-grabbing" mechanism for

older adults, where the symbol signified that the sentences that they were reading were important

and required more attention. However, despite altered reading times in negative sentences with a

symbol, young adults did not demonstrate an increase in their comprehension accuracy. One

possibility is that young adults may be less experienced with the use of negation symbols, unlike

older adults who take more prescription medications and therefore see the negation symbol

presented in conjunction with text more often. As a result, young adults may be less experienced

at incorporating additional information into their situation model representation of the text.

Another possibility is that older adults spent more time processing the symbol and the

sentences in general, given that all of their reading times were slower than young adults. This

additional processing time in conjunction with older adults' experience at incorporating









additional information into a situation model may have increased their comprehension relative to

Experiment 1. A final possibility is that the increased attention may have released some working

memory resources, allowing more resources to be allocated toward reading than other mental

processes. For example, some participants commented that they were trying to predict what they

would be asked as they read the sentences. This allocation of resources toward comprehension

may have helped readers to focus more clearly on creating a memory representation of the whole

text, as opposed to focusing resources on a portion of the text that may not have had anything to

do with the subsequent comprehension question. For young adults, this assistance was not

important for releasing additional working memory resources because they were not as limited as

older adults. This explanation is consistent with theories on working memory (e.g., Just &

Carpenter, 1992), which suggest that when enough working memory resources are available to

process a particular piece of text, that information will be understood better than when not

enough resources are available for processing. The possibility remains that a release of working

memory resources may come not only from context, which Just and Carpenter (1992) suggest,

but also from increased awareness of the situation.

The potential impact of the symbol as an "attention-grabbing" mechanism highlights the

importance of working memory in these experiments, which is supported by the regression

analyses. Working memory was the most significant predictor of reading comprehension

performance in both experiments, although some interesting differences emerged between the

two experiments. For Experiment 1, working memory was the only variable that was a

significant predictor of comprehension for both negative and non-negative sentences. In the

present experiment, vocabulary and age also explained variance in reading comprehension but

only for negative sentences presented with a symbol. Individuals with larger vocabularies are









more experienced with language and may better be able to incorporate the meaning of a symbol

with the meaning of a sentence. Similarly, age also reflects experience, as older adults are more

experienced in seeing symbols combined with text in other contexts, e.g., prescription

medication labels. However, for the sentences without a symbol, only operation span

significantly accounted for variance in comprehension accuracy, and more variance was

accounted for by operation span in Experiment 1 than in Experiment 2. The reason for this result

may be linked to the idea that presence of a symbol increased attention overall, releasing

working memory resources and making working memory less important for comprehension of

these sentences. Interestingly, the variance accounted for by operation span was especially low

for non-negative symbol sentences in Experiment 2, where the symbol really has no meaning to

assist in comprehension. This result suggests that these sentences are easy to read, requiring no

incorporation of the symbol's meaning as well as no negation. Therefore, working memory

would not be taxed when reading these sentences, and the capacity of these resources would not

constrain comprehension.

Another difference that emerged between Experiment 1 and Experiment 2 was the

influence of the probe word on comprehension accuracy. While the probe words had no effect on

comprehension accuracy for negative sentences in Experiment 1, the verbatim probe words

increased comprehension relative to the new probes for negative sentences in Experiment 2. This

difference in results suggests that while readers paid more attention to reading the sentences in

Experiment 2, they may also have been able to use the increased attention to prevent

incorporating information that was unnecessary for the situation model representation. In

addition, the verbatim word may have cued the reader to remember the wording used in the

sentence, which combined with increased attention could increase comprehension accuracy. New









and related probes did not have this effect because they were not present in the text and would

less specifically refer back to a portion of the text that the reader would need to answer the

question.

The second hypothesis, that benefits from a symbol would be more dramatic at the

situation model level, was also not supported. By examining probe word naming times, it was

evident that the presence of a symbol did not specifically affect any level of representation, as it

did not influence naming times at all. The meaning of the symbol was presented to provide a

context and some pre-existing knowledge about the text and therefore should have been

incorporated as such in the situation model representation of the text (Kintsch, 1998), but it was

not, evidenced by the lack of interaction between symbol and probe word type on probe naming

times. This result suggests that readers were unable to understand the meaning/use of the symbol

in relation to the text and therefore could not successfully incorporate the meaning of the symbol

with the meaning of the text. Informal support for this idea comes from comments made by

several participants that they were unsure of what to do with the symbol. Another possibility is

that the symbol did not provide enough information to create a context for the sentence it

described, so that it could not assist in creating an appropriate memory representation of the text.

In terms of levels of representation, this possibility means that the symbol would not be included

in the situation model representation of the text, as it did not serve as a form of pre-existing

knowledge related to the text (e.g., Kintsch, 1998).

More generally, the results of Experiment 1 and Experiment 2 displayed remarkable

consistency. The stability of the findings suggests that the effects of negation on comprehension

are real (e.g., Cornish, 1971, Experiment 1; Cornish & Wason, 1970; Hoosain, 1973; Just &

Carpenter, 1971; Kaup, 2001; Kaup et al., 2004; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup









et al., in press; MacDonald & Just, 1989; Sherman, 1973) and extend to more real-world types of

texts. Negation slowed probe naming times and decreased comprehension accuracy, suggesting

that negation is a complex structure that causes readers difficulty in processing and remembering

negated concepts. This experiment showed that a negation symbol helped both young and older

adult readers to better process all sentences, not just negated ones. Perhaps other compensatory

strategies can help readers to specifically process negation more effectively.











Table 3-1. Means and standard deviations (st. dev.) of demographic measures.
Age group


Young
Mean St. dev
Years of education** 13.36 1.39
Health rating (out of 10) 8.22 1.39
Vocabulary (out of 25)** 13.32 3.33
Forward digit span 7.22 1.19
Backward digit span 4.87 1.16
Hours spent writing** 2.78 1.88
Hours spent reading 3.28 1.92
Hours watching TV 2.30 1.72
Hours doing crosswords* .28 .76
Spelling training (out of 10)** 6.17 2.08
Spelling ability now 6.33 1.53
(out of 10)
Operation span** .90 .09
Congruent Stroop** 22.04 5.74
Incongruent Stroop** 20.68 3.39
Note: indicates that the age differences were significant at the p
the age differences were significant at the p <.01 level


Older


Mean
16.83
7.98
20.98
7.05
5.15
1.55
2.97
2.15
.62
7.31
6.32


.85
37.22
45.74
<.05 level,


St. dev.
3.46
1.70
2.86
1.31
1.41
1.18
1.72
1.43
.94
2.32
1.94


.10
14.12
11.41
** indicates that


Table 3-2. Means and standard deviations for overall reading times (msec)
Age group Sentence type Symbol Mean Standard deviation
Young Negative Symbol 434.76 101.24
No symbol 434.01 100.76
Non-negative Symbol 439.96 111.83
No symbol 439.42 106.65
Older Negative Symbol 589.42 183.18
No symbol 589.25 191.28
Non-negative Symbol 591.32 180.37
No symbol 588.82 182.49









Table 3-3. Means and standard deviations for target word reading times (msec)
Age group Sentence type Symbol Mean Standard deviation
Young Negative Symbol 429.60 120.05
No symbol 425.24 113.01
Non-negative Symbol 437.40 128.57
No symbol 430.73 119.82
Older Negative Symbol 613.56 220.08
No symbol 619.93 246.10
Non-negative Symbol 611.44 223.05
No symbol 606.52 217.89


Table 3-4. Means and standard deviations for post-target word reading times (msec)
Age group Sentence type Symbol Mean Standard Deviation
Young Negative Symbol 470.10 181.15
No symbol 451.84 122.17
Non-negative Symbol 466.13 143.89
No symbol 482.30 161.03
Older Negative Symbol 641.07 241.00
No symbol 615.32 204.95
Non-negative Symbol 634.52 228.87
No symbol 624.04 216.83


Table 3-5. Means and standard deviations for modifier and target word reading times for
negative sentences only (msec)
Age group Word type Symbol Mean Standard deviation
Young Modifier Symbol 386.35 101.85
No symbol 383.60 98.32
Target Symbol 429.60 120.05
No symbol 425.24 113.01
Older Modifier Symbol 545.86 176.20
No symbol 545.56 193.48
Target Symbol 613.56 220.08
No symbol 619.93 246.10









Table 3-6. Means and standard deviations of probe word naming times (msec)


Age group
Young


Sentence type
Negative






Non-negative






Negative






Non-negative


Probe type Symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol


Mean
759.86
757.62
793.65
778.96
808.83
801.54
737.37
739.05
785.55
770.13
788.28
810.71
817.62
815.01
844.19
831.07
854.78
872.56
737.47
802.21
845.49
827.86
869.75
840.74


Table 3-7. Means and standard deviations for comprehension accuracy (%)
Age group Sentence type Symbol Mean Standard deviation
Young Negative Symbol 69.38 16.45
No symbol 70.32 15.10
Non-negative Symbol 75.60 11.70
No symbol 75.68 13.51
Older Negative Symbol 67.46 16.57
No symbol 68.02 15.80
Non-negative Symbol 73.05 13.39
No symbol 73.34 12.90


Std. deviation
109.85
128.55
160.64
113.06
130.03
134.94
111.60
103.54
119.49
119.51
124.67
139.90
182.71
182.20
195.82
178.14
201.88
222.61
111.60
174.52
189.63
170.98
223.51
197.78


Older











Table 3-8. Means and standard deviations for comprehension accuracy by probe type (%)


Sentence type
Negative


Non-negative


Negative


Non-negative


Age group
Young


Probe type Symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol
Verbatim Symbol
No symbol
Related Symbol
No symbol
New Symbol
No symbol


Mean
71.27
69.16
66.25
70.45
66.74
67.55
80.31
80.89
75.36
76.11
67.70
72.41
69.77
69.51
69.01
70.16
65.17
65.33
75.15
80.31
75.19
73.48
68.89
6756


Table 3-9. Results of regression analysis on negative sentence presented with symbol
comprehension accuracy
Variable 13 p
Operation span .20 < .03
Vocabulary .58 < .01
Age -.49 < .01
Stroop effect -.07 .47
Years of education -.17 .12
Hours spent reading .03 .69


Std. deviation
21.16
20.48
21.32
24.61
26.01
20.04
16.70
25.47
18.98
21.89
22.14
20.44
21.95
22.55
22.89
25.73
23.90
21.67
18.77
16.78
14.47
16.34
19.76
24 19


Older









Table 3-10. Results of regression analysis on negative sentence presented without a symbol
comprehension accuracy
Variable 13 p
Operation span .31 < .01
Stroop effect -.02 .81
Age -.02 .82
Vocabulary .14 .11
Years of education .01 .96
Hours spent reading .11 .21


Table 3-11. Results of regression analysis on non-negative sentence presented with symbol
comprehension accuracy
Variable 13 p
Operation span .18 <.05
Stroop effect -.06 .51
Age -.04 .66
Vocabulary .05 .58
Years of education .04 .67
Hours spent reading .06 .53


Table 3-12. Results of regression analysis on non-negative sentence presented without symbol
comprehension accuracy
Variable 13 p
Operation span .28 < .01
Stroop effect .05 .58
Age -.03 .75
Vocabulary .14 .13
Years of education .07 .43
Hours spent reading .05 .61












Sentence Type x Symbol Interaction for Post-Target Reading
Times

Symbol
o No symbol


580

S570
E
560

E 550
i-
0540

530

520

510

500


Negative


Non-negative


Sentence Type


Figure 3-1. Post-target reading times by sentence type and presence of symbol.









CHAPTER 4
GENERAL DISCUSSION

The present research has offered insight into reading under circumstances that increased

the difficulty of processing, comprehending, and retrieving information. Specifically, the present

research investigated how young and older adults were influenced by negation (e.g., "no," "not,"

or "never") while reading real-world types of texts, and how individual differences, i.e., working

memory differences, contributed to age-related changes in reading. This research also explored

the effects of a compensatory strategy on comprehension of negated text. Several general

conclusions can be drawn and are discussed below.

Consistent with previous research on negation, naming and comprehension following

reading were affected by the presence of negation in text. Similar findings occurred in both

experiments, suggesting that these effects of negation on reading are stable. The present research

extended these findings to online measures of reading, although negation speeded target reading

times, especially in young adult readers. Except for this age difference in reading times, negation

similarly influenced young and older adults' comprehension across multiple measures when

reading text, contrary to the predictions of IDH (e.g., Hasher & Zacks, 1988). MacDonald and

Just (1989) suggested that when a person reads a negative sentence, the negated concept must be

inhibited because it is no longer appropriate to the situation described by the text. The IDH

predicted that older adults would have more difficulty discarding inappropriate information that

has been gated into working memory, which in this case would be the concept that was negated.

Although older adults named the probes more slowly than young adults and had poorer

comprehension accuracy overall, the presence of negation did not exacerbate these age

differences. Thus, inhibition of negative concepts was not especially problematic for older

adults, which has several implications.









First, negation may not always be processed in terms of inhibition; instead, negation can be

processed in other ways, such as by transforming the negated concept into its alternate meaning

(e.g., Gough, 1965; Mehler, 1963). Transformations would require more working memory

resources to process negation because the reader needs to hold the affirmative version of the

sentence in memory and then manipulate it to its opposite based on the negative modifier.

Support for this idea comes from the importance of working memory in the present experiments,

as shown by the regression analyses. Second, given that older adults demonstrated more

difficulty with inhibition than young adults as evidenced by the Stroop test, but did not

demonstrate a difficulty with inhibition during reading negation, it appears that the IDH may not

account for older adults' comprehension of negation. The present research therefore demonstrates

another circumstance under which older adults do not show an exacerbated difficulty with

inhibition (e.g., Burke & Harrold, 1988; James et al., 1998).

The lack of inhibition observed in the present experiments may also be related to the

stimuli. MacDonald and Just (1989) suggested that inhibition was important for understanding

negation, as a reader would have to de-activate the negated concept in order to have a correct

representation of the sentence. However, the negated sentences in their experiments always took

the same form, e.g., "Elizabeth baked some bread but no cookies." In these sentences, the

presence of "cookies" is denied, so it needs to be removed from the memory representation of

the text in order to accurately represent what was present. In the present experiments, negation

was presented in different ways, e.g., "He could not part with his school or the people in it." In

sentences like this one, the meaning of the negation is not to deny the presence of an item but

instead is to change the meaning of the word it modifies. These differences in how negation was









used in the experimental sentences could have contributed to the lack of inhibition observed in

the present research.

The present research also demonstrated that negation did not have a greater influence at the

situation model level, compared with the surface level of representation, even though negation

did affect memory for the text overall. In addition, there were no specific circumstances (e.g., the

situation model level) where older adults were able to offset the difficulty produced by negation

based on their extensive experience with processing information as a gist representation. An

examination of the probe naming times demonstrated that all probes were equivalently slowed

following a negative sentence relative to a non-negative sentence, suggesting that the different

levels of representation were not differentially affected by the presence of negation. This result is

contrary to Kaup (2001), who demonstrated that negation affected the situation model level of

representation but not the surface level. Perhaps the effects of negation on different levels of

representation are task-dependent; Kaup (2001) used recognition as an activation measure,

whereas the present research used probe naming. As a task involving explicit memory,

recognition may more directly tap into the memory representations that are affected by the

negative concept, whereas probe naming, an implicit task, may be less sensitive.

With respect to compensatory strategies, the presence of a symbol did not help a reader to

form a correct representation of the negated text, at any level, contrary to predictions derived

from the CI Model (Kintsch, 1998). Kintsch (1998) proposed that a reader will initially activate

all of the information presented, including the text and any related information, and then

activation is constrained to include only the information necessary to the meaning of the text,

rather than the exact wording, creating the situation model representation. The information

included in the representation includes pre-existing knowledge about the topic the reader is









reading, which in the present research would have included the symbol presented before and

during reading. According to Just and Carpenter (1992), providing a context can release working

memory resources for comprehension. Combining these two theories, this contextual information

may guide the activation of the information presented during reading, resulting in a well-

understood and well-represented situation model. However, the symbol in the present experiment

did not assist in providing a context for readers, nor did it appear to be integrated with the

information presented in the text. These findings suggest that (a) either the symbol did not

provide enough information to serve as compensation in the form of a context, or (b) readers

were unable to use the symbol effectively in that capacity. The first possibility would suggest

that readers were unaware of the meaning of the symbol, but this option seems unlikely, given

that the circle with the line through it is the universal symbol for "do not." The second possibility

implies that readers were unable to integrate the meaning of the symbol with the meaning of the

negated phrase in the sentence. This suggestion seems more plausible because the readers were

not explicitly instructed about what to do with the symbol; therefore, they either may not have

seen the connection between the meaning of the text and the meaning of the symbol or did not

know that they were supposed to connect the two.

Although a symbol did not have specific effects on the processing of negated text, there

was some evidence that the presence of a symbol may have increased older adults' attention to

the reading task, as shown in post-target reading times. Furthermore, the age differences in

comprehension accuracy that appeared in Experiment 1 disappeared in Experiment 2, even when

no symbol was present. Research on pictorial descriptors given with medication instructions

indicate that people prefer when a picture is presented than when it is not, because the picture

reinforces the information presented in the text (e.g., Sojourner & Wogalter, 1997). A preference









for instructions including a depiction of the instruction may encourage readers to pay more

attention to those instructions, which may seem more important because of the added symbol.

This same logic can be applied to the present research such that readers may have generally

increased their attention to the sentences because a symbol was present. However, this resulted in

increasing overall accuracy because once readers recognized that the sentences may be

important, they continued to pay closer attention to all of the sentences, even when no symbol

was present. This increased attention was especially relevant for older adult readers, who are

more experienced with the negation symbol from seeing it on prescription labels and have

already realized its importance.

The present research also demonstrated the impact of negation when reading real-world

texts, which is a significant extension of previous work on negation. Some of the previous

research on negation dealt with text that more closely resembled logic statements (e.g., "The dot

is not red;" see Cornish, 1971), perhaps in the attempt to make the experimental sentences focus

only on negation and avoid other types of complexities that may interact with the negation.

However, these are not the ways in which readers normally encounter negation. By using a more

ecological approach (e.g., Neisser, 1978) and by having participants read sentences from actual

fiction stories, the present research replicated the detrimental effects of negation that were

demonstrated in previous experiments. The similarities among results indicate that the

difficulties associated with understanding negation while reading are real and are not a function

of difficulties in logical reasoning. The present research demonstrates the importance of

continuing to use real-world texts to explore comprehension in the world in which we encounter

it. In addition, the present research demonstrates that older adults did not show specific deficits









in comprehension of negation with these types of texts, which may have positive implications for

their ability to function in real-world situations.

Limitations

There are several limitations of the present experiments that are worthy of discussion.

First, the text that was used was very rich in context, where many details described by the text

were given to depict a vivid situation. As a result, older adults may not have shown as much

decline in comprehension because they were reading text under more optimal circumstances than

they may have encountered in previous research. Previous research on reading indicates that

older adults are better able to store and represent text in their memory in the situation model

representation because this level is the most useful representation (e.g., Radvansky et al., 2001).

It is possible that the stimuli used in the present research lent themselves to easily forming a rich

and meaningful representation, even when negation was present. Given more detail and more

description of events than sentences used in previous research on negation, the sequence of

events in the sentences used here were more easily visualized, and older adult readers might have

more easily drawn upon their experience and information in the text to form a situation model

representation. The additional depth of information to the sentences in the present research may

have put older adults at an advantage, shielding potential age differences in comprehension of

negation that would have appeared if the exact wording of the text had been required to show

comprehension.

Second, previous research on negation has demonstrated that various types of negation

may lead to different results (e.g., Just & Carpenter, 1971; Sherman, 1973). For example, the

addition of the word "not" makes comprehension more difficult than the addition of another type

of negative word, like "few" or the prefix "un-." The present research used the word "not"

because it was expected to make comprehension more difficult than the other types, allowing the









possibility of finding age differences. However, the word "not" may have multiple functions,

serving as both a word that denies existence (e.g., "She some baked bread but no cookies.") and

as a word that changes meaning (e.g., "He could not part with his school or the people in it.").

These differing functions may both cause difficulties in comprehension but for different reasons.

Where the function of the word "not" is to deny existence of an object, processing most likely

involves inhibition. Where the function of the word "not" is to change the meaning to its

opposite, processing is likely more working-memory intensive because of the required

transformation (e.g., Mehler, 1963; Gough, 1965). Given these potential processing differences,

it is evident that "not" as a structure in influencing comprehension may be more complex than

previously thought.

Third, the time spent attending to the symbol was not recorded, so there is no indication of

whether or not readers were spending enough time processing the symbol for the symbol to serve

as compensation. In addition, the symbol was present during the entire sentence, and there was

no way to measure whether people looked back at the symbol periodically. While there is no

standard amount of time that a reader should spend processing the pictorial information when it

is presented, a longer amount of time may be related to the usefulness of the symbol. Readers

who spent longer looking at the symbol, or exerted more effort to incorporate the meaning of the

symbol into the text that they were reading, may have improved their own comprehension.

Furthermore, age differences in time spent processing the symbol are also possible and could be

linked to the lack of age differences observed in comprehension accuracy when symbols were

used in the experiment. However, this information is not available from the measures gathered

from the present research.









Fourth, a comparison in the present research has been made between instructions (such as

medication instructions used with a symbol) and fictional text. The limitation here is that these

are two very different kinds of texts. For instructions, it is often the case that the exact wording is

very important to the message, whereas for fictional text and other types of narratives, the

message can be remembered and stored as a situation model or just the gist. Due to the different

nature of these types of text, the process by which a reader may read and store the information,

as well as incorporate the meaning from a symbol, may be different. For instructions, where the

wording is important to the message, the symbol may be mapped on to a specific word. For

fictional or narrative text, where the wording is not as important to the message, the meaning

from the symbol may simply be incorporated into the situation model of the text. Therefore,

although the present experiments used more real-world texts, their generalizability to texts like

medication instructions may be limited.

Finally, in the second experiment, participants were not given explicit instruction on how

to interpret the symbol or how to incorporate the symbol into the meaning of the sentence; they

were just told to look at the symbol before they started to read the sentence. After completing the

experiment, some participants reported not knowing what to do with the symbol or why it was

there. These post-hoc reports suggest that the implicit nature of the symbol may not have been

sufficient to provide compensation in reading these types of sentences. Giving some deliberate

instruction on how to consciously use the symbol, rather than implicit ones, may have increased

the impact that it could have had on compensation. For example, explicit instructions to use the

negation symbol may have encouraged readers to develop a compensatory reading strategy to

help them improve their comprehension of the negation.









Future Directions

Future research on negation and aging should continue to focus on three important aspects

of these topics that were explored here: (1) circumstances where a reader might have more or

less difficulty in understanding negation, (2) circumstances where a reader or a writer may

improve comprehension of text involving negation, and (3) exploring and applying these

investigations within the context of real-world types of texts (e.g., non-fiction, instructions,

jokes).

One direction for future research is to further break down the different types of negation

that may cause difficulty in comprehension. While some breakdown has occurred in the past

(e.g., Just & Carpenter, 1971; Sherman, 1973), the present research has indicated that additional

exploration may be necessary. It seems fundamentally different to negate a noun (e.g.,

"cookies") than to negate an adjective (e.g., "bright."). One argument is that negative modifiers

may differentially impact targets as a function of their parts of speech. Negating a noun often

indicates that it is not present and would be processed by deactivating the negated concept.

However, negating an adjective would simply change the meaning, and would be processed by

transforming the sentence into its negative form. Given that older adults have demonstrated

difficulty with inhibition in the Stroop test, they may be more differentially impacted by a type

of negation that evokes inhibition (e.g., negating a noun), but not differentially impacted by a

type of negation that involves a change of meaning or transformation (e.g., negating an

adjective).

Another direction for future research involves a further examination of the use of symbols

on comprehension. Given that readers prefer the presentation of symbols in an instructional

situation (e.g., Sojourner & Wogalter, 1997), there is some indication that some people may find

them helpful. Future research should investigate whether more time processing the symbol or









whether conscious strategies for using the symbol will improve assistance from the symbol. In

addition, given the difference between instructional text and fiction, future research may also be

able to determine whether symbols, when understood by the reader, may have a larger impact on

one type of text over the other. Depending on the strategy given, mapping the symbol onto a

particular word would assist instructional text better than narrative text, whereas instructing

participants to incorporate the meaning of the symbol with the meaning of the text may better

benefit narrative text.

A final direction for future research is the investigation of additional mechanisms for

compensation in reading negation. Given that older adults have demonstrated that they pause

more often while reading complex sentences, leading to better retrieval of the information they

read (e.g., Stine, 1990), and that they are also able to apply a strategy of pausing more often

during reading to improve their comprehension (e.g., Noh et al., 2006), an exploration of pausing

during reading negation seems logical. Whether individuals who spontaneously pause after

reading a negation have better comprehension accuracy than individuals who do not pause could

be assessed. For readers who do not naturally adjust their reading via pauses, a training strategy

could be employed to investigate whether explicit, effortful strategies can improve

comprehension of negation, e.g., instructing readers to pause after reading a negative phrase.

Since negative sentences can be categorized as another type of complex sentence, pausing after

reading the negated concept may provide time for wrap-up and to complete processing of this

difficult construction.

The exploration of negation comprehension in the present research speaks to the

importance of this type of construction in text. Given that readers of all ages have difficulty

understanding text that contains negative words, both readers and writers need to be sensitive to









the use of this construction. Readers need to learn how to adjust their reading strategies when

encountering this construction. Writers should avoid using negation to minimize comprehension

difficulties whenever possible. Negation appears in many everyday situations, oftentimes without

our awareness because it is a small, simple word, but people should be aware that little things

can have big effects on comprehension.









APPENDIX A
SAMPLE STIMULI

Sample negative sentence:

As Joel looked around, he thought about what leaving his school would mean, and he knew that

he could not part with the school and the people in it.

Sample non-negative sentence:

As Joel looked around, he thought about what leaving his school would mean, and he knew that

he could part with the school and the people in it.

Sample probes and frequencies:

Verbatim probe part (11)

Related probe split (26)

New probe mash (4)

Sample comprehension question:

What did Joel know about his ability to leave his school and the people in it?









APPENDIX B
EXPERIMENT 1 INSTRUCTIONS

In this experiment, you will be asked to read sentences, say aloud specific words, and answer
questions about the sentences that you have read. It is important that you read in order to
understand the gist of what you are reading, but it is not necessary that you memorize each
sentence.

The reading task is different from how you normally read. You will be silently reading
sentences one word at a time, where you will control your own reading speed. The words in the
paragraphs will show up as dashes (--------) on the screen, and you will press the "Space Bar"
key to make a word appear. Each time you press the "Space Bar", the previous word will return
to dashes, and the next word will appear.

The sentences you will be reading have been taken from a story about magic, so you may notice
some unusual events happening. Just read the sentences at your normal pace, and do your best to
understand what you are reading, as you will be asked comprehension questions about them.
Remember, you do not need to memorize the sentences word-for-word, but you will be asked to
remember the main ideas of the sentences.

After reading the sentences, you will see a single word on the screen. You need to say this word
aloud, as soon as it appears. Say this word as quickly and as naturally as you can, but be careful
not to make any extraneous noises, such as coughing or clearing your throat that may be picked
up by the microphone. The main point is to say this word aloud as soon as possible after seeing
it.

After saying the word out loud, press the "Enter" key, and a question will appear on the screen.
Read the question and say your answer out loud so that the experimenter can write down your
answer. You may answer using one or more words, whatever you feel is necessary to answer the
question. In addition, at this time, please feel free to give any comments about whether or not
you feel your answer was hindered by an awkward construction of the sentence. However, you
do not need to give comments if you do not have any, simply wait until the experimenter has
finished writing, and press the "Enter" key and continue on to the next sentence.

This pattern of tasks (reading, saying words, and answering questions) will repeat throughout the
experiment, so be sure that you understand what tasks you will be asked to do. If you have
questions, please ask the experimenter now. Otherwise, press the "Enter" key, and you will be
given several practice trials to get familiar with the tasks.









APPENDIX C
EXPERIMENT 2 INSTRUCTIONS

In this experiment, you will be asked to read sentences, say aloud specific words, and answer
questions about the sentences that you have read. It is important that you read in order to
understand the gist of what you are reading, but it is not necessary that you memorize each
sentence.

The reading task is different from how you normally read. You will be silently reading
sentences one word at a time, where you will control your own reading speed. The words in the
paragraphs will show up as dashes (--------) on the screen, and you will press the "Space Bar"
key to make a word appear. Each time you press the "Space Bar", the previous word will return
to dashes, and the next word will appear.

In addition, you will sometimes see a symbol presented before a sentence. The symbol may or
may not seem relevant in helping you to understand the sentence, but it is important to look at
the symbol before starting to read the sentence. After looking at the symbol, you will press the
"Space Bar" to begin reading the sentence.

The sentences you will be reading have been taken from a story about magic, so you may notice
some unusual events happening. Just read the sentences at your normal pace, and do your best to
understand what you are reading, as you will be asked comprehension questions about them.
Remember, you do not need to memorize the sentences word-for-word, but you will be asked to
remember the main ideas of the sentences.

After reading the sentences, you will see a single word on the screen. You need to say this word
aloud, as soon as it appears. Say this word as quickly and as naturally as you can, but be careful
not to make any extraneous noises, such as coughing or clearing your throat that may be picked
up by the microphone. The main point is to say this word aloud as soon as possible after seeing
it.

After saying the word out loud, press the "Enter" key, and a question will appear on the screen.
Read the question and say your answer out loud so that the experimenter can write down your
answer. You may answer using one or more words, whatever you feel is necessary to answer the
question. In addition, at this time, please feel free to give any comments about whether or not
you feel your answer was hindered by an awkward construction of the sentence. However, you
do not need to give comments if you do not have any, simply wait until the experimenter has
finished writing, and press the "Enter" key and continue on to the next sentence.

This pattern of tasks (reading, saying words, and answering questions) will repeat throughout the
experiment, so be sure that you understand what tasks you will be asked to do. If you have
questions, please ask the experimenter now. Otherwise, press the "Enter" key, and you will be
given several practice trials to get familiar with the tasks.









LIST OF REFERENCES


Burke, D. M. (1997). Language, aging, and inhibitory deficits: Evaluation of a theory. Journal of
Geronotology: Psychological Sciences, 52B, P254-P264.

Burke, D. M., & Harrold, R. M. (1988). Automatic and effortful semantic processes in old
age:Experimental and naturalistic approaches. In L.L. Light & D. M. Burke (Eds.),
Language, memory and aging (pp. 100-116). New York: Cambridge University Press.

Connelly, S. L., Hasher, L., & Zacks, R. T. (1991). Age and reading: The impact of distraction.
Psychology and Aging, 6, 533-541.

Conway, A.R.A., Kane, M.J., Bunting, M.F., Hambrick, D.Z., Wilhelm, O., & Engle, R.W.
(2005). Working memory span tasks: A methodological review and user's guide.
Psychonomic Bulletin & Review, 12, 769-786.

Cornish, E. R. (1971). Pragmatic aspects of negation in sentence evaluation and completion
tasks. British Journal ofPsychology, 62, 505-511.

Cornish, E. R., & Wason, P. C. (1970). The recall of affirmative and negative sentences in an
incidental learning task. Quarterly Journal ofExperimental Psychology, 22, 109-114.

Dowse, R., & Ehlers, M.S. (2002). The influence of education on the interpretation of
pharmaceutical pictograms for communicating medicine instructions. International
Journal ofPharmacy Practice, 11, 11-18.

Francis, W.N., & Kucera, H. (1982). Frequency analysis of English usage: Lexicon and
Grammar. Boston: Houghton and Mifflin.

Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Mini-mental state: A practical method
for grading the state of patients for the clinician. Journal of Psychiatric Research, 12, 189-
198.

Gough, P. B. (1965). Grammatical transformations and speed of understanding. Journal of
Verbal Learning and Verbal Behavior, 4, 107-111.

Grown-ups are turning to teen books. (2005, September 10). The Gainesville Sun, p. 3D.

Hamm, V. P., & Hasher, L. (1992). Age and the availability of inferences. Psychology and
Aging, 7, 56-64.

Hasher, L., & Zacks, R.T. (1988). Working memory, comprehension, and aging: A review and a
new view. In G.H. Bower (Ed.), The psychology of learning and motivation (Vol. 22, pp.
193-225). New York: Academic Press.

Hoosain, R. (1973). The processing of negation. Journal of Verbal Learning and Verbal
Behavior, 12, 618-626.









James, L.E., Burke, D.M., Austin, A., & Hulme, E. (1998). Production and perception of
"verbosity" in younger and older adults. Psychology andAging, 13, 355-367.

Just, M.A., & Carpenter, P.A. (1971). Comprehension of negation with quantification. Journal of
Verbal Learning and Verbal Behavior, 10, 244-253.

Just, M.A., & Carpenter, P.A. (1992). A capacity theory of comprehension: Individual
differences in working memory. Psychological Review, 99, 122-149.

Kaup, B. (2001). Negation and its impact on the accessibility of text information. Memory and
Cognition, 29, 960-967.

Kaup, B., Dijkstra, K., & Ludtke, J. (2004). Resolving anaphors after reading negative sentences.
Presented at the 44th annual meeting of the Psychonomic Society, Minneapolis,
Minnesota.

Kaup, B., Ludtke, J., & Zwaan, R. A. (2005). Effects of negation, truth value, and delay on
picture recognition after reading affirmative and negative sentences. Proceedings of the
27th annual conference of the Cognitive Science Society.

Kaup, B., & Zwaan, R.A. (2003). Effects of negation and situational presence on the
accessibility of text information. Journal of Experimental Psychology: Learning. Memory,
and Cognition, 29, 439-446.

Kaup, B., Zwaan, R. A., & Ludtke, J. (in press). The experiential view of language
comprehension: How is negation represented? In F. Schmalhofer & C. Perfetti (Eds.),
Higher level language processes in the brain. Inference and comprehension processes.
Mahwah, NJ: Erlbaum.

Keele, S. (1972). Attention demands of memory retrieval. Journal ofExperimental Psychology,
73, 245-248.

Kemper, S. (1987). Syntactic complexity and elderly adults' prose recall. Experimental Aging
Research, 13, 47-65.

Kemtes, K. A., & Kemper, S. (1997). Younger and older adults' on-line processing of
syntactically ambiguous sentences. Psychology andAging, 12, 362-371.

Kintsch, W. (1998). Comprehension: A paradigm for cognition. New York: Cambridge
University Press.

Light, L. L. (1988). Language and aging: Competence versus performance. In J.E. Birren & V.L.
Bengston (Eds.), Emergent ithi, i e' of aging (pp. 177-213). New York: Springer
Publishing.

MacDonald, M. C., & Just, M. A. (1989). Changes in activation levels with negation. Journal of
Experimental Psychology: Leuiri1g. Memory, and Cognition, 15, 633-642.









MacKay, D. G., & Abrams, L. (1996). Language, memory, and aging: Distributed deficits and
the structure of new-versus-old connections. In J.E. Birren & K.W. Schaie (Eds.),
Handbook ofpsychology of aging (pp. 251-256). San Diego: Academic Press.

MacLeod, C.M. (1991). Half a century of research on the Stroop effect: An integrative review.
Psychological Bulletin, 109, 163-203.

May, C. P., Zacks, R. T., Hasher, L., & Multhaup, K. S. (1999). Inhibition in the processing of
garden-path sentences. Psychology andAging, 14, 304-313.

Mehler, J. (1963). Some effects of grammatical transformations on the recall of English
sentences. Journal of Verbal Learning and Verbal Behavior, 2, 346-351.

Neisser, U. (1978). Memory: What are the important questions? In M.M. Gruneberg, P.E.
Morris, & R.N. Sykes (Eds.), Practical aspects of memory (pp. 3-24). New York:
Academic Press.

Newsome, M.R., & Glucksberg, S. (2002). Older adults filter irrelevant information during
metaphor comprehension. Experimental Aging Research, 28, 253-267.

Noh, S.R., Shake, M.C., Joncich, A.D., Hindin, S.B., & Stine-Morrow, E.A.L. (2006, April).
Age differences in the effects of instruction on resource allocation in reading. Poster
presented at the 11th biennial Cognitive Aging Conference, Atlanta, GA.

Norman, S., Kemper, S., & Kynette, D. (1992). Adults' reading comprehension: Effects of
syntactic complexity and working memory. Journals of Gerontology: Psychological
Sciences, 47, P258-P265.

Norman, S., Kemper, S., Kynette, D., Cheung, H., & Anagnopoulos, C. (1991). Syntactic
complexity and adults' running memory span. Journal of Geronotology: Psychological
Sciences, 46, P346-P351.

Radvansky, G. A., Copeland, D. E., Berish, D. E., & Dijkstra, K. (2003). Aging and situation
model updating. Agin,. Neuropsychology, and Cognition, 10, 158-166.

Radvansky, G. A., & Curiel, J. M. (1998). Narrative comprehension and aging: The fate of
completed goal information. Psychology andAging, 13, 69-79.

Radvansky, G. A., Zacks, R. T., & Hasher, L. (2005). Age and inhibition: The retrieval of
situation models. Journal of Gerontology: Psychological Sciences, 60B, P276-P278.

Radvansky, G. A., Zwaan, R. A., Curiel, J. M., & Copeland, D. E. (2001). Situation models and
aging. Psychology and Aging, 16, 145-160.

Salthouse, T.A. (1996). The processing-speed theory of adult age differences in cognition.
Psychological Review, 103, 403-428.









Sherman, M. A. (1973). Bound to be easier? The negative prefix and sentence comprehension.
Journal of Verbal Learning and Verbal Behavior, 12, 76-84.

Sichel, J.L., & Chandler, K.A. (1969). The color-word interference test: The effects of varied
color-word combinations upon verbal response latency. Journal ofPsychology, 72, 219-
231.

Sojourner, R.J., & Wogalter, M.S. (1997). The influence of pictorials on evaluations of
prescription medication instructions. Drug Information Journal, 31, 963-972.

Stine, E. A. L. (1990). On-line processing of written text by younger and older adults.
Psychology and Aging, 5, 68-78.

Stine-Morrow, E. A. L., Loveless, M. K., & Soederberg, L. M. (1996). Resource allocation in
on-line reading by younger and older adults. Psychology and Aging, 11, 475-486.

Stine-Morrow, E. A. L., Ryan, S., & Leonard, J. S. (2000). Age differences in on-line syntactic
processing. Experimental Aging Research, 26, 315-322.

Stroop, J.R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental
Psychology, 18, 643-662.

Turner, M.L., & Engle, R.W. (1989). Is working memory capacity task dependent? Journal of
Memory and Language, 28, 127-154.

Waters, G. S., & Caplan, D. (1996). Processing resource capacity and the comprehension of
garden path sentences. Memory & Cognition, 24, 342-355.

Waters, G. S., & Caplan, D. (2001). Age, working memory and on-line syntactic processing in
sentence comprehension. Psychology and Aging, 16, 128-144.

Zwaan, R. A., & Radvansky, G. A. (1998). Situation models in language comprehension and
memory. Psychological Bulletin, 123, 162-185.

Zwaan, R. A., & Singer, M. (2003). Text comprehension. In A.C. Graesser, M.A. Gernsbacher,
& S.R. Goldman (Eds.), Handbook of discourse processes (pp. 83-121). Mahwah, NJ:
Lawrence Erlbaum Associates.









BIOGRAPHICAL SKETCH

Sara J. Margolin was born in Trumbull, Connecticut, in 1981, but relocated with her family

to Boynton Beach, Florida, at age 10. She graduated from Santaluces Community High School

as Salutatorian in 1999, and decided to continue her education at the University of Florida. She

began her studies at the university in the fall semester of 1999, and completed her Bachelor of

Science degree in 2002, graduating with honors with a major in psychology, a minor in

mathematics, and an invitation to be inducted into Phi Beta Kappa. Upon graduating, Sara

immediately enrolled in the graduate program at the University of Florida to continue her studies

in cognitive psychology. Studying language, memory, and their interaction with aging, she

graduated with her Master of Science degree in 2005, and continued working toward her Ph.D. in

cognitive psychology. During this time, Sara has also worked toward earning a Certificate in

gerontology.

After finishing her degree, Sara began her career as a faculty member at the State

University of New York, College at Brockport. She is living in Rochester, New York with her

husband and two dogs.





PAGE 1

1 NEGATION COMPREHENSION AND AGING: THE ROLE OF INHIBITION AND THE IMPACT OF COMPENSATION DURING READING By SARA J. MARGOLIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Sara J. Margolin

PAGE 3

3 To those in my life who have encouraged, s upported, inspired, and helped me achieve my goalsIt is with gratitude that I complete this work, and dedica te it to my husband, my family, and my dearest friends.

PAGE 4

4 ACKNOWLEDGMENTS I thank the young and older adul ts who participated in my experiments; their help and generosity are deeply appreciated. A special thank you goes to Jennifer Eastman, Max Thomas, Giovanna Morini, and Lindsey Bartlett for their assistance in stimuli development and data collection. I gratefully thank my committee members for their thoughtful insights into guiding me through this project. Above all, I thank Li se Abrams, whose attention, comments, and guidance made this project possible and the best it could possibly be.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............10 CHAPTER 1 INTRODUCTION..................................................................................................................12 The Impact of Negation on Reading Comprehension............................................................15 Theories of Language Comprehension and Aging.................................................................17 Specific Aims.................................................................................................................. ........21 2 EXPERIMENT 1................................................................................................................... .23 Hypothesis 1................................................................................................................... ........23 Hypothesis 2................................................................................................................... ........24 Methods........................................................................................................................ ..........26 Participants................................................................................................................... ...26 Design......................................................................................................................... .....26 Materials...................................................................................................................... ....26 Procedure...................................................................................................................... ...30 Results........................................................................................................................ .............31 Demographics, Operation Span, and Stroop Test...........................................................31 Reading Times.................................................................................................................32 Overall reading times...............................................................................................33 Target word reading times........................................................................................33 Post-target word reading times.................................................................................34 Modifier versus target reading times for negative sentences only...........................34 Probe Naming Times.......................................................................................................35 Comprehension Accuracy...............................................................................................36 The Relationship between Reading Time s, Naming Times, and Comprehension Accuracy......................................................................................................................38 Discussion..................................................................................................................... ..........39 3 EXPERIMENT 2................................................................................................................... .48 Hypothesis 1................................................................................................................... ........49 Hypothesis 2................................................................................................................... ........50 Methods........................................................................................................................ ..........51 Participants................................................................................................................... ...51

PAGE 6

6 Design......................................................................................................................... .....51 Materials...................................................................................................................... ....51 Procedure...................................................................................................................... ...52 Results........................................................................................................................ .............53 Demographics, Operation Span, and Stroop Test...........................................................53 Reading Times.................................................................................................................53 Overall reading times...............................................................................................54 Target word reading times........................................................................................54 Post-target word reading times.................................................................................55 Modifier versus target reading times for negative sentences only...........................55 Probe Naming Times.......................................................................................................56 Comprehension Accuracy...............................................................................................56 Discussion..................................................................................................................... ..........59 4 GENERAL DISCUSSION.....................................................................................................71 Limitations.................................................................................................................... ..........76 Future Directions.............................................................................................................. ......79 APPENDIX A SAMPLE STIMULI...............................................................................................................82 B EXPERIMENT 1 INSTRUCTIONS......................................................................................83 C EXPERIMENT 2 INSTRUCTIONS......................................................................................84 LIST OF REFERENCES............................................................................................................. ..85 BIOGRAPHICAL SKETCH.........................................................................................................89

PAGE 7

7 LIST OF TABLES Table page 2-1 Length and frequency averages for each type of probe word............................................44 2-2 Means and standard deviations (s t. dev.) of demographic measures.................................44 2-3 Means and standard deviations for overall reading times (msec)......................................44 2-4 Means and standard deviations fo r target word reading times (msec)...............................45 2-5 Means and standard deviations fo r post-target reading times (msec)................................45 2-6 Means and standard deviations for modi fier and target word reading times for negative sentences only (msec)..........................................................................................45 2-7 Means and standard deviations of probe word naming times (msec)................................45 2-8 Means and standard deviations for comprehension accuracy (%).....................................46 2-9 Means and standard deviations for comprehension accuracy with both sentence type and probe type (%)......................................................................................46 2-10 Results of regression analysis on negative sentence comprehension accuracy.................46 2-11 Results of regression analysis on non-negative sentence comprehension accuracy....................................................................................................................... ......46 3-1 Means and standard deviations (s t. dev.) of demographic measures.................................65 3-2 Means and standard deviations for overall reading times (msec)......................................65 3-3 Means and standard deviations fo r target word reading times (msec)...............................66 3-4 Means and standard deviations for post-target word reading times (msec).......................66 3-5 Means and standard deviations for modi fier and target word reading times for negative sentences only (msec)..........................................................................................66 3-6 Means and standard deviations of probe word naming times (msec)................................67 3-7 Means and standard deviations for comprehension accuracy (%).....................................67 3-8 Means and standard deviations for comprehension accuracy by probe type (%)..............68 3-9 Results of regression analysis on ne gative sentence presented with symbol comprehension accuracy....................................................................................................68

PAGE 8

8 3-10 Results of regression analysis on nega tive sentence presented without a symbol comprehension accuracy....................................................................................................69 3-11 Results of regression analysis on non -negative sentence presented with symbol comprehension accuracy....................................................................................................69 3-12 Results of regression analysis on non-neg ative sentence presented without symbol comprehension accuracy....................................................................................................69

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Target word reading times as a f unction of age group and sentence type.........................47 2-2 Comprehension accuracy for questions after each probe t ype as a function of sentence type.................................................................................................................. ...47 3-1 Post-target reading times by sent ence type and presence of symbol.................................70

PAGE 10

10 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy NEGATION COMPREHENSION AND AGING: THE ROLE OF INHIBITION AND THE IMPACT OF COMPENSATION DURING READING By Sara J. Margolin August 2007 Chair: Lise Abrams Major: Psychology The present research investigated the infl uence of several factors on young and older adults' reading comprehension of re al-world texts. The first experi ment determined the effects of negation on comprehension of written sentences and examined the level of text representation at which negation most severely impacts compre hension. The second experiment assessed the effectiveness of a specific compensatory st rategy, the use of a negation symbol, in comprehending negation. Using online, activatio n, and content measures of comprehension, participants read sentences modified from ficti on novels. Participants read sentences one word at a time at their own pace, then immediately named a probe word that was verbatim from the sentence, related to the senten ce, or unrelated. Participants then answered a comprehension question about the sentence they had read. Th e results showed that negation affected comprehension in both age groups, such that se ntence reading times were faster for negative sentences, probe naming times were slower fo llowing negative sentences, and comprehension accuracy was lower for negative sentences, re lative to non-negative sentences. Negation equivalently slowed verbatim and related prob es' naming times, suggesting that negation had similar effects on the situation model and surface le vels of representation. However, these effects of negation were not exacerbated with age, al though older adults' comprehension overall was

PAGE 11

11 poorer than young adults. With respect to comp ensation, providing the negation symbol during reading did not facilitate the pr ocessing of negation specifically but instead encouraged readers to focus more attention on all of the sentences. These findings suggest that age and inhibition are less important in influencing negation comprehe nsion than working memory. Furthermore, the lack of age deficits in negation comprehensi on has positive implicati ons for older adults' functioning in a world where negatio n is present in many contexts, such as prescription labels and road signs.

PAGE 12

12 CHAPTER 1 INTRODUCTION The present research explores a topic that co ncerns an activity people engage in every day through their interactions with other people and the world: la nguage comprehension. In tasks such as reading or conversing with friends and co-workers, processing and understanding the language with which we are presented becomes cr itical for functioning. However, there are some significant differences between comprehens ion of spoken versus written language. Comprehension of spoken language is aided by various cues, such as feedback provided from a conversational partner, or prosodic cues, such as intonation. In contrast, reading is dependent on the reader and his/her cognitive capacities. The pro cess of reading is also complicated because it involves various levels of comprehension (e.g., surface level, text base level, situation level) to reach full understanding of the text (e.g., Kint sch, 1998). Given that reading is important for functioning under many real-world circumstances (e.g., menus, road signs, voting, medication labels, and instruction manuals), language co mprehension through reading becomes important for study, and the variables that may impact the ability to perform such a task need to be examined. Research investigating language comprehe nsion through reading has used various experimental tasks to study the processes that take place during reading, the information from the text that is stored in memory, and the kinds of representations that are formed from the information presented in the text. Zwaan and Sing er (2003) reviewed vari ous techniques in the literature, including "online" measures lik e self-paced reading and the moving-window technique, "activation" measures like lexical decision, recognition, and naming, and "content" measures like the think-aloud protocol, question an swering, and sentence completion, to describe their use in examining readi ng through different means.

PAGE 13

13 Online measures of reading comprehension (e.g., Kemtes & Kemper, 1997; Stine, 1990; Stine-Morrow, Loveless, & Soeder berg, 1996) test reading as th e process is occurring. Selfpaced reading presents segments of text (e.g., wor d, phrase, or sentence) on a computer screen at a rate controlled by the reader. The reader can m ove from one segment of text to the next by pressing a key on the keyboard. For the self-paced reading technique only the segments of text presented for reading are present on the screen, while no other information about the length of the upcoming words or length of the passage is given. This presentation is different from the moving-window technique, another type of self -paced reading, where an entire passage is presented on the screen with al l of the letters replaced with da shes, which give clues about the length of the words to be read. The dependent measure of both self-p aced reading and the moving-window technique has traditionally been reading times between key presses. Longer response times provide evidence for increased difficulty in reading that piece of text. In activation measures (e.g., Hamm & Hasher, 1992; Radvansky & Curiel, 1998; Radvansky, Zacks, & Hasher, 2005), the accessibility of the information that is learned from the text is tested. The term "activation" describes how accessible the texts meaning, or occasionally verbatim text, is in memory. If the meaning or verbatim text is readily available (i.e., still activated), then that piece of information is considered an important part of the mental representation of the text. One activation measure often used is the naming measure. In the naming measure, a word, either from the text or se mantically related to a word from the text, is presented after reading. The part icipant is required to say alo ud the word shown. The dependent measure for activation tasks is the time require d to say the presented word. The faster the response times, the more available the word is in the mental representation of the text, indicating its importance to the situation described in the text.

PAGE 14

14 Content techniques (e.g., Radvansky, Copela nd, Berish, & Dijkstra, 2003; Radvansky, Curiel, Zwaan, & Copeland, 2001) measure the kind of representations that are being formed or have been formed about the text (depending on when the test is given), as well as the information that is contained in that representation. A common content measure involves memory and includes free or cued recall. Reca ll tests require participants to either reconstruct the text that they read from memory (i.e., free recall) or to fi ll in a missing part of the text given a portion of the text as a cue (i.e., cued recall). Unlike the other measures, content techniques focus on accuracy, where inaccurate responses are evidence th at the text may not be stored clearly or correctly in memory. In the aging literature, older adults' abilities have been examined through a variety of these methodologies, often combining methodologies to expand the scope of exploration. General findings from online measures suggest that ol der and young adults alloca te their reading time differently (e.g., Connelly, Hasher, & Zacks, 1991; Kemtes & Kemper, 1997; Stine, 1990; StineMorrow et al., 1996; Waters & Caplan, 2001). Olde r adults are more likely to pause at both clause boundaries (i.e., the end of a single idea in a sent ence, often indicated by the presence of a comma) and sentence boundaries (indicated by a period), where young adults will typically pause only at sentence boundaries. The different stra tegies for older adults' allocation of reading time do not suggest that older adu lts were less able to read and understand text, but rather that their approach in doing so was different. In activation techniques (e.g., Hamm, & Hasher, 1992; May, Zacks, Hasher, & Multhaup, 1999; Radvans ky & Curiel, 1998; Radvansky et al., 2003; Radvansky et al., 2001; Radvansky et al. 2005;), older adults have s hown little, if any, deficits in comprehension (for a review, see MacKay & Ab rams, 1996). While older adults may be slower to access information from the text, their accuracy in accessing that information is similar to

PAGE 15

15 young adults. In content measures (e.g., Kemtes & Kemper, 1997; Stine, 1990; Stine-Morrow et al., 1996), older adults show some deficits relati ve to young adults in that they often recall less specifics from the text, although th ey usually can retrieve the gist of the text processed during reading. However, these techniques rely on retrieval from memory, so age differences in content measures may actually result from age-related dec lines in memory, rather than difficulties in comprehension (e.g., Stine-Morrow et al., 1996). Using these techniques, research on language comprehension has examined many variables that influence a readers comp rehension of text (e.g., syntac tic complexity, negation, lexical ambiguity, morphological complexity) and provides evidence of how and when difficulties occur in the reading process. Studies of the above variables give insight into the mechanics involved in the reading process through an exam ination of constructions that na turally occur in language and that may increase difficulty in understanding. One specific focus of past research has been on young adults comprehension of negation in text ; however, older adults comprehension of negation has not been explored. The examina tion of negation as a variable in language comprehension is important for older adults because of the impact it may have on reading prescription medication labels, road signs, and even voting forms, all of which may contain negation. The Impact of Negation on Reading Comprehension The general findings in the nega tion literature are that sentences that include negation take longer to process during reading and that young a dult readers show poorer comprehension when sentences contain negation compared to sentence s that do not contain ne gation, i.e., non-negative or affirmative sentences (e.g., Cornish, 1971, E xperiment 1; Cornish & Wason, 1970; Hoosain, 1973; Just & Carpenter, 1971; Kaup, 2001; Kaup, Dijkstra, & Ludtke, 2004; Kaup, Ludtke, & Zwaan, 2005; Kaup & Zwaan, 2003; Kaup, Zwaan, & Ludtke, in press; MacDonald & Just,

PAGE 16

16 1989; Sherman, 1973). Both activation and content m easures have shown that the presence of negation slows responses times to and decrease s accuracy of comprehending negated concepts (Cornish, 1971; Cornish & Wason, 1970; Just & Carpenter, 1971, Experiment 1; Kaup, 2001; Kaup et al., 2004; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup et al., in press; MacDonald & Just, 1989, Sherman, 1973). For example, MacDonald and Just (1989) exam ined the impact of negation on comprehension of a sentence via two activation measures, probe recognition (where participants indicated whether th ey recognized a word as having been in the sentence they had just read) and probe naming (where participants said the probed word out loud). Results showed that participants res ponses to both recognizing and naming a probe were slower if the probe word had been negated in the sentence they ha d read. These results s uggest that negation may reduce the availability of negated concepts in the mental representation by deactivating the negated concept. Further examination of the ava ilability of negated text has been examined by Kaup (2001). In two experiments, participants completed a probe recognition test, where the probe word was part of a definite/indefinite noun phrase from the sentence (Experiment 1; e.g., "the table" or "a table") or was a creation/destruction verb used in the sentence (Experiment 2; e.g., "bake" or "burn"). In the first experiment, the meaning of the probe word (e.g., "table") was the same, but the specificity of the word before it (e.g., "the" vs. "a") differed between the two examples. In the second experiment, the meaning of the probe word s was different. The results for Experiment 1 showed longer response times and lower accuracy for recognizing probe words that had been negated relative to non-negated probe words, but the type of noun phrase (d efinite vs. indefinite) had no effect on the speed or accuracy of responses The results of Experiment 2 showed slower recognition times for negated probes than non-ne gated probes, with a greater difference for

PAGE 17

17 creation verbs. Together, these findings s uggest that negation has greater effects on comprehension of text at the level of mean ing rather than the specific word level. Similar effects of negation have been found with content measures. For example, Cornish and Wason (1970) examined negation's effect on recall, i.e., whether ne gative sentences would be recalled less than positive sentences. Results showed that affirmative sentences (e.g., "It is bright") were recalled more often than the negati ve sentences (e.g., "It is not bright"), and more errors were made on the negative sentences. These errors often included changes in syntax but not meaning (e.g., recalling "It is not bright" as "It is dull"); the negation word was deleted, changing the sentence structure, and was instead included in the meaning of another word (i.e., "not bright" is recalled as "dull"). The impli cations of these results are that negation not only makes ideas less accessible during reading, but it al so makes them difficult to store and retrieve from memory. Theories of Language Comprehension and Aging Previous research on reading comprehension has been framed in terms of several theories. The general theories of comprehension give an overview of how information from reading can be represented in memory and how comprehens ion may progress independent of age (e.g., Just & Carpenter, 1992; Kintsch, 1998). The agi ng theories then focus on how language comprehension can be influenced by variab les such as working memory, which may fundamentally change across the lifespa n (e.g., Hasher & Zacks, 1988; Light, 1988). One theory of comprehension, a capacity theo ry of language comprehension (e.g., Just & Carpenter, 1992), explains how working me mory resources are critical for language comprehension during reading. The principal elem ent of this theory is that the amount of information that can be held in working memo ry, i.e., capacity, influences the degree of comprehension during reading. However, there is an inherent limit to how much working

PAGE 18

18 memory can hold. Difficulties arise when the cap acity needed to comprehend a sentence exceeds an individual's working memory capacity. Cont ext may relieve the burden on working memory to some degree. When sentences are read in cont ext, i.e., a framework of text involving related ideas, there is some amount of pre-activation of ideas (i.e., prep aration for what ideas are to come next), which lessens the amount of capacity needed to process subsequent ideas. This theory's basis in working memory capacity has impli cations in the aging literature such that older adults are believed to have smaller working memory capacities (e.g., Light, 1988; Waters & Caplan, 2001), and thus may show decrements in processing sentences that tax their working memory limitations. Given that negation seems a pervasive construc t that is likely to cause difficulty in comprehension, it is important to understand how it is processed. Negation may tax working memory, as processing negation may involve more steps than reading non-negated text (e.g., activating and then deactiv ating a particular concept). MacDona ld and Just (1989) proposed that when we read negation, all ideas that are negated become deactivated. For example, after reading the statement "Elizabeth baked some bread but no cookies," "cookies" b ecomes deactivated and therefore is removed from working memory. Additional time and processing resources are needed to deactivate the negated concept, whic h makes processing of negation more difficult. Various types of negation may differentially burden working memory. The type of negation that shows the most de trimental effect on reading comp rehension is the negation that involves changes in syntax (i.e., adding "not" into the sentence; Just & Carpenter, 1971, Experiment 1; Sherman, 1973). For example, Just and Carpenter (1971) asked participants to verify sentences in comparison with a presented picture and examined three types of negation: explicit negatives, implicit syntactic negatives, and semantic negatives. Explicit negatives were

PAGE 19

19 syntactically negativ e because they overtly included the ne gative by use of a negative modifier (e.g., "The dots are not red"). Implicit syntactic negatives were negatives that did not include a negative modifier but described a subset of items that were nega ted (e.g., "Few of the dots are red") and included a word that can be used in di fferent syntactic structures; for example, in the sentence Few of the dots are red, "few" can either indicate that a few dots are red or that some other amount of dots are not red. Semantic negatives also refer to a subset of items but do not contain a syntactic property for marking the nega tion (e.g., "A minority of the dots are red"). The results showed that response times were long er following explicit and implicit syntactic negatives relative to semantic negatives, suggesti ng that syntactic negatives are more difficult for comprehension than other types of negatives and that negation can be viewed as a syntactically complex structure. Another theory of comprehension is the C onstruction Integration (CI) Model (Kintsch, 1998). In this model, a network is formed duri ng reading, containing a "t ext representation" of the newly read information. For the CI Model, there are three levels of repr esentation that a text can take in memory. The first of these repr esentations is the "surface level," where the representation is only th e verbatim words themselves. The s econd level is the "textbase," which represents the words as well as their meanings The highest level of representation is the "situation model," including the meaning of the te xt as well as conclusions and inferences drawn from the text. While all readers use the surface leve l, textbase level, and situation model level for processing and storage of text information, there are some age differences in the representations on which readers tend to rely. There is evidence th at older adults are be tter able to understand and store text at the situation le vel than at the surface and textba se level, while young adults are better able to store text at th e surface and textbase levels (e .g., Radvansky et al., 2001). Given a

PAGE 20

20 limited working memory capacity, older adults ma y continue only store the meaning/situation model instead of the lower levels of repres entation. Young adults ma y tend to store exact wording as well as the situation model representation because they have additional working memory resources, which leaves more room to store additional information which may increase the specificity of the memory representation of the text. This theory involving different levels of re presentation of text (e.g., Kintsch, 1998; Zwaan & Radvansky, 1998) can be applied to the compre hension of negation. In the propositional nodal network, the ideas in the sentence that were activated during read ing are subsequently constrained to deactivate unneeded information. In the case of negation, the negated word is not needed in the situation that is described. For example in the sentence, "Elizabeth baked some bread but no cookies," "cookies" are not neces sary for representing what happened in the situation described in the text. In other words, not all of the concepts that are presented in the text are necessary for the situation model and theref ore do not need to remain as highly activated, allowing a more accurate representation of the situ ation to be created. The CI Model as applied to negation can explain some of the previously-r eviewed studies. For example, the results of Kaup (2001), where negation had an impact on co mprehension only when the meaning of the text was changed, suggest that negation affects the s ituation model represen tation of the text rather than the surface level. The results from Cornish and Wason (1970) also suggest that the meaning of the negation is more important than specific wording, and that readers remember the gist or the situation model represen tation instead of the exact words. A final theory, related to Just and Carpente r's capacity theory of working memory that relates working memory difficulties to aging, is the Inhibition Deficit (ID) Theory (Hasher & Zacks, 1988). This theory proposes that inhibi tion from working memory is the focus of

PAGE 21

21 problems in comprehension in aging. When ina ppropriate information is held in working memory, unnecessary amounts of working memory capacity are taken up and are not available for appropriate information processing. If inappr opriate information gets into working memory, it is the duty of the inhibitory mechanism to deactivate that informati on (i.e., get it out of working memory). However, if the inhibitory mechanism is not worki ng properly, inappropriate information will not get de-activated as quickly, and the resulting capacity remaining to process the appropriate information from the senten ce is diminished. Hasher and Zacks (1988) specifically proposed that older a dults have inefficient inhibitory mechanisms. Therefore, older adults may not be able to efficiently or appr opriately deactivate a ne gated concept that is unimportant to the situation mode l representation of the text. Specific Aims In light of the potential for difficulty in ol der adults comprehension of negation, the aims of the present studies are to: (1 ) determine the effect of negati on on older adults comprehension, (2) assess the function of inhi bition in negation processing, (3 ) demonstrate the impact of negation on multiple levels of text representa tion, and (4) test the effectiveness of a compensatory strategy in facili tating young and older adults comp rehension of negation. All of these aims were tested within th e context of texts that readers may encounter in real life, as opposed to texts that have been created by an ex perimenter, a technique not previously used in negation research. Many studies of negation assume that the simple laboratory experiments extend to real-life situations. For example, Cornish (1971, Experiment 1) had participants compare negated sentences to a multi-colored circ le (e.g., participants verified the sentence The dot is not red in reference to a picture they saw). This task se ems very different from the way that negation is encountered in everyday life, such as through reading a novel, newspaper, or even a prescription label. These experiments ma y actually be testing l ogic in that comparing

PAGE 22

22 sentences to pictures may be more remini scent of if-then reasoning than reading comprehension. As Neisser (1978) suggested, a more ecological approach should be taken toward laboratory research because no practical imp act can be made from research that is not at least partly based in reality. By using logic sentences (e.g., "The dot is not red"), previous experiments have not sufficiently asked or answ ered questions of comprehension, such as: How does negation affect our reading? How does negati on affect our representation of that text? How does negation affect our retention and retr ieval of the information we have read? The present research will demonstrate whet her young and older adults are affected by negation when reading more real-world types of te xts. The next chapter will describe the first in a series of two experiments, which use onlin e, activation, and conten t techniques to more thoroughly assess negation's influe nce on reading comprehension.

PAGE 23

23 CHAPTER 2 EXPERIMENT 1 The purposes of Experiment 1 were to de termine the effects of negation on young and older adults' comprehension of negation and to exam ine the level of text re presentation at which negation most severely impacts comprehensi on. Specific hypotheses are presented below. Hypothesis 1 Older adults will have more difficulty proces sing negation than young adults because of an increased difficulty with inhibition. As proposed by the IDH, older adults will have trouble inhibiting the concept that is modified by the negation, resulting in the concept's inclusion in their memory representation of the text. As an olde r reader takes in the information from the text, he/she will be less likely to disregard the negate d concept and will include the concept in their memory representation of the text as if it were never negated. Representing the negated concept (which should not be represented because it is not relevant in th e situation described by the text) takes up some working memory resources. Accordi ng to Just and Carpenter's capacity theory of working memory, using working memory res ources for processing negation would cause problems storing the rest of the appropriate/rel evant content of the te xt in working memory, especially when that text is difficult to process an d would require a lot of resources. The result is that for older adults, whose working memory resources may already be limited (e.g., Light, 1988), an inhibition deficit will resu lt in incorrectly re presenting the negated concept in memory and prevent appropriate processing of the text as a whole. The al ternative hypothesis is that older adults will not show a difficulty with inhibition as it relates to negation processing, contrary to the IDH. Burke (1997) suggested th at older adults can demonstrate inhibition to the same degree as young adults, such as in completing se ntences (e.g., Burke & Harrold, 1988), metaphor comprehension (e.g., Newsome & Glucksberg, 2002), or picture descriptions (e.g., James, Burke,

PAGE 24

24 Austin, & Hulme, 1998). Therefore, older adults may represent negation appropriately and be able to prevent the negative concept from being included in their mental representation of the text. The potential for increased difficulty in nega tion processing for older adults compared to young adults may be caused by factors other than i nhibition, even if working memory limitations are still relevant. Both Gough (1965) and Mehler (1963) suggested that ne gation, as well as any type of complex sentence, can be processed in te rms of its simpler form (called its kernel) and then transformed back to the original, more comp lex, form later for retrieval or verification. In terms of negation, this idea is that a negative sentence would be processed in terms of its affirmative counterpart and then marked with a footnote that indicates that the sentence was negative. Later, both the sentence and its footnote are retrieved, a nd the sentence is reversed into its negative interpretation once again. Older adults would have difficulty in this instance because they would need to hold both the affirmative form of the sentence and the note about its transformation in their working memory. This additional information would take up working memory resources and cause difficulty, not be cause the reader was unable to inhibit the necessary information to make an appropriate re presentation, but rather because older adults' working memory capacities are limited. Regard less of the mechanism for making negation difficult to process, limited working memory resources for older adults would be counterproductive in processing negation. Hypothesis 2 Older adults will show the most difficulty in representing negation at the situation model level of representation. The situa tion model, i.e., the highest leve l and most meaningful level of text representation, is the level of text where information is stored and used for many tasks, including learning. In relation to aging, there is evidence that ol der adults more often understand

PAGE 25

25 and store text at the situation le vel than at the surface and textbase level, while young adults store text at the surface and textbase levels (e.g., Ra dvansky et al., 2001). Given that older adults represent information from text at the situation m odel level, this level may be the most affected by any difficulty in representing and remember ing the information presented in the text. The research on young adults negation comprehension suggests that negation most adversely affects the situation model represen tation of text (Kaup, 2001) In addition, older adults most easily remember text from the situ ation model because it is the most meaningful representation, and older adul ts limited working memory capacity prevents them from necessarily retaining multiple representations of the text (e.g., Radvansky et al., 2001). Combining these results with the prediction that older adults may have more difficulty correctly understanding negation than young adults, older a dults difficulty in processing negation is expected to be most evident at the situation model level. To measure the availability of that level of representation in a readers memory for that text, activation measures will use words representing the situation model level (i.e., words that show the gist meaning of the text) as well as the surface level (i.e., words verbatim from the text). An alternative hypothesis can be made based on older adults expertis e in using situation model representations and their mo re extensive practice at processi ng at this higher and usually important level (Radvansky et al., 2001). Given that older adults primarily st ore and retrieve text at the situation model le vel, they may be more experienced at using this level of representation. Thus, older adults may be able to offset, i.e., co mpensate for, a difficulty that would occur at the situation model level with their extensive experience. The result would be that older adults would be relatively unimpaired by the negation at the situation model level, compared to processing at the surface and textbase levels. Wh ile this prediction also supports the idea that

PAGE 26

26 older adults tend to process and st ore text at the situation model le vel, it differs in that older adults will be able to use their experience to offs et their difficulties with negation at levels in which they best process information from text. Methods Participants Fifty-four young (40 female and 14 male) and 54 older adults (35 female and 19 male) were tested. The young adults ranged from 18 to 23 ( M = 19.09, SD = 1.22) years of age and were recruited from introductory psychology cla sses at the university. The older adults ranged from 64 to 87 ( M = 74.59, SD = 5.98) years of age. The older adults were community-dwelling older adults, recruited from churches, clubs, and libraries surrounding the uni versity. Participants were fluent speakers of American English and reported that they not previously read the Harry Potter series by J.K. Rowling or seen the corresponding movies. Design The present experiment used a 2 x 2 x 3 f actorial design with ag e group (young and older), sentence type (negative and non-negative), and probe type (verbatim, related, and new) as factors. Sentence type and probe type were within-subjects factors, and age group was a between-subjects factor. The de pendent variables were the reading times for words in the sentence, the naming time for the probe words, and comprehension question accuracy. Reading times measured online processing (i.e., processi ng during reading), naming times measured the activation levels of the probed words, and comp rehension question accuracy indicated how much and how accurately a reader retain ed the content of the text speci fically relating to the negation. Materials One hundred and two experimental sentences and 30 filler sentences were used in the present experiment. Experimental sentences cam e from the Harry Potter series by J.K.

PAGE 27

27 Rowling. As reported by Laurie May of the Alachua County Libr ary System, this series was selected as text that both young and older adults would like to read, based on observations from library employees and rates of check-out (L May, personal communi cation, June 1, 2006). In addition, an article in the Gaines ville Sun newspaper reported H arry Potter as one of the young adult books that other age groups are beginning to enjoy reading ( Grown-ups are turning to teen books, 2006). Sentences were modified to include or exclude negative modifiers, and changes in grammar were made as necessary. Names of characters were also changed to obscure the source of the text and to avoid participants bringing preexisting id eas about what they would be reading. Each sentence (e.g., Trying very hard to not imagine what Katie might look like if they found her, Adrian led the way forward. ) contained a target word, e.g., imagine. Each target was preceded by either a negative modifier (e.g., not imagine) or no modifier (e.g., imagine). A negative modifier was defined as a word that de notes the absence of the word that it modifies. Target words were of varying parts of speech, including verbs in 63.7% of sentences, nouns in 10.8% of sentences, adverbs used in 1% of sent ences, and adjectives in 24.5% of sentences. Modifiers were no, not, and never, with no used in 11.7% of sentences (7.7 % modifying verbs, 84.6% modifying nouns, and 7.7% modifying adjectives), not in 87.3% of sentences (71.6 % modifying verbs, 1.1% modifyi ng adverbs, and 27.3% modifying adjectives), and never in 1.0% of sentences (100% modifying verbs). Each sentence was associated with three possi ble probes for the naming task. The verbatim probes were identical to the targ ets, e.g., imagine. The related probes were semantically related to the target, e.g., dream specifically synonyms of the target word, as defined by a thesaurus, which could replace the target in the text without a change in meaning. The new probes were words that were not present in the sentence, were unrelated to the situa tion described by the text

PAGE 28

28 (e.g., adjust), and could not replace the target word in the sentence without a significant change in meaning. All probe types were relati vely high in frequency (e.g., Francis & Kucera, 1982), and were similar in length (Table 2-1). Th e presentation of each type of probe word was counterbalanced across participan ts, along with negative modifiers and no modifiers, creating 6 versions of stimuli with 17 sentences in each co ndition. Each sentence was also associated with one comprehension question, specifically focusing on the target and its representation in the situation model of the text. Appe ndix A displays sample stimuli, including sentences in both the negative and non-negative conditions, their corr esponding probes (verbatim, related, and new), and comprehension questions. In addition, 30 easy, non-negative filler sentences, with probes of all three types and comprehension questions, were presented to build confidence in less skilled readers who might have had more difficulty with the experimental comp rehension questions and also to hide the pattern of que stions always referring to the negative portion of the sentence. Sentences, probes, and comprehension questions were presented on a computer screen via a Gateway E-series Pentium 4, 1.8 GHz PC-compa tible computer using a program written in Visual Basic 5.0. Naming times were measured using a Microsoft Multimedia Control sound recorder written into Visual Basic 5.0. The sound recorder began recordi ng at the onset of the probe word and stopped recording when particip ant finished saying the probe and pressed the "enter" key. Each sound file was manually ex amined using Audacit y, a sound recording and analysis program that displays the waveform of the sounds recorded, and the time between probe presentation and naming onset fo r each probe word was extracted. Young and older participants were tested on several measures of cognitive ability, including a 25-item multiple-choice vocabulary test forward and backward digit span tests, a Stroop interference task (e.g., MacLeod, 1991; Str oop, 1935) and an Operation Span task (Turner

PAGE 29

29 & Engle, 1989) to provide measures of sema ntic knowledge, inhibition, and working memory capacity. Participants also completed ques tionnaires on other background demographics, including questions about their education, read ing, and writing habits. In addition, older participants completed a test of mental status known as the Mini Mental State Examination (Folstein, Folstein, & McHugh, 1972), a 30-point questionnaire surveying older adults orientation, attention an d calculation, recall, and language pr ocessing abilities, where all older participants scored at least 27 out of 30, M = 28.83, SD = 1.24. For the Stroop interference task, participants sa w a color word written in a particular color on a computer screen. Underneath the word, there were four boxes: a red box, a blue box, a yellow box, and a green box. The task was for partic ipants to click the mouse on the box that matched the color in which the word was written. A motor response was used instead of verbal responding, similar to Keele (1972), in order to fu lly computerize the task and more precisely time the sessions. The first 20 trials were congrue nt, such that the color and the word matched (e.g., the participant saw the word red written in red). The next 20 trials were incongruent, where the color and the word did not match (e.g., the pa rticipant saw the word red written in yellow). Trials were blocked by type to reduce confusion from transitioning from one type of stimulus to another, and congrue nt trials were presented first to simulate the control first method originally used by Stroop (1935). The time to complete each block1 was recorded, and the difference between the incongruent and congrue nt blocks was calculated as a measure of the Stroop effect. A larger Stroop Effect i ndicates more difficulty with inhibition. For the Operation Span task, participants were presented with a math equation to verify (e.g., Is (9/3)-2=3?), followed by a word to remember. These two types of materials were 1 Individual word timing has been investigated in previous studies an d found to yield similar amounts of Stroop interference as the timing of blocks (e.g., Sichel & Chandler, 1969).

PAGE 30

30 presented several times (i.e., a math equation, th en a word, then a math equation, then a word) until participants were presented with a number of words to remember, ranging from two to five, with set size presented in the same random order for all participants (stimuli was the same as was used in Turner & Engle, 1989). Participants we re then prompted to recall out loud the words from that set, i.e., the group of words presented for recall. This process was repeated for 12 sets (three sets of each size from two to five word s), the proportion of words recalled from each set was calculated, and the average of these proporti ons was calculated to make each participants final operation span score. A lthough there are other scoring me thods, this method was chosen because it was a straightforward measure of how many items participants could recall (Conway, Kane, Bunting, Hambrick, Wilhelm, & Engle, 20 05 present a summary of scoring techniques). Procedure Upon beginning the experiment, participants were asked to sign an informed consent form, giving their approval to continue with the study. Older adults also completed the MMSE screening at this time. Then, participants receiv ed instructions on the e xperimental tasks (to be described below) and completed th ree practice trials. Appendix B pr esents the instructions that were seen by the participants. Following the practice trials, the experimenter answered any questions the participant may have had, and da ta collection began for each of 132 trials (102 experimental and 30 filler trials randomized for pr esentation). Each trial consisted of the reading task, the naming task, and one comprehension que stion. The tasks were completed as follows. The first task was the self-paced reading task, where participants read sentences one word at a time by pressing the space ba r with their dominant hand to move from one word to the next. Sentences appeared on the screen with letters replaced by dashes while preserving spaces and punctuation. With each key press, one word wa s revealed while the previous word returned to dashes, and the reading time for each word wa s recorded by the computer. After reading the

PAGE 31

31 final word in each sentence, a 500 msec pause was presented, and participants were presented with one of three possible probe words (verbatim related, and new) for the naming task. The 500 msec delay was determined by p ilot testing young adu lts on various delay presentations (i.e., 250, 500, 750, 1000, and 1250 msec) to determine the l ongest amount of time possible to show the probe while still getting an effect of negation (i.e., a longer naming time for negatively modified probes compared to non-modified probes ). The delay was presented to make a negation effect able to be detected in both age groups, si nce older adults tend to be slower in processing than young adults in many domains (e.g., Salthous e, 1996). Upon presentation, participants said aloud the word shown. The multimedia sound reco rder recorded the sound, and the naming latency, i.e., the time between pr obe presentation and the onset of the participants voice, was extracted. After naming, participants presse d the enter key to immediately show a comprehension question. This question was answered verbally, using one or more words, and the experimenter recorded the answers. After answ ering this question and pressing the enter key, the next sentence appeared on the screen. The process was repeated for all 132 sentences. When the reading, naming, and comprehensi on trials were comp leted, participants completed the background questionnaires and cognitive tests, as described earlier. After these cognitive tests, participants were debriefed and thanked for their particip ation. Older adults were given monetary compensation, and young adults were given course credit. Results Demographics, Operation Span, and Stroop Test Independent samples t-tests on the demogra phic measures (Table 2-2 shows means and standard deviations) indicat ed that older adults had more years of education, t (106) = 9.21, p < .01, had greater vocabulary scores, t (104) = 15.20, p < .01, and reported spending marginally more time watching television, t (105) = 1.94, p < .06, and doing crossword puzzles, t (105) =

PAGE 32

32 1.74, p < .09, than young adults. Young adults had ma rginally larger forward digit spans than older adults, t (105) = 1.92, p < .06, and reported spending more time writing, t (104) = 5.79, p < .01, than older adults. Both young and older adul ts rated themselves similarly on health, p > .59 their spelling training while they were in school, p > .14, and their spelling ability now, p s > .27. In addition, both age groups reported sp ending similar amounts of time reading, p > .35, and had similar backward digit spans, p > .20. For the Operation Span test, young adults reca lled a higher proportion of words than older adults, t (105) = 4.10, p < .01, suggesting that older adults had smaller working memory capacities. For the Stroop task, an Age Group x Trial Type (congruent and incongruent) ANOVA was conducted on mean response time for each block of trials, resulting in a main effect of age group, F (1, 106) = 125.13, MSE = 105.93, p < .01, a main effect of trial type, F (1, 106) = 34.66, MSE = 41.20, p < .01, and an interaction, F (1, 106) = 24.95, MSE = 41.20, p < .01. Follow-up tests on the interaction indicated that young adults had similar re sponse times for congruent and incongruent trials, p > .53, but older adults had longer re sponse times on incongruent trials relative to congruent trials, p < .01. The lack of Stroop interference for young adults may be due to the use of a motor task, which tends to have smaller Stroop interference effects (e.g., Keele, 1972). Reading Times For all participants' reading times, reading times exceeding plus or minus three standard deviations from the mean of each age group were excluded, so as not to include instances where participants skipped over words, stopped during reading for a brea k, or paused to ask a question. For young adults, this technique eliminated 1.9% of the data, and for older adults, this technique eliminated 1.7% of the data. Tests of reading tim es measure the difficulty that older adults may have in comparison to young adults in reading on-line. Longer reading times indicate difficulty

PAGE 33

33 with that portion of the sentence, and provide a specific time frame during reading, where young and older adults may differ. These tests will provi de a test of Hypothesis 1, investigating whether additional time is necessary for processing negation because it is a difficult sentence construction. Overall reading times To examine the speed with which readers in each age group were reading the words in the presented sentences, a 2 (Age Group) x 2 (Sen tence Type) ANOVA was conducted (Table 2-3 shows means and standard deviations) on the mean reading time per word, averaged across all words in a sentence. Results showed a main effect of age group, F (1, 106) = 35.92, MSE = 31385.87, p < .01, where older adults ( M = 603.02) spent more time reading each word than young adults ( M = 458.54). In addition, this analysis reve aled a main effect of sentence type, F (1, 106) = 5.33, MSE = 115.32, p < .02, where participants spent less time reading words in negative sentences ( M = 529.09) than non-negative sentences ( M = 532.46). The interaction between age group and sentence type was not significant, p > .22. Target word reading times To distinguish specific locations within each se ntence that participants may have increased or decreased their reading speed, analyses of the different types of words in each sentence were conducted. First, a 2 (Age Group) x 2 (Sentenc e Type) ANOVA was conducted on the time spent reading the target word, i.e., the word that was modified by the negation in negative sentences or was not modified in the non-negative senten ces (Table 2-4 shows means and standard deviations). A main effect of age group was revealed, F (1, 106) = 30.79, MSE = 44719.88, p < .01, such that older adults ( M = 609.53) spent more time readi ng target words than young adults ( M = 449.85). While the main effect of sentence type was not significant, p > .13, a marginal Age Group x Sentence Type intera ction (Figure 2-1) was shown, F (1, 106) = 3.02, MSE =

PAGE 34

34 839.01, p < .09. Further analyses on this interacti on revealed that young adults spent less time reading the target word when it was in a negativ e sentence than when it was in a non-negative sentence, p < .02, whereas older adults ha d equivalent target reading times regardless of sentence type, p > .88. For both negative and nonnegative sentences, older adults read the target word slower than did young adults, p s < .01, but the age group difference was slightly larger for negative sentences. Post-target word reading times A 2 (Age Group) x 2 (Sentence Type) ANOV A was conducted on post-target reading times, the word immediately following the targ et (Table 2-5 shows means and standard deviations). Results showed only a main effect of age group, F (1, 106) = 41.52, MSE = 28325.21, p < .01, where older adults ( M = 604.18) read after-targ et words slower than young adults ( M = 456.60). The main effect of sentence type and the interaction were not significant, p s > .68. Modifier versus target reading ti mes for negative sentences only To compare the reading times of modifier word s with the reading times of the words they modify (i.e., the target words), a 2 (Age Gr oup) x 2 (Word Type) ANOVA was performed on the reading times of the modifiers and the target wo rds in negative sentences only (Table 2-6 shows means and standard deviations). Only negative sentences were evaluated because there were no modifier words in non-negative sentences. The an alysis revealed a main effect of age group, F (1, 106) = 31.87, MSE = 40046.29, p < .01, and a main effect of word type, F (1, 106) = 96.19, MSE = 1715.21, p < .01. These main effects were moderated by an Age Group x Word Type interaction, F (1, 106) = 5.16, MSE = 1715.21, p < .03, such that reading times for targets were slower than for modifiers for both age groups, bu t this difference was la rger for older adults, p s < .01.

PAGE 35

35 Probe Naming Times Any inaccurate responses, including those wh ere a participant laughed, stuttered, coughed, cleared their throat, or pressed the "enter" ke y before they said the probe word, were not included, eliminating 17.3% of potential response s from young adults a nd 16.0% of potential responses from older adults. Similar to reading times, outliers in participants' naming times (i.e., the latency between the onset of the probe word and the time the participant began saying the word) were accounted for by excluding naming times plus or minus three standard deviations from the mean of each age group. This method excluded 1.8% of the data for each age group. Probe naming times test the activation level of pr obe words represented in different levels of representation of the text, which provide a test of Hypothesis 2. A 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) ANOVA was conducted on mean naming times to determine the availability of the probe words in memory after having just read a negative or non-negative senten ce (Table 2-7 shows means and standard deviations). This analysis revealed a main effect of age group, F (1, 106) = 4.31, MSE = 97162.48, p < .04, where older adults had longer naming times ( M = 789.10) overall than did young adults ( M = 738.26). In addition, a main effect of probe type, F (1, 106) = 36.02, MSE = 2603.96, p < .01, showed that both verbatim ( M = 740.39) and related probes ( M = 770.07) had faster naming times than new probes ( M = 780.57), p s < .01. Verbatim probes were also named more quickly than related probes, p < .04. Finally, there was a marginally si gnificant main effect of sentence type, F (1, 106) = 2.72, MSE = 2339.81, p < .10, where probes followi ng negative sentences ( M = 766.81) had slower naming times than probes presented after non-negative sentences ( M = 760.54). No interactions among any of the variables were significant, p s > .19.

PAGE 36

36 Comprehension Accuracy Comprehension accuracy was coded to indicate whether an answer was either correct or incorrect. Correct responses included answers th at used either the verbatim words from the sentence or synonyms of the words in sentence, ju st as long as the answer captured the meaning or the gist of what happened in the sentence. Answers that were the opposite or different from what happened, off the topic from what happened, or indicated th at the participant did not know the answer, were marked as incorrect. On ra re occasions when there were uncertainties, additional coders were consulted, and a 2/3 major ity ruled the answer as correct or incorrect. Tests of comprehension accuracy provide an analys is of the content meas ure, testing Hypothesis 1 and examining the storage of the in formation read by young and older adults. To examine participants comprehension, a 2 (Age Group) x 2 (Sentence Type) ANOVA was conducted on mean accuracy in answering th e comprehension questions, which focused on the portion of the sentence containing the ta rget (Table 2-8 shows means and standard deviations). This analysis reve aled a main effect of age group, F (1, 106) = 7.31, MSE = .023, p < .01, such that older adults had poorer accuracy overall ( M = 68.0%) on the comprehension questions than did the young adults ( M = 73.5%). In addition, a main effect of sentence type, F (1, 106) = 19.29, MSE = .007, p < .01, showed that participants' accuracy was poorer for negative sentences ( M = 68.3%) than for nonnegative sentences ( M = 73.3%). The interaction of the two variables was not significant, p > .53. To determine the extent of the impact of working memory capacity and inhibition on comprehension accuracy in contributing to th e main effect of age group shown above, two ANCOVA analyses were conducted with age group and sentence type as variables, as in the above ANOVA. The first of these analyses includ ed Operation Span score as a covariate. No significant differences based on age group ( Molder adjusted = 69.2%, Myoung adjusted = 72.1%), F (1,

PAGE 37

37 104) = 1.90, MSE = .02, p > .17, nor an interaction of age group and sentence type, F < 1, were revealed, showing that the age difference in co mprehension accuracy disappeared when working memory between the age groups was equated. The second ANCOVA analys is included Stroop Effect score (the measure of inhibition) as a covariate. Results for this analysis showed a marginal main effect of age group, F (1, 105) = 3.74, MSE = .02, p < .06, such that older adults still had lower comprehension accuracy ( Madjusted = 68.6%) than young adults ( Madjusted = 73.0%). Unlike the ANCOVA with working memory as a cova riate, this analysis showed that age group differences remained in comprehension when th e measure of inhibition was equated across age groups. The interaction of age group and sentence type was not significant, p s > .92. To determine whether the presentation of a probe word for naming (between reading and answering the comprehension question) could influence comprehension accuracy, a 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) ANOVA was conducted on me an percent accuracy in answering the comprehension qu estions (Table 2-9 shows means and standard deviations). As probe type is the only new variable in this an alysis, only effects invo lving probe type were presented. A main effect of probe type was revealed, F (2, 212) = 16.72, MSE = .015, p < .01, such that accuracy was lower for comprehens ion questions that came after new probes ( M = 66.8%) relative to verbatim probes ( M = 73.5%) and related probes ( M = 71.5%), p s < .01. Comprehension accuracy for questions that cam e after verbatim probes was marginally higher than for questions that came after related probes, p < .093. Finally, a Sentence Type x Probe Type interaction (Figure 2-2) was shown, F (2, 212) = 4.52, MSE = .015, p < .01. Further analyses on this interaction demonstrated that comprehension accuracy was poorer for negative sentences than for non-negative sentences when the questions came after verbatim probes, p < .01, or after related probes, p < .01, but there was no difference in comprehension of negative

PAGE 38

38 and non-negative sentences following new probes, p > .34. Within sentence type, there was an effect of probe type for non-negative sentences, p < .01, such that accuracy after saying verbatim probes and related probes was similar, p > .17, but accuracy for these questions was significantly greater than accuracy for questi ons presented after new probes, p s < .01. For negative sentences, comprehension accuracy was similar following the three different probe types, p > .15. A series of regression analys es were conducted to further examine the influence of working memory on comprehension accuracy, and to determine if any other variables may also have an impact on comprehension. Stepwise regr essions were conducted using either operation span, Stroop effect, or age as th e first variable entere d because of the theore tical implications of each of these variables on comprehension. The remaining variables entered were vocabulary score, years of education, and time spent read ing each day. The model that predicted the most variance for negative and non-negati ve sentences entered operation sp an first; for both types of sentences, only operation span score accounted for a significant amount of variance (13.1% for negative sentences and 16.5% for non-negative se ntences). After accounting for this variable, no other variables entered into th e regression model. Results fr om the regression for negative sentences are presented in Table 2-9, and results from the regression for non-negative sentences are presented in Table 2-10. The Relationship between Reading Times, Na ming Times, and Comprehension Accuracy Pearson correlations were conducted to dete rmine the relationships between modifier reading times, target reading times, probe naming times, and comprehension accuracy, separately for non-negative and negative sentences. For yo ung adults, a significant positive correlation was demonstrated for target word reading times and probe naming times for both negative sentences, r (54) = .27, p < .05, and non-negative sentences, r (54) = .36, p < .01. Young adults also demonstrated a marginally significant correlati on for modifier reading times and probe naming

PAGE 39

39 times, r (54) = .26, p < .06, in negative sentences. However, no significant correlations occurred with young adults comprehension accuracy, p s > .70. In contrast to young adults, older adults showed no significant correlations between any of the measures, p s > .17. Discussion The results of this experiment showed th at negation detrimentally impacted reading comprehension via two of the three measures used (probe naming times, comprehension accuracy), consistent with previous work in negation comprehension (e.g., Cornish, 1971, Experiment 1; Cornish & Wason, 1970; Hoos ain, 1973; Just & Carpenter, 1971; Kaup, 2001; Kaup et al., 2004; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup et al., in press; MacDonald & Just, 1989; Sherman, 1973). The pr esence of negation within sent ences decreased comprehension accuracy and marginally increased naming times of a probe word, independent of whether that word had actually appeared in the preceding sentence. However, both young and older adults were adversely affected by negation to the same degree, and negation did not differentially affect one level of representation over a nother, for either age group. The first hypothesis, which pr edicted that older adults would have more difficulty processing negation than young adults, was not su pported. There were several indicators that older adults did not have an increased diffi culty with inhibition when processing negated sentences, as proposed by the IDH (e.g., Hasher & Zacks, 1988). First, th e covariance analyses showed that controlling for inhibition (using th e Stroop interference task ) did not eliminate age differences in comprehension accuracy. Second, the regression analyses showed that inhibition did not significantly explain any of the variance in comprehendi ng negative sentences. Third, although probe naming times were marginally sl owed by negation, the ne gation effect did not interact with age group. Finally, overall reading times showed no interaction between age group and sentence type, and in fact, negation speeded rather than slowed overall reading times for

PAGE 40

40 both groups relative to non-negative sentences. The lack of increased inhibition problems during reading for older adults occurred despite older adults having more difficu lty with inhibition in general, as measured by the Stroop task, in supp ort of the alternative to the first hypothesis. Therefore, it can be concluded that the older adults did not demonstr ate a difficulty with inhibition in reading negation in this experiment and may not have a difficulty as it relates to negation processing, contrary to the IDH. Although negation did not differentially affect older adults in the majority of tasks administered here, several measures did reve al a differential impact on young and older adults' reading times. With respect to ta rget reading times, young adults in creased their reading speed of the target word when it was modified by a negati ve modifier relative to no modifier, whereas older adults' target reading times were equivale nt in both types of sentences. Young adults may have been sensitive to the fact that negation is more difficult to process and therefore slowed their reading times after encountering a negation word. Young adults' target reading times also correlated with their probe naming times for both negative and non-negative sentences, whereas older adults did not show this relationship. Thes e results suggest that target words that were easier to read were also easier to represent in the memory re presentation, allowing young adults faster access to the representation that they had stored. In contrast, olde r adults' reading times were less indicative of the ease with which they accessed the repres entation they created, possibly because they did not adapt their readi ng times when more difficult circumstances (i.e., negation) arose. Another age difference occurr ed when focusing on negative sentences only, where young adults and older adults read the modifier word more quickly than the target word, but the difference between the two words' reading times was greater for older adults. Similar to the way a person may read the words the, it, or and, a reader may also skim over the word

PAGE 41

41 not based on its length, as thes e shorter words tend to be less meaningful than other words in the sentence. Older adults may be especially like ly to read these words quickly because of their limited working memory capacity; they are aware of their limits and instead focus on words that they think are more meaningful and can help facilitate their comprehension, such as the target. There was also evidence of deficits in olde r adults reading compre hension independent of negation, consistent with other research on aging (e.g., Connelly et al., 1991; Kemtes & Kemper, 1997; Stine, 1990; Stine-Morrow et al., 1996). Olde r adults spent more time than young adults reading both negative an d non-negative sentences, suggesting th at they require more processing time than do young adults (e.g., Salthouse, 1996). Th e results on comprehension accuracy also showed declines for older adults, who had poor er overall accuracy than young adults, despite spending more time reading the sentences. In fact, neither young nor older adults showed a correlation between reading times and compre hension accuracy. Content measures require retrieving information from memory after readin g the sentences has been completed, which may put older adults at a disadvant age because of their smaller working memory capacities (e.g., Light, 1988; Waters & Caplan, 2001). Consistent w ith this view, operation span as a covariate removed the effect of age on comprehension a ccuracy. Furthermore, the regression analyses demonstrated that working memory accounted fo r the most variance in comprehension accuracy. These findings suggest that work ing memory, rather than age pe r se, was more critical for comprehension accuracy in the present experime nt, although one would have expected it to be especially relevant in compre hending negative sentences, whic h did not occur. While working memory is a primary determinant of comprehens ion accuracy (e.g., Just & Carpenter, 1992), the regression analyses indicate that there is consid erable variance unaccounte d for, suggesting that

PAGE 42

42 many other factors may be useful in understand ing negation, such as attention, experience with negation, and the appropriateness of the negation used in the text. The second hypothesis predicted that older ad ults would show the most difficulty in representing negation at the situ ation model level of representa tion. However, given that older adults did not show more difficulty with nega tion than did young adults, the question becomes whether one level of representation (e.g., Kintsc h, 1998) was affected more than another by negation for either age group. This hypothesis wa s not supported, as naming times for related probes (the ones representative of the situation mode l) were faster than new probes for both negative and non-negative sentences If negation had increased the difficulty in representing the situation model, naming times would have been slowest for probes representing this level of processing. Therefore, negation ha d equivalent effects at surface and situation model levels of representation, for both age groups, which also c ontradicts the alternativ e hypothesis that older adults would be unaffected by negation at the s ituation model level. However, independent of negation, the probe naming time analysis revealed th at readers did have di fferential access to the levels of representation, evidenced by longer na ming times for new and related probes, compared to verbatim probes. This result suggests that fo r all sentences, both the gist and exact wording information were retained in memory represen tations, but the exact wo rding remained more highly activated. Two possibilitie s could explain why exact word ing was more highly activated than the gist. The first is that the reader did not form the s ituation model representation (e.g., Kintsch, 1998) because there was not enough time to do so between reading and probe word naming. This possibility seems unlikely given that reading occurs at a fairly fast rate, and a reader is used to quickly cons olidating information between cons ecutive sentences of a text. The second possibility is that the constraining of activation that occurs after reading may have

PAGE 43

43 included only the exact wording from the text b ecause these words were the most effective in communicating the ideas in the sentence. Given the richness and complexity of the text used in the present experiment's sentences, the exact wordi ng may have been preferab le in order to retain the meaning of the message, as opposed to simple r sentences used in prev ious research where a gist representation involving a character (e.g., Elizabeth) and the items he/she created (e.g., bread) would be sufficient. Unexpectedly, the presence of the probe word s for naming also influenced comprehension accuracy. Readers accuracy was greater followi ng verbatim and related probes than new probes for non-negative sentences, whereas comprehens ion accuracy was equivalent for negative sentences independent of the probe. For non-nega tive sentences, the probe word was more easily incorporated into the situation model of the text as additional information when it was relevant to the text (i.e., verbatim or related) than when it was not (e.g., new). However, correctly representing negative sentences meant excluding it from the situ ation model; ther efore, none of the probes would be relevant to the text and should not be incorporated in to the situation model. If the probe word was incorporated into the s ituation model of a nega tive sentence, then the representation of the sentence w ould now be incorrect. Thus, readers were sensitive to the presence of the probe word, potentially because of the time at which it was presented (i.e., between the sentence and the comp rehension question) and let it influence the representation of the text that they created.

PAGE 44

44 Table 2-1. Length and frequency averages for each type of probe word. Probe type Average frequency Average length Verbatim 204.39 6.53 Related 154.43 6.90 New 131.51 6.56 Table 2-2. Means and standard deviations (st. dev.) of demographic measures. Age group Young Older Mean St. dev Mean St. dev. Years of education* 13.07 1.27 16.78 2.67 Health rating (out of 10) 7.74 1.51 7.90 1.50 Vocabulary (out of 25)* 13.3 3.22 21.32 2.09 Forward digit span*** 7.56 1.41 7.06 1.28 Backward digit span 4.98 1.51 5.32 1.19 Hours spent writing* 2.57 1.53 1.20 .79 Hours spent reading 3.11 1.89 2.81 1.38 Hours watching TV*** 1.96 1.47 2.58 1.84 Hours doing crosswords*** .23 .50 .49 .97 Spelling training (out of 10) 6.42 1.72 7.08 2.68 Spelling ability now 6.54 1.36 6.15 2.17 (out of 10) Operation span* .90 .08 .82 .12 Congruent Stroop* 22.33 5.26 33.63 8.29 Incongruent Stroop* 23.11 6.10 43.14 12.68 Note: indicates that the age differences were significant at the p < .01 level; ** indicates that the age differences were significant at the p < .05 level; *** indicates that the age differences were marginally signific ant at the p < .10 level. Table 2-3. Means and standard deviations for overall reading times (msec) Age group Sentence type Mean reading time Std. deviation Young Negative 457.75 106.53 Non-negative 459.31 103.16 Older Negative 600.42 143.82 Non-negative 605.61 142.58

PAGE 45

45 Table 2-4. Means and standard deviations fo r target word reading times (msec) Age group Sentence type Mean reading time Std. deviation Young Negative 443.42 120.95 Non-negative 456.27 114.32 Older Negative 609.96 187.24 Non-negative 609.10 168.42 Table 2-5. Means and standard deviations for post-target reading times (msec) Age group Sentence type Mean reading time Std. deviation Young Negative 456.64 100.17 Non-negative 456.57 91.21 Older Negative 602.23 139.56 Non-negative 606.14 142.63 Table 2-6. Means and standard deviations for modifier and target wo rd reading times for negative sentences only (msec) Age group Word type Mean Std. deviation Young Modifier word 400.95 107.78 Target word 443.42 120.95 Older Modifier word 541.88 149.07 Target word 609.96 187.24 Table 2-7. Means and standard deviations of probe word naming times (msec) Age group Sentence type Probe t ype Mean Std. deviation Young Negative Verbatim 714.30 105.65 Related 746.09 111.98 New 761.83 106.54 Non-negative Verbatim 709.24 110.77 Related 739.40 117.44 New 758.69 120.55 Older Negative Verbatim 775.47 161.19 Related 798.65 159.83 New 804.52 152.50 Non-negative Verbatim 762.56 146.63 Related 796.13 156.45 New 797.25 144.80

PAGE 46

46 Table 2-8. Means and standard deviations for comprehension accuracy (%) Age Group Sentence type Mean Std. deviation Young Negative 70.68 12.20 Non-negative 76.40 9.52 Older Negative 65.87 15.24 Non-negative 70.14 10.95 Table 2-9. Means and standard deviations for comprehension accuracy with both sentence type and probe type (%) Age group Sentence type Probe t ype Mean Std. deviation Young Negative Verbatim 72.02 18.11 Related 69.24 17.71 New 68.32 13.51 Non-negative Verbatim 81.58 13.64 Related 77.11 13.52 New 70.95 13.33 Older Negative Verbatim 66.95 17.45 Related 66.39 18.60 New 63.80 19.17 Non-negative Verbatim 73.56 17.13 Related 73.44 12.68 New 64.28 14.44 Table 2-10. Results of regression analysis on negative sentence comprehension accuracy Variable p Operation span .34 < .01 Stroop effect .04 .68 Age .01 .93 Vocabulary .02 .81 Years of education .14 .13 Hours spent reading .06 .50 Table 2-11. Results of regression analysis on non-negative sentence comprehension accuracy Variable p Operation span .42 < .01 Stroop effect .13 .18 Age .16 .11 Vocabulary .14 .14 Years of education .11 .23 Hours spent reading .02 .80

PAGE 47

47 Mar g inal A g e x Sentence T y pe Interaction for Tar g et Reading Times 350 400 450 500 550 600 650 Negative SentencesNon-Negative Sentences Sentence TypeReading Times (msec) Young Adults Older Adults Figure 2-1. Target word reading times as a function of age group and sentence type. Sentence T y pe x Probe T y pe Interaction for Comprehension Accuracy50 55 60 65 70 75 80 85VerbatimRelatedNew Probe TypeAccuracy (%) Negative Non-negative Figure 2-2. Comprehension accuracy for questions after each probe type as a function of sentence type.

PAGE 48

48 CHAPTER 3 EXPERIMENT 2 Experiment 1 demonstrated that negation detrimentally impacted reading at multiple points in the reading process, for both young and older adults, similar to other complex structures such as syntactic complexity (e.g., Kemper, 1987; No rman et al., 1992; Norman, Kemper, Kynette, Cheung, & Anagnopoulos, 1991; Stine-Morrow, Ry an, & Leonard, 2000; Waters & Caplan, 1996). However, there is evidence that compensatory strategies, such as slowing reading times at specific points in the reading process, can incr ease recall by encouraging wrap-up and integration of the previously-read clause or portion of text (e.g., Stine, 19 90; Noh, Shake, Joncich, Hindin, & Stine-Morrow, 2006; Stine-Morro w et al., 2000). For example, Noh et al. (2006) instructed older adult readers to sl ow down when they reached a comma in the text (indicat ing the end of a clause), and this change in th eir reading strategy led to improve d recall of the text. Compensation can also occur spontaneously without instruction; Stine (1990) measured reading times and recall and found that older adult readers, who slowed th eir reading at clause bou ndaries instead of only sentence boundaries, had perfect recall. These studi es suggest that compensatory strategies help to reduce the burden on working memory during reading; therefore, it seems likely that comprehension of negation may also benefit fr om a compensatory strategy. A compensatory strategy may make available some additional working memory resources for processing negation, resulting in appropriate activation and de -activation levels of negated and non-negated concepts from the text. One problem with the above compensatory stra tegies is that the reader must actively commit to using the strategy, which may or may not occur in real-life re ading situations. The present experiment explored the possibility of compensation during reading that is offered within the texts themselves. For example, a compensatory strategy may be seen in the prescription

PAGE 49

49 medication industry. Sometimes appearing on prescr iption labels are instru ctions that include negation (e.g., Do not take this medication w ith aspirin) accompanied by a symbol (e.g., a circle with a line through it) These symbols may serve the purpose of helping readers, particularly people with increased difficulties in comprehension, to understand the instructions that pertain to the medication they are taking. Howe ver, if the symbol is difficult to interpret or inapplicable to the instructions they are give n, more working memory resources may be used, resulting in an increased difficu lty in comprehension. Research i nvolving the use of pictures in the interpretation of medication in structions has demonstrated that people preferred instructions given with pictures to demons trate the instruction compared to text-only instructions (e.g., Sojourner & Wogalter, 1997). However, Dowse a nd Ehlers (2002) showed that pictures should accompany, but not replace, textual instruction, as misinterpretation of pict orial descriptions can occur. The purpose of Experiment 2 was to assess th e effectiveness of a specific compensatory strategy, the use of a negation symbol, in co mprehension of negation. Both age groups had greater difficulty understanding ne gation relative to non-negated sentences in Experiment 1, suggesting that there is a need to facilitate the pro cessing of negation. Hypothesis 1 Symbolic compensation will release working memory resources and assist in comprehension of negation. Given that working me mory capacity is particularly vulnerable to the changes that accompany aging, compensation is expected to have a greater impact on how older adults process negation. Just and Carp enter (1992) more specifically defined the mechanism underlying compensation as activation of relevant ideas. They proposed that when sentences are read within a br oader context (e.g., a whole paragraph focused on a particular topic), that context can serve as a framework of ideas surroundi ng the information presented in

PAGE 50

50 the sentence, creating preactivation of the ideas to come next This pre-activation prepares the reader for the concepts that will be presente d in the sentence, leaving more working memory resources available for processing and facilitating comprehension. In the present experiment, a symbol is used to provide a context or framework, prior to reading each sentence. If the symbol provides a context for negation, th en the pre-activation provided by the symbol could release some worki ng memory resources to assist young and older adults processing of the negate d text. Therefore, compensation may assist in the process of reading and the representation crea ted by the text. With respect to probe naming times, a symbol is expected to facilitate negation processing such that naming times for negatively-modified probe words will not differ from naming times of probes not modified, relative to sentences without a symbol. For comprehension accuracy, the pr esence of a symbol is expected to improve comprehension for negative sentences, resul ting in no difference be tween negative and nonnegative sentences, contrary to sent ences presented w ithout a symbol. Hypothesis 2 Benefits from the presence of a symbol are e xpected be more dramatic at the situation model level. Although Experiment 1 did not de monstrate differences among the levels of representation at which negation impacted reading comprehension in probe naming, the symbol in Experiment 2 provides additional informati on that can function as pre-existing knowledge about the text, creating the possibi lity for compensation to impact comprehension at higher levels (the situation model level) th an others. According to rese arch on older adults reading comprehension, text representations at the situ ation model level tend to be the most easily remembered for readers of this age group (e.g., Radvansky et al., 2001). Given that older adults process information at this high level, whic h also includes information inferred from or previously learned about the topi c in the text (e.g., Kintsch, 1 998), older adults may be more

PAGE 51

51 experienced at incorporating outsi de information into the informa tion that they are reading, thus creating a better situation model representation when a symbol is presented than when it is not. Specifically, the predicted lack of an effect of negation on probe naming times described in Hypothesis 1 is expected for verb atim as well as related probes. Methods Participants Sixty young (45 female and 15 male) and 60 olde r adults (33 female and 27 male) were tested. Young and older adult partic ipants were recruited from sim ilar sources as in Experiment 1, although none had previously participated in Experiment 1. Young adults ranged in age from 18 to 24 years ( M = 19.32, SD = 1.33), while older adults ranged in age from 63 to 88 years ( M = 75.46, SD = 5.81). Participants of both age groups were fluent speakers of American English and had not read the books nor watched the movies from the Harry Potter series. Design The present experiment used a 2 x 2 x 3 x 2 factorial design with age group (young and older), sentence type (negative and non-negative), probe type (verbatim, related, and new), and symbol (symbol and no symbol) as factors. Sent ence type, probe type, and symbol were withinsubjects factors, whereas age gr oup was a between-subjects factor As in Experiment 1, reading time served as an online measure of pro cessing, naming time measured activation, and comprehension accuracy measured rete ntion of the content of the text. Materials The same 132 sentences (both negative and nonnegative, filler and experimental), probe words (verbatim, related, and new), and comprehe nsion questions from Experiment 1 were used in Experiment 2, with one modification. Half of the experimental sentences were presented with a symbol, which differed as a function of modifi er. For sentences with a negative modifier, the

PAGE 52

52 symbol was a circle with a lin e through it, denoting absence. For sentences without a negative modifier, the symbol was an open circle. Unlike negative sentences, the symbol for non-negative sentences was neutral and was used so that symbol s did not only appear for negative sentences. Both symbols were approximately 1.5 inches in diameter and were white with a thin, black outline. As in Experiment 1, the Operation Sp an test, Stroop test, and other demographic measures obtained were also administered. Procedure The procedure for Experiment 2 was identical to Experiment 1, except the reading task was modified to allow the symbol to be presented. Pa rticipants were instructed that they may see some symbols presented before the text, that the symbol may or may not seem important to them, but to make sure they looked at the symbol before beginning to read the sentence (Appendix C shows exact instructions). In the symbol condition, the symbol appeared directly to the left of the sentence, with dashes that represented the words of the sentence to the right. After looking at the symbol, participants pressed the s pace bar to read the first word of the sentence. Participants continued to read each senten ce at a self-paced rate by pressi ng the space bar. The symbol remained on the screen for the entire duration of reading the sentence a nd disappeared with the dashes and the last word when the participant finished reading the sentence. As in Experiment 1, after reading each senten ce, participants were presented with a probe word for naming after a delay of 500 msec. The probe word was verbatim from the passage, related to the passage, or not presented at all. After naming the probe wor d, participants pressed the enter key, and a comprehension question wa s presented. Participants answered each question aloud and pressed the enter ke y to move on to the next sentence.

PAGE 53

53 Results Demographics, Operation Span, and Stroop Test One older participant's data was excluded because of improper data recording. Independent samples t-tests on the demographics measures (T able 3-1 shows means and standard deviations) revealed that older adults reported more years of education, t (116) = 7.13, p < .01, had higher vocabulary test scores, t (117) = 13.48, p < .01, spent more time doing crossword puzzles, t (117) = 2.14, p < .04, and had more rigorous spelling training, t (117) = 2.82, p < .01, than young adults. Young adults reported spending more time writing each day th an did older adults, t (117) = 4.27, p < .01. Young and older adults did not differ on their reports of time spent reading, p > .36, time spent watching television, p > .61, ratings of their spelling ability, p > .97, or ratings of health, p > .41. Young and older adults also did not differ on either forward digit span, p > .47, or backward digit span, p > .23. For the Operation Span test, young adults r ecalled a higher proportion of words than did older adults, t (117) = 2.85, p < .01, demonstrating young adults larger working memory capacities. For the Stroop task, an Age Group x Trial Type (congruent or incongruent) ANOVA was conducted on the mean response time to complete each block of trials. A main effect of age group, F (1, 117) = 182.14, MSE = 132.27, p < .01, and a main effect of trial type, F (1, 117) = 14.22, MSE = 53.54, p < .01, were revealed. These main effects were moderated by an Age Group x Trial Type interaction was revealed, F (1, 117) = 27.15, MSE = 53.54, p < .01, where older adults showed a Stroop eff ect with slower times for inc ongruent than congruent trials, p < .01, but young adults had equivalent response times for the two blocks, p > .31. Reading Times For all participants' reading times, reading times exceeding plus or minus three standard deviations from the mean of each age group were excluded, so as not to include instances where

PAGE 54

54 participants skipped over words, stopped during reading for a brea k, or paused to ask a question. This method removed 1.5% of young adults readi ng time data and 1.5% of older adults reading time data. Reading time analyses were conducted to test part of Hypothesis 1, where the impact of a symbol could be more important in how ol der adults process text during reading, compared to young adults. Overall reading times To examine the average speed of reading th e sentences, a 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol) ANOVA was conducted on the reading times averaged over all types of words in the sentences (Table 3-2 shows means and standard deviations). Th is analysis revealed a main effect of age group, F (1, 117) = 31.23, MSE = 88807.42, p < .01, where older adults ( M = 589.70) read more slowly overall than young adults ( M = 437.04). In addition, a marginal main effect of sentence type, F (1, 117) = 3.40, MSE = 357.26, p < .08, was revealed such that readers read negative sentences ( M = 511.86) more quickly than non-negative sentences ( M = 514.88). No other main effects or in teractions were significant, p s > .19. Target word reading times To further break down where readers spent more time and less time during reading, reading times of different types of words were analyzed. A 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol) ANOVA was conducted on target word reading times (Table 3-3 shows means and standard deviations). Results from this an alysis showed a main effect of age group, F (1, 117) = 31.77, MSE = 124218.38, p < .01. In addition, a marginal inte raction of sentence type and age group was revealed, F (1, 117) = 2.73, MSE = 2261.77, p < .10, such that older adults read target words slower than young adults for both types of sentences, but the age difference was larger for negative sentences, p s < .01. The interaction did not emerge when examined within each age group, as targets in negative sentences were r ead equivalently to ta rgets in non-negative

PAGE 55

55 sentences for both young adults, p > .28, and older adults, p > .21. No other main effects, p > .60, or interactions, p > .46, were revealed. Post-target word reading times Next, a 2 (Age Group) x 2 (Sentence Type ) x 2 (Symbol) ANOVA was conducted on posttarget reading times (Table 3-4 shows means a nd standard deviations), the word that came immediately after the target word. A ma in effect of age group was revealed, F (1, 117) = 23.77, MSE = 129994.86, p < .01, such that older adults read post-target words slower than young adults. In addition, sentence type and symbol showed a significant inte raction (Figure 3-1), F (1, 117) = 3.80, MSE = 4837.91, p < .05, where readers took longer to read post-target words in negative sentences in the presence of a symbol than without a symbol, p < .04, but when reading non-negative sentences, post-target reading times were equivalent whether a symbol was present or not, p > .74. No other effects were significant, p s > .19. Modifier versus target reading ti mes for negative sentences only To examine how much processing time readers us ed for modifier words compared to target words, a 2 (Age Group) x 2 (Word Type) x 2 (Symbol) ANOVA was conducted on negative sentences only (Table 3-5 shows means and standard deviations). This analysis revealed a main effect of age group, F (1, 117) = 35.15, MSE = 103718.99, p < .01, where older adults ( M = 581.23) read more slowly than young adults ( M = 406.20), and a main effect of word type, F (1, 117) = 122.47, MSE = 3127.90, p < .01, such that target words ( M = 522.08) were read more slowly than modifier words ( M = 465.34). These main effects were moderated by an Age Group x Word Type interaction, F (1, 117) = 7.77, MSE = 3127.90, p < .01. Follow-up tests on this interaction revealed that both young and older adults read the ta rget words more slowly than modifier words, but the differen ce was larger for older adults, p s < .01. No other effects were significant, p s > .52.

PAGE 56

56 Probe Naming Times Recordings from probe naming where a partic ipant coughed, cleared their throat, laughed, made a comment, or pressed the enter key befo re they said the probe word were discarded. This process removed 29.5% of potential respon ses from young adults, but only 11.4% of older adults potential responses. Similar to reading ti me data, participants naming times outside of three standard deviations above the mean for each age group were discarded, which eliminated 2.0% of young adults naming time data and 1.6% of older adults naming time data. Probe word naming time analyses were conducted to examine the impact of symbol on different levels of text representation, as pred icted in Hypothesis 2. A 2 (Age Group) x 2 (Sentence Type) x 3 (Probe Type) x 2 (Symbol) ANOVA was conducted on mean naming times (Table 3-6 sh ows means and standard deviations). This analysis revealed a main effect of age group, F (1, 117) = 4.69, MSE = 246601.37, p < .03, a main effect of sentence type, F (1, 117) = 6.61, MSE = 5213.40, p < .01, and a main effect of probe type, F (1, 117) = 41.46, MSE = 7823.85, p < .01. Young adults ( M = 777.64) named probes more quickly than older adults ( M = 835.07), and probes follo wing negative sentences ( M = 811.31) were named more slowly than probe s following non-negative sentences ( M = 801.40). New probes ( M = 830.90) were named more sl owly than related probes ( M = 809.61), which were named more slowly than verbatim probes ( M = 778.55), with all differences significant, p < .01. The main effect of symbol was not significant, p > .22, and no interactions among any of the variables were significant, p > .12. Comprehension Accuracy To examine comprehension accuracy, a 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol) ANOVA was conducted (Table 3-7 shows means a nd standard deviations). This analysis revealed only a main effect of sentence type, F (1, 117) = 35.83, MSE = .01, p < .01, where

PAGE 57

57 comprehension accuracy for negative sentences ( M = 68.8%) was poorer than for non-negative sentences ( M = 74.4%). No main effect of age group, p > .31, or symbol, p > .53, was revealed, and no interactions were significant, p > .75. Comprehension accuracy analyses tested the content of the memory representation, and tested Hypothesis 1, where assist ance of a symbol is predicted to help the storage of information. Given that the probe words were presente d between reading and presentation of the comprehension question, a 2 (Age Group) x 2 (S entence Type) x 3 (Probe Type) x 2 (Symbol) ANOVA was conducted on mean comprehension accu racy to examine the impact of probe words on comprehension of the information read in the sentence (Table 3-8 shows means and standard deviations). As probe type was the only new variable in this analysis, only effects involving probe type are reported. A main ef fect of probe type was demonstrated, F (1, 115) = 17.06, MSE = .03, p < .01. However, this main effect was moderated by its interaction with sentence type, F (2, 230) = 2.46, MSE = .03, p < .03. Follow-up analyses on this interaction revealed that for non-negative sentences, accuracy was greater following verbatim probes than related probes, which was grea ter than new probes, with all differences significant, p s < .01. For negative sentences, the only significant effect was that accuracy following verbatim probes was better than accuracy following new probes, p < .05; accuracy following related probes was not different from either verbatim probes, p > .58, or new probes, p > .11. The main effect of age group was not significant, p > 59, nor was the main effect of symbol, p > .25, or other interactions, p > .31. To determine whether readers were better able to use the symbol and integrate it with their memory representation of the text, data were spli t into thirds, and a 2 (Age Group) x 2 (Sentence Type) x 2 (Symbol) x 3 (Group Order) ANOVA was conducted on comprehension accuracy. It is

PAGE 58

58 important to note that because fillers were randomly distributed throughout the stimuli presentation, these groups have different numbers of experimental stimuli. With respect to differences in group order, a main effect was revealed, F (2, 234) = 4.35, MSE = .04, p < .01, such that comprehension accuracy was greater for the last third of the comprehension questions compared to the first third of comprehension questions, p < .01, and comprehension accuracy was marginally greater for the middle section of comprehension questions than the first third of comprehension questions, p < .07. However, there was no difference between the second and third sections of questions, p > .32. This effect of group order di d not significantly interact with age group, p > .10, symbol, p > .97, or sentence type, p > .34. A series of stepwise regression analyses we re conducted to examine the influence of operation span, Stroop effect, age, vocabulary scor e, years of education, and time spent reading each day on comprehension accuracy of each type of sentence: (1 ) negative with a symbol, (2) negative without a symbol, (3) non-negative w ith a symbol, and (4) non-negative without a symbol. Variables entered in this order presen t the best fit model, although operation span, Stroop effect, and age were each entered first in every combination with the remaining variables, similar to Experiment 1. Results are presented in Table 3-9, Table 3-10, Table 3-11, and Table 312. Results showed that for negative sentence s presented with a symbol, a model containing operation span, vocabulary score, and age accounted for 20.2% of the variance in comprehension accuracy, with operation span accounting for 10.2% of the variance, vocabulary accounting for an additional 4.2%, and then age accounting fo r an additional 5.8%. For negative sentences without a symbol, non-negative se ntences with a symbol, and nonnegative sentences without a symbol, only operation span accounted for a sign ificant amount of variance: 8.7%, 2.6%, and 6.8%, respectively.

PAGE 59

59 Discussion The present experiment attempted to offset the difficulties in comprehension caused by negation by exploring the impact of presentation of a symbol as a compensatory strategy. The results showed that the presence of a sym bol did not increase comprehension for negated sentences specifically, for either age group. Effects of negation we re minimal for reading times but emerged for probe naming times and comprehension accuracy. Within comprehension accuracy, readers did improve sign ificantly over the course of the experimental session. However, relative to Experiment 1, some change s in negation processing did occur, specifically readers post-target reading times in negative sentences were slowed by the presence of a symbol, and young adults' target reading times did not slightly decrease in negative sentences compared to non-negative sentences. Furtherm ore, older adults' comprehension accuracy increased (i.e., was not poorer than young adults' comprehension accuracy) in this experiment, suggesting that having a symbol present for some sentences may have been beneficial to reading by generally increasing a ttention during reading. Although the first hypothesis, that symbolic compensation will release working memory resources and assist in comprehension of ne gation, was not supported, there was evidence to suggest that the presentation of the symbol may have had some effects on the reading process. First, for sentences with a negati ve modifier, reading times of th e post-target word were slowed by presentation of a symbol relative to sentences without a symbol, an eff ect that did not occur for non-negative sentences. This result suggests that the readers may have been sensitive to the fact that spending more time processing the meaning of the negation symbol with the information presented in the text might help them integrate that information. It is worth noting that the location of this additional reading time was specific to the post-target word; no effects of symbol were seen on target reading times. These results indicate that readers waited until after

PAGE 60

60 the negated phrase to incorporate th e meaning of the symbol into what they had read, resulting in longer reading times immediately af ter the target word, where they were able to pause for wrapup and integration (e.g., Stine, 1990; Stine-Morro w, 1996). Another piece of evidence supporting an influence of the symbol is the young adults' equivalence in reading times for negated and nonnegated targets, unlike Experiment 1 where young adu lts were slightly faster on negated targets. The symbol may have made young adults more awar e of the nature of the sentences and realize that they needed to slow down to process the negation more thoroughly. This additional effort during re ading may have been especially beneficial for older adults comprehension accuracy, as they had equivalent accuracy to young adults, unlike in Experiment 1 where older adults' comprehension was poorer than young adults. While slower reading times did not directly lead to better comprehension accuracy, e.g., older adu lts always had slower reading times than young adults in Experiment 1 and still had poorer accuracy, the symbol presented in the present experiment may have se rved as an attention-grabbing mechanism for older adults, where the symbol signified that the sentences that they were reading were important and required more attention. Howeve r, despite altered reading times in negative sentences with a symbol, young adults did not demonstrate an increase in their comprehension accuracy. One possibility is that young adults may be less experi enced with the use of negation symbols, unlike older adults who take more prescription medi cations and therefore s ee the negation symbol presented in conjunction with text more often. As a result, young adults may be less experienced at incorporating additional information into their situation model representation of the text. Another possibility is that older adults spent more time processing the sy mbol and the sentences in general, given that all of their reading times were slower than young adults. This additional processing time in c onjunction with older adults' experience at incorporating

PAGE 61

61 additional information into a situ ation model may have increased their comprehension relative to Experiment 1. A final possibility is that the increased attention may have released some working memory resources, allowing more resources to be allocated toward read ing than other mental processes. For example, some participants comment ed that they were trying to predict what they would be asked as they read the sentences. This allocation of resources toward comprehension may have helped readers to focus more clearly on creating a memory repr esentation of the whole text, as opposed to focusing resources on a portion of the text that may not have had anything to do with the subsequent comprehension questi on. For young adults, this assistance was not important for releasing additional working memory resources because they were not as limited as older adults. This explanation is consistent with theories on working memory (e.g., Just & Carpenter, 1992), which suggest that when enough working memory resources are available to process a particular piece of text, that inform ation will be understood better than when not enough resources are available for processing. The po ssibility remains that a release of working memory resources may come not only from cont ext, which Just and Carpenter (1992) suggest, but also from increased awareness of the situation. The potential impact of the symbol as an attention-grabbing mechanism highlights the importance of working memory in these expe riments, which is supported by the regression analyses. Working memory was the most signi ficant predictor of reading comprehension performance in both experiments, although some interesting differences emerged between the two experiments. For Experiment 1, working memory was the only variable that was a significant predictor of comprehension for bot h negative and non-negative sentences. In the present experiment, vocabulary and age also ex plained variance in read ing comprehension but only for negative sentences presen ted with a symbol. Individuals with larger vocabularies are

PAGE 62

62 more experienced with language and may better be able to incorporate th e meaning of a symbol with the meaning of a sentence. Similarly, age also reflects experience, as older adults are more experienced in seeing symbols combined with text in other contexts e.g., prescription medication labels. However, for the senten ces without a symbol, only operation span significantly accounted for variance in comprehension accuracy, and more variance was accounted for by operation span in Experiment 1 than in Experiment 2. The reason for this result may be linked to the idea that presence of a symbol increased attention overall, releasing working memory resources and making working me mory less important for comprehension of these sentences. Interestingly, the variance acco unted for by operation span was especially low for non-negative symbol sentences in Experiment 2, where the symbol really has no meaning to assist in comprehension. This result suggests th at these sentences are easy to read, requiring no incorporation of the symbols meaning as we ll as no negation. Therefore, working memory would not be taxed when reading these sentences, and the capacity of these resources would not constrain comprehension. Another difference that emerged between Experiment 1 and Experiment 2 was the influence of the probe word on comprehension ac curacy. While the probe words had no effect on comprehension accuracy for negative sentences in Experiment 1, the verbatim probe words increased comprehension relative to the new probes for negative sentences in Experiment 2. This difference in results suggests that while readers paid more attention to reading the sentences in Experiment 2, they may also have been able to use the increased attention to prevent incorporating information that was unnecessary for the situat ion model representation. In addition, the verbatim word may have cued the reader to remember the wording used in the sentence, which combined with increased atte ntion could increase comp rehension accuracy. New

PAGE 63

63 and related probes did not have this effect because they were not present in the text and would less specifically refer back to a portion of the te xt that the reader would need to answer the question. The second hypothesis, that bene fits from a symbol would be more dramatic at the situation model level, was also not supported. By examining probe word naming times, it was evident that the presence of a symbol did not specifically affect any level of representation, as it did not influence naming times at all. The mean ing of the symbol was presented to provide a context and some pre-existing knowledge about the text and therefore should have been incorporated as such in the s ituation model representation of the text (Kintsch,1998), but it was not, evidenced by the lack of interaction between symbol and probe word type on probe naming times. This result suggests that readers were una ble to understand the meaning/use of the symbol in relation to the text and ther efore could not successfu lly incorporate the m eaning of the symbol with the meaning of the text. Informal suppor t for this idea comes from comments made by several participants that they were unsure of wh at to do with the symbol. Another possibility is that the symbol did not provide enough informa tion to create a context for the sentence it described, so that it could not assi st in creating an appropriate memo ry representation of the text. In terms of levels of representation, this possibili ty means that the symbol would not be included in the situation model representa tion of the text, as it did not serve as a form of pre-existing knowledge related to the text (e.g., Kintsch, 1998). More generally, the results of Experiment 1 and Experiment 2 displayed remarkable consistency. The stability of the findings suggest s that the effects of negation on comprehension are real (e.g., Cornish, 1971, Experiment 1; Cornish & Wason, 1970; Hoosain, 1973; Just & Carpenter, 1971; Kaup, 2001; Kaup et al., 2004 ; Kaup et al., 2005; Kaup & Zwaan, 2003; Kaup

PAGE 64

64 et al., in press; MacDonald & Ju st, 1989; Sherman, 1973) and extend to more real-world types of texts. Negation slowed probe naming times a nd decreased comprehension accuracy, suggesting that negation is a complex structure that causes readers difficulty in pr ocessing and remembering negated concepts. This experiment showed th at a negation symbol helped both young and older adult readers to better process all sentences, no t just negated ones. Perhaps other compensatory strategies can help readers to specific ally process negation more effectively.

PAGE 65

65 Table 3-1. Means and standard deviations (st. dev.) of demographic measures. Age group Young Older Mean St. dev Mean St. dev. Years of education** 13.36 1.39 16.83 3.46 Health rating (out of 10) 8.22 1.39 7.98 1.70 Vocabulary (out of 25)** 13.32 3.33 20.98 2.86 Forward digit span 7.22 1.19 7.05 1.31 Backward digit span 4.87 1.16 5.15 1.41 Hours spent writing** 2.78 1.88 1.55 1.18 Hours spent reading 3.28 1.92 2.97 1.72 Hours watching TV 2.30 1.72 2.15 1.43 Hours doing crosswords* .28 .76 .62 .94 Spelling training (out of 10)** 6.17 2.08 7.31 2.32 Spelling ability now 6.33 1.53 6.32 1.94 (out of 10) Operation span** .90 .09 .85 .10 Congruent Stroop** 22.04 5.74 37.22 14.12 Incongruent Stroop** 20.68 3.39 45.74 11.41 Note: indicates that the age differences were significant at the p < .05 level; ** indicates that the age differences were signi ficant at the p < .01 level Table 3-2. Means and standard deviations for overall reading times (msec) Age group Sentence type Symbol Mean Standard deviation Young Negative Symbol 434.76 101.24 No symbol 434.01 100.76 Non-negative Symbol 439.96 111.83 No symbol 439.42 106.65 Older Negative Symbol 589.42 183.18 No symbol 589.25 191.28 Non-negative Symbol 591.32 180.37 No symbol 588.82 182.49

PAGE 66

66 Table 3-3. Means and standard deviations fo r target word reading times (msec) Age group Sentence type Symbol Mean Standard deviation Young Negative Symbol 429.60 120.05 No symbol 425.24 113.01 Non-negative Symbol 437.40 128.57 No symbol 430.73 119.82 Older Negative Symbol 613.56 220.08 No symbol 619.93 246.10 Non-negative Symbol 611.44 223.05 No symbol 606.52 217.89 Table 3-4. Means and standard deviations fo r post-target word r eading times (msec) Age group Sentence type Symbol Mean Standard Deviation Young Negative Symbol 470.10 181.15 No symbol 451.84 122.17 Non-negative Symbol 466.13 143.89 No symbol 482.30 161.03 Older Negative Symbol 641.07 241.00 No symbol 615.32 204.95 Non-negative Symbol 634.52 228.87 No symbol 624.04 216.83 Table 3-5. Means and standard deviations for m odifier and target word reading times for negative sentences only (msec) Age group Word type Symbol Mean Standard deviation Young Modifier Symbol 386.35 101.85 No symbol 383.60 98.32 Target Symbol 429.60 120.05 No symbol 425.24 113.01 Older Modifier Symbol 545.86 176.20 No symbol 545.56 193.48 Target Symbol 613.56 220.08 No symbol 619.93 246.10

PAGE 67

67 Table 3-6. Means and standard deviations of probe word naming times (msec) Age group Sentence type Probe type Symbol Mean Std. deviation Young Negative Verbatim Symbol 759.86 109.85 No symbol 757.62 128.55 Related Symbol 793.65 160.64 No symbol 778.96 113.06 New Symbol 808.83 130.03 No symbol 801.54 134.94 Non-negative Verbatim Symbol 737.37 111.60 No symbol 739.05 103.54 Related Symbol 785.55 119.49 No symbol 770.13 119.51 New Symbol 788.28 124.67 No symbol 810.71 139.90 Older Negative Verbatim Symbol 817.62 182.71 No symbol 815.01 182.20 Related Symbol 844.19 195.82 No symbol 831.07 178.14 New Symbol 854.78 201.88 No symbol 872.56 222.61 Non-negative Verbatim Symbol 737.47 111.60 No symbol 802.21 174.52 Related Symbol 845.49 189.63 No symbol 827.86 170.98 New Symbol 869.75 223.51 No symbol 840.74 197.78 Table 3-7. Means and standard deviations for comprehension accuracy (%) Age group Sentence type Symbol Mean Standard deviation Young Negative Symbol 69.38 16.45 No symbol 70.32 15.10 Non-negative Symbol 75.60 11.70 No symbol 75.68 13.51 Older Negative Symbol 67.46 16.57 No symbol 68.02 15.80 Non-negative Symbol 73.05 13.39 No symbol 73.34 12.90

PAGE 68

68 Table 3-8. Means and standard deviations for comprehension accuracy by probe type (%) Age group Sentence type Probe type Symbol Mean Std. deviation Young Negative Verbatim Symbol 71.27 21.16 No symbol 69.16 20.48 Related Symbol 66.25 21.32 No symbol 70.45 24.61 New Symbol 66.74 26.01 No symbol 67.55 20.04 Non-negative Verbatim Symbol 80.31 16.70 No symbol 80.89 25.47 Related Symbol 75.36 18.98 No symbol 76.11 21.89 New Symbol 67.70 22.14 No symbol 72.41 20.44 Older Negative Verbatim Symbol 69.77 21.95 No symbol 69.51 22.55 Related Symbol 69.01 22.89 No symbol 70.16 25.73 New Symbol 65.17 23.90 No symbol 65.33 21.67 Non-negative Verbatim Symbol 75.15 18.77 No symbol 80.31 16.78 Related Symbol 75.19 14.47 No symbol 73.48 16.34 New Symbol 68.89 19.76 No symbol 67.56 24.19 Table 3-9. Results of regression analysis on nega tive sentence presented with symbol comprehension accuracy Variable p Operation span .20 < .03 Vocabulary .58 < .01 Age .49 < .01 Stroop effect .07 .47 Years of education .17 .12 Hours spent reading .03 .69

PAGE 69

69 Table 3-10. Results of regression analysis on negative sentence pres ented without a symbol comprehension accuracy Variable p Operation span .31 < .01 Stroop effect .02 .81 Age .02 .82 Vocabulary .14 .11 Years of education .01 .96 Hours spent reading .11 .21 Table 3-11. Results of regression analysis on non-negative sentence presented with symbol comprehension accuracy Variable p Operation span .18 <.05 Stroop effect .06 .51 Age .04 .66 Vocabulary .05 .58 Years of education .04 .67 Hours spent reading .06 .53 Table 3-12. Results of regression an alysis on non-negative sentence pr esented without symbol comprehension accuracy Variable p Operation span .28 < .01 Stroop effect .05 .58 Age .03 .75 Vocabulary .14 .13 Years of education .07 .43 Hours spent reading .05 .61

PAGE 70

70 Sentence Type x Symbol Interaction for Post-Target Reading Times 500 510 520 530 540 550 560 570 580 590 600 NegativeNon-negative Sentence TypeReading Times (msec) Symbol No symbol Figure 3-1. Post-target reading times by sentence type and presence of symbol.

PAGE 71

71 CHAPTER 4 GENERAL DISCUSSION The present research has offered insight into reading under circumst ances that increased the difficulty of processing, comprehending, and re trieving information. Sp ecifically, the present research investigated how young and older adults were influenced by nega tion (e.g., no, not, or never) while reading real-w orld types of texts, and how i ndividual differences, i.e., working memory differences, contributed to age-related changes in readi ng. This research also explored the effects of a compensatory strategy on co mprehension of negated text. Several general conclusions can be drawn and are discussed below. Consistent with previous research on negation, naming and comprehension following reading were affected by the presence of negati on in text. Similar findings occurred in both experiments, suggesting that these effects of ne gation on reading are stable The present research extended these findings to online measures of re ading, although negation sp eeded target reading times, especially in young adult r eaders. Except for this age diffe rence in reading times, negation similarly influenced young and older adults' comprehension across multiple measures when reading text, contrary to the predictions of IDH (e.g., Hasher & Zacks, 1988). MacDonald and Just (1989) suggested that when a person reads a negative sentence, the negated concept must be inhibited because it is no longe r appropriate to the situation described by the text. The IDH predicted that older adults would have more di fficulty discarding inappropriate information that has been gated into working memory, which in th is case would be the concept that was negated. Although older adults named the probes more slowly than young adults and had poorer comprehension accuracy overall, the presence of negation did not exacerbate these age differences. Thus, inhibition of negative concep ts was not especially problematic for older adults, which has several implications.

PAGE 72

72 First, negation may not always be processed in terms of inhibition; instead, negation can be processed in other ways, such as by transforming the negated concept into its alternate meaning (e.g., Gough, 1965; Mehler, 1963). Transformations would require more working memory resources to process negation because the reader needs to hold the affirmative version of the sentence in memory and then manipulate it to its opposite based on the negative modifier. Support for this idea comes from the importance of working memory in the present experiments, as shown by the regression analyses. Second, gi ven that older adults demonstrated more difficulty with inhibition than young adults as evidenced by the Stroop test, but did not demonstrate a difficulty with i nhibition during reading negation, it appears that the IDH may not account for older adults' comprehension of negati on. The present research therefore demonstrates another circumstance under which older adults do not show an exacerbated difficulty with inhibition (e.g., Burke & Harrol d, 1988; James et al., 1998). The lack of inhibition observed in the pres ent experiments may also be related to the stimuli. MacDonald and Just (1989) suggested that inhibition was im portant for understanding negation, as a reader would have to de-activate the negated concep t in order to have a correct representation of the sentence. Ho wever, the negated sentences in their experiments always took the same form, e.g., Elizabeth baked some br ead but no cookies. In these sentences, the presence of cookies is denied, so it needs to be removed from the me mory representation of the text in order to accurately represent what was present. In the present experiments, negation was presented in different ways, e .g., He could not part with his school or the people in it. In sentences like this one, the meaning of the nega tion is not to deny the presence of an item but instead is to change the meaning of the word it modifies. These differences in how negation was

PAGE 73

73 used in the experimental sentences could have co ntributed to the lack of inhibition observed in the present research. The present research also demonstrated that ne gation did not have a gr eater influence at the situation model level, compared with the su rface level of representation, even though negation did affect memory for the text overall. In ad dition, there were no specific circumstances (e.g., the situation model level) where older adults were able to offset the diffi culty produced by negation based on their extensive experien ce with processing information as a gist representation. An examination of the probe naming times demonstrat ed that all probes were equivalently slowed following a negative sentence rela tive to a non-negative sentence, suggesting that the different levels of representation were not differentially a ffected by the presence of negation. This result is contrary to Kaup (2001), who demo nstrated that negation affected the situation model level of representation but not the surface level. Perhaps th e effects of negation on different levels of representation are task-dependent; Kaup (2001) used recognition as an activation measure, whereas the present research used probe na ming. As a task involving explicit memory, recognition may more directly tap into the memo ry representations that are affected by the negative concept, whereas probe naming, an implicit task, may be less sensitive. With respect to compensatory strategies, the pr esence of a symbol did not help a reader to form a correct representation of the negated text, at any level, contrary to predictions derived from the CI Model (Kintsch, 1998). Kintsch (1998) proposed that a reader will initially activate all of the information presented, including the text and any related information, and then activation is constrained to include only the in formation necessary to the meaning of the text, rather than the exact wording, creating the si tuation model representation. The information included in the representation in cludes pre-existing knowledge a bout the topic the reader is

PAGE 74

74 reading, which in the present research would ha ve included the symbol presented before and during reading. According to Just and Carpenter (1992), providing a context can release working memory resources for comprehension. Combining th ese two theories, this contextual information may guide the activation of the information presented during reading, resulting in a wellunderstood and well-represented situation model. Ho wever, the symbol in the present experiment did not assist in providing a cont ext for readers, nor did it appear to be integrated with the information presented in the text. These findings suggest that (a) eith er the symbol did not provide enough information to serve as compensati on in the form of a context, or (b) readers were unable to use the symbol e ffectively in that capacity. The first possibility would suggest that readers were unaware of the meaning of the symbol, but this option seems unlikely, given that the circle with the line through it is the universal symbol for do not. The second possibility implies that readers were unable to integrate the meaning of the symbol with the meaning of the negated phrase in the sentence. This suggestion seems more plausible be cause the readers were not explicitly instructed about wh at to do with the symbol; theref ore, they either may not have seen the connection between the meaning of the te xt and the meaning of the symbol or did not know that they were supposed to connect the two. Although a symbol did not have specific effect s on the processing of negated text, there was some evidence that the presen ce of a symbol may have increas ed older adults attention to the reading task, as shown in post-target readi ng times. Furthermore, the age differences in comprehension accuracy that appeared in Experime nt 1 disappeared in Experiment 2, even when no symbol was present. Research on pictorial descriptors given with medication instructions indicate that people prefer when a picture is presented than wh en it is not, because the picture reinforces the information presented in the text (e.g., Sojourner & Wogalter, 1997). A preference

PAGE 75

75 for instructions including a de piction of the instruction may encourage readers to pay more attention to those instructions, which may seem more important because of the added symbol. This same logic can be applied to the present research such that readers may have generally increased their attention to the sentences because a symbol was present. However, this resulted in increasing overall accuracy because once read ers recognized that the sentences may be important, they continued to pay closer attentio n to all of the sentences, even when no symbol was present. This increased attention was especi ally relevant for older adult readers, who are more experienced with the negation symbol fr om seeing it on prescription labels and have already realized its importance. The present research also demonstrated the impact of negation when reading real-world texts, which is a significant extension of pr evious work on negation. Some of the previous research on negation dealt with te xt that more closely resembled logic statements (e.g., The dot is not red; see Cornish, 1971), perhaps in the atte mpt to make the experimental sentences focus only on negation and avoid other ty pes of complexities that ma y interact with the negation. However, these are not the ways in which read ers normally encounter ne gation. By using a more ecological approach (e.g., Neisser, 1978) and by having participants read sentences from actual fiction stories, the present research replicated the detrimental effects of negation that were demonstrated in previous experiments. The similarities among results indicate that the difficulties associated with understanding negation while reading are real and are not a function of difficulties in logical reasoning. The presen t research demonstrates the importance of continuing to use real-world text s to explore comprehension in th e world in which we encounter it. In addition, the presen t research demonstrates that older ad ults did not show specific deficits

PAGE 76

76 in comprehension of negation with these types of texts, which may have positive implications for their ability to function in real-world situations. Limitations There are several limitations of the present experiments that are worthy of discussion. First, the text that was used was very rich in context, where many details described by the text were given to depict a vivid s ituation. As a result, older adu lts may not have shown as much decline in comprehension because they were read ing text under more optimal circumstances than they may have encountered in previous researc h. Previous research on reading indicates that older adults are better able to store and represent text in thei r memory in the situation model representation because this level is the most useful representation (e.g., Radvansky et al., 2001). It is possible that the stimuli used in the present research lent th emselves to easily forming a rich and meaningful representation, even when negati on was present. Given mo re detail and more description of events than sentences used in previous research on negation, the sequence of events in the sentences used here were more eas ily visualized, and older adult readers might have more easily drawn upon their experience and informa tion in the text to form a situation model representation. The additional depth of information to the sentences in the present research may have put older adults at an advantage, shieldi ng potential age differences in comprehension of negation that would have appeared if the exact wording of the te xt had been required to show comprehension. Second, previous research on negation has demo nstrated that various types of negation may lead to different results (e.g., Just & Ca rpenter, 1971; Sherman, 1973). For example, the addition of the word not makes comprehension mo re difficult than the a ddition of another type of negative word, like few or the prefix un. The present research used the word not because it was expected to make comprehension mo re difficult than the other types, allowing the

PAGE 77

77 possibility of finding age differences. However, the word not may have multiple functions, serving as both a word that denies existence (e .g., She some baked bread but no cookies.) and as a word that changes meaning (e.g., He could not part with his school or the people in it.). These differing functions may both cause difficultie s in comprehension but for different reasons. Where the function of the word not is to deny existence of an object, processing most likely involves inhibition. Where the function of the wo rd not is to change the meaning to its opposite, processing is likely more workingmemory intensive because of the required transformation (e.g., Mehler, 1963 ; Gough, 1965). Given these poten tial processing differences, it is evident that not as a structure in infl uencing comprehension may be more complex than previously thought. Third, the time spent attending to the symbol wa s not recorded, so there is no indication of whether or not readers were spending enough time pr ocessing the symbol for the symbol to serve as compensation. In addition, the symbol was present during the en tire sentence, and there was no way to measure whether people looked back at the symbol periodically. While there is no standard amount of time that a reader should spe nd processing the pictorial information when it is presented, a longer amount of time may be related to the us efulness of the symbol. Readers who spent longer looking at the sy mbol, or exerted more effort to incorporate the meaning of the symbol into the text that they were read ing, may have improved their own comprehension. Furthermore, age differences in time spent proces sing the symbol are also possible and could be linked to the lack of age differences observed in comprehension accuracy when symbols were used in the experiment. However, this informati on is not available from the measures gathered from the present research.

PAGE 78

78 Fourth, a comparison in the present research ha s been made between instructions (such as medication instructions used with a symbol) and fictional text. The limitation here is that these are two very different kinds of text s. For instructions, it is often th e case that the exact wording is very important to the message, whereas for fic tional text and other t ypes of narratives, the message can be remembered and stored as a situa tion model or just the gist. Due to the different nature of these types of text, the process by wh ich a reader may read and store the information, as well as incorporate the meaning from a symbol may be different. For instructions, where the wording is important to the message, the symbol may be mapped on to a specific word. For fictional or narrative text, where the wording is not as important to the message, the meaning from the symbol may simply be incorporated in to the situation model of the text. Therefore, although the present experiments used more real-wor ld texts, their generali zability to texts like medication instructions may be limited. Finally, in the second experiment, participants were not given explic it instruction on how to interpret the symbol or how to incorporate th e symbol into the meaning of the sentence; they were just told to look at the symbol before they started to read the sentence. After completing the experiment, some participants reported not know ing what to do with the symbol or why it was there. These post-hoc reports sugge st that the implicit nature of the symbol may not have been sufficient to provide compensati on in reading these types of se ntences. Giving some deliberate instruction on how to consciously use the symbol rather than implicit ones, may have increased the impact that it could have had on compensation. For example, explicit in structions to use the negation symbol may have encouraged readers to develop a compensatory reading strategy to help them improve their comprehension of the negation.

PAGE 79

79 Future Directions Future research on negation and aging should co ntinue to focus on th ree important aspects of these topics that were explored here: (1) ci rcumstances where a reader might have more or less difficulty in understanding negation, (2) circ umstances where a read er or a writer may improve comprehension of text involving ne gation, and (3) exploring and applying these investigations within the context of real-wor ld types of texts (e.g., non-fiction, instructions, jokes). One direction for future research is to furt her break down the different types of negation that may cause difficulty in comprehension. Wh ile some breakdown has occurred in the past (e.g., Just & Carpenter, 1971; Sherman, 1973), the pr esent research has indicated that additional exploration may be necessary. It seems funda mentally different to negate a noun (e.g., cookies) than to negate an ad jective (e.g., bright.). One argume nt is that negative modifiers may differentially impact target s as a function of their parts of speech. Negating a noun often indicates that it is not presen t and would be processed by deac tivating the negated concept. However, negating an adjective would simply change the meaning, and would be processed by transforming the sentence into its negative form. Given that older adults have demonstrated difficulty with inhibition in the Stroop test, th ey may be more differentially impacted by a type of negation that evokes inhibition (e.g., negati ng a noun), but not differe ntially impacted by a type of negation that involves a change of meaning or tr ansformation (e.g., negating an adjective). Another direction for future research involves a further examination of the use of symbols on comprehension. Given that read ers prefer the presentation of symbols in an instructional situation (e.g., Sojourner & Woga lter, 1997), there is some indication that some people may find them helpful. Future research should investig ate whether more time pr ocessing the symbol or

PAGE 80

80 whether conscious strategies for using the symbol will improve assistance from the symbol. In addition, given the difference between instructional text and fiction, future research may also be able to determine whether symbols, when unders tood by the reader, may have a larger impact on one type of text over the ot her. Depending on the strategy given, mapping the symbol onto a particular word would assist in structional text bette r than narrative text, whereas instructing participants to incorporate the meaning of the symbol with the meaning of the text may better benefit narrative text. A final direction for future research is the investigation of additional mechanisms for compensation in reading negation. Given that olde r adults have demonstrated that they pause more often while reading complex sentences, lead ing to better retrieval of the information they read (e.g., Stine, 1990), and that they are also ab le to apply a strategy of pausing more often during reading to improve their comprehension (e .g., Noh et al., 2006), an exploration of pausing during reading negation seems logical. Whether individuals who spontan eously pause after reading a negation have better co mprehension accuracy than indi viduals who do not pause could be assessed. For readers who do not naturally adjust their reading via pauses, a training strategy could be employed to investig ate whether explicit, effort ful strategies can improve comprehension of negation, e.g., instructing reader s to pause after reading a negative phrase. Since negative sentences can be categorized as another type of complex sentence, pausing after reading the negated concept may provide time fo r wrap-up and to complete processing of this difficult construction. The exploration of negation comprehension in the present research speaks to the importance of this type of construction in text. Given that readers of all ages have difficulty understanding text that contains negative words, both readers and wr iters need to be sensitive to

PAGE 81

81 the use of this construction. Read ers need to learn how to adjust their reading strategies when encountering this construction. Writers should avoid using nega tion to minimize comprehension difficulties whenever possible. Negation appears in many everyday situations, oftentimes without our awareness because it is a small, simple wor d, but people should be aware that little things can have big effects on comprehension.

PAGE 82

82 APPENDIX A SAMPLE STIMULI Sample negative sentence: As Joel looked around, he thought about what leaving his school would mean, and he knew that he could not part with the school and the people in it. Sample non-negative sentence: As Joel looked around, he thought about what leaving his school would mean, and he knew that he could part with the sc hool and the people in it. Sample probes and frequencies: Verbatim probe part (11) Related probe split (26) New probe mash (4) Sample comprehension question: What did Joel know about hi s ability to leav e his school and the people in it?

PAGE 83

83 APPENDIX B EXPERIMENT 1 INSTRUCTIONS In this experiment, you will be asked to read sentences, say aloud specific words, and answer questions about the sentences that you have read. It is importa nt that you read in order to understand the gist of what you are reading, but it is not necessary that you memorize each sentence. The reading task is different from how you nor mally read. You will be silently reading sentences one word at a time, where you will control your own reading speed. The words in the paragraphs will show up as da shes (--------) on the screen, and you will press the Space Bar key to make a word appear. Each time you press the Space Bar, the previous word will return to dashes, and the next word will appear. The sentences you will be reading have been take n from a story about magic, so you may notice some unusual events happening. Just read the senten ces at your normal pace, and do your best to understand what you are reading, as you will be asked comprehension questions about them. Remember, you do not need to memorize the sentences word-for-word, but you will be asked to remember the main ideas of the sentences. After reading the sentences, you will see a single word on the scree n. You need to say this word aloud, as soon as it appears. Say th is word as quickly and as natu rally as you can, but be careful not to make any extraneous noises, such as coug hing or clearing your throat that may be picked up by the microphone. The main point is to say this word aloud as soon as possible after seeing it. After saying the word out loud, press the Enter key, and a ques tion will appear on the screen. Read the question and say your answer out loud so that the experiment er can write down your answer. You may answer using one or more words, whatever you feel is n ecessary to answer the question. In addition, at this time, please feel free to give a ny comments about whether or not you feel your answer was hindered by an awkwar d construction of the sentence. However, you do not need to give comments if you do not ha ve any, simply wait until the experimenter has finished writing, and press the Enter key and continue on to the next sentence. This pattern of tasks (reading, saying words, and answering questions) will repeat throughout the experiment, so be sure that you understand what tasks you will be asked to do. If you have questions, please ask the experimenter now. Othe rwise, press the Enter key, and you will be given several practice trials to get familiar with the tasks.

PAGE 84

84 APPENDIX C EXPERIMENT 2 INSTRUCTIONS In this experiment, you will be asked to read sentences, say aloud specific words, and answer questions about the sentences that you have read. It is importa nt that you read in order to understand the gist of what you are reading, but it is not necessary that you memorize each sentence. The reading task is different from how you nor mally read. You will be silently reading sentences one word at a time, where you will control your own reading speed. The words in the paragraphs will show up as da shes (--------) on the screen, and you will press the Space Bar key to make a word appear. Each time you press the Space Bar, the previous word will return to dashes, and the next word will appear. In addition, you will sometimes see a symbol presented before a sentence. The symbol may or may not seem relevant in helping you to understa nd the sentence, but it is important to look at the symbol before starting to read the sentence. After looking at the sy mbol, you will press the Space Bar to begin reading the sentence. The sentences you will be reading have been take n from a story about magic, so you may notice some unusual events happening. Just read the senten ces at your normal pace, and do your best to understand what you are reading, as you will be asked comprehension questions about them. Remember, you do not need to memorize the sentences word-for-word, but you will be asked to remember the main ideas of the sentences. After reading the sentences, you will see a single word on the scree n. You need to say this word aloud, as soon as it appears. Say th is word as quickly and as natu rally as you can, but be careful not to make any extraneous noises, such as coug hing or clearing your throat that may be picked up by the microphone. The main point is to say this word aloud as soon as possible after seeing it. After saying the word out loud, press the Enter key, and a ques tion will appear on the screen. Read the question and say your answer out loud so that the experiment er can write down your answer. You may answer using one or more words, whatever you feel is necessary to answer the question. In addition, at this time, please feel free to give a ny comments about whether or not you feel your answer was hindered by an awkwar d construction of the sentence. However, you do not need to give comments if you do not ha ve any, simply wait until the experimenter has finished writing, and press the Enter key and continue on to the next sentence. This pattern of tasks (reading, saying words, and answering questions) will repeat throughout the experiment, so be sure that you understand what tasks you will be asked to do. If you have questions, please ask the experimenter now. Othe rwise, press the Enter key, and you will be given several practice trials to get familiar with the tasks.

PAGE 85

85 LIST OF REFERENCES Burke, D. M. (1997). Language, aging, and inhi bitory deficits: Evaluation of a theory. Journal of Geronotology: Psychological Sciences 52B P254P264. Burke, D. M., & Harrold, R. M. (1988). Automa tic and effortful semantic processes in old age:Experimental and naturali stic approaches. In L.L. Light & D. M. Burke (Eds.), Language, memory and aging (pp. 100). New York: Cambridge University Press. Connelly, S. L., Hasher, L., & Zacks, R. T. (1991) Age and reading: The impact of distraction. Psychology and Aging 6 533. Conway, A.R.A., Kane, M.J., Bunting, M.F., Ha mbrick, D.Z., Wilhelm, O., & Engle, R.W. (2005). Working memory span tasks: A methodological review and users guide. Psychonomic Bulletin & Review 12 769. Cornish, E. R. (1971). Pragmatic aspects of negation in sentence evaluation and completion tasks. British Journal of Psychology 62 505. Cornish, E. R., & Wason, P. C. (1970). The recall of affirmative and negative sentences in an incidental learning task. Quarterly Journal of Experimental Psychology 22 109. Dowse, R., & Ehlers, M.S. (2002). The influe nce of education on the interpretation of pharmaceutical pictograms for communi cating medicine instructions. International Journal of Pharmacy Practice 11 11. Francis, W.N., & Kucera, H. (1982). Frequency analysis of E nglish usage: Lexicon and Grammar Boston: Houghton and Mifflin. Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Minimental state: A practical method for grading the state of patients for the clinician. Journal of Psychiatric Research 12 189 198. Gough, P. B. (1965). Grammatical transf ormations and speed of understanding. Journal of Verbal Learning and Verbal Behavior 4 107. Grownups are turning to teen books. (2005, Se ptember 10). The Gainesville Sun, p. 3D. Hamm, V. P., & Hasher, L. (1992). Age and the availability of inferences. Psychology and Aging 7 56. Hasher, L., & Zacks, R.T. (1988). Working memo ry, comprehension, and ag ing: A review and a new view. In G.H. Bower (Ed.), The psychology of le arning and motivation (Vol. 22, pp. 193). New York: Academic Press. Hoosain, R. (1973). The processing of negation. Journal of Verbal Learning and Verbal Behavior 12 618.

PAGE 86

86 James, L.E., Burke, D.M., Austin, A., & Hu lme, E. (1998). Production and perception of "verbosity" in younger and older adults. Psychology and Aging 13 355. Just, M.A., & Carpenter, P.A. (1971). Compre hension of negation with quantification. Journal of Verbal Learning and Verbal Behavior 10 244. Just, M.A., & Carpenter, P.A. (1992). A capac ity theory of comprehension: Individual differences in working memory. Psychological Review 99 122. Kaup, B. (2001). Negation and its impact on the accessibility of text information. Memory and Cognition 29 960. Kaup, B., Dijkstra, K., & Ludtke, J. (2004). Resolv ing anaphors after read ing negative sentences. Presented at the 44th annual meeting of the Psychonomic Society, Minneapolis, Minnesota. Kaup, B., Ludtke, J., & Zwaan, R. A. (2005). Ef fects of negation, truth value, and delay on picture recognition after reading affirmative and negative sentences. Proceedings of the 27th annual conference of th e Cognitive Science Society. Kaup, B., & Zwaan, R.A. (2003). Effects of negation and situational presence on the accessibility of text information. Journal of Experimental Psychology: Learning, Memory, and Cognition 29 439. Kaup, B., Zwaan, R. A., & Ludtke, J. (in pr ess). The experientia l view of language comprehension: How is negation represented? In F. Schmalhofer & C. Perfetti (Eds.), Higher level language processes in the brai n: Inference and comprehension processes. Mahwah, NJ: Erlbaum. Keele, S. (1972). Attention de mands of memory retrieval. Journal of Experimental Psychology 73 245. Kemper, S. (1987). Syntactic complexity and elderly adults' prose recall. Experimental Aging Research 13 47. Kemtes, K. A., & Kemper, S. (1997). Younge r and older adults online processing of syntactically ambiguous sentences. Psychology and Aging 12 362. Kintsch, W. (1998). Comprehension: A paradigm for cognition. New York: Cambridge University Press. Light, L. L. (1988). Language and aging: Competen ce versus performance. In J.E. Birren & V.L. Bengston (Eds.), Emergent theories of aging (pp. 177). New York: Springer Publishing. MacDonald, M. C., & Just, M. A. (1989). Ch anges in activation le vels with negation. Journal of Experimental Psychology: Learning, Memory, and Cognition 15 633.

PAGE 87

87 MacKay, D. G., & Abrams, L. (1996). Language, me mory, and aging: Dist ributed deficits and the structure of newversusold connections In J.E. Birren & K.W. Schaie (Eds.), Handbook of psychology of aging (pp. 251). San Diego: Academic Press. MacLeod, C.M. (1991). Half a century of research on the Stroop effect: An integrative review. Psychological Bulletin 109 163. May, C. P., Zacks, R. T., Hasher, L., & Multhaup, K. S. (1999). Inhibition in the processing of gardenpath sentences. Psychology and Aging 14 304. Mehler, J. (1963). Some effects of grammati cal transformations on the recall of English sentences. Journal of Verbal Learning and Verbal Behavior 2 346. Neisser, U. (1978). Memory: What are the im portant questions? In M.M. Gruneberg, P.E. Morris, & R.N. Sykes (Eds.), Practical aspects of memory (pp. 3). New York: Academic Press. Newsome, M.R., & Glucksberg, S. (2002). Older adults filter irreleva nt information during metaphor comprehension. Experimental Aging Research 28 253. Noh, S.R., Shake, M.C., Joncich, A.D., Hindin, S.B., & StineMorrow, E.A.L. (2006, April). Age differences in the effects of instruc tion on resource allocati on in reading. Poster presented at the 11th biennial Cognitive Aging Conference, Atlanta, GA. Norman, S., Kemper, S., & Kynette, D. (1992). Adults' reading comprehension: Effects of syntactic complexity and working memory. Journals of Gerontology: Psychological Sciences 47 P258P265. Norman, S., Kemper, S., Kynette, D., Che ung, H., & Anagnopoulos, C. (1991). Syntactic complexity and adults running memory span. Journal of Geronotology: Psychological Sciences 46 P346P351. Radvansky, G. A., Copeland, D. E., Berish, D. E ., & Dijkstra, K. (2003). Aging and situation model updating. Aging, Neuropsychology, and Cognition 10 158. Radvansky, G. A., & Curiel, J. M. (1998). Narrati ve comprehension and aging: The fate of completed goal information. Psychology and Aging 13 69. Radvansky, G. A., Zacks, R. T., & Hasher, L. (2005). Age and inhibition: The retrieval of situation models. Journal of Gerontology: Psychological Sciences 60B P276P278. Radvansky, G. A., Zwaan, R. A., Curiel, J. M., & Copeland, D. E. (2001). Situation models and aging. Psychology and Aging 16 145. Salthouse, T.A. (1996). The processingspeed theo ry of adult age differences in cognition. Psychological Review 103 403.

PAGE 88

88 Sherman, M. A. (1973). Bound to be easier? Th e negative prefix and sentence comprehension. Journal of Verbal Learning and Verbal Behavior 12 76. Sichel, J.L., & Chandler, K.A. (1969). The color word interference test: The effects of varied colorword combinations upon verbal response latency. Journal of Psychology 72 219 231. Sojourner, R.J., & Wogalter, M.S. (1997). Th e influence of pictorials on evaluations of prescription medication instructions. Drug Information Journal 31 963. Stine, E. A. L. (1990). Online processing of written text by y ounger and older adults. Psychology and Aging, 5, 68. StineMorrow, E. A. L., Loveless, M. K., & So ederberg, L. M. (1996). Resource allocation in online reading by younger and older adults. Psychology and Aging 11 475. StineMorrow, E. A. L., Ryan, S., & Leonard, J. S. (2000). Age differences in online syntactic processing. Experimental Aging Research 26 315. Stroop, J.R. (1935). Studies of interf erence in serial verbal reactions. Journal of Experimental Psychology 18 643. Turner, M.L., & Engle, R.W. (1989). Is working memory capacity task dependent? Journal of Memory and Language 28 127. Waters, G. S., & Caplan, D. (1996). Processing resource capacity and the comprehension of garden path sentences. Memory & Cognition 24 342. Waters, G. S., & Caplan, D. (2001). Age, work ing memory and online syntactic processing in sentence comprehension. Psychology and Aging 16 128. Zwaan, R. A., & Radvansky, G. A. (1998). Situa tion models in language comprehension and memory. Psychological Bulletin 123 162. Zwaan, R. A., & Singer, M. (2003). Text compre hension. In A.C. Graesser, M.A. Gernsbacher, & S.R. Goldman (Eds.), Handbook of discourse processes (pp. 83). Mahwah, NJ: Lawrence Erlbaum Associates.

PAGE 89

89 BIOGRAPHICAL SKETCH Sara J. Margolin was born in Trumbull, Conn ecticut, in 1981, but relocat ed with her family to Boynton Beach, Florida, at age 10. She gra duated from Santaluces Community High School as Salutatorian in 1999, and decided to continue he r education at the University of Florida. She began her studies at the university in the fall semester of 1999, and completed her Bachelor of Science degree in 2002, graduating with honors with a major in psychology, a minor in mathematics, and an invitation to be inducte d into Phi Beta Kappa. Upon graduating, Sara immediately enrolled in the gradua te program at the University of Florida to continue her studies in cognitive psychology. Studying language, memor y, and their interaction with aging, she graduated with her Master of Science degree in 2005, and continue d working toward her Ph.D. in cognitive psychology. During this tim e, Sara has also worked towa rd earning a Ce rtificate in gerontology. After finishing her degree, Sara began her career as a faculty member at the State University of New York, College at Brockport. She is living in Rochester, New York with her husband and two dogs.