Second Language Reading: The Interrelationships among Text Adjuncts, Students' Proficiency Levels and Reading Strategies

Permanent Link: http://ufdc.ufl.edu/UFE0011351/00001

Material Information

Title: Second Language Reading: The Interrelationships among Text Adjuncts, Students' Proficiency Levels and Reading Strategies
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011351:00001

Permanent Link: http://ufdc.ufl.edu/UFE0011351/00001

Material Information

Title: Second Language Reading: The Interrelationships among Text Adjuncts, Students' Proficiency Levels and Reading Strategies
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011351:00001

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 E20101124_AAAAEH INGEST_TIME 2010-11-25T02:05:42Z PACKAGE UFE0011351_00001
25927 F20101124_AADBJQ park_h_Page_006.QC.jpg
6318 F20101124_AADBKF park_h_Page_021thm.jpg
6610 F20101124_AADBKG park_h_Page_022thm.jpg
6325 F20101124_AADBJR park_h_Page_006thm.jpg
20915 F20101124_AADBJS park_h_Page_007.QC.jpg
6587 F20101124_AADBKH park_h_Page_024thm.jpg
5217 F20101124_AADBJT park_h_Page_007thm.jpg
6535 F20101124_AADBKI park_h_Page_026thm.jpg
1560 F20101124_AADBJU park_h_Page_008thm.jpg
6431 F20101124_AADBKJ park_h_Page_027thm.jpg
20377 F20101124_AADBJV park_h_Page_009.QC.jpg
22145 F20101124_AADBKK park_h_Page_028.QC.jpg
11374 F20101124_AADBJW park_h_Page_010.QC.jpg
22494 F20101124_AADBKL park_h_Page_029.QC.jpg
19634 F20101124_AADBJX park_h_Page_011.QC.jpg
6429 F20101124_AADBLA park_h_Page_044thm.jpg
6374 F20101124_AADBKM park_h_Page_029thm.jpg
5360 F20101124_AADBJY park_h_Page_011thm.jpg
6511 F20101124_AADBLB park_h_Page_046thm.jpg
25081 F20101124_AADBKN park_h_Page_030.QC.jpg
14582 F20101124_AADBJZ park_h_Page_012.QC.jpg
22008 F20101124_AADBLC park_h_Page_047.QC.jpg
6372 F20101124_AADBKO park_h_Page_031thm.jpg
6453 F20101124_AADBLD park_h_Page_047thm.jpg
24920 F20101124_AADBKP park_h_Page_033.QC.jpg
6261 F20101124_AADBLE park_h_Page_049thm.jpg
6524 F20101124_AADBKQ park_h_Page_034thm.jpg
6602 F20101124_AADBLF park_h_Page_050thm.jpg
24579 F20101124_AADBKR park_h_Page_036.QC.jpg
6654 F20101124_AADBLG park_h_Page_051thm.jpg
23670 F20101124_AADBLH park_h_Page_052.QC.jpg
6672 F20101124_AADBKS park_h_Page_036thm.jpg
6486 F20101124_AADBLI park_h_Page_052thm.jpg
22737 F20101124_AADBKT park_h_Page_037.QC.jpg
6496 F20101124_AADBLJ park_h_Page_053thm.jpg
6498 F20101124_AADBKU park_h_Page_037thm.jpg
20791 F20101124_AADBLK park_h_Page_054.QC.jpg
23171 F20101124_AADBKV park_h_Page_038.QC.jpg
6125 F20101124_AADBLL park_h_Page_054thm.jpg
6510 F20101124_AADBKW park_h_Page_038thm.jpg
5904 F20101124_AADBMA park_h_Page_069thm.jpg
20544 F20101124_AADBLM park_h_Page_056.QC.jpg
23652 F20101124_AADBKX park_h_Page_040.QC.jpg
18732 F20101124_AADBMB park_h_Page_070.QC.jpg
5752 F20101124_AADBLN park_h_Page_056thm.jpg
24331 F20101124_AADBKY park_h_Page_041.QC.jpg
5313 F20101124_AADBMC park_h_Page_070thm.jpg
6552 F20101124_AADBLO park_h_Page_057thm.jpg
6392 F20101124_AADBKZ park_h_Page_043thm.jpg
6645 F20101124_AADBMD park_h_Page_071thm.jpg
17701 F20101124_AADBLP park_h_Page_058.QC.jpg
5534 F20101124_AADBME park_h_Page_073thm.jpg
5209 F20101124_AADBLQ park_h_Page_058thm.jpg
23796 F20101124_AADBMF park_h_Page_074.QC.jpg
6658 F20101124_AADBLR park_h_Page_060thm.jpg
23734 F20101124_AADBMG park_h_Page_077.QC.jpg
21872 F20101124_AADBLS park_h_Page_061.QC.jpg
6489 F20101124_AADBMH park_h_Page_077thm.jpg
22083 F20101124_AADBMI park_h_Page_078.QC.jpg
5805 F20101124_AADBLT park_h_Page_061thm.jpg
6353 F20101124_AADBMJ park_h_Page_078thm.jpg
22966 F20101124_AADBLU park_h_Page_062.QC.jpg
23175 F20101124_AADBMK park_h_Page_079.QC.jpg
6388 F20101124_AADBLV park_h_Page_062thm.jpg
6404 F20101124_AADBML park_h_Page_079thm.jpg
5748 F20101124_AADBLW park_h_Page_066thm.jpg
21246 F20101124_AADBMM park_h_Page_080.QC.jpg
24264 F20101124_AADBLX park_h_Page_067.QC.jpg
9780 F20101124_AADBNA park_h_Page_090.QC.jpg
6056 F20101124_AADBMN park_h_Page_080thm.jpg
6651 F20101124_AADBLY park_h_Page_067thm.jpg
3104 F20101124_AADBNB park_h_Page_090thm.jpg
22520 F20101124_AADBMO park_h_Page_081.QC.jpg
20180 F20101124_AADBLZ park_h_Page_069.QC.jpg
5980 F20101124_AADBNC park_h_Page_091thm.jpg
6365 F20101124_AADBMP park_h_Page_081thm.jpg
24537 F20101124_AADBND park_h_Page_092.QC.jpg
21326 F20101124_AADBMQ park_h_Page_082.QC.jpg
6697 F20101124_AADBNE park_h_Page_092thm.jpg
6019 F20101124_AADBMR park_h_Page_082thm.jpg
24295 F20101124_AADBNF park_h_Page_093.QC.jpg
5487 F20101124_AADBMS park_h_Page_083thm.jpg
6633 F20101124_AADBNG park_h_Page_093thm.jpg
24493 F20101124_AADBMT park_h_Page_084.QC.jpg
6418 F20101124_AADBNH park_h_Page_094thm.jpg
23379 F20101124_AADBNI park_h_Page_095.QC.jpg
21974 F20101124_AADBMU park_h_Page_085.QC.jpg
6415 F20101124_AADBNJ park_h_Page_095thm.jpg
6217 F20101124_AADBMV park_h_Page_085thm.jpg
21841 F20101124_AADBNK park_h_Page_096.QC.jpg
21543 F20101124_AADBMW park_h_Page_086.QC.jpg
23571 F20101124_AADBNL park_h_Page_097.QC.jpg
24973 F20101124_AADBMX park_h_Page_087.QC.jpg
24620 F20101124_AADBOA park_h_Page_111.QC.jpg
6311 F20101124_AADBNM park_h_Page_097thm.jpg
6829 F20101124_AADBMY park_h_Page_087thm.jpg
6782 F20101124_AADBOB park_h_Page_111thm.jpg
24014 F20101124_AADBNN park_h_Page_098.QC.jpg
21780 F20101124_AADBMZ park_h_Page_089.QC.jpg
22792 F20101124_AADBOC park_h_Page_112.QC.jpg
F20101124_AADBNO park_h_Page_098thm.jpg
3875 F20101124_AADBOD park_h_Page_114thm.jpg
24395 F20101124_AADBNP park_h_Page_099.QC.jpg
18513 F20101124_AADBOE park_h_Page_115.QC.jpg
6670 F20101124_AADBNQ park_h_Page_099thm.jpg
11357 F20101124_AADBOF park_h_Page_116.QC.jpg
23832 F20101124_AADBNR park_h_Page_100.QC.jpg
20963 F20101124_AADBOG park_h_Page_118.QC.jpg
24511 F20101124_AADBNS park_h_Page_101.QC.jpg
21020 F20101124_AADBOH park_h_Page_119.QC.jpg
6454 F20101124_AADBNT park_h_Page_103thm.jpg
21948 F20101124_AADBOI park_h_Page_120.QC.jpg
8500 F20101124_AADBNU park_h_Page_104.QC.jpg
6246 F20101124_AADBOJ park_h_Page_120thm.jpg
6253 F20101124_AADBOK park_h_Page_121thm.jpg
2620 F20101124_AADBNV park_h_Page_104thm.jpg
5831 F20101124_AADBOL park_h_Page_122thm.jpg
22053 F20101124_AADBNW park_h_Page_106.QC.jpg
21649 F20101124_AADBPA park_h_Page_141.QC.jpg
15871 F20101124_AADBOM park_h_Page_124.QC.jpg
23045 F20101124_AADBNX park_h_Page_107.QC.jpg
F20101124_AADBPB park_h_Page_142.QC.jpg
5213 F20101124_AADBON park_h_Page_124thm.jpg
22425 F20101124_AADBNY park_h_Page_108.QC.jpg
25038 F20101124_AADBPC park_h_Page_145.QC.jpg
7904 F20101124_AADBOO park_h_Page_126.QC.jpg
F20101124_AADBNZ park_h_Page_109.QC.jpg
22350 F20101124_AADBPD park_h_Page_146.QC.jpg
2798 F20101124_AADBOP park_h_Page_126thm.jpg
8524 F20101124_AADBOQ park_h_Page_127.QC.jpg
6218 F20101124_AADBPE park_h_Page_146thm.jpg
8797 F20101124_AADBOR park_h_Page_128.QC.jpg
4056 F20101124_AADBPF park_h_Page_147thm.jpg
2894 F20101124_AADBOS park_h_Page_130thm.jpg
107851 F20101124_AADBPG UFE0011351_00001.mets FULL
2667 F20101124_AADBOT park_h_Page_131thm.jpg
2937 F20101124_AADBOU park_h_Page_132thm.jpg
8552 F20101124_AADBOV park_h_Page_133.QC.jpg
9584 F20101124_AADBOW park_h_Page_135.QC.jpg
9532 F20101124_AADBOX park_h_Page_137.QC.jpg
3191 F20101124_AADBOY park_h_Page_138thm.jpg
6549 F20101124_AADBOZ park_h_Page_140.QC.jpg
9665 F20101124_AADANA park_h_Page_138.QC.jpg
22736 F20101124_AADANB park_h_Page_032.QC.jpg
103691 F20101124_AADANC park_h_Page_106.jp2
1053954 F20101124_AADAND park_h_Page_061.tif
156032 F20101124_AADANE park_h_Page_008.jp2
36361 F20101124_AADANF park_h_Page_055.jpg
F20101124_AADANG park_h_Page_082.tif
F20101124_AADANH park_h_Page_102thm.jpg
115844 F20101124_AADANI park_h_Page_041.jp2
F20101124_AADANJ park_h_Page_090.tif
109140 F20101124_AADANK park_h_Page_095.jp2
116270 F20101124_AADAMV park_h_Page_087.jp2
F20101124_AADANL park_h_Page_147.tif
F20101124_AADAMW park_h_Page_102.tif
23662 F20101124_AADAOA park_h_Page_102.QC.jpg
1370 F20101124_AADANM park_h_Page_003thm.jpg
73765 F20101124_AADAMX park_h_Page_099.jpg
10417 F20101124_AADAOB park_h_Page_117.QC.jpg
38714 F20101124_AADANN park_h_Page_010.jpg
7991 F20101124_AADAMY park_h_Page_004.QC.jpg
F20101124_AADAOC park_h_Page_013.tif
72035 F20101124_AADANO park_h_Page_052.jpg
47739 F20101124_AADAMZ park_h_Page_139.jp2
23997 F20101124_AADAOD park_h_Page_023.QC.jpg
18771 F20101124_AADANP park_h_Page_005.QC.jpg
6458 F20101124_AADAOE park_h_Page_074thm.jpg
6838 F20101124_AADANQ park_h_Page_030thm.jpg
3128 F20101124_AADAOF park_h_Page_055thm.jpg
6851 F20101124_AADANR park_h_Page_143thm.jpg
58000 F20101124_AADAOG park_h_Page_105.jpg
39468 F20101124_AADANS park_h_Page_128.jp2
F20101124_AADAOH park_h_Page_081.tif
106305 F20101124_AADANT park_h_Page_032.jp2
109856 F20101124_AADAOI park_h_Page_077.jp2
112001 F20101124_AADANU park_h_Page_100.jp2
5962 F20101124_AADAOJ park_h_Page_086thm.jpg
10079 F20101124_AADANV park_h_Page_129.QC.jpg
F20101124_AADAOK park_h_Page_100.tif
6277 F20101124_AADAOL park_h_Page_112thm.jpg
23590 F20101124_AADANW park_h_Page_018.QC.jpg
93887 F20101124_AADAOM park_h_Page_056.jp2
69664 F20101124_AADANX park_h_Page_075.jpg
74019 F20101124_AADAPA park_h_Page_045.jpg
32529 F20101124_AADAON park_h_Page_129.jpg
14663 F20101124_AADANY park_h_Page_125.QC.jpg
23182 F20101124_AADAPB park_h_Page_059.QC.jpg
3957 F20101124_AADAOO park_h_Page_008.QC.jpg
2974 F20101124_AADANZ park_h_Page_133thm.jpg
70414 F20101124_AADAPC park_h_Page_112.jpg
6190 F20101124_AADAOP park_h_Page_118thm.jpg
23549 F20101124_AADAPD park_h_Page_034.QC.jpg
6200 F20101124_AADAOQ park_h_Page_089thm.jpg
F20101124_AADAPE park_h_Page_062.tif
69822 F20101124_AADAOR park_h_Page_025.jpg
5180 F20101124_AADAPF park_h_Page_068thm.jpg
F20101124_AADAPG park_h_Page_054.tif
118075 F20101124_AADAOS park_h_Page_030.jp2
25271604 F20101124_AADAPH park_h_Page_005.tif
22166 F20101124_AADAOT park_h_Page_031.QC.jpg
6740 F20101124_AADAPI park_h_Page_084thm.jpg
6607 F20101124_AADAOU park_h_Page_017thm.jpg
24115 F20101124_AADAPJ park_h_Page_024.QC.jpg
111097 F20101124_AADAOV park_h_Page_093.jp2
6343 F20101124_AADAPK park_h_Page_032thm.jpg
66199 F20101124_AADAOW park_h_Page_089.jpg
6319 F20101124_AADAPL park_h_Page_107thm.jpg
19062 F20101124_AADAQA park_h_Page_083.QC.jpg
6360 F20101124_AADAPM park_h_Page_039thm.jpg
F20101124_AADAOX park_h_Page_040.tif
23233 F20101124_AADAQB park_h_Page_064.QC.jpg
24268 F20101124_AADAPN park_h_Page_022.QC.jpg
F20101124_AADAOY park_h_Page_042.tif
8011 F20101124_AADAQC park_h_Page_134.QC.jpg
22706 F20101124_AADAPO park_h_Page_048.QC.jpg
F20101124_AADAOZ park_h_Page_103.tif
38129 F20101124_AADAQD park_h_Page_114.jpg
66647 F20101124_AADAPP park_h_Page_086.jpg
63597 F20101124_AADAQE park_h_Page_125.jp2
1051979 F20101124_AADAPQ park_h_Page_006.jp2
74352 F20101124_AADAQF park_h_Page_060.jpg
6249 F20101124_AADAPR park_h_Page_075thm.jpg
71907 F20101124_AADAQG park_h_Page_061.jpg
77797 F20101124_AADAPS park_h_Page_142.jpg
3111 F20101124_AADAQH park_h_Page_137thm.jpg
70196 F20101124_AADAPT park_h_Page_009.jpg
6787 F20101124_AADAQI park_h_Page_042thm.jpg
63214 F20101124_AADAPU park_h_Page_069.jpg
6611 F20101124_AADAQJ park_h_Page_045thm.jpg
70120 F20101124_AADAPV park_h_Page_038.jpg
F20101124_AADAQK park_h_Page_144thm.jpg
71593 F20101124_AADAPW park_h_Page_079.jpg
F20101124_AADAQL park_h_Page_095.tif
15928 F20101124_AADAPX park_h_Page_123.QC.jpg
93530 F20101124_AADAQM park_h_Page_069.jp2
F20101124_AADARA park_h_Page_074.tif
23913 F20101124_AADAQN park_h_Page_046.QC.jpg
22406 F20101124_AADAPY park_h_Page_071.QC.jpg
F20101124_AADARB park_h_Page_023.tif
8366 F20101124_AADAQO park_h_Page_130.QC.jpg
2709 F20101124_AADAPZ park_h_Page_134thm.jpg
6901 F20101124_AADARC park_h_Page_033thm.jpg
F20101124_AADAQP park_h_Page_113.tif
24448 F20101124_AADARD park_h_Page_050.QC.jpg
22913 F20101124_AADAQQ park_h_Page_063.QC.jpg
76475 F20101124_AADARE park_h_Page_088.jp2
F20101124_AADAQR park_h_Page_088.tif
102662 F20101124_AADARF park_h_Page_120.jp2
93091 F20101124_AADAQS park_h_Page_089.jp2
6258 F20101124_AADARG park_h_Page_014thm.jpg
23492 F20101124_AADAQT park_h_Page_051.QC.jpg
23785 F20101124_AADARH park_h_Page_019.QC.jpg
84409 F20101124_AADAQU park_h_Page_007.jpg
109391 F20101124_AADARI park_h_Page_014.jp2
F20101124_AADAQV park_h_Page_055.tif
74441 F20101124_AADARJ park_h_Page_092.jpg
6140 F20101124_AADAQW park_h_Page_096thm.jpg
3674 F20101124_AADARK park_h_Page_116thm.jpg
12191 F20101124_AADAQX park_h_Page_114.QC.jpg
F20101124_AADARL park_h_Page_018.tif
F20101124_AADAQY park_h_Page_115.tif
73331 F20101124_AADASA park_h_Page_098.jpg
108808 F20101124_AADARM park_h_Page_048.jp2
98721 F20101124_AADASB park_h_Page_013.jp2
23815 F20101124_AADARN park_h_Page_026.QC.jpg
65114 F20101124_AADAQZ park_h_Page_013.jpg
F20101124_AADASC park_h_Page_010.tif
F20101124_AADARO park_h_Page_067.tif
68271 F20101124_AADASD park_h_Page_110.jpg
70430 F20101124_AADARP park_h_Page_017.jpg
4565 F20101124_AADASE park_h_Page_123thm.jpg
99785 F20101124_AADARQ park_h_Page_075.jp2
23608 F20101124_AADASF park_h_Page_094.QC.jpg
22595 F20101124_AADARR park_h_Page_103.QC.jpg
22919 F20101124_AADASG park_h_Page_049.QC.jpg
102828 F20101124_AADARS park_h_Page_096.jp2
104474 F20101124_AADASH park_h_Page_047.jp2
106581 F20101124_AADART park_h_Page_020.jp2
122388 F20101124_AADASI park_h_Page_142.jp2
9281 F20101124_AADARU park_h_Page_136.QC.jpg
112405 F20101124_AADASJ park_h_Page_027.jp2
6700 F20101124_AADARV park_h_Page_019thm.jpg
6441 F20101124_AADASK park_h_Page_048thm.jpg
109014 F20101124_AADARW park_h_Page_017.jp2
F20101124_AADASL park_h_Page_001.tif
3200 F20101124_AADARX park_h_Page_139thm.jpg
5806 F20101124_AADATA park_h_Page_141thm.jpg
6156 F20101124_AADASM park_h_Page_028thm.jpg
17600 F20101124_AADARY park_h_Page_068.QC.jpg
26045 F20101124_AADATB park_h_Page_144.QC.jpg
F20101124_AADASN park_h_Page_032.tif
32634 F20101124_AADARZ park_h_Page_138.jpg
100561 F20101124_AADATC park_h_Page_082.jp2
F20101124_AADASO park_h_Page_121.tif
4855 F20101124_AADATD park_h_Page_003.jp2
28043 F20101124_AADASP park_h_Page_130.jpg
6738 F20101124_AADATE park_h_Page_101thm.jpg
24191 F20101124_AADASQ park_h_Page_045.QC.jpg
65623 F20101124_AADATF park_h_Page_054.jpg
72671 F20101124_AADASR park_h_Page_093.jpg
F20101124_AADATG park_h_Page_085.tif
F20101124_AADASS park_h_Page_033.tif
F20101124_AADATH park_h_Page_028.tif
70788 F20101124_AADAST park_h_Page_049.jpg
1360 F20101124_AADATI park_h_Page_002thm.jpg
110247 F20101124_AADASU park_h_Page_051.jp2
21890 F20101124_AADATJ park_h_Page_121.QC.jpg
6669 F20101124_AADASV park_h_Page_059thm.jpg
22314 F20101124_AADATK park_h_Page_110.QC.jpg
23498 F20101124_AADASW park_h_Page_044.QC.jpg
6490 F20101124_AADATL park_h_Page_040thm.jpg
6153 F20101124_AADASX park_h_Page_110thm.jpg
F20101124_AADATM park_h_Page_053.tif
6262 F20101124_AADASY park_h_Page_142thm.jpg
13384 F20101124_AADAUA park_h_Page_147.QC.jpg
49025 F20101124_AADATN park_h_Page_113.jpg
19340 F20101124_AADASZ park_h_Page_066.QC.jpg
40484 F20101124_AADAUB park_h_Page_147.jpg
F20101124_AADATO park_h_Page_076thm.jpg
3112 F20101124_AADAUC park_h_Page_135thm.jpg
3208 F20101124_AADAUD park_h_Page_010thm.jpg
34266 F20101124_AADATP park_h_Page_139.jpg
71432 F20101124_AADAUE park_h_Page_097.jpg
F20101124_AADATQ park_h_Page_038.tif
74215 F20101124_AADBAA park_h_Page_022.jpg
23948 F20101124_AADAUF park_h_Page_027.QC.jpg
6625 F20101124_AADATR park_h_Page_100thm.jpg
73333 F20101124_AADBAB park_h_Page_023.jpg
111313 F20101124_AADAUG park_h_Page_046.jp2
22652 F20101124_AADATS park_h_Page_075.QC.jpg
74004 F20101124_AADBAC park_h_Page_024.jpg
69200 F20101124_AADAUH park_h_Page_032.jpg
20619 F20101124_AADATT park_h_Page_122.QC.jpg
73324 F20101124_AADBAD park_h_Page_026.jpg
72850 F20101124_AADAUI park_h_Page_113.jp2
F20101124_AADATU park_h_Page_094.tif
73314 F20101124_AADBAE park_h_Page_027.jpg
4769 F20101124_AADAUJ park_h_Page_113thm.jpg
F20101124_AADATV park_h_Page_051.tif
67485 F20101124_AADBAF park_h_Page_028.jpg
112691 F20101124_AADAUK park_h_Page_111.jp2
6690 F20101124_AADATW park_h_Page_145thm.jpg
70146 F20101124_AADBAG park_h_Page_029.jpg
4423 F20101124_AADAUL park_h_Page_125thm.jpg
10023 F20101124_AADATX park_h_Page_139.QC.jpg
77172 F20101124_AADBAH park_h_Page_030.jpg
6337 F20101124_AADAVA park_h_Page_020thm.jpg
68534 F20101124_AADAUM park_h_Page_047.jpg
6291 F20101124_AADATY park_h_Page_109thm.jpg
55118 F20101124_AADAVB park_h_Page_068.jpg
16405 F20101124_AADAUN park_h_Page_113.QC.jpg
F20101124_AADATZ park_h_Page_142.tif
69760 F20101124_AADBAI park_h_Page_031.jpg
22802 F20101124_AADAVC park_h_Page_020.QC.jpg
99066 F20101124_AADAUO park_h_Page_109.jp2
76791 F20101124_AADBAJ park_h_Page_033.jpg
105651 F20101124_AADAVD park_h_Page_006.jpg
2947 F20101124_AADAUP park_h_Page_136thm.jpg
72600 F20101124_AADBAK park_h_Page_034.jpg
70819 F20101124_AADAVE park_h_Page_063.jpg
43555 F20101124_AADAUQ park_h_Page_135.jp2
70476 F20101124_AADBAL park_h_Page_035.jpg
6482 F20101124_AADAVF park_h_Page_064thm.jpg
22511 F20101124_AADAUR park_h_Page_025.QC.jpg
72981 F20101124_AADBBA park_h_Page_065.jpg
86050 F20101124_AADAVG park_h_Page_105.jp2
110738 F20101124_AADAUS park_h_Page_034.jp2
61040 F20101124_AADBBB park_h_Page_066.jpg
71060 F20101124_AADBAM park_h_Page_037.jpg
74579 F20101124_AADAVH park_h_Page_101.jpg
3350 F20101124_AADAUT park_h_Page_117thm.jpg
71420 F20101124_AADBBC park_h_Page_072.jpg
72277 F20101124_AADBAN park_h_Page_039.jpg
F20101124_AADAVI park_h_Page_106.tif
65849 F20101124_AADAUU park_h_Page_123.jp2
57559 F20101124_AADBBD park_h_Page_073.jpg
73634 F20101124_AADBAO park_h_Page_040.jpg
23495 F20101124_AADAVJ park_h_Page_072.QC.jpg
46373 F20101124_AADAUV park_h_Page_123.jpg
73105 F20101124_AADBBE park_h_Page_074.jpg
75328 F20101124_AADBAP park_h_Page_041.jpg
24355 F20101124_AADAVK park_h_Page_042.QC.jpg
44278 F20101124_AADAUW park_h_Page_012.jpg
68609 F20101124_AADBBF park_h_Page_078.jpg
74812 F20101124_AADBAQ park_h_Page_042.jpg
10621 F20101124_AADAVL park_h_Page_055.QC.jpg
112758 F20101124_AADAUX park_h_Page_026.jp2
65652 F20101124_AADBBG park_h_Page_080.jpg
69227 F20101124_AADBAR park_h_Page_043.jpg
114274 F20101124_AADAVM park_h_Page_042.jp2
F20101124_AADAUY park_h_Page_145.tif
65588 F20101124_AADBBH park_h_Page_082.jpg
F20101124_AADAWA park_h_Page_139.tif
72774 F20101124_AADBAS park_h_Page_046.jpg
F20101124_AADAVN park_h_Page_048.tif
23513 F20101124_AADAUZ park_h_Page_057.QC.jpg
76272 F20101124_AADBBI park_h_Page_084.jpg
6158 F20101124_AADAWB park_h_Page_106thm.jpg
69741 F20101124_AADBAT park_h_Page_048.jpg
20924 F20101124_AADAVO park_h_Page_091.QC.jpg
45224 F20101124_AADAWC park_h_Page_138.jp2
75358 F20101124_AADBAU park_h_Page_050.jpg
F20101124_AADAVP park_h_Page_003.tif
75974 F20101124_AADBBJ park_h_Page_087.jpg
6422 F20101124_AADAWD park_h_Page_072thm.jpg
71746 F20101124_AADBAV park_h_Page_051.jpg
46958 F20101124_AADAVQ park_h_Page_116.jp2
55718 F20101124_AADBBK park_h_Page_088.jpg
71587 F20101124_AADAWE park_h_Page_021.jpg
64625 F20101124_AADBAW park_h_Page_056.jpg
F20101124_AADAVR park_h_Page_075.tif
65414 F20101124_AADBCA park_h_Page_118.jpg
29950 F20101124_AADBBL park_h_Page_090.jpg
88850 F20101124_AADAWF park_h_Page_115.jp2
69477 F20101124_AADBAX park_h_Page_057.jpg
51468 F20101124_AADAVS park_h_Page_124.jpg
66857 F20101124_AADBCB park_h_Page_120.jpg
64109 F20101124_AADBBM park_h_Page_091.jpg
F20101124_AADAWG park_h_Page_141.tif
53948 F20101124_AADBAY park_h_Page_058.jpg
41354 F20101124_AADAVT park_h_Page_136.jp2
67827 F20101124_AADBCC park_h_Page_121.jpg
72108 F20101124_AADBBN park_h_Page_094.jpg
70936 F20101124_AADAWH park_h_Page_053.jpg
71090 F20101124_AADBAZ park_h_Page_064.jpg
6598 F20101124_AADAVU park_h_Page_023thm.jpg
63316 F20101124_AADBCD park_h_Page_122.jpg
71872 F20101124_AADBBO park_h_Page_095.jpg
5777 F20101124_AADAWI park_h_Page_009thm.jpg
7921 F20101124_AADAVV park_h_Page_131.QC.jpg
44214 F20101124_AADBCE park_h_Page_125.jpg
67584 F20101124_AADBBP park_h_Page_096.jpg
F20101124_AADAWJ park_h_Page_133.tif
24188 F20101124_AADAVW park_h_Page_060.QC.jpg
28156 F20101124_AADBCF park_h_Page_127.jpg
73596 F20101124_AADBBQ park_h_Page_100.jpg
95121 F20101124_AADAWK park_h_Page_122.jp2
5150 F20101124_AADAVX park_h_Page_088thm.jpg
25878 F20101124_AADBCG park_h_Page_131.jpg
72406 F20101124_AADBBR park_h_Page_102.jpg
71569 F20101124_AADAWL park_h_Page_059.jpg
78251 F20101124_AADAVY park_h_Page_015.jp2
28006 F20101124_AADBCH park_h_Page_133.jpg
69388 F20101124_AADAXA park_h_Page_103.jpg
25399 F20101124_AADBBS park_h_Page_104.jpg
65597 F20101124_AADAWM park_h_Page_119.jpg
108897 F20101124_AADAVZ park_h_Page_107.jp2
26684 F20101124_AADBCI park_h_Page_134.jpg
36669 F20101124_AADAXB park_h_Page_127.jp2
68120 F20101124_AADBBT park_h_Page_106.jpg
48327 F20101124_AADAWN park_h_Page_114.jp2
29742 F20101124_AADBCJ park_h_Page_136.jpg
73282 F20101124_AADAXC park_h_Page_019.jpg
70254 F20101124_AADBBU park_h_Page_107.jpg
37233 F20101124_AADAWO park_h_Page_130.jp2
F20101124_AADAXD park_h_Page_007.jp2
69188 F20101124_AADBBV park_h_Page_108.jpg
32553 F20101124_AADAWP park_h_Page_104.jp2
32150 F20101124_AADBCK park_h_Page_137.jpg
115114 F20101124_AADAXE park_h_Page_036.jp2
64472 F20101124_AADBBW park_h_Page_109.jpg
22092 F20101124_AADAWQ park_h_Page_043.QC.jpg
22142 F20101124_AADBCL park_h_Page_140.jpg
6006 F20101124_AADAXF park_h_Page_119thm.jpg
75008 F20101124_AADBBX park_h_Page_111.jpg
4820 F20101124_AADAWR park_h_Page_015thm.jpg
113692 F20101124_AADBDA park_h_Page_019.jp2
71706 F20101124_AADBCM park_h_Page_141.jpg
F20101124_AADAXG park_h_Page_030.tif
34521 F20101124_AADBBY park_h_Page_116.jpg
23105 F20101124_AADAWS park_h_Page_076.QC.jpg
110912 F20101124_AADBDB park_h_Page_021.jp2
87705 F20101124_AADBCN park_h_Page_143.jpg
3027 F20101124_AADAXH park_h_Page_003.QC.jpg
32007 F20101124_AADBBZ park_h_Page_117.jpg
18818 F20101124_AADAWT park_h_Page_073.QC.jpg
114527 F20101124_AADBDC park_h_Page_022.jp2
86128 F20101124_AADBCO park_h_Page_144.jpg
23795 F20101124_AADAXI park_h_Page_065.QC.jpg
111350 F20101124_AADAWU park_h_Page_141.jp2
111945 F20101124_AADBDD park_h_Page_023.jp2
85085 F20101124_AADBCP park_h_Page_145.jpg
31134 F20101124_AADAXJ park_h_Page_135.jpg
22807 F20101124_AADAWV park_h_Page_017.QC.jpg
115287 F20101124_AADBDE park_h_Page_024.jp2
72560 F20101124_AADBCQ park_h_Page_146.jpg
71097 F20101124_AADAXK park_h_Page_071.jpg
F20101124_AADAWW park_h_Page_009.tif
105850 F20101124_AADBDF park_h_Page_025.jp2
24766 F20101124_AADAYA park_h_Page_126.jpg
26501 F20101124_AADBCR park_h_Page_001.jp2
72015 F20101124_AADAXL park_h_Page_077.jpg
29017 F20101124_AADAWX park_h_Page_128.jpg
103388 F20101124_AADBDG park_h_Page_028.jp2
5393 F20101124_AADBCS park_h_Page_002.jp2
72198 F20101124_AADAXM park_h_Page_044.jpg
18394 F20101124_AADAWY park_h_Page_088.QC.jpg
105958 F20101124_AADBDH park_h_Page_029.jp2
3422 F20101124_AADAYB park_h_Page_129thm.jpg
28951 F20101124_AADBCT park_h_Page_004.jp2
68743 F20101124_AADAXN park_h_Page_085.jpg
8623 F20101124_AADAWZ park_h_Page_132.QC.jpg
106323 F20101124_AADBDI park_h_Page_031.jp2
104742 F20101124_AADAYC park_h_Page_061.jp2
1051980 F20101124_AADBCU park_h_Page_009.jp2
23302 F20101124_AADAXO park_h_Page_021.QC.jpg
115311 F20101124_AADBDJ park_h_Page_033.jp2
18767 F20101124_AADAYD park_h_Page_105.QC.jpg
987566 F20101124_AADBCV park_h_Page_010.jp2
F20101124_AADAXP park_h_Page_065thm.jpg
105766 F20101124_AADBDK park_h_Page_035.jp2
28366 F20101124_AADAYE park_h_Page_132.jpg
90949 F20101124_AADBCW park_h_Page_011.jp2
104097 F20101124_AADAXQ park_h_Page_081.jp2
23285 F20101124_AADAYF park_h_Page_053.QC.jpg
63368 F20101124_AADBCX park_h_Page_012.jp2
9835 F20101124_AADAXR park_h_Page_003.jpg
107413 F20101124_AADBEA park_h_Page_062.jp2
109000 F20101124_AADBDL park_h_Page_037.jp2
2793 F20101124_AADAYG park_h_Page_127thm.jpg
93955 F20101124_AADBCY park_h_Page_016.jp2
2976 F20101124_AADAXS park_h_Page_128thm.jpg
106754 F20101124_AADBEB park_h_Page_063.jp2
107906 F20101124_AADBDM park_h_Page_038.jp2
22872 F20101124_AADAYH park_h_Page_035.QC.jpg
112556 F20101124_AADBCZ park_h_Page_018.jp2
74838 F20101124_AADAXT park_h_Page_067.jpg
108777 F20101124_AADBEC park_h_Page_064.jp2
111573 F20101124_AADBDN park_h_Page_039.jp2
F20101124_AADAYI park_h_Page_092.tif
2318 F20101124_AADAXU park_h_Page_140thm.jpg
112827 F20101124_AADBED park_h_Page_065.jp2
106396 F20101124_AADBDO park_h_Page_043.jp2
1051969 F20101124_AADAYJ park_h_Page_005.jp2
F20101124_AADAXV park_h_Page_059.tif
89198 F20101124_AADBEE park_h_Page_066.jp2
110870 F20101124_AADBDP park_h_Page_044.jp2
57208 F20101124_AADAYK park_h_Page_083.jpg
F20101124_AADAXW park_h_Page_089.tif
114773 F20101124_AADBEF park_h_Page_067.jp2
107373 F20101124_AADBDQ park_h_Page_049.jp2
70705 F20101124_AADAYL park_h_Page_062.jpg
61302 F20101124_AADAXX park_h_Page_016.jpg
82167 F20101124_AADBEG park_h_Page_068.jp2
23550 F20101124_AADAZA park_h_Page_039.QC.jpg
115323 F20101124_AADBDR park_h_Page_050.jp2
F20101124_AADAXY park_h_Page_037.tif
80121 F20101124_AADBEH park_h_Page_070.jp2
F20101124_AADAZB park_h_Page_111.tif
110489 F20101124_AADBDS park_h_Page_052.jp2
F20101124_AADAYM park_h_Page_114.tif
F20101124_AADAXZ park_h_Page_018thm.jpg
105428 F20101124_AADBEI park_h_Page_071.jp2
109731 F20101124_AADBDT park_h_Page_053.jp2
6227 F20101124_AADAYN park_h_Page_025thm.jpg
110057 F20101124_AADBEJ park_h_Page_072.jp2
25393 F20101124_AADAZC park_h_Page_143.QC.jpg
98799 F20101124_AADBDU park_h_Page_054.jp2
6523 F20101124_AADAYO park_h_Page_063thm.jpg
75191 F20101124_AADBEK park_h_Page_073.jp2
74544 F20101124_AADAZD park_h_Page_036.jpg
50176 F20101124_AADBDV park_h_Page_055.jp2
5699 F20101124_AADAYP park_h_Page_016thm.jpg
110165 F20101124_AADBEL park_h_Page_074.jp2
71362 F20101124_AADAZE park_h_Page_076.jpg
102359 F20101124_AADBDW park_h_Page_057.jp2
5639 F20101124_AADAYQ park_h_Page_105thm.jpg
114250 F20101124_AADBFA park_h_Page_101.jp2
6612 F20101124_AADAZF park_h_Page_041thm.jpg
78445 F20101124_AADBDX park_h_Page_058.jp2
F20101124_AADAYR park_h_Page_047.tif
108014 F20101124_AADBFB park_h_Page_102.jp2
99199 F20101124_AADBEM park_h_Page_078.jp2
109464 F20101124_AADAZG park_h_Page_076.jp2
106474 F20101124_AADBDY park_h_Page_059.jp2
6040 F20101124_AADAYS park_h_Page_013thm.jpg
103809 F20101124_AADBFC park_h_Page_103.jp2
106070 F20101124_AADBEN park_h_Page_079.jp2
133500 F20101124_AADAZH park_h_Page_145.jp2
112557 F20101124_AADBDZ park_h_Page_060.jp2
113125 F20101124_AADAYT park_h_Page_045.jp2
106312 F20101124_AADBFD park_h_Page_108.jp2
94638 F20101124_AADBEO park_h_Page_080.jp2
5384 F20101124_AADAZI park_h_Page_115thm.jpg
F20101124_AADAYU park_h_Page_084.tif
104104 F20101124_AADBFE park_h_Page_110.jp2
79864 F20101124_AADBEP park_h_Page_083.jp2
70157 F20101124_AADAZJ park_h_Page_081.jpg
141473 F20101124_AADAYV park_h_Page_144.jp2
106133 F20101124_AADBFF park_h_Page_112.jp2
115983 F20101124_AADBEQ park_h_Page_084.jp2
F20101124_AADAZK park_h_Page_057.tif
62849 F20101124_AADAYW park_h_Page_115.jpg
44610 F20101124_AADBFG park_h_Page_117.jp2
102353 F20101124_AADBER park_h_Page_085.jp2
72648 F20101124_AADAZL park_h_Page_005.jpg
6244 F20101124_AADAYX park_h_Page_035thm.jpg
97952 F20101124_AADBFH park_h_Page_118.jp2
95878 F20101124_AADBES park_h_Page_086.jp2
F20101124_AADAZM park_h_Page_143.tif
56641 F20101124_AADAYY park_h_Page_070.jpg
98090 F20101124_AADBFI park_h_Page_119.jp2
39324 F20101124_AADBET park_h_Page_090.jp2
113521 F20101124_AADAZN park_h_Page_040.jp2
6209 F20101124_AADAYZ park_h_Page_108thm.jpg
101558 F20101124_AADBFJ park_h_Page_121.jp2
97643 F20101124_AADBEU park_h_Page_091.jp2
157317 F20101124_AADAZO UFE0011351_00001.xml
112722 F20101124_AADBEV park_h_Page_092.jp2
75770 F20101124_AADBFK park_h_Page_124.jp2
108969 F20101124_AADBEW park_h_Page_094.jp2
31986 F20101124_AADBFL park_h_Page_126.jp2
108473 F20101124_AADBEX park_h_Page_097.jp2
25439 F20101124_AADAZR park_h_Page_001.jpg
F20101124_AADBGA park_h_Page_008.tif
45593 F20101124_AADBFM park_h_Page_129.jp2
110267 F20101124_AADBEY park_h_Page_098.jp2
10167 F20101124_AADAZS park_h_Page_002.jpg
F20101124_AADBGB park_h_Page_011.tif
113465 F20101124_AADBEZ park_h_Page_099.jp2
24124 F20101124_AADAZT park_h_Page_004.jpg
F20101124_AADBGC park_h_Page_012.tif
34889 F20101124_AADBFN park_h_Page_131.jp2
13162 F20101124_AADAZU park_h_Page_008.jpg
F20101124_AADBGD park_h_Page_014.tif
38397 F20101124_AADBFO park_h_Page_132.jp2
63551 F20101124_AADAZV park_h_Page_011.jpg
F20101124_AADBGE park_h_Page_015.tif
36749 F20101124_AADBFP park_h_Page_133.jp2
71328 F20101124_AADAZW park_h_Page_014.jpg
F20101124_AADBGF park_h_Page_016.tif
35794 F20101124_AADBFQ park_h_Page_134.jp2
51579 F20101124_AADAZX park_h_Page_015.jpg
F20101124_AADBGG park_h_Page_017.tif
45379 F20101124_AADBFR park_h_Page_137.jp2
73505 F20101124_AADAZY park_h_Page_018.jpg
F20101124_AADBGH park_h_Page_019.tif
26129 F20101124_AADBFS park_h_Page_140.jp2
69558 F20101124_AADAZZ park_h_Page_020.jpg
F20101124_AADBGI park_h_Page_020.tif
132747 F20101124_AADBFT park_h_Page_143.jp2
F20101124_AADBGJ park_h_Page_021.tif
119256 F20101124_AADBFU park_h_Page_146.jp2
F20101124_AADBGK park_h_Page_022.tif
56782 F20101124_AADBFV park_h_Page_147.jp2
F20101124_AADBGL park_h_Page_024.tif
F20101124_AADBFW park_h_Page_002.tif
F20101124_AADBHA park_h_Page_049.tif
F20101124_AADBGM park_h_Page_025.tif
F20101124_AADBFX park_h_Page_004.tif
F20101124_AADBHB park_h_Page_050.tif
F20101124_AADBGN park_h_Page_026.tif
F20101124_AADBFY park_h_Page_006.tif
F20101124_AADBHC park_h_Page_052.tif
F20101124_AADBFZ park_h_Page_007.tif
F20101124_AADBHD park_h_Page_056.tif
F20101124_AADBGO park_h_Page_027.tif
F20101124_AADBHE park_h_Page_058.tif
F20101124_AADBGP park_h_Page_029.tif
F20101124_AADBHF park_h_Page_060.tif
F20101124_AADBGQ park_h_Page_031.tif
F20101124_AADBHG park_h_Page_063.tif
F20101124_AADBGR park_h_Page_034.tif
F20101124_AADBHH park_h_Page_064.tif
F20101124_AADBGS park_h_Page_035.tif
F20101124_AADBHI park_h_Page_065.tif
F20101124_AADBGT park_h_Page_036.tif
F20101124_AADBHJ park_h_Page_066.tif
F20101124_AADBGU park_h_Page_039.tif
F20101124_AADBHK park_h_Page_068.tif
F20101124_AADBGV park_h_Page_041.tif
F20101124_AADBHL park_h_Page_069.tif
F20101124_AADBGW park_h_Page_043.tif
F20101124_AADBHM park_h_Page_070.tif
F20101124_AADBGX park_h_Page_044.tif
F20101124_AADBIA park_h_Page_096.tif
F20101124_AADBHN park_h_Page_071.tif
F20101124_AADBGY park_h_Page_045.tif
F20101124_AADBIB park_h_Page_097.tif
F20101124_AADBHO park_h_Page_072.tif
F20101124_AADBGZ park_h_Page_046.tif
F20101124_AADBIC park_h_Page_098.tif
F20101124_AADBID park_h_Page_099.tif
F20101124_AADBHP park_h_Page_073.tif
F20101124_AADBIE park_h_Page_101.tif
F20101124_AADBHQ park_h_Page_076.tif
F20101124_AADBIF park_h_Page_104.tif
F20101124_AADBHR park_h_Page_077.tif
F20101124_AADBIG park_h_Page_105.tif
F20101124_AADBHS park_h_Page_078.tif
F20101124_AADBIH park_h_Page_107.tif
F20101124_AADBHT park_h_Page_079.tif
F20101124_AADBII park_h_Page_108.tif
F20101124_AADBHU park_h_Page_080.tif
F20101124_AADBIJ park_h_Page_109.tif
F20101124_AADBHV park_h_Page_083.tif
F20101124_AADBIK park_h_Page_110.tif
F20101124_AADBHW park_h_Page_086.tif
F20101124_AADBIL park_h_Page_112.tif
F20101124_AADBHX park_h_Page_087.tif
F20101124_AADBJA park_h_Page_131.tif
F20101124_AADBIM park_h_Page_116.tif
F20101124_AADBHY park_h_Page_091.tif
F20101124_AADBJB park_h_Page_132.tif
F20101124_AADBIN park_h_Page_117.tif
F20101124_AADBHZ park_h_Page_093.tif
F20101124_AADBJC park_h_Page_134.tif
F20101124_AADBIO park_h_Page_118.tif
F20101124_AADBJD park_h_Page_135.tif
F20101124_AADBIP park_h_Page_119.tif
F20101124_AADBJE park_h_Page_136.tif
F20101124_AADBJF park_h_Page_137.tif
F20101124_AADBIQ park_h_Page_120.tif
F20101124_AADBJG park_h_Page_138.tif
F20101124_AADBIR park_h_Page_122.tif
F20101124_AADBJH park_h_Page_140.tif
F20101124_AADBIS park_h_Page_123.tif
F20101124_AADBJI park_h_Page_144.tif
F20101124_AADBIT park_h_Page_124.tif
F20101124_AADBJJ park_h_Page_146.tif
F20101124_AADBIU park_h_Page_125.tif
2581 F20101124_AADBJK park_h_Page_001thm.jpg
F20101124_AADBIV park_h_Page_126.tif
325822 F20101124_AADBJL park_h.pdf
F20101124_AADBIW park_h_Page_127.tif
4121 F20101124_AADBKA park_h_Page_012thm.jpg
7701 F20101124_AADBJM park_h_Page_001.QC.jpg
F20101124_AADBIX park_h_Page_128.tif
21268 F20101124_AADBKB park_h_Page_013.QC.jpg
3233 F20101124_AADBJN park_h_Page_002.QC.jpg
F20101124_AADBIY park_h_Page_129.tif
22958 F20101124_AADBKC park_h_Page_014.QC.jpg
2599 F20101124_AADBJO park_h_Page_004thm.jpg
F20101124_AADBIZ park_h_Page_130.tif
16497 F20101124_AADBKD park_h_Page_015.QC.jpg
4980 F20101124_AADBJP park_h_Page_005thm.jpg




Copyright 2005 by Heenam Park


To my mother, Jungrye Yu


iv ACKNOWLEDGMENTS I would like to thank Dr. Ratree Wayla nd and Dr. Theresa A. Antes for their insights and comments, which contributed imme asurably to this study. I would also like to thank the students who participated in th is study and the instructors at the English Language Institute who allowed me to come to their classes to recruit participants. Finally, I would like to thank Dr. Caroline Wiltshire and Dr. Joaquim Camps for their comments.


v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION........................................................................................................1 2 REVIEW OF THE LITERATURE..............................................................................4 Effects of Text Adjuncts...............................................................................................4 Types of Reading Processes..................................................................................4 Discussion of Schema Theory...............................................................................7 Different Types of Text Adjuncts........................................................................12 Reading Comprehension Measurements.............................................................17 Interaction of Text Adjuncts, Text Types and L2 Learners................................18 Learners’ Proficiency in L2 Reading..................................................................21 Reading Strategy Studies.....................................................................................24 Foreign Language Reading Anxiety...........................................................................26 Types of Anxiety.................................................................................................27 General Discussion of L2 Reading A nxiety versus General L2 Anxiety............30 Anxiety Measures................................................................................................36 Types of measures and limits of these measures..........................................36 Foreign Language Classroom Anxiet y Scale (FLCAS)/Foreign Language Reading Anxiety Scale (FLRAS)...........................................................39 Summary.....................................................................................................................41 Significance of Current Study....................................................................................41 Statement of Purpose..................................................................................................42 Research Questions and Hypotheses..........................................................................42 Research Questions.............................................................................................42 Hypotheses..........................................................................................................43 3 METHODOLOGY.....................................................................................................44 Participants.................................................................................................................44 Materials.....................................................................................................................4 6


vi English Proficiency Test......................................................................................45 Background Questionnaire..................................................................................47 Reading Texts......................................................................................................50 Procedure....................................................................................................................50 Scoring........................................................................................................................ 53 Statistical Analyses.....................................................................................................56 4 RESULTS...................................................................................................................57 Results of Test of Adolescent and Adult Language (TOAL-3)..................................57 Standard Score for TOAL-3................................................................................58 Standard Scores for TOAL-3 by Subgroups.......................................................60 Recall Protocols..........................................................................................................62 Effects of Text Structures according to Proficiency Level.................................66 Relationship between Proficiency and Recall Scores.........................................67 Varying Effects of Text Adjuncts according to Text Structure...........................68 Relationship between Reading Prof iciency and Reading Ability.......................71 Summary.....................................................................................................................72 Background Information Questionnaire.....................................................................73 Relationship between FLRAS and Other Factors...............................................73 Relationship between Reading Strategy Questionnaire (RSQ) and Proficiency Level................................................................................................................75 Correlation between L1 reading abilit y and L2 reading score on TOAL-3.77 Correlation between L1 reading ability and L2 recall score........................77 Correlation between Length of Resi dence (LOR) and Other Factors.................78 5 DISCUSSION.............................................................................................................79 Interrelationship between L2 Pr oficiency and Text Adjunct......................................80 Interaction between Text Type and Text Adjunct within Each Proficiency Level....81 Role of English Proficiency in L2 Reading................................................................85 Relationship between ESL Learners’ Anxiet y Levels and their Proficiency Levels..86 Relationship between Students’ An xiety Level and Reading Ability........................87 Relationship between ESL Learners’ R eading Strategies and their Reading Comprehension Ability..........................................................................................88 Relationship between Reading St rategies and Proficiency........................................89 Correlation between LOR and Other Factors.............................................................90 Relationship between LOR and Students’ Anxiety Level...................................90 Relationship between LOR and St udents’ Proficiency Level.............................90 Relationship between LOR and Reading Ability................................................91 Interviews with Participants.......................................................................................91 6 CONCLUSION...........................................................................................................93 Principal Findings.......................................................................................................93 Pedagogical Implications............................................................................................97


vii Research Implications.................................................................................................98 Limitations of Current Study......................................................................................99 APPENIX A BACKGROUND QUESTIONNAIRE.....................................................................102 Part I Language Background....................................................................................102 Part II........................................................................................................................ 103 Part III....................................................................................................................... 103 B SUMMARY OF TEXTS..........................................................................................105 The Olympic Games.................................................................................................106 On Being Fat in America..........................................................................................107 Supertankers.............................................................................................................108 Closing Down Nuclear Power Plants.......................................................................109 Nuclear Energy versus Solar Energy........................................................................110 C TEXT ADJUNCTS...................................................................................................111 The Olympic Games (High and Low)......................................................................111 Vocabulary Group.............................................................................................111 Expanded Framework Group............................................................................111 On Being Fat in America (Low)...............................................................................111 Vocabulary Group.............................................................................................111 Expanded Framework Group............................................................................112 Supertankers (High)..................................................................................................112 Vocabulary Group.............................................................................................112 Expanded Framework Group............................................................................112 Closing Down Nuclear Power Plants (High)............................................................112 Vocabulary Group.............................................................................................112 Expanded Framework Group............................................................................113 Nuclear Energy versus Solar Energy (Low).............................................................113 Vocabulary Group.............................................................................................113 Expanded Framework Group............................................................................113 D RECALL...................................................................................................................114 E SCORING TEMPLATE...........................................................................................115 The Olympic Games (Idea Units: 100).....................................................................115 On Being Fat in America (Idea Units: 142)..............................................................117 Supertankers (Idea Units: 104).................................................................................121 Closing Down Nuclear Powe r Plants (Idea Units: 98).............................................123 Nuclear Energy versus Sola r Energy (Idea Units: 91)..............................................126


viii LIST OF REFERENCES.................................................................................................129 BIOGRAPHICAL SKETCH...........................................................................................135


ix LIST OF TABLES Table page 3-1 Participants’ average length of re sidence in the U.S. and average age....................45 3-2 Characteristics of participants..................................................................................46 3-3 Types of FLRAS and sp ecific questions asked........................................................49 3-4 Types of reading strategies and specific questions asked........................................49 3-5 Examples of subtests................................................................................................54 4-1 Guidelines for descri bing standard score.................................................................58 4-2 Total standard score and Mean standard score.........................................................58 4-3 TOAL-3 total standard score by subtest for H group...............................................58 4-4 TOAL-3 total standard score by subtest for L group...............................................59 4-5 Mean and SD of total standard score in H group.....................................................61 4-6 Mean and SD of total standard score in L group......................................................61 4-7 Standard score in H group (ANOVA)......................................................................61 4-8 Standard score in L group (ANOVA)......................................................................61 4-9 Reading text and number of idea units.....................................................................63 4-10 Mean (%) of subgroups fo r “tightly-structured” passa ges and for “all passages” (H group)..................................................................................................................63 4-11 Mean (%) of subgroups fo r “tightly-structured” passa ges and for “all passages” (L group)..................................................................................................................63 4-12 Mean (%) for “tightly-structured” pa ssages and for “all passages” (H group)........66 4-13 Mean (%) for “tightly-structured” pa ssages and for “all passages” (L group)........66 4-14 Correlations between TOAL-3 and readings...........................................................68


x 4-15 Mean number and percentage of idea un its recalled correctly per passage and per text adjunct in H group.............................................................................................68 4-16 Mean number and percentage of idea un its recalled correctly per passage and per text adjunct in L group.............................................................................................69 4-17 ANOVAs on the readings in H group......................................................................71 4-18 ANOVA on the readings in L group........................................................................71 4-19 Average score by item in the FLRAS for H group..................................................73 4-20 Average score by item in the FLRAS for L group...................................................74 4-21 Average score by item in the RSQ for H group.......................................................76 4-22 Average score by item in the RSQ for L group........................................................76 4-23 Average score of item 9 on the RSQ for both groups..............................................77


xi 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 SECOND LANGUAGE READING: THE INTERRELATIONSHIPS AMONG TEXT ADJUNCTS, STUDENTS’ PROFICIENC Y LEVELS AND READING STRATEGIES By Heenam Park August 2005 Chair: Ratree Wayland Cochair: Theresa A. Antes Major Department: Linguistics According to the psycholinguistic view of reading, readers use their background knowledge to interpret texts. This backgr ound knowledge has been termed “schemata”. Distinct from the notion of schema is research on the role of text adjuncts in secondlanguage (L2) reading. Text ad juncts are pre-reading info rmation such as pictures, definition lists, or text struct ure information regarding the te xt provided to help students understand the reading passage. Several research ers found that the facilitative effects of different text adjuncts on L2 reading may va ry as a function of l earners’ proficiency and of text types. The primary goal of my study was to investigate the effects of different types of text adjuncts on L2 reading co mprehension. Specifically, my study examined whether L2 learners from various proficiency levels benefit from different types of text adjuncts. I also examined the relationship between English as a Second Language (ESL) learners’ anxiety levels and th eir proficiency levels. Finall y, I examined the relationship between ESL learners’ readi ng strategies and their read ing comprehension ability. I


xii hypothesized that Low-Level ESL learners will benefit more from a text adjunct that provides a definition list than from one that provides information on the text type, while the opposite would be true for High-Level ESL learners. Research participants were 61 heterogeneous ESL students. Participants we re divided into a hi gh proficiency group and a low proficiency group based on the class pl acement. Participants in each proficiency level were further divided into two treat ment groups (vocabul ary group and expanded framework group) and a control group. After reading a passage, students were asked to write down as much as they could remember from the passage for their recall protocol, without referring back to it. Results show ed that the high-proficiency ESL students benefited most, as hypothesized, from the expa nded framework text adjunct when they read in an L2. Results also showed that, c ontrary to expectations, the low-proficiency ESL students benefited from both the e xpanded framework text adjunct and the vocabulary text adjunct.


1 CHAPTER 1 INTRODUCTION Most research on reading, whether first language (L1) or second language (L2), is centered around the issue of whether reading is a decoding process or a psycholinguistic synthesizing process. In a d ecoding process, readers constr uct the meaning of the text being read by merely decoding the graphi c forms in the text. A psycholinguistic synthesizing process of reading, on the othe r hand, involves an interaction between the text and the readers. In the psycholinguist ic view, readers are believed to use their background knowledge (both linguistic knowledg e and knowledge of the world) when interpreting the text. This background knowledge has been called “schemata.” According to schema theory (Carrell, 1984a), reading comprehension is a result of the reader activating the appropriate schemata when inte racting with a text. Two types of schema have been identified in the literature: conten t schema and formal schema. Content schema refers to the readers’ background knowledge of the topic or the content of the passage being read. Formal schema refers to the readers’ background knowledge of the overall organization or discourse structur e of the text (for example, whether it is a narrative or an expository text). Both types of schema have been shown to positively affect reading comprehension in L2. Distinct from (but related to) the notion of schema, is research on the role of text adjuncts (i.e., pictures, prefatory statements titles) in L2 reading. Several researchers (Hudson, 1982; Johnson, 1982; Lee & Riley, 1990; Taglieber et al., 1988, among others) found that the facilitative effects of text adjuncts on L2 reading may vary as a function of


2 learners’ proficiency and of text types (e.g., na rrative or expository). However, there have been inconsistent results on th e effects of text adjuncts in relation to the learners’ proficiency levels and the type s of text adjuncts used. Johns on (1982) and Taglieber et al. (1988) argue that giving learners vocabulary words before they read is not as effective as giving them background knowledge about a par ticular reading passage. In contrast, Laufer (1992) claims that vocabulary knowledge is essential, particularly for low-level ESL students. Another factor shown to aff ect reading comprehension is reading anxiety. Previous research shows that language anxiety is re lated to learners’ be liefs; and, as Horwitz (1987, 1988) notes, learners’ beliefs are related to the strategies they actually use. In this sense, language anxiety research should be a ssociated with learners ’ beliefs or strategy studies. Many researchers (Horwitz et al., 1986; MacIntyre & Gardner, 1989, 1994a) have shown that anxiety often has a ne gative effect on langu age learning. Horwitz (1986), for example, found that higher scores on the Foreign Language Classroom Anxiety Scale (FLCAS) were significantly corr elated with lower actual final grades as well as expected grades. Much research has been done on the role of anxiety in language learning with native speakers of English l earning other languages. However, there has been little research on the effect of anxiety in langua ge learning among English as a Foreign Language (EFL) students. Another important factor in L2 reading is the reading strategies used by L2 learners. In the past, reading strategies were known as skills Much reading research has empirically investigated reading strategies actually used by successful and unsuccessful L2 readers. When reading English, ESL learne rs’ proficiency is rela ted to their reading


3 strategies. Unfortunately, many researchers do not measure le arners’ proficiency levels reliably. My study included 61 heterogeneous ESL stude nts. The participants were divided into a high-proficiency group and a low-pr oficiency group, based on their composite scores on the Test of Adoles cent and Adult Language (TOAL). Participants in each proficiency level were further divided into two treatment groups (a vocabulary group and an expanded framework group) and a control group. The vocabulary groups received a definition list before they read each passa ge. Participants in the expanded framework groups received information on text types befo re they read each pa ssage. Control groups had no treatment. After reading each passage, a ll participants were asked to complete a recall protocol. The main purpose of my study was to invest igate the effects of different types of text adjuncts on L2 reading comprehension. Specifically, my study aimed to determine whether L2 learners from various profic iency levels benefit from vocabulary or framework text adjuncts. I also examined the interaction between anxiety level and degree of proficiency. Moreover, I examined the difference in reading strategies used by low and high proficiency L2 learners, and their interactions with text adjuncts.


4 CHAPTER 2 REVIEW OF THE LITERATURE Effects of Text Adjuncts In traditional views of se cond-language (L2) reading comprehension, “meaning is conceived to be ‘in’ the text to have a separate, independe nt existence from the reader” (Carrell, 1984c, p. 332). However, many recent L2 reading researchers advocate schema theory (Bernhardt, 1984; Brantmeier, 2005; Carrell, 1984a, 1984b, 1984c; Lee & Riley, 1990; Meyer & Freedle, 1984; Rudell & Speaker 1985). Schema theory research argues for the importance of background knowledge in (a psycholinguistic model of) the reading process. Second-language reading can be diffi cult for various reasons, including readers’ failure to access appropriate schemata (e.g., co ntent and formal schema). Text adjuncts are, therefore important because they may help L2 readers activate the appropriate schemata and, thus, enhance L2 reading. Text adjuncts are mostly used as pre-reading activities in L2 reading. The ne xt sections discuss text adj uncts and related factors. Types of Reading Processes There are three major read ing processes: Bottom-up, t op-down, and interactive. This section discusses each of the three reading processes and the differences between text adjuncts and reading pro cesses. In my study, processes and adjuncts are the central components that were observed. A bottom-up (or text-driven) approach starts with letters, words, and then sentences (Urquhart & Cyril, 1998). According to this approach, a word should take longer to recognize than a single letter. However, one criticism of a bottom-up approach is that


5 recognizing a word often takes less time than recognizing each of its individual letters. Moreover, there have been no definite findi ngs to indicate whet her one stage of the process is over before the next stage begins. In contrast, in a top-down (or reader-dri ven) approach, reader s start with their background knowledge (whole text) and make pred ictions about the text, and then verify their predictions by using text data (words ) in the text (Urquhart & Cyril, 1998). The main shortcoming of a top-down approach is th at it is impossible to see how a reader can begin by engaging the text as a whole, then proceed to paragraphs, then to individual sentences, ending with single letters. In other words, this process (background knowledge, then paragraphs, then sentences, th en single letters) may not represent what really happens. The shortcomings of mutually exclusive bottom-up and top-down explanations lend support to a more interactive approach. From a psycholinguistic viewpoint, the interactive process is the most effective becau se reading represents decoding a text while using a reader’s background knowledge. Research in schema theory shows that reading comprehension is an interactive process between the reader and the te xt, because the text provides the reader with a framework for co mprehending the ideas that are held within the discrete word, sentence, and paragraph le vel linguistic units, but the reader must recognize the rhetorical structur e of the text in order for the ideas to have any coherence (Brantmeier, 2005; Carrell, 1983, 1984a; Carr ell & Eisterhold, 1988; Eskey, 1988; Meyer & Freedle, 1984; Young, 1993). According to Es key (1988), the interactive model does not assume the primacy of top-down proces sing skills, but rather posits a constant interaction between bottom-up and top-down pr ocessing in reading. In this view, good


6 readers are both good decoders and good interpre ters of texts: “To properly achieve both fluency and accuracy, developing readers must, therefore, work at perfecting both their bottom-up recognition skills and their top-down interpretation strategies” (Eskey, 1988, p. 95). Eskey (1988) suggests that good r eading can result only from a constant interaction between information achieve d by bottom-up decoding and information provided by top-down analysis. According to Eskey (1988), a major virtue of the interactive model is that it directs our atte ntion to both the top-down and bottom-up skills that fluent and accurate reading demands. Although Eskey believes that an interactive appro ach is the most effective, he also emphasizes the importance of bottom-up decoding and criticizes some researchers for overemphasizing higher-level strategies (predi cting from context or using schemata and other kinds of background knowledge). In his opinion, language is a major problem in second language reading; even educated guessing at meaning is no substitute for accurate decoding. He also adds that language is a ki nd of schema too, although one that for fluent native users may be activated automaticall y. Eskey admits, however, that successful comprehension is much more than simple decoding, nonetheless he suggests that decoding is also a cognitive process, invol ving bottom-up as well as top-down skills, and successful reading cannot be achieved without it. Fluent reading involves both skillful decoding and relating the information so obtaine d to the reader’s pr ior knowledge of the subject and the world. In Eskey’s view, such lower-level skills as the rapid an d accurate identification of lexical and grammatical forms are not merely obstacles to be cleared on the way to higherlevel “guessing game” strategies, but skills to be mastered as a necessary means of taking much of the guesswork out of reading comprehension (p. 98).


7 He concludes that an interactive model of reading provides the most persuasive account of this reciprocal per ceptual/cognitive process. On the other hand, Goodman (1971) emphasi zes “the act of the construction of meaning as being an on-going, cyclical pr ocess of sampling from the input text, predicting, testing and confirmi ng or revising those predicti ons, and sampling further” (Carrell & Eisterhold, 1988, p. 74). In th is model, the efficient reader does not use all of the textual cues. However, Eskey (1 986, 1988) questions Goodman’s position and cautions that we risk converting r eading into a real guessing game. Discussion of Schema Theory Many reading researchers believe that read ing is an interactiv e process, requiring text-based features and a reader’s background knowledge. In other wo rds, in reading, the reader interacts with the text. Bernhardt (1984 ) believes that the interpretation of the text does not depend on text-based features alone but rather on the information that the readers bring with them to the text. The ps ycholinguistic view of reading advocated by Clark (1980) agrees with Bernhardt (1984) that readers use their background knowledge (both linguistic knowledge and knowledge of the world) in interpreting the text. The role of background knowledge in language comprehension has been formalized as schema theory. Schema theory holds that any pa rticular text does not have meaning by itself; rather, a text only provi des directions for listeners or readers as to how they should retrieve or construct meaning from thei r own, previously acquired knowledge. Such knowledge is called the readers’ back ground knowledge; the previously acquired knowledge structures are known as schemata Thus, according to schema theory, readers activate appropriate background knowledge agai nst which they try to give a text a consistent interpretation (Carre ll, 1984c). Carrell suggests that what is understood from a


8 text is a function of the part icular schema activated at th e time of processing the text.L2 readers often do not succeed in comprehending a text because of their failure to access the appropriate schema. There are two main types of schemata: content schemata and formal schemata. Content schemata comprise background knowle dge of the content area of a text. For example, a teacher can give students a pictur e of a cell for an expository text about cell biology. Formal schemata comprise background knowledge of the formal rhetorical organizational structures of different types of text. For example, a teacher can give students a rhetorically oriented framework through text adjuncts. L2 readers often do not process a text the same way the L1 readers do. In general, L2 readers tend to be linguistically bound to a text. That is, L2 readers may process the literal language of the text, but they may not make the nece ssary connections between the text and the appropriate background informati on needed to comprehe nd the text (Carrell, 1983). Neither advanced nor high-intermediate ESL readers seem to use context or textual clues frequently. As Carrell (1983, p. 199) states, “T hey are not efficient topdown processors, making appropriate predictions based on context, no r are they efficient bottom-up processors, building up a mental repr esentation of the text based on the lexical information in the text.” Moreover, fluent readers of a passage in their native language recover quickly from wrong guesses about textual meaning. Smith (1994) argues that any reader can be a fluent reader or a poor reader based on his/her background knowledge of a text. However, ESL readers may not recover as native spea kers do, because ESL readers may extract inaccurate information from the text and this may lead to inaccurate predictions for


9 meaning. Also, ESL readers lack background knowledge compared to native speakers (Johnson, 1982). From a schema theory viewpoi nt, activation of inappropriate schemata can also impede reading comprehension (R udell & Speaker, 1985). Hen ce, the successful activation of an appropriate schema is very important. Thus, to ac tivate the appropriate content schema and formal schema, text adjunc ts may be useful in L2 reading, as they could enhance reading comprehension. Text adjuncts are pre-reading information (s uch as pictures, defi nition lists, or text structure information regarding the text) prov ided to help students understand the reading passage. Text adjuncts serve to activate schemata in several ways. First, text adjuncts help to activate content schemata. For example, vocabulary text adjuncts enable advanced learners to make bridging in ferences. Furthermore, backgr ound text adjuncts help to activate content schemata regarding tw o different types of background knowledge: cultural and topic. According to Carrell and Ei sterhold (1988), L2 r eading can be difficult due to many culturally loaded concepts pr esupposed by a text. To overcome this problem, a teacher can give students a preview involvi ng a key concept that is culturally loaded. Second, text adjuncts can help to activate formal schemata. For example, rhetorically oriented text adjuncts may help activate L2 readers’ prior knowledge of the rhetorical structure of the text to be read. Hudson’s (1982) study showed an interacti on between overall linguistic proficiency in ESL and content-induced schematic effect s in ESL reading comprehension. As such, Hudson’s study suggests the faci litating effects on comprehens ion of explicitly inducing content schemata through pre-r eading activities, as compared to two other methods of inducing schemata (vocabulary activities a nd read-reread activiti es) especially at


10 beginning and intermediate proficiency leve ls. Hudson argues that this activation of appropriate schemata should be a purpose of pr e-reading activities, and that text adjuncts could enhance reading comprehension when give n as pre-reading activities to L2 readers. Additionally, Carrell and Eisterhold (1988) s uggest that previewing (e.g., definition lists) is an important activity in the reading cl assroom, which sometimes involves teaching a culturally loaded key concept. They conclude that EFL/ESL readers should become more aware that reading is a highly interactive process between themselves and their prior background knowledge, on the one hand; and the te xt itself, on the othe r. Carrell argues that an instructor should teach L2 readers to process texts interactively. That is to read texts using appropriate knowledge but cons tantly checking and restructuring that knowledge to fit the details encounter ed during reading (Carrell, 1984c). When a reader interprets a text, s/he may fi nd a certain type of text adjunct helpful. There is a difference between “reading pro cesses” and “text adjuncts.” The reader processes a text in a bottom-up, top-down, or interactive way. We need to distinguish text adjuncts from processing t ypes in order not to confuse them with other terms such as the bottom-up processing or top-down processi ng modes, or content or formal schemata. Carrell (1984c) argues that vocabulary text adjuncts should be considered a type of background knowledge, because they can lead a reader to connect the vocabulary with pre-existing concepts and know ledge. According to Carrell a nd Eisterhold (1988), lessproficient students tend to use a bottomup processing mode, and more-proficient students tend to use a top-down processing mode In other words, less-proficient students tend to use more textual items when they read, and more-profici ent students tend to depend on their background knowledge when tr ying to understand a passage. So, it seems


11 that for Carrell and Eisterhold, certain types of text adjuncts tend to facilitate certain types of processing for L2 read ers of certain proficiency level. For example, a vocabulary text adjunct will faci litate bottom-up processing among less proficient readers, but topdown processing among more proficient readers. Carrell and Eisterhold’s implications mi ght not be entirely correct, though. The processing of a certain type of text adjunct doe s not necessarily facilitate a certain type of processing. For example, a vocabulary list te xt adjunct, describing various food items appearing in the passage, may activate some r eaders’ prior experiences in a restaurant, and, thus, activating their background knowledge and tri ggering top-down processing. While, other readers of equal proficiency leve l who lack such prior experiences may use the vocabulary as a key for further decodi ng the text (i.e., bo ttom-up processing). In determining processing type, the reader ’s background is a more important factor than the type of text adjunct used (Bernhardt, 1984). For a text adj unct to be effective it must be integrated into the reader’s pre-existing knowledge a nd other pre-reading activities meant to build b ackground knowledge (Carrell, 1984c ). That is to say, any given text adjunct may trigger different types of processing. In summation, text adjuncts most affect readers’ comprehension by helping them to activate appropriate schemata. Several studies give evidence that text adjuncts aid comprehension through the activation of appropr iate schemata. Proficiency level affects text adjuncts inconsistently. Hudson (1982) states that certai n types of text adjuncts are best suited to certain proficiency levels In Hudson’s study, usi ng a vocabulary list was more effective for advanced learners than fo r beginning and intermediate learners. On the other hand, Taglieber et al.’s (1988) result s showed that vocabulary pre-teaching was


12 significantly less effective than the other tw o strategies for average-ability learners. Moreover, in Johnson’s (1982) study, exposure to meanings of the target vocabulary words did not significantly affect advanced students’ reading comprehension. A text adjunct can also affect comprehension with regards to the rhetori cal organization of a text. A fluent reader may recognize a text’s rhetorical structure before s/he reads. Therefore, the explicit teaching of text’s rhet orical structure as a type of formal text adjunct can aid reading comprehension. Before the reader decodes and interprets a text, text adjuncts help to activate appropria te schemata, allowing the reader to better comprehend a text. For example, the reader given a text adjunct such as background knowledge can better understand the text because the reader already has an appropriate schema activated before reading the passage, and this schema lets the reader comprehend the text more correctly. Moreover, this schema helps the reader to guess accurately, when guessing meanings in the text. Different Types of Text Adjuncts Several reading-research studi es provide evidence of the effects of text adjuncts on reading comprehension (Carrell, 1983; Hudson, 1982; Lee, 1986a, 1986b; Lee & Riley, 1990; Taglieber, Johnson & Yarbrough, 1988). H udson (1982) investigated the effects of three different types of text adjuncts in L2 reading comprehension on beginning, intermediate, and advanced ESL students who we re proficient readers in their L1. Each group was given three graded reading passages. The first type of text adjunct that Hudson tested was pictorial text adjunc t. In this condition, subjects were given a set of pictures about the general topic of the passage and then were asked a set of questions about each of the pictures. The pictures represented a type of background text adjunct. After that, each subject wrote self-generated predictions of what s/he expected to find in the reading


13 passage. The second type of text adjunct wa s vocabulary instructi on. In this condition, subjects were given a list of vocabulary items which would appear in the reading passage. The third method was read-test/read-test. In this condition, subjects read the passage silently, and then took the same reading co mprehension test. The procedure was then repeated one more time. Hudson’s results showed that beginning and intermediate students scored higher in the pre-reading condition (pictures) than in the other two conditions. In his study, advanced readers had less trouble processing visual information and changing schemata than did lower-level r eaders, and advanced readers were able to bring more non-visual “behind the eye-ball” knowledge when reading in the L2 than beginning and intermediate readers were. H udson concluded that a schema can override language proficiency as a factor in L2 read ing comprehension; th e linguistic ceiling is only one determinant of reading comprehens ion. In other words, Hudson suggests that text adjuncts help readers override their lingui stic limitations, particularly at the lower levels of proficiency. Like Hudson, Taglieber et al (1988) investigated the effects on Brazilian EFL students’ reading comprehension of three di fferent types of text adjuncts: pictorial context, vocabulary pre-teaching, and pre-quest ioning. The subjects were enrolled in the sixth semester of EFL at the college level. Ta glieber et al. stated that the subjects had average levels of English proficiency. Thei r results showed that vocabulary pre-teaching was significantly less effective than the othe r two strategies. They argued that although knowledge of words’ meanings was nece ssary for comprehension, heightened background knowledge from the other preread ing activities enabled students to use


14 context to construct meaning successfully for the passages even when all of the words were unknown. Vocabulary pre-teaching is a widely used type of text adjunct. However, the results of vocabulary pre-teaching are inconsiste nt. As mentioned above, Hudson (1982) found that the use of a vocabulary list was found to be more effectiv e for advanced learners than for beginning and intermediate learners Johnson (1982), on the other hand, found no facilitative effect of voca bulary pre-teaching among her subj ects. Specifically, Johnson’s results suggest that ESL students’ comprehens ion of a passage on th e topic of Halloween was affected more by prior experience in th e American culture than by knowledge of the vocabulary in the text. In this study, the s ubjects were advanced ESL students at the university level. There were divided into four groups Group 1 read the passage without a vocabulary list (to study before reading or to refer to whil e reading). Group 2 studied the definitions of the target words before reading but was not able to refer to the list while reading. Group 3 read the passage with the ta rget words glossed in the passage. Group 4 studied the target vocabulary words before r eading, with the definitions of the target words glossed in the passage. Written recall was significantly better for the familiar than for the unfamiliar information in the passage. Moreover, subjects of all four groups more accurately recognized information from the familiar section of the passage compared with the unfamiliar section. However, exposur e to the meaning of difficult vocabulary words in the passage did not seem to affect the comprehension of ESL readers. In other words, the effects of vocabulary difficulty on reading comprehension were not as obvious as the effects of background knowledge. Johnson argues that familia rity with the topic of the passage, combined with background knowle dge of the idea allows the reader to


15 construct meaning successfully for unfamiliar vo cabulary words. She explains that, “the normal redundancy in a text may enable read ers to cope with unfamiliar words without too much disruption in their understanding. Reader s seem to be able to construct a text from memory based on inferences made while reading” (p. 514). Based on these findings, Johnson concludes that “knowledge obtained from real experien ces in the foreign culture is effective for reading comprehension of a passage on that topic” (p. 514), and is more important than knowledge of vocabulary. When an instructor in an L2 classroom teaches, s/he must be concerned with the following aspects regarding vocabulary pre-te aching. First, as Taglieber et al. (1988) note, the instructor should teach the voca bulary important to th e text, not vocabulary important to the lesson (p. 457). Second, accord ing to Carrell (1984c), merely presenting new or unfamiliar vocabulary items does not gua rantee the induction of new schemata. Lee and Riley (1990) investigated the eff ect of giving novice French L2 learners information about the rhetorical structure of an expository pros e passage as a text adjunct. They considered two types of rhetorical or ganization: problem-solution and a collection of descriptions. Subjects were divided into three treatment gr oups based on the type of formal schema text adjunct they rece ived: no framework, minimal framework, and expanded framework. The minimal-framework gr oup was told either that their passage presented “a collection of descriptions” or “a problem and two solutions.” For the “collection of descript ions” passage, the expanded framew ork group was told that their passage presented “a collection of descriptions of some activit ies that the French engage in during their leisure time th at demonstrate an interest in both personal enrichment and communication with others”. For the problem-s olution passage, the expanded-framework


16 group was told that their passage presented th e problem of “amateur prospectors pillaging archeological treasures in France and that two possible solutions would be discussed.” Both minimal and expanded frameworks were presented in the s ubjects’ native language (English). To determine rhetorical structure, Lee and Riley used Meyer’s (1975) contentstructure analysis, which provi des a hierarchical descripti on of the text. The top-level idea units represent the main idea being collect ively described (collect ion of descriptions text) or presenting the problem and its solution(s) (problem-solution text). Lee and Riley found that only the expanded framework impr oved recall of the ove rall content of the collection of descriptions passage; however, the expanded framework had less effect on the problem-solution passage, improving recall sc ores of only the top-level ideas of that passage. Lee and Riley concluded that an e xpanded framework helps improve readers’ comprehension of a loosely-structured passa ge (collection descri ption), but not of a tightly-structured one (problem-solution). From a review of previous research it beco mes apparent that the facilitative effect of different types of text adj uncts (vocabulary list and framework) may vary as a function of learners’ proficiency levels. In this re gard, Carrell (1984c) argued that pre-reading activities should differ accordi ng to proficiency levels. Carrell and Eisterhold (1988) also suggest a relationship between proficiency level and the use of adjuncts. Specifically, they asse rt that illustrations may be particularly appropriate for students with minimal language skills. Background information and previewing, on the other hand, are particularly important for less-proficient language students.


17 Reading Comprehension Measurements There are several reading-comprehension measurements in L2 reading research: cloze tests, multiple-choice questions, true-false questions, open-ended questions, and recall. The cloze test may not be a reliable measurement of reading comprehension because the processes involved do not reflect actual reading-comprehension processes. Moreover, the type of reading task may infl uence how readers interact with the passage (Wolf, 1993). Bernhardt (1991) suggests that for scores on a cloze test to make high correlations with scores on other measures, exact word scoring must be used (pp. 196197). Multiple-choice questions are often used for reading comprehension because the task is familiar to subjects and easy to admi nister and score. However, with multiplechoice questions, test items can sometimes be answered without reading the passage. Another critique of multiple-choice tests is th at learners do not always need to carefully read all of the readi ng passage to answer the questions Instead, test-takers can often depend on “the recognition of a few key words in the passage or on clues from the other questions, and on syntactic or semantic relati onships between the stem and the choices to perform better than they would merely by chance” (Wolf, 1993, pp. 474-475). Wolf suggests in her study that multiple-choice items in Spanish may assess students’ ability to guess rather than their abili ty to construct meaning. Wolf (1993) states that test-takers approach open-ended questions by paying attention to a few surface features of the passage. Furthermore, Bernhardt (1991) says that open-ended questions limit the range or t ype of responses possi ble; and can affect test-takers’ comprehension of the passage by providing more information about the content.


18 In regards to a recall prot ocol, Bernhardt (1991) argues that “a free recall measure provides a purer measure of comprehensi on, uncomplicated by linguistic performance and tester interference” (p. 200). She also asserts that the recall protocol is a more reliable and valid measurement of comprehension because it does not delimit the response or influence comprehension. As Wolf (1993) said, no “perfect” test of comprehension exists because comprehension varies according to a variety of interactions among reader-based components (such as students’ background knowledge or linguistic prof iciency) and textbased components (such as difficulty level of a text). Interaction of Text Adjuncts, Text Types and L2 Learners Text types also play a role in L2 reading. For expository texts, Meyer and Freedle (1984) recognize five different t ypes of rhetorical organization: collection of descriptions, causation, problem/solution, and comparison. Acco rding to Meyer and Freedle, each of these types represents a different abstract sc hema of ways in which writers organize and readers understand topics. Certain types of rhetorical organiza tion of expository text are recalled differently from ot her types (Carrell, 1984b; Me yer, Brandt & Bluth, 1980; Meyer & Freedle, 1984). For example, Carre ll (1984b) investigates the effects of rhetorical organization of di fferent types of expository text on ESL readers. Results showed that recall was better for the texts with the three more tightly organized discourse structures (comparison, causation, and problem/s olution) than for those that were more loosely organized. Carrell found that the more highly organized types of discourse are generally more facilitative of recall than the less organize d collection of descriptions. Similarly, Meyer and Freedle’s (1984) study show ed that the different “top-level” overall organizational types affect reading comprehe nsion in English as a native language as well. Meyer and Freedle found be tter recall of discourse when it was organized with the


19 comparison, causation, and problem/solution st ructures than with the collection of descriptions structures. Research on discourse or text comprehe nsion has shown that comprehension is determined by the rhetorical organization of a text as well as by the local effects of sentences or paragraphs (Carre ll, 1984a). Carrell’s study inves tigates the effects of story structure or narrative rhetorical organizati on on reading recall in ESL. According to Carrell, a story schema may be considered as the rhetorical structur e or the grammar of a story. In this study, Carrell divi des stories into two types: st andard and interleaved. In the standard version, three simple two-episode st ories were constructed. Each standard story was then rearranged to create an interleaved version. In th e interleaved version, following the Setting, the five basic nodes of each episode were presented in interleaved fashion: two Beginnings, two Reactions, two Attempts two Outcomes, and two Endings. Carrell notes that “It has even been characterized as a grammar which generates a tree structure consisting of labeled nodes, the constituents of a story” (p. 90). The results indicate that a greater number of nodes were remembered for the standard stories than for the interleaved stories. According to Carrell, whereas native speaking a dults do not have to devote more effort to encoding, nonnative reader s must dedicate more effort to linguistic encoding, and, therefore, have le ss effort to dedicate to the sorting of interleaved input into ideal schematic form. Carrell concl udes that in schema theory, reading comprehension is a role of the readers’ pr ocessing and activati on of the appropriate formal and content schemata in interaction wi th a text (along with the linguistic cues provided in the text by the author). She suggests that L2 reading can be regarded partly as


20 the process of acquiring appropria te new formal and content sc hemata and of learning to activate the appropriate sche mata during comprehension. Carrell (1984a) pointed out th at empirical research has shown the effects of story schema in L1 comprehension, but that no co mparable research has been conducted to investigate the role of story schema in L2 comprehension. The reason for this lack of studies might be because literary text is cu lturally embedded. Frequently, L2 learners have problems related to the absence of a ppropriate generalized information assumed by the writer and possessed by a reader sharing that writer’s cultural background (Carrell & Eisterhold, 1988). Research on training with expository text structure shows that reading comprehension can be significantly facilita ted by explicitly te aching readers about expository text structure and by teaching divers e strategies for iden tifying and employing that structure during the readi ng process. Carrell’s (1985) st udy found that training on the top-level rhetorical organizati on of expository texts significa ntly increased the amount of information that ESL students could recall. In the area of narrative pr ose, Carrell (1984a) has demonstrated the effects of a simple na rrative schema in ESL. In a study comparing L1 and L2 reading comprehension of literary texts, Fecteau (1999) found that for all students, literal comprehension was superior in the L1, but amount of recall may also have been affected by factors such as succe ssful activation of a ppropriate schema and processing of key details. Relationship between proficiency and text adjuncts. Previous research also shows that low and high proficiency learners be nefit from different types of text adjuncts. Specifically, decoding is more frequently us ed among low-proficiency learners. This is


21 due, perhaps, to the fact that for them, word -by-word decoding is not automatic, and that is why their reading is not fa st. Therefore, for these lear ners, linguistic knowledge, e.g., syntactic knowledge, may be more important be cause they have not reached the threshold level of competence in L2. Laufer and Sim ( 1985) observe that “L2 reading is a function of both reading strategies and foreign language competence, since below a certain proficiency level the strategies are not likel y to be applied” (p. 406). Even if they do transfer their L1 reading stra tegies, low-proficiency learners may not use them due to factors such as inadequate knowledge of syntact ic features of the L2. On the other hand, for high-proficiency learners, construction of meaning is more frequently used than decoding. These learners can decode automatical ly because of their L2 proficiency levels. Learners’ Proficiency in L2 Reading In ESL reading studies, learne rs’ English proficiency levels their strategy use, and their reading comprehension have been show n to be interconnected. As noted by Clarke (1980), the role of language pr oficiency is greater than had previously been assumed. Kern (1989) concurs, noting that language pr oficiency factors are ce rtainly an important element in determining L2 reading compre hension. More importantly, a relationship between learners’ overall L2 proficiency has been shown to relate to their reading comprehension (e.g., Block, 1986; Carre ll, 1991; Clark, 1979, 1980; Kern, 1989; Taillefer, 1996). In a study by Carrell (1991), L2 proficiency level was found to be more important than L1 reading skills on L2 r eading comprehension. Al so, Clarke’s (1979) ‘short circuit hypothesis’ emphasizes L2 prof iciency. In his hypothesis, Clarke argues that learners cannot transfer their L1 r eading strategies until they reach a certain proficiency level of L2. He e xplains that learner’s low pr oficiency short-circuits L2 reading comprehension. One of the major clai ms of the present study is that it is


22 necessary to measure L2 learners’ proficienc y before generalizing any results to the ESL population as a whole. The techniques for a ssessing L2 proficiency in L2 reading research are discussed in the next section. Problems with assessing L2 proficiency. Thomas (1994) reviews advantages and disadvantages of the four major techniques fo r assessing the proficie ncy of L2 learners: impressionistic judgments, inst itutional status, use of inhouse assessment instruments, and standardized tests. Additionally, he addresses the question of why target language proficiency is sometimes a poorly controlled factor in research on SLA. First, when researchers use impressionistic judgments for assessing the pr oficiency of L2 learners, they typically measure how long a learner ha s lived in an L2-speaking environment. However, this information may not be revealing, because having residency in an L2speaking environment does not necessarily mean th at the learner will at tain any particular level of proficiency. Wolf (1993) also point s out that the problem with many subject pools is the inconsistent manner in which s ubjects are categorized into L2 experience levels. As Wolf notes, L2 learners’ profic iency cannot always be categorized by amount of classroom instruction since learners are often placed into interm ediate and advanced courses according to their scores on L2 entrance and proficiency exams. Naiman, Frohlich, and Todesco (1975) did not give a prof iciency test to students; they explained that they either knew the students persona lly or depended on tr usted recommendations. However, personal impressions about student s may not provide an accurate measure of learners’ proficiency. In light of this, this study has a lack of generalizability. A second traditional way of determining L2 proficiency is to rely on learner’s institutional status. An example of L2 prof iciency defined by inst itutional status is


23 provided by Lee and Riley (1990): “[All participants] were enro lled in the third semester of the French language program at the University of Illinois” (p. 28). However, as pointed out by Thomas (1994), “[i]nstitutions differ greatly in the standards by which they assign a given status to individuals, and in the rigid ity with which those standards are maintained” (p. 317). Such practice, thus, threatens the external validity (generalizability) of th e research results. A third method used to determine learners’ pr oficiency in L2 research is by relying on locally developed and administered test s. Thomas (1994) commented on this method that researchers rarely provi ded clear information about th e form of either in-house placement tests or in-house resear ch-internal tests. He also crit icized the fact that in many instances, researchers did not report the content of their te sts with enough detail for readers to understand. A fourth method is the use of standardized tests. The advantage of this method is that it enhances the generalizability of the re search results. For example, Thomas (1994) observes that “a score of 90 on the Michigan test or 500 on the TOEFL serves as a recognizable benchmark, enhancing the generaliz ability of the research results” (p. 324). However, according to Thomas (1994), some researchers reported only the minimum or maximum scores achieved by the total sample. Additionally, it has been noted that “the content of st andardized tests is available for public scrutiny, and their validity is subject to ongoing investigation” (Thomas, 1994, p. 324). However, a researcher needs to be cau tious when he/she chooses subsections of a standardized test. For L2 proficiency, Yano, Long, and Ro ss (1994) chose the Structure Subtest Form A of the CELT (Comprehensive English Language Test ). It consists of


24 items that assess grammatical knowledge and th at deal with choice of verbs and verb forms. Moreover, the reason for the use of th e CELT Structure Test in their study was to provide a covariate for the reading comp rehension measures. However, the CELT subtests that were used in this study may not be appropriate for m easuring students’ L2 overall proficiency. Indeed, many researcher s use L2 proficiency tests in their SLA research, but sometimes they underestimate the importance of those test s. For this reason, their results may not be generalizable or accurate. Reading Strategy Studies Research on metacognition has shown that readers’ metacognitive awareness of strategies is related to their reading proficiency (Bar nett, 1988; Carrell, 1989; Carrell, Pharis & Liberto, 1989). Carrell (1989) inves tigated the relationship between L1 and L2 readers’ metacognitive awareness and read ing ability. The results showed that the “global,” top-down types of reading strategies (e.g., te xt-gist, background knowledge, and text organization) were not significantly related to L1 reading performance for either the Spanish L1 group or the English L1 group. Interestingly, “local” reading strategies (e.g., focusing on grammar structures, sound-le tter, word-meaning, and text details) tended to be negatively correlated with L1 reading performance. The ESL group of more advanced proficiency levels tended to use more “global” or top-down strategies in their perceptions of effective a nd difficulty-causing reading st rategies (Carrell, 1989). Several researchers have identified appare nt relationships betw een certain types of reading strategies and successful or unsu ccessful foreign or se cond language reading. Young (1993) investigated foreign language read ers’ cognitive processes (strategy use, i.e., top-down and bottom-up) when they read authentic and edited texts. Young pointed out that much research on strategy use in L2 reading explores the basis for successful and


25 unsuccessful reading comprehe nsion. Most L1 and L2 stra tegy research suggests that readers who focus on reading as a decodi ng process rather than as a meaningconstruction process tend to be less successful readers. The results of Clarke’s (1980) study also showed that successful readers fo cused on the meaning of the text. Anderson (1991), on the other hand, investigated indivi dual differences in strategy use in second language reading and testing. In this study, relati ng sentences from one part of the text to another and monitoring affective feelings about the text are strategies that participants reported frequently. Anderson suggested that strategic reading is not only a matter of knowing what strategy to use, but also knowi ng how to use a strategy successfully and how to orchestrate it with other strategies. In the next section, research methodologies used for examining reading strategy use are discussed. Ways to examine reading strategies. There are several methods used for examining reading strategy use. First, resear chers can interview students directly about their use of strategies. An a dvantage of this method is that a researcher can ask students directly about the strategies and techniques that they might have used in learning an L2. However, there are several disadvantages to simply conducting interviews. Interviews between the language learners and their teacher s as well as observation of overt learning behavior may lead to the identification of only the conscious aspects of learning, although the language learning process involves bot h conscious and unconscious components. Also, as noted by Naiman, Frohlich and Todesc o (1975), “there are i nherent difficulties involved in asking people about their language learning expe riences and techniques when they are not directly involved in the proce ss of language learning at the time” (p. 69).


26 Questionnaires are easy and convenient to ad minister and do not require students to verbalize the strategies that they use. Furthermore, the questionnaire such as the one employed in Padron and Waxman’s (1988) st udy can help students develop greater awareness of the strategies they em ploy during reading. However, there are disadvantages and limitations to this method. Since stude nts answers are based on the questionnaire, we really do not know about other techniques. A dditionally, as Carrell (1989) notes, Barnett’s (1988) questionnaire was developed and scored in terms of predetermined “correct” responses. That is to say, the predetermined opinions by the researcher about the effectiveness of different reading strategies affected the students’ responses. The questionnaire used in Ruscio lelli’s (1995) study, how ever, also includes “good” and “bad” re ading techniques. Verbal reports, another method for gene rating student feedback, are used to examine the strategies that learners use when reading and when completing comprehension tests. An advantage of verbal reports is that verbal reports provide a detailed description of the comprehension strategies used by students (Block, 1986). However, the verbal reports’ item type can a ffect how learners interact with a text, “in that to answer some items, test-takers had to reread portions of the passage, while for others, test-takers only had to match surface features from the question to the item” (Wolf, 1993, p. 474). For these reasons, ve rbal reports can produce misleading, or inaccurate information. Foreign Language Reading Anxiety There has been much research on FL read ing anxiety. Connected with FL anxiety, reading anxiety is worth inve stigating due to the relati onship between ESL learners’ language anxiety and multiple factors, such as their proficiency le vel and their reading


27 comprehension ability. It is also meaningful to examine the relationship between their anxiety and their background. This section star ts with subtypes of anxiety in the L2 literature. It is followed by a discussion of the effect of a nxiety on L2 reading. Then, the ways in which anxiety affects the L2 read ing process are discussed. Specifically, the effect of anxiety on different reading processes, such as i nput, processing, and output is discussed. Finally, this section e nds with anxiety measurements. Types of Anxiety There are two broad perspectives on the nature of anxiety, which can be categorized as trait and s ituation-specific anxiety (M acIntyre, 1999; MacIntyre & Gardner, 1989, 1991c). Trait anxiety is “a f eature of an individual’s personality and therefore is both stable over time and a pplicable to a wide range of situations” (MacIntyre, 1999, p. 28). Trait anxiety is a pe rmanent predisposition to be anxious. For example, if a student feels anxious genera lly, he has trait anxi ety. Situation-specific anxiety is similar to trait anxiet y, except that it applies only to specific types of situations. According to MacIntyre (1999), examples of si tuation-specific anxiet ies are stage fright, test anxiety, math anxiety, a nd language anxiety. Each situat ion is different, so a person may be nervous in a test, but not on a stag e, or while learning math or a language. Language anxiety is a type of situation-spec ific anxiety and include s the categories of oral, written, and reading anxiety. Speaking in an L2 has been cited as the most anxiety-provoking experience for language learners (Horwitz, Horwitz & Cope 1986; Koch & Terrell, 1991; Price, 1991). Specifically, in Koch and Terrell (1991), stude nts reported oral pres entations, skits, roleplaying, defining words in Spanish, dealing with situations, and charades to be the most anxiety-producing activities. The research on la nguage anxiety suggest s the existence of a


28 relationship between anxiety a nd language learning performan ce. In a study investigating the relationship between L2 classroom anxi ety and L2 writing anxiety, Cheng, Horwitz and Schallert (1999) found that L2 writing anxiety is a language-skill-specific anxiety, whereas classroom anxiety is a more general type of anxiety about learning an L2 in general. Additionally, Saito, Garza, and Horw itz (1999) found that reading in an L2 can also be anxiety-provoking to some students. The interpretation of anxiety (i.e., trait and situation-sp ecific anxiety) has to be conducted using an appropriate measure for the appropriate type of anxiety. For example, to measure language anxiety, we must use a scale that measures situation-specific language anxiety rather than one measuri ng trait anxiety. Using an inappropriate measurement can lead to contradictory re sults. Scovel’s (1978) review of anxiety research showed contradictory results, bot h within and across studies. Young (1991) also pointed out that the relationship between anxiet y and L2 performance is not consistent (p. 426). However, among the sixteen studies de scribed by Young, only three used a scale just for language anxiety. M acIntyre (1999) argues that the results of Young’s survey are inconsistent because she mixed all the different types of anxiety together. In other words, the inconsistency may be due to the fact th at only three of the st udies she produced had distinguished language anxiety as a separate t ype of anxiety. In orde r to be understood as a variable in this study, it is important to define language anxiety separately from other types of anxiety before we can describe its effects (i.e., its relationships to performance). For example, Horwitz (1986) focused her research only on classroom anxiety and used the FLCAS (Foreign Language Classroom Anxiet y Scale) and a Test Anxiety scale to measure the effects of such anxiety on students’ test-taking performance.


29 As MacIntyre (1999) notes, language anxiety is a type of situation-specific anxiety, and, therefore, research on language anxi ety should utilize measures of anxiety experienced in L2 contexts ( p. 29). MacIntyre and Gardner (1 991c) argue that L2 anxiety is best studied with situati on-specific measures. Using the s ituation-specific approach, the process through which a given situation produc es anxiety can be investigated. These researchers suggest that more meaningful and consistent results have emerged from situation-specific measures (MacIntyre & Gardner, 1991c). MacIntyre and Gardner (1994b) define language anxiety as “the feeling of tension a nd apprehension specifically associated with second language contexts, in cluding speaking, listening, and learning” (p. 284). Horwitz et al. (1986) add that L2 anxiety is not simp ly the combination of fears such as communication apprehension, test anxiety, and fear of negative evaluation transferred to L2 learning. According to Horw itz et al., L2 anxiety is unique because it involves learners’ self-concepts of being able to communicate competently and to present themselves genuinely. They define anxiety as “a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language lear ning arising from the uniqueness of the language learning proce ss” (Horwitz, Horwitz & Cope, 1986, p. 128). Several studies support the claim that L2 a nxiety does indeed exist and is prevalent (MacIntyre and Gardner, 1989, 1991b). MacIntyr e and Gardner (1989) found in a study of 104 native English-speaking students that th ey reported significantly more anxiety in their French class than in math or English class. MacI ntyre and Gardner (1991b) used factor analysis to investigat e the relationships among various anxiety scales. There were three scales of anxiety. The first factor wa s general anxiety. This general anxiety was found to include most of th e anxiety scales, such as trait anxiety, communication


30 apprehension, interpersonal a nxiety, and others. The second f actor was state anxiety. The third factor was language a nxiety. The results suggest that language anxiety can be distinguished from other kinds of anxiety, and these results are consistent with studies by Horwitz (1986) and MacIntyre and Gardner ( 1989). Notably, in thes e studies, general anxiety had no correlation with learning. Therefore, MacIntyre and Gardner (1989, 1991b) claimed that language anxiety was distin ct from more genera l types of anxiety and that performance in the L2 was negativ ely correlated with la nguage anxiety but not with more general types of anxiety. According to MacI ntyre(1999) and Saito et al. (1999), language anxiety deve lops from negative experiences particularly early in language learning. General Discussion of L2 Reading Anxiety versus General L2 Anxiety General L2 anxiety has also been shown to be different from L2 reading anxiety. Specifically, whereas general L2 anxiety has been found to be independent of target language, levels of reading anxiety were found to be different according to target language and were related to the specific wr iting systems. Sellers (2000) argued that the Reading Anxiety Scale (RAS) in her study meas ured reading anxiety validly and reliably. She claimed that the RAS identif ies reading anxiety as a distinct variable in L2 learning which correlates positively w ith the scores on the FLCAS fo r her subjects. This suggests that learners with higher leve ls of L2 anxiety also tended to have higher levels of L2 reading anxiety and vice versa. However, in her study, the positive correlation between RAS and FLCAS does not mean that RAS affects FLCAS. Reading anxiety has a correlation with lo wer reading scores. One finding in the correlational study by Saito et al. (1999) indicated that students with higher levels of L2 anxiety tended to also have higher levels of L2 reading a nxiety and vice versa. In their


31 study, students with higher levels of reading anxiety received significantly lower grades than students with lower anxiety levels. Th is relationship was also obtained for general L2 anxiety. Saito et al. (1999), thus, suggested that L2 readin g anxiety, like other types of L2 anxiety, has a negative co rrelation to final grades. St udents who perceived reading their target language as rela tively difficult had significantl y higher levels of reading anxiety than those who perceived it as somewhat difficult, followed by those who perceived reading as relatively easy. However, one criticism of fi nal grade measurement is that the final grade may not be the best measure of a student’s performance because it is a student’s total score in a class; therefore, it would be better to use multiple measurements, as Sellers (2000) did. Although each meas ure of comprehension (e.g., multiple-choice questions or written recall) has disadvantages, using two different comprehension assessment measures may allow for a more accurate assessment of students’ comprehension. Sellers also used tw o different proficiency levels in her study. This is useful for looking at the differences between proficiency levels in terms of anxiety levels. There are several classroom situations wher e reading anxiety might be present, and effect L2 reading performance. For exampl e, reading aloud may also cause anxiety. A substantial number of students feel partic ularly anxious when asked to read aloud. Several researchers argue that language anxiety takes up processing capacity (e.g., Lee, 1999; MacIntyre & Gardner, 1994a, 1994b; Sellers, 2000; Tobias, 1986); for example, Lee (1999) states that “[l]anguage anxiety takes up processing capacity, thereby diminishing language learners’ reading perfor mance” (p. 50). According to Lee (1999),


32 when less processing capacity is availabl e, reading processes will not take place automatically, and, hence, reading is slowed down. As a result, comprehension suffers. MacIntyre and Gardner (1994a, 1994b) inve stigated the effects of anxiety on cognitive processing and found that such eff ects may be quite far-reaching. For example, highly anxious readers may expend part of thei r mental energy thinki ng about things that are completely irrelevant to the reading activ ity. As a result, the r eading process cannot take place automatically or efficiently. In c ontrast, a less anxious reader does not expend energy on these task-irrelevant thoughts, and therefore, has more mental energy to contribute to the reading pr ocess itself (Sellers, 2000). Tobias (1986) claims that anxiety in gene ral may affect learning in all three stages. At the input stage, anxiety can take aw ay from attending to new information and encoding it. At the processing stage, it can interfere with the organization and assimilation of this new information. Finally, at the output stage, anxiety can interfere with the retrieval and production of previously learned material. MacIntyre and Gardner (1994b) develope d three different language anxiety measures: input anxiety, processing anxiety, and output anxiety. Specifically, the input stage is meant to illustrate th e learner’s first experiences with a given stimulus at a given time. The processing stage involves the cogni tive operations performed on the subject matter: organization, storage, and assimilati on of the material. Output involves the production of previously learne d material. Their findings showed that anxious students seem to have difficulty holding distinct verb al items in short-term memory. This may explain why anxious students have trouble comprehending long sentences (Horwitz et al., 1986). These students may also recall less in formation on a reading comprehension test.


33 MacIntyre and Gardner’s (1994b) findings also suggest that, with anxious students, only a small number of verbal statements ente r the processing stage. For example, in MacIntyre and Gardner’s (1994b) study, the Pa ragraph Translation test score in the processing stage demonstrated that anxious st udents were not able to translate a passage as well as their more relaxed counterparts di d. In fact, the poetic na ture of the passage used in this study required students to gue ss at the meanings of some terms, and it appears that the more anxious students did not guess as often as the more relaxed ones. This may reflect anxious students’ unwillingn ess to risk an incorrect or incomplete translation; that is, they ma y avoid responding in order to avoid guessing. The result of this study clearly showed that anxiety affects language learning at each of the three stages, and that the effects appear cumulative. In particular, anxious students tended to take more time for processing. Therefore, when time was limited, performance was hindered at the output stage. Consequently, an xiety occurring at an earlier stage in the reading process may limit the learner’s ability to store relevant information in his/her memory. Thus, he/she may recall less info rmation on a reading comprehension test. An additional aspect of anxiety that ha s been studied in terms of the reader’s “affective state” has to do w ith the reader’s L2 cultural knowledge. In the interactive reading process advocated by Rudell and Sp eaker (1985), a reader’s affective state (including the reader’s interests, attitudes, and values that de cide goals for the reading of a passage) may be altered by text variables such as content and form. These variables might influence the affective state of the re ader. Additionally, in hi s affective model of reading, Mathewson (1985) argues, that if c ontent is culturally unfamiliar, rather than neutral or familiar, reading materials take long er to read, are more difficult to recall, and


34 the readers’ comprehension decreased. He furt her observes that “[c]ulturally compatible reading materials resulted in greater amounts and accuracy of recall than did culturally incompatible materials” (Mathewson, 1985, p. 851). This observation is supported by a study by Saito et al. (1999), which found that students were nervous when they had to read about cultural topics with which they were unfamiliar. According to the study, students felt that the cultural topics should be explained be fore being presented in the target language. These results imply that stude nts are less anxious if they have cultural background knowledge about wh at they are reading. Concerning comprehension, reading anxi ety correlates to the amount of information that readers recall as well as to specific information they recall. Sellers (2000), for example, found that more hi ghly anxious students tended to recall less passage content than did participants w ho experienced minimal anxiety. Not only did highly anxious learners recall less passage conten t, they also recalled fewer ‘high’ pausal units. In her study, a pausal unit is one th at has a pause at each end, during normally paced oral reading. High-level units represen t central ideas, and contribute significantly to the main idea of the passage. Sellers argues that remembering important information is more demanding and requires more mental capacity, because in the processing of important information, readers must organize, interpret, and interrelate the information. Connected with comprehension, anxiety a ffects readers’ ability to focus on a reading task. Sellers (2000) used an instrument called the Cognitive Interference Questionnaire to assess the number of off-task thoughts of each student while reading. Results from the Cognitive Interference Qu estionnaire indicated that highly anxious students tended to experience more off-task, interfering thoughts than their less-anxious


35 counterparts. In other words, highly anxious readers were mo re distracted by interfering thoughts and were less able to concentrate on the task at hand, wh ich affected their comprehension of the reading passage. Students’ beliefs about readi ng are also related to readi ng anxiety. For example, if a student believes that reading is a linear proces s, he believes that every word has the same importance and feels that he must look up ever y new word. Approaching a text linearly is not helpful to comprehension, and therefore, his reading is not efficient (fluent). This lack of efficiency and comprehension may lead him to be anxious (Lee, 1999). Additionally, in the area of L2 learning, Young (1991) cl aims that learner beliefs about language learning are one major source of foreign langu age anxiety. She states that unrealistic beliefs and expectations can lead to frustrat ion and anxiety. Horwitz et al. (1986) mention examples of learners’ beliefs concerni ng the importance of correctness and the unacceptability of guessing, which may also tend to lead to anxiety. Consequently, anxiety affects fluency in terms of speed and particular types of in formation that readers recall and comprehend. Text factors may also affect anxiety levels in reading. For example, inappropriate level of text difficulty may in crease students’ reading anxiet y level. Teachers should pay careful attention to the selec tion of texts, and should choose authentic materials that have an appropriate level of diffi culty; if they choose material s that are too difficult, the anxiety level of some students may increase and they may avoid r eading. Consideration of the reading passage’s level is important, because if the level of the passage is too high for readers, a researcher cannot measure re aders’ comprehension accurately. In other


36 words, anxiety levels of readers may change according to the level of a reading passage or the topic of a text. Readers’ motivation and purpos e also affect anxiety leve ls in reading. Anxieties about language learning may be related to students’ previ ous language learning experiences, their motivations, and thei r self-concepts abou t language learning. Experience in language learning is especially helpful in developing re alistic expectations for and approaches to reading in a second la nguage. Therefore, Saito et al. (1999) argue that teachers should pay partic ular attention to reading strategy instruction with beginners. The knowledge that a nxiety and difficulties are possi ble is often reassuring for many students who feel al one in their reactions. Anxiety Measures Types of measures and limits of these measures There have been several methods used to measure anxiety (e.g., diary, interview, essay, observation, and questionnaire). Since af fective variables are usually not directly observable, data are often based on “inferen ces made by an observe r concerning how the person really feels or thinks or would beha ve under certain circum stances” (Oller, 1979, p. 11). Bailey (1983), in a study investigati ng the relationship between competitiveness and anxiety in adult L2 learning, discusses some advantages of the di ary studies. First, he states that a diary study may re veal hidden affective variables. For example, in a diary of a student in his study, the student saw her inability to compet e with her more proficient classmates as causing her a great deal of anxiety. Second, the diar y studies can provide developmental data. Third, diaries can give researchers insights on many students even within a uniform classroom. Fina lly, he states that “the diar y studies allow us to see the classroom experience as a dynamic and co mplex process through the eyes of the


37 language learner” (p. 98). However, the major criticism of Bailey’s study is that diaries are subjective. The researcher interprets an xiety based only on what students wrote. Price (1991) used student interviews to examine the question of L2 anxiety from the perspective of the anxi ous learner. This method has several disadvantages. For example, if the interview is conducted in the L2 and an interviewee has low language proficiency, he/she may not understand exactly wh at the researcher’s questions are. Also, the questions asked are usually specific questions of inte rest to the researcher, and, therefore, the students’ responses are limited to these types of questions. Because of this, an interview measure can be biased. However, this method has some advantages. The interaction between the resear cher and interviewee can po ssibly elicit more detailed information about the learner’s feelings and thoughts than is availa ble from a diary or essay. MacIntyre and Gardner (1991a) used an e ssay, along with other instruments, to measure anxiety. Each student was asked to re call an experience that required the use of their French skills, in which they felt either very relaxed and conf ident or very nervous and apprehensive. However, this essay method can also be biased, because judges’ subjective opinions are involved in the analysis of the essays. Accord ing to Patten (2000), human judgments are subject to bi ases and other sources of erro rs. This is especially true of qualitative measures like those described above. Zbornik and Wallbrown (1991) conducted re search with children reading in their L1. Zbornik and Wallbrown used the Read ing Anxiety Scale (RAS) questionnaire to measure three domains of reading anxiety: the fears of curiosity, aggression, and independence. However, as Oller (1979) poi nted out, questionnaires have much built-in


38 bias. Bias can also include not giving all possible ideas, cau ses, or outcomes equal chance of being examined or realized. The questionnaire is biased because it limits the number of possible responses a subject can provide. In addition, items on the questionnaire can influence subjects’ choice, due to how an ite m is worded. For example, if the items are only positively worded, the subjects may pos itively respond to the items automatically. Therefore, a researcher should be very car eful with positive and negative wordings in his/her design of questions to avoid subj ects choosing all one scale. Furthermore, respondents may be trying to say what they th ink a questioner or interviewer wants them to say, so that the questione r or interviewer does not get true responses (Oller, 1979). Specifically, Oller (1979) argues that high pr oficiency learners may anticipate how the researcher wants them to respond and thus respond in that way. Oller suggests that students’ proficiency might also bias the resu lts of the questionnaire. According to Oller, students’ responses may be more closely rela ted to the ability to determine intelligently what the “correct” answer is than with thei r actual behavior and feelings. For example, high proficiency learners may have low a nxiety scores because they know what an investigator wants (Oller, 1979; Oller & Perk ins, 1978). In other words, high achievers tend to acquire positive affective tendencies as a result of doing well, whereas low achievers might acquire negativ e attitudes. Additionally, que stionnaires generally use a Likert scale. One of the disadva ntages of using a Li kert scale is that the steps on the scale may not be equal and may miss options that subjects want. However, as Oller (1979) notes, the necessary relia nce of affective measures on self-re ports is an entirely inevitable weakness: “With affective variables the m easurement problem is considerably more complex and is necessarily inferential and indirect ” (p. 18). Oller states that while many


39 sorts of behaviors can be directly observed, attitudes can only be inferred from behaviors and statements of the person in question. Despite the above-mentioned weaknesses, questionnaires are commonly used in correlational studies. The use of correlationa l studies to measure reading anxiety is subject to many limitations and problems. Correlational studies cannot demonstrate a causal relationship, and therefore, lack inte rnal validity. However, correlational studies are used to measure anxiety because it is very difficult to quantify something as abstract as anxiety. Kleinmann’s (1977) study emphasized that anxi ety itself is not a simple, unitary construct that can be comfortably quantified into eith er “high” or “low” amounts. There are some suggestions for partially compensating for the disadvantages of correlational studies. As s uggested by Bailey (1983), one way of minimizing the problems of inference in research on affectiv e variables is to ask readers directly how they feel or think, either in carefully word ed questionnaires or in terviews, instead of relying only on observation. By doing so, resear chers can better get data on students’ attitudes and feelings. Additionally, Scove l (1978) explained the advantage of questionnaires and self-reports in that they are much more precise in focusing in on anxiety than physical measures lik e perspiration, heart rate etc. Foreign Language Classroom Anxiety Sc ale (FLCAS)/Foreign Language Reading Anxiety Scale (FLRAS) In L2 research, the FLRAS, a questionnaire, is the central instrument that has been used for measuring reading anxiety. The FLRAS was developed to measure anxiety related to L2 reading. It contains 20 Likert scale items scored on a 5-point scale. The theoretical range of the FL RAS scale is from 20 to 100. The FLRAS items include various aspects of reading, students’ percepti ons of reading difficulties in their target


40 language, and their perceptions of the relative difficulty of reading as compared to the difficulty of other language skil ls (Saito et al., 1999). In th is sense, the FLRAS is valid because the items are the ones which readi ng anxiety is supposed to be measured. Additionally, the FLCAS, a nother questionnaire commonl y used in correlational studies, also holds some value as a measur ement of classroom anxiety. It contains 33 Likert scale items also scored on a 5-point scale. The items reflect communication apprehension, test-anxiety, and fear of negative evaluation in the L2 classroom (Horwitz et al., 1986). Horwitz (1986) reported that internal consis tency using Cronbach’s alpha was 0.93, based on a sample of 108 subjects. Test -retest reliability ove r a period of eight weeks was r = 0.83 ( p = 0.001), for 78 subjects. In addi tion, higher scores on the FLCAS were associated with lower expected gr ades in the foreign language class ( r = -0.52, p = 0.001), and lower actual final grades ( r = -0.49, p = 0.003 in Spanish class; r = -0.54, p = 0.001 in French class). Although Sellers (2000) attempted to invest igate the relationship between students’ L2 reading anxiety and their reading comp rehension, and the relationship between the students’ reading anxiety and th e reading process itself, there are several criticisms of and limitations to her study. First, her level divisi on was based on semester or course levels, rather than a standardized test. Conse quently, results of Sellers’ study lacks generalizability. Another critic ism centers on the author’s se lection of text material. Sellers did not consider the leve l of text although she mentione d this as a limitation of her study. The consideration of the reading passage’s level might be important because if the level of the passage is too high for stude nts, a researcher cannot measure their comprehension accurately. Even though traine d native speakers of Spanish divided the


41 reading passage into pausal units, the selection of text based only on general and familiar topics to university students is not suffici ent. In other words, anxiety levels of participants may vary according to the level of reading passages or the topic of the text. The limitation of this study is also related to the limitation of co rrelational studies in general. In other words, si nce correlational stud ies cannot identify a causal factor, she cannot interpret her findings that the RAS (Reading Anxiety Scale) affects the FLCAS and her study therefore also lacks internal validity. Summary In summary, previous research has shown that L2 learners differing in their degree of proficiency may employ different modes of processing while reading. L2 learners with relatively lower levels of pr oficiency tend to use a bottom-up mode of processing, while high proficiency L2 learners prefer a top-down mode of pr ocessing. Moreover, results of previous research on the effect of text adjuncts on L2 read ing comprehension suggest that the facilitative effect of a text adjunct varies as a function of readers’ proficiency level, as well as of text type. Recent studies also show ed that anxiety level may affect L2 reading comprehension. That is, highly anxious student s tend to recall less passage content than do participants who experience minimal anxiety. Significance of Current Study Lee and Riley (1990) examined the effect of text adjuncts for the comprehension of expository prose. They found that the expanded framework helped improve novice foreign language readers’ comprehension of a loosely-structured type of passage (collection of descriptions), but not of a tig htly-structured type (problem-solution). One of the purposes of the present study is to re plicate the results of Lee and Riley (1990). Specifically, the present study examines whethe r the use of an expanded framework helps


42 improve ESL learners’ comprehension of a loos ely-structured text more than that of a tightly-structured text. The present study employs three te xts that have different structures in order to compare the effects of text adjuncts according to text structure. There have been contradictory results on th e effects of text adjuncts depending on ESL learners’ proficiency levels. Therefore, the present study also inves tigates the effects on L2 reading comprehension of two different type s of text adjuncts (i .e., definition lists and frameworks) in different proficiency le vels (Low and High). By measuring the participants’ English proficiency reliably, a relationship between pr oficiency levels and reading anxiety levels can be examined. Statement of Purpose The main purpose of the present study is to investig ate the effects of differing types of text adjuncts on L2 reading compre hension. Specifically, this research seeks to answer the question of whether L2 learners fr om various proficiency levels benefit from the provision of vocabulary and framework text adjuncts. The interaction between anxiety level and degree of proficiency level is explore d. Moreover, the difference in reading strategies used by low and high pr oficiency L2 learners is also examined. Research Questions and Hypotheses Research Questions The following research questi ons guided this study: 1. Do high and low proficiency ESL learners benefit from different types of text adjuncts (definition list, framework)? 2. Is there an interaction between text type and text adjunct within each proficiency level? 3. Is there any relationship between ESL learne rs’ anxiety levels a nd their proficiency levels?


43 4. Is there any relationship between ESL learners’ reading strategies at each proficiency level and their r eading comprehens ion ability? Hypotheses 1. It is hypothesized that the ‘expanded fr amework’ sub-group will have lower recall scores than the ‘vocabulary’ sub-group in the Low proficiency (L) group, because the expanded framework sub-group will lack the vocabulary from the passage they need to understand the passage. 2. It is also hypothesized th at the ‘expanded framework’ sub-group that receives information on the text types will do better than the ‘vocabulary’ sub-group that receives a vocabulary list in the High proficiency (H) group. Because of their proficiency level, they will not need passage vocabulary, but they will need information about the rhetori cal structure of an expository text to fully understand the passage like a native speaker would.


44 CHAPTER 3 METHODOLOGY Participants The subjects who participated in my st udy consisted of a heterogeneous group of 61 University of Florida ESL (English as a Second Language) student s, representing 15 different languages (Table 3-2). They were recruited from the English Language Institute (ELI), the Academic Spoken English (ASE ) program and the Scholarly Writing (SW) program at the University of Florida in th e spring semester of 2003. Students at the ELI are international students with variou s language backgrounds, and their English proficiency level is generally lower than th e students in the ASE and SW programs. In addition, students at the ELI are divided in to several levels based on a standardized placement test, which is administered at th e beginning of the semester. Most of the students at the ELI have lived in the U.S. for only a short time, and have not taken the TOEFL or the GRE prior to thei r admission to this program. Students in the ASE program, however, are international students who are attending graduate school at th e University of Florida. Typica lly, they are older, and have stayed in the United States for a longer period of time, and are taking ASE classes because their graduate student assistan tships depend on successful completion ASE examinations. ASE classes are divided into three levels (I, II, and III), based on proficiency. English proficiency levels of students in the ASE program are generally higher than ELI students and Un like ELI students, all of the ASE students have taken the TOEFL and GRE before they matriculated in to the University of Florida. Like ASE


45 students, students in the Sc holarly Writing Program are in ternational grad uate students who need to take English writi ng classes as a result of scor ing the written component of their entrance examination. For the purposes of my study I recruited students from high intermediate and advanced classes at the ELI as well as all ASE and Scholarly Writing Program competency levels. Students who volunteer ed from the ASE and Scholarly Writing Program were instructed to individually co me to my office. In the case of the ELI students, I was granted permission by their inst ructors to use their class time to collect data. Due to variations in proficiency, the pa rticipants were divided into two groups Higher, and Lower where all of the participants in the Lower group were part of the ELI. The average length of residence in the U.S. for the Higher group was 2 years, 2 months; and that for the Lower group was only 4 months (Table 3-1). The length of residency has a strong correlation to profic iency, and is, therefore, a key variable in this study. Table 3-1. Participants’ average length of residence in the U.S. and average age Group Av. length of residence in the U.S. (m) Av. age (y) High (N = 31) 26.55 29 Low (N = 30) 4.67 25 English Proficiency Test In order to divide students into higher and lower proficiency levels, I used the Test of Adolescent and Adult Language (TOAL). For measuring overall proficiency, it is beneficial to use a test that includes the four language sk ills (listening, speaking, reading, and writing). As Alderson, Krahnke and Stansf ield state, “Proficiency tests measure the test taker’s overall ability in English along a broad scale” (1987, p. iv). A standardized test is a more reliable assessment of L2 proficiency than more subjective assessments,


46 such as impressionistic judgments, in-house te sts, or institutional status. It was assumed that the students in the ELI were Lower pr oficiency students and that those in the ASE and Scholarly Writing programs were Higher prof iciency students. I originally based the placement of students into Higher and Lower groups based on their program affiliation. To ensure that these groupings matched actual student proficiency I measured the Table 3-2. Characterist ics of participants Characteristic Number of participants Language background Arabic 1 Chinese 9 Chinese/Taiwanese 1 Farsi 1 Hebrew 1 Japanese 5 Korean 24 Mandarin 1 Punjabi/Hindi 1 Russian 1 Spanish 9 Thai 3 Tigrinya 1 Turkish 2 Vietnamese 1 Gender Male 36 Female 25 Age range 19 to 42 years Materials students’ overall English pr oficiency in a reliable way using the TOAL to ensure the accuracy of this assumption. The TOAL-3 (Test of Adolescent and Adult Language-Third Edition) is a standardized test that measures a student’s ove rall English proficiency level. The test has an internal consistency, test-retest, and inter-rater reliability, and the reliability coefficients exceeded 0.80 indicating that te st error was minimal The TOAL-3 is not biased with regard to race or genderand is designed for student s between the ages of 18 to


47 24 (Hammill, Brown, Larsen & Wiederholt, 1994 ). This standardized test has been created to test language skill s of both native speakers of English and ESL students, making it one of the most difficult standardiz ed exams for ESL students. The TOAL is extremely thorough, in that it includes te n different subsections : listening, speaking, reading, writing, spoken language, written language, vocabulary, grammar, receptive language, and expressive language in this st udy. For this study, four subtests of the TOAL-3 were used: listening/vocabulary, r eading/vocabulary, reading/grammar, and writing/grammar. I administered these subtests of the TOAL-3 to all participants. I included only the students who finished all four subsections of th e TOAL and the three reading sections. For my study, I chose the TOAL-3 over the TOEFL exam. Even though the TOEFL exam is a rigorous and reliable inst rument for testing stude nt proficiency, there are several reasons why I did not choose th e TOEFL as a proficiency test. First, many ESL students have taken a TOEFL-type test be fore they entered the U.S. Even though a test itself may be known to be reliable, it may not be reliable enough because many students are accustomed to merely memorizing th e test format, and, as a result, the test score may not correlate to a student’s Eng lish proficiency. To avoid this problem, a researcher could use the st udent’s past TOEFL score. This, however, may not be indicative of the student’s Eng lish proficiency at the time of the experiment. Therefore, for the reason of consistency and immediacy, the TOAL was chosen as a test of the ESL students’ overall English proficiency. Background Questionnaire In addition to the standardized test, a background questionnaire, which consisted of the following parts: (1) participants’ L1 and L2 backgrounds, as well as their English


48 learning experiences; (2) some items on readi ng anxiety when readi ng English; and (3) some items on reading strategies that they employ when reading English (Appendix A) was administered to all participants. The pur pose of part I (participants’ L1 and L2 background) was to create a correlation be tween the participants’ backgrounds with factors, such as proficiency levels, anxiety levels, and r eading ability. In order to accurately evaluate a subject’s ‘reading anxi ety’ items, I employed the Foreign Language Reading Anxiety Scales (FLRAS), which is developed by Saito, Garza, and Horwitz (1999). Likewise for the purpose of identifyi ng a subject’s ‘reading strategies’ items, I used the Reading Strategy Questionnaire (RSQ) (Carrell, 198 9; Pritchard, 1990; Rusciolelli, 1995).The FLRAS reveals the subjec t’s preferred reading strategies within a field of three catagories: t op-down reading strategies, botto m-up reading strategies, and one reading belief (Table 3-3). Item 9 on th e RSQ was “Rate your reading skill in your native language,” and was scored on a fivepoint Likert scale, ranging from (1) very good, to (2) fairly good, (3) average, (4) be low average, and (5) bad. This kind of questioning provides a useful function of dete rmining the reader’s L1 proficiency, which registers as a key f actor in this study. Since FLRAS was designed for native speake r-participants, it has to be modified for ESL participants for use in this study. Th e FLRAS consisted of six items and the RSQ consisted of nine items. The two questionnaires were scored on a five -point Likert scale, ranging from (1) ‘strongly disagree’ to (5) ‘s trongly agree.’ To insure the clarity and comprehensiveness of the FLRAS, the items on the questionnaire were pilot-tested with 18 ESL students at the English Language Inst itute at the University of Florida (Park, 2001). For the five items (items 1-5) in th e FLRAS, a higher score signifies a higher


49 degree of anxiety. Through the evaluation of anxiety, I am able to examine the relationship between learners’ proficiency levels and their anxiety levels. I can also investigate possible relationships between learners’ anxiety level and their reading comprehension ability. Through the RSQ, I am ab le to identify which reading strategies are used by fluent readers and which read ing strategies are em ployed by non-fluent readers. Table 3-3. Types of FLRAS and specific questions asked Types of FLRAS Item Reading anxiety (Items 1-5) 1. When reading English, I often understand the words but I don’t always understand what the author is saying. 2. When I’m reading, I am nervous if I don’t know the topic. 3. I get upset when I don’t understand the grammar in a sentence when reading English. 4. When reading English, if I don’t understand every word, I get nervous and confused. 5. I feel very uncomfortable when a teacher asks me to read aloud in class. Classroom activity (Item 6) 6. I like to do group work in English class. Table 3-4. Types of reading strate gies and specific questions asked Types of reading strategies Item Top-down reading strategies 1. When reading in English, I skip the words I don’t know and continue reading. (Items 1, 2, 3, 7) 2. I read the title and imagin e what the article might be about. 3. I start reading and try to figure out the meaning as I go along. 7. I think it's helpful if a teacher gives me information about a text before I read Bottom-up reading strategies 4. I usually translate into my native language when I read an English passage. (Items 4, 5, 6) 5. I read every wo rd and look up the ones I don’t know. 6. I usually pronounce words silently, and when I don't know how to pronounce a word, I think I won't get the right meaning. Reading belief (Item 8) 8. I think readi ng out loud to myself is helpful.


50 Reading Texts For the main experiment, all participants were asked to read three ‘expository’ texts written by native speakers of English that were not simplified or edited in any way. In general, an expository text does not contain as much culturally embedded background knowledge as a literary text. Therefore, it is ea sier to control for the text effect with an expository text than with a narrative text. The texts used by previous researchers we re used in this present study (Carrell, 1985, 1992; Connor, 1984; Meyer, Brandt & Bl uth, 1980; Oh, 1990). First, by using the same texts, results of this study can be direc tly compared with previ ous results. As such, I am able to interpret the reading results according to students’ proficiency levels. All participants from both groups were asked to read a loosely-st ructured text, titled “The Olympic Games” (a collection of descript ions). Since both groups read the passage, I could compare the effect of text adjuncts in each group for this type of passage. Next, I selected a set of texts specific to the partic ipants’ proficiency leve l. I chose the texts “Supertankers” and “Closing down nuclear pow er plants for the group with the higher proficiency. These readings em ploy tightly-structured heuris tics, consisting of problem and solution, and comparison/contrast format s, respectively. For th e lower proficiency group, I chose “On being fat in America” and “Nuclear energy versus solar energy.” These texts also are written with tightly-s tructured heuristic or ganization, problem and solution, and comparison, respectively (Appendix B). Therefore, each student read a total of three passages, two of whic h were proficiency specific. Procedure At the outset of this study, I divided each proficiency level into three sub-groups: two treatment groups and a cont rol group. I pilot-tested the te xts with 6 students (3 ELI


51 students and 3 international gr aduate students) in March 2003. The purpose of the pilot study was to test the feas ibility of my procedure, as well as to check the difficulty of the texts. The pilot study did not reveal a ny major flaws in the overall methodology; therefore, all students who participat ed underwent the same procedure. The testing of the participants took pl ace over four meetings. As I mentioned before, the ELI instructors allowed me to us e their classrooms to conduct the study for the Lower group. Members of the Higher group me t with me four times individually in my office. On the first day of treatment, I gave the participants informed consent forms and the background questionnaire. On the sec ond day of treatment, I administered the TOAL-3. Due to the unique conditions of the ELI cl assroom modifications had to be made to the TOAL-3 testing. For example, ELI readi ng/writing classes are scheduled in two-hour blocks, and it is difficult to continue the test for more than two hours. Additionally, the speaking subtest is supposed to be measured with each student, so it is not best to administer the test with a group of part icipants. Therefore, for feasibility and methodological reasons, I chose four subt ests of which included writing. On the third day, I gave the students th e “The Olympic Games” passage. Students were asked to read and reread the passage for 20 minutes. After all the participants finished with their reading, they were allowe d to write the recall pr otocol for 20 minutes. Accordingly, they had 40 minutes total. The ‘vocabulary’ groups received a definition list before they read the passage. I selected vo cabulary words that I considered essential to comprehending the reading. Participants in the ‘expanded framework’ sub-groups received written information on the text type s (collection of desc riptions, problem and


52 solution, and comparison/contrast) before they read the passage (Appendix C). For example, if the rhetorical st ructure of the text employed a problem/solution heuristic, the expanded framework explained that the text presented a problem and suggested a solution for the problem at the end. I modeled the expa nded framework in this study after Lee and Riley (1990); i.e., information about the rhetorical structur e of an expository text was used as a text adjunct. After the participants finished reading the passage, for the recall protocol I read the instructi ons, explaining that they should write down as much as they could remember from the passage without refe rring back to the passage. The participants were encouraged to use complete sentences a nd to use the words in the passage or their own words (Appendix D). On the fourth day, I tested the Lower group using the passage “On being fat in American” and the Higher group using the passa ge “Nuclear energy ve rsus solar energy.” The procedure was repeated exactly the same as that used on the third day. The students were asked to write in Englis h for all recall protocols, which is an important factor in my study. Lee (1986a, 1986b) found that reporti ng of recall in the native language produced a more accurate asse ssment of subjects’ true understanding than did recall reported in the foreign la nguage. His suggestion, however, works only for a homogeneous group when I understand the L1. In the present study, learners had many different language backgrounds, and, therefor e, were asked to recall in English, which would have had an effect on their performance. Another important factor in my procedure was that before the participants began reading, they were told that they would be as ked to immediately recal l the text after they were finished. By using the immediate recall protocol, I was able to control for memory


53 effects. Some concern may be raised with re gard to the short-term memory issue when students recall. However, if a student does not comprehend a text, it is likely that he/she will have difficulty remembering its content. Th erefore, the recall protocol is not greatly affected by short-term memory (Smith, 1994) (Note that the memory effect was not tested in this study.) Scoring All questions in the subtests of the TOAL that were used in the present study were scored as 0 or 1. The listen ing/vocabulary, reading/vocabul ary, and reading/grammar are multiple-choice question-type tests. There are two correct answers for each item in all of the three tests, and to receive a score of 1, bot h answers had to be correct. Because of this fact, it is not easy to answer correctly if a subject simply guesses. Also, if a subject has one right answer and one wrong answer for a question, he or she receives a score of 0. For example, in the listening/vocabulary test participants were asked to choose two pictures which were representative of two m eanings of a homophone (i.e., palm) that was produced (Table 3-5). In the listening/vocabul ary test, students were supposed to choose two pictures which were most representative of the word spoken. This section could be difficult if the students knew only one meani ng of the word. In the reading/vocabulary test, they were supposed to choose two answ ers which were most like the words in the example. In the reading/grammar test, they were supposed to choos e two sentences that had the most similar meaning. Unlike the first three sections, the f ourth section focused on writing/grammar where participants were aske d to combine a set of sentences into one sentence. Guidelines for the writing/grammar section included examples of permissible and impermissible answers for each item. For in stance, item 5 consisted of the sentences, “The girl looked frightened” and “She wasn’t frightened.” Permissi ble answers included


54 “The girl (She) looked scared (frightened) but she wasn’t,” or “Even though she looked frightened, she wasn’t.” An example of an im permissible answer is: “She was frightened but didn’t look it.” Each student ’s overall English proficienc y score was the sum of the four subtests’ raw score. Table 3-5. Examples of subtests Subtest Example 1 (Listening/Vocabulary) [palm] 5 (Reading/Vocabulary) [red, green, blue] A. yellow B. circle C. orange D. light 6 (Reading/Grammar) A. Sam plays. B. Sam will not play. C. Sam has played. D. Sam is playing. E. Sam is going to play. 8 (Writing/Grammar) We ate lunch. It was an hour ago. Since the writing/grammar subtest was an open-ended question, the answers were varied. For the scoring of this subtest, I di scussed students’ answer s with a native speaker of English to ensure that they would be scor ed fairly. The native speaker was 24 years old at the time of the study and had worked as a teaching assistant in the Academic Spoken English program at the University of Florid a for 2.5 years. After I listed all of the students’ answers that were not in the guidelines, the native speaker and I went over the responses item by item. A student’s score repr esents the sum of co rrect answers on these four subtests. For scoring of the recalls, I followed Carre ll’s (1992) study by di viding the text into “idea units,” because it is important to have a small idea unit in th is study. Having small idea units prevents cases of participants producing half-correct and half-incorrect


55 responses in one idea unit. I did not weigh th e idea units by level of importance, because the purpose of the reading comprehension test was to observe total number of idea units recalled, not type of idea un its recalled. Basically, each idea unit consisted of a single clause (main or subordinate, including adverbial and relativ e clauses). Each infinitive, gerund, and conjunction was cons idered a separate idea un it. Moreover, prepositional phrases were identified as separate idea uni ts. If an idea was complex, it was broken into smaller idea units. Students’ scores consisted of the total number of idea units that they successfully recalled. The following are the num ber of idea units for each of the five texts: “The Olympic Games” (100); “Supertanke rs” (104); “Closing down nuclear power plants” (98); “On being fat in America” (142); “N uclear energy versus solar energy” (91). For my study I did not choose to employ Meyer’s (1985) recall protocol scoring system, which has been used frequently in previous research. Meyer’s system identifies the structural characteristics as well as the lexi cal units of a passage, and for this reason is important for researchers. Moreover, the pr ocedure helps to meas ure the relationship between passage type and level of performa nce. Because the system arranges the idea units hierarchically, its use provides a dem onstration of not only which lexical and relational units are recalled, but also from which portion of the structure those units are remembered. However, there are disadvantag es to the Meyer recall protocol system. First, the Meyer recall protocol requires a c onsiderable amount of time, and hence is not efficient. Also, it requires sufficient expert ise to develop instruments (Bernhardt, 1991, pp. 201-203). For these reasons, Carrell’s system had better applica tions to my study. After the scoring templates were finalized for each text, I graded a total of 183 recall protocols. The scoring template of each reading passage was determined based on


56 the same rules. Generally, I chose flexible scoring. If a subject paraphrased and the meaning of the idea unit was the same as in the original idea unit, it was considered correct. More importantly I was concerned w ith ideas, not grammar. Hence, a different usage of an auxiliary verb or unimportant gr ammatical mistakes were considered correct idea units. I decided that rephrasing of an idea unit was acceptable response before grading, because the central c oncern is having a subject and a verb together in an idea unit. I checked with a native speaker in order to obtain native speaker intuition on the accuracy of the participants’ responses. When there was a conflict, we discussed the issue and came to an agreement on an analysis. Statistical Analyses A one-way analysis of variance (ANOVA) was performed to determine statistical significance of the two proficiency groups, (H igher and Lower), in terms of the TOAL. An ANOVA was also performed to determin e statistical signifi cance of the subgroups, (“control”, “vocabulary” and “expanded fram ework”) in each proficiency group. In order to examine the relationship between students’ anxiety levels and recall scores, TOAL scores and recall, and LOR and recall, a Pe arson product-moment correlation coefficient was performed. Descriptive statistics were used to describe each subgroup’s responses on the FLRAS and the RSQ.


57 CHAPTER 4 RESULTS This chapter includes the results of th e TOAL-3, the recall protocols, and the participants’ background information. For th e TOAL-3, descriptive statistics and the results of a one-way analysis of varian ce (ANOVA) of the results are presented. The relationship between the students’ English pr oficiency, as measured by the TOAL-3 and their reading ability, as measured by the recall is examined. Also, the relationship between the students’ reading proficiency in English, as measured by the subtest 6 (Reading/Grammar) and their reading ability, as measured by the recall is examined. Finally, this chapter includes the results of the questionnaires, i.e., the FLRAS, the Reading Strategy Questionnaire (RSQ), and the relationship between the results of these questionnaires and other factors. Information concerning the st udents’ length of residence is also discussed. Results of Test of Adolescent and Adult Language (TOAL-3) The purpose of administering the TOAL-3 was to systematically measure the students’ proficiency levels. Four subtests were administered in the study: Subtest 1 (Listening/Vocabulary), subtest 5 (Reading/ Vocabulary), subtest 6 (Reading/Grammar) and subtest 8 (Writing/Grammar). A total raw sc ore (120) is the sum of the four subtests’ raw scores. The mean percentage of accura te responses of the Higher (H) group was 65.65% and that of the Lower (L) group was 48.25%. In order to determine if there is a statis tically significant difference between the H and the L groups in terms of the TOAL-3 ra w score, a one-way analysis of variance


58 (ANOVA) was performed. As expected, the re sults were statistically significant ([ F (1, 59) = 47.473, p = 0.000]). This result confirmed that participants in the H group are significantly more pr oficient than those in the L group. Standard Score for TOAL-3 The TOAL-3 standard scores provide an indication of a person’s subtest performance. The mean of the TOAL-3subtes t standard scores is always 10 and the standard deviation is 3. A signifi cant aspect of this system is that a standard score on one subtest may be compared with a standard sc ore on any other subt est (Hammill et al, 1994, p. 36). Table 4-1 gives the guide lines for describing standard scores and the guidelines apply to each subtest. Table 4-1. Guidelines for describing standard score Standard score Description 17-20 Very superior 15-16 Superior 13-14 Above average 8-12 Average 6-7 Below average 4-5 Poor 1-3 Very poor Table 4-2. Total standard score and Mean standard score Group Total standard score (61) Mean standard scoreMean percentage High (N = 31) 1096/1891 35.35 57.96 Low (N = 30) 762/1830 25.40 41.64 Table 4-3. TOAL-3 total standa rd score by subtest for H group Subtest (# of total standard score) Lowest Highest Mean SD 1 (16) 3 12 7.52 1.75 5 (15) 4 12 7.71 1.62 6 (14) 7 14 10.65 1.87 8 (16) 6 12 9.48 1.61


59 Table 4-4. TOAL-3 total standard score by subtest for L group Subtest (# of total standard score) Lowest Highest Mean SD 1 (16) 1 9 5.07 2.20 5 (15) 3 10 5.77 1.59 6 (14) 3 12 7.70 2.32 8 (16) 2 11 6.87 2.62 The total standard score (61) is the sum of the four s ubtests’ standard scores. The total maximum score is 16 for subtest 1, 15 for subtest 5, 14 for subtest 6, and 16 for subtest 8. As shown in Table 4-2, the mean pe rcentage for the standard score of the H group was 57.96% and 41.64% for the L group. An ANOVA showed that there was no significant difference among the subgroups (control, vocabulary, and expanded framework group) on the total standard score in either the H or the L group. However, there was a significant difference between the two proficiency groups, on the total standard score ([ F (1, 59) = 47.559, p = 0.000]). For example, an ANOVA showed that there was a significant difference between the proficiency groups on subtest 1’s standard score ([ F (1, 59) = 23.302, p = 0.000]), on subtest 5’s standard score ([ F (1, 59) = 22.378, p = 0.000]), on subtest 6’s standard score ([ F (1, 59) = 28.849, p = 0.000]), and on subtest 8’s standard score ([ F (1, 59) = 22.227, p = 0.000]). The lowest score for the TOAL-3 total standard score for the H group was 27 and the highest score was 46 out of 61. The lowest score for the TOAL-3 total score for the L group was 9 and the highest score was 36 out of 61 (H group, M: 35.35, SD: 4.52; L group, M: 25.40, SD: 6.59). The TOAL-3 presented normative tables for subtests for scoring based on the standard scores and percentiles. Both H and L groups had the highest mean standard score of 10.65 and 7.70, respectively, for subtes t 6 (reading/grammar). Also, both groups had the lowest mean standard score of 7.52 and 5.07, respectively, for subtest 1


60 (listening/vocabulary). The H group had a mu ch higher mean score (9.48) than the L group (6.87) in subtest 8 writing/grammar (Tables 4-3 and 4-4). The results show the average standard scores for both groups. Fo r the H group, the mean standard score was 8.84 (SD: 1.13), which is in the range of 8–12 ‘average’ according to Table 4-1. For the L group, the mean standard score was 6.35 (SD: 1.65), which is in the range of 6–7 ‘below average’. Standard Scores for TOAL-3 by Subgroups Within each group, the mean and standard deviation of each subgroup was evaluated based on standard scores. Of the H subgroups, the control group had the highest score, followed by the expanded fr amework group, which in turn was followed by the vocabulary group (Table 4-5). The mean and standard deviation of subgroups in the L proficiency group were also calculated in Table 4-6. The expanded framework group had the highest score, followed by the vocabulary group, and the control group had the lowest score. In order to investigate statistical significance among the subgroups at each proficiency level, a one-way ANOVA wa s performed. There was no statistically significant difference among the subgroups in either the H or L proficiency group ( p = 0.282 and p = 0.082, respectively). Therefore, I assu me that the subgroups’ proficiency levels are the same. I discuss the interrelations hip of students’ profic iency level and text adjuncts in detail in the discussion chapter. Tables 4-7 and 4-8 give the ANOVAs fo r the H and L subgroups on each of the TOAL-3’s subtests. There were no statis tically significant differences among the subgroups for each subtest in the H group. In the L group, there were no statistically significant differences among the subgroups for any subtest except for subtest 6 (Reading/Grammar) ([ F (2, 27) = 5.145, p = 0.013]). A post-hoc Bonferroni test revealed


61 that both the vocabulary group and th e expanded framework group performed significantly better than the control group ( p = 0.018 and p = 0.046, respectively). Table 4-5. Mean and SD of total standard score in H group Subgroup Mean SD Control 36.55 5.07 Expanded framework 35.90 4.31 Vocabulary 33.50 3.89 Table 4-6. Mean and SD of total standard score in L group Subgroup Mean SD Expanded framework 27.27 6.34 Vocabulary 27.00 6.20 Control 21.33 6.12 Table 4-7. Standard sc ore in H group (ANOVA) Subtest df MS F p 1 Between groups 2 2.62 0.85 0.439 Within groups 28 3.09 Total 30 5 Between groups 2 4.73 1.92 0.165 Within groups 28 2.46 Total 30 6 Between groups 2 0.99 0.27 0.765 Within groups 28 3.68 Total 30 8 Between groups 2 6.40 2.76 0.080 Within groups 28 2.32 Total 30 Table 4-8. Standard score in L group (ANOVA) Subtest df MS F p 1 Between groups 2 2.12 0.42 0.660 Within groups 27 5.02 Total 29 5 Between groups 2 2.00 0.78 0.469 Within groups 27 2.57 Total 29 6 Between groups 2 21.57 5.15 0.013 Within groups 27 4.19 Total 29 8 Between groups 2 11.47 1.75 0.192 Within groups 27 6.54 Total 29


62 Recall Protocols This section examines raw scores and mean percentages of recalls for all texts, which are listed in Table 4-9 along with the number of idea units in each. The first text, “The Olympic Games,” read by both profic iency groups, yields crucial comparable results. Table 4-10 presents mean percenta ge of subgroups for the H group and Table 411 presents mean percentage of subgroups for the L group. An ANOVA showed that there was a statistically significant difference in recall scores between the two proficiency groups for this passage ([ F (1, 59) = 58.857, p = 0.000]), suggesting that the reading ability of the H group is significantly higher than the reading ability of the L group. For this loosely-structured text, the expanded framework subgroups for both H and L groups benefited most from the text adjunct. The expanded framework group for H had the highest score, while the ot her treatment group, the voca bulary group, had the lowest score. However, an ANOVA showed that th ere was no significant difference among the three groups in the H group ([ F (2, 28) = 3.126, p = 0.060]). In contrast, the two treatment groups for the L group had a better sc ore than the control group (Table 4-11). An ANOVA showed that there was a signifi cant difference among the three groups in the L group ([ F (2, 27) = 6.040, p = 0.007]). A post-hoc Bonferroni test revealed that the vocabulary group recalled significantly more than the control group ( p = 0.024), and that the expanded framework group also recalled sign ificantly better than the control group ( p = 0.010). Therefore, for the loosely-structur ed text, both the expanded framework text adjunct and the vocabulary list text adjunc t were helpful for the L group students. For the ‘problem and solution’ passage, each proficiency group read a text chosen to match students’ proficiency level. The hi gher proficiency student s read “Supertankers”


63 Table 4-9. Reading text and number of idea units Group Reading text Text type Number of idea units Both groups The Olympic games Collection of descriptions 100 High Supertankers Problem/solution 104 Closing down nuclear power plants Comparison/contrast 98 Low On being fat in America Problem/solution 142 Nuclear versus solar en ergyComparison/contrast 91 Table 4-10. Mean (%) of subgroups for “ti ghtly-structured” passages and for “all passages” (H group) Group Subgroup P/S C/C Mean of tightlystructured passages C/D Mean of all passages H C 41.52 28.57 35.05 43.36 37.82 V 47.21 33.98 40.60 37.80 39.66 E 63.08 54.39 58.74 59.90 59.12 P/S: problem/solution text, C/C: comparison/contrast text, C/D: collection of descriptions text; C: control group, V: vocabular y group, E: expanded framework group Table 4-11. Mean (%) of subgroups for “ti ghtly-structured” passages and for “all passages” (L group) Group Subgroup P/S C/C Mean of tightlystructured passages C/D Mean of all passages L C 11.03 10.26 10.65 6.56 9.28 V 16.76 20.44 18.60 16.40 17.87 E 20.49 17.08 18.79 17.36 18.31 and the lower proficiency st udents read “On being fat in America.” For the tightlystructured text read by the H group, “Supe rtankers,” the expanded framework group again had the highest reading score, follo wed by the vocabulary group, and the control group with the lowest score. However, th ere was no significant difference among the three groups. For the tightly-structured text read by the L group, “On being fat in America,” the expanded framework group also had the highest reading score, followed by


64 the vocabulary group, and the control group had the lower score. However, again, there was no significant difference among the three groups. For the “comparison/contrast” passage, each group of students again read a text chosen according to their proficiency leve ls. The H group students read ‘Closing down nuclear power plants’ and the L group students read “Nuclear versus solar energy.” For the tightly-structured text, “Closing down nuclear power plants,” the expanded framework group once again had the highest reading score, followed by the vocabulary group, followed by the control group. In this case, there was a signi ficant difference among the three groups ([ F (2, 28) = 5.662, p = 0.009]). A post-hoc Bonferroni test revealed that the expanded framework group re called significantly more than the control group ( p = 0.010), but not more than the vocabulary group. For the tightly-structured text, “Nuclear versus solar energy,” the vocabulary group had the highest recall score, followed by the expanded framework group, followed by the control group. There was a significant difference among the three groups ([ F (2, 27) = 4.207, p = 0.026]). A post-hoc Bonferroni test rev ealed that the vocabulary group recalled significantly more than the control group ( p = 0.024). For the H group, the expanded framework group scored the highest average in both tightly-structured passages: th e problem/solution and compar ison/contrast passages. For the L group, the expanded framework group be nefited most for the problem/solution passage but the vocabulary group benefited most for the comparison/contrast passage. In the H group, an ANOVA showed that ther e was a significant difference among the subgroups ([ F (2, 59) = 7.185, p = 0.002]). A post-hoc Bonferroni test revealed that the expanded framework group did better than the control group ( p = 0.002), and the


65 vocabulary group ( p = 0.025). In other words, for the tightly-structured passages, the expanded framework text adjunct was more help ful than the vocabulary list text adjunct in the H group. In the L group, an ANOVA show ed that there was also a significant difference among the subgroups on the tightly-structured texts ([ F (2, 57) = 4.038, p = 0.023]). A post-hoc Bonferroni test reveal ed that the expanded framework group did better than the control group ( p = 0.031), but not better th an the vocabulary group. Within the H group, the expanded framework group achieved the highest percentage score for all passages. For th e tightly-structured texts, the expanded framework group scored the highest with th e vocabulary group placing second. However, for the loosely-structured text, the contro l group recalled more than the vocabulary group. When all texts are considered together, the expanded framework groups showed the highest mean percentage score among three groups. In the H group, an ANOVA showed that there was significance between th e subgroups on all texts together ([ F (2, 90) = 10.005, p = 0.000]). A post-hoc Bonferroni test revealed that the expanded framework group did better than the control group ( p = 0.000), and better than the vocabulary group ( p = 0.001). Within the L group, the expanded framework groups showed the highest percentage score for both the collection of de scriptions passage and the problem/solution passage. For the comparison/contrast text however, the vocabulary group achieved the highest percentage score. When all text s are considered together, the expanded framework groups showed the highest perc entage score, although there was not a significant difference between the vocabular y group and the expanded framework group. An ANOVA showed that there was a signi ficant difference among the subgroups on all


66 texts together ([ F (2, 87) = 7.856, p = 0.001]). A post-hoc Bonferr oni test revealed that the expanded framework group did better than the control group ( p = 0.001), and that the vocabulary group also did bett er than the control group ( p = 0.005). Effects of Text Structures according to Proficiency Level Table 4-12 presents the mean recall percen tage of all passages for the H group and Table 4-13 presents the mean recall percen tage of all passages for the L group. The students in both groups recalled the most for the problem/solution text. Table 4-12. Mean (%) for “tigh tly-structured” passages and for “all passages” (H group) Group P/S C/C Mean percentage of tightlystructured passages C/D Mean percentage of all passages H 50.60 38.98 44.79 47.02 45.53 Table 4-13. Mean (%) for “tightly-structured” passages and for “all passages” (L group) Group P/S C/C Mean percentage of tightlystructured passages C/D Mean percentage of all passages L 16.09 15.93 16.01 13.44 15.15 In order to determine a statistically si gnificant difference among the three text types, the collections of descriptions, problem/solution, or comparison/contrast in each proficiency group, a one-way ANOVA was pe rformed. There was a statistically significant difference among the text types in the H group ([ F (2, 90) = 3.376, p = 0.039]). A post-hoc Bonferroni test rev ealed that the mean recall score of problem/solution text was significantly higher than that of the comp arison/contrast text (Mean: 52.32, 37.87, respectively) ( p = 0.035). Moreover, there was also a statistically significant difference among the text types in the L group ([ F (2, 87) = 7.215, p =


67 0.001]). A post-hoc Bonferroni test revealed th at the mean recall of problem/solution text was significantly higher than that of the collection of descript ions text (Mean: 23.30, 13.80, respectively) ( p = 0.003), and that the mean recall of problem/solution text was significantly higher than th at of the comparison/contrast text (Mean: 23.30, 14.70, respectively) ( p = 0.007). The overall results reveal that ther e was no difference between the looselystructured text and the tightly -structured texts in the H gr oup. However, participants in the Lower proficiency group recalled signifi cantly more idea units from the tightlystructured text than from the loosel y-structured text type (Mean: 13.80, 19.00, respectively) ([ F (1, 88) = 4.303, p = 0.041]). Relationship between Proficiency and Recall Scores In order to examine a relationship between the students’ proficiency levels and their recall scores, a Pearson product-moment correl ation coefficient was performed. Table 414 gives the summary of the co rrelations between the TOAL-3 and all readings. Overall, a moderate correlation was found between participants’ raw TOAL-3 scores and the number of ideas units they recalled from th e collection of description passage. However, this correlation was significant only among partic ipants in the higher proficiency group ( r = 0.443, p = 0.013) but not among those in the lower proficiency group ( r = 0.618, p = 0.000). There was no relationship between the st udents’ English proficiency, as shown by the TOAL-3, and the recall scores of the problem/solution passages for either group. However, there were correlations for the comparison/contrast pa ssages for both groups ( r = 0.398, p = 0.027 for the H group, and r = 0.472, p = 0.009 for the L group). Since there were significant correlations between the TOAL -3 scores and only some of the reading


68 scores, as shown in Table 4-14, a student’s English proficiency doe s not always affect reading in a second language. Table 4-14. Correlations between TOAL-3 and readings Group C/D (%) P/S (%) C/C (%) H 44 22 40 p = 0.013 p = 0.027 L 13 29 47 p = 0.009 Both groups 62 p = 0.000 Tables 4-15 and Table 4-16 are summary tabl es of the mean number and percentage of idea units recalled per passage and per treatme nt type. Standard deviations are given in parentheses. Since there were no statistica lly significant differences among subgroups for either H or L group, the mean percentage of each subgroup per passage in the following tables informs the text adjunct effect. Table 4-15. Mean number and percentage of idea units recalled correctly per passage and per text adjunct in H group Control Vocabulary Expanded framework Row mean # % # % # % # % C/D (N = 100) 43.36 43.36 37.80 37.8059.90 59.90 47.02 47.02 (22.92) (15.01) (22.64) (22.02) P/S (N = 104) 43.18 41.52 49.10 47.2165.60 63.08 52.63 50.60 (22.35) (15.91) (27.35) (23.66) C/C (N = 98) 28.00 28.57 33.30 33.9853.30 54.39 38.2 38.98 (14.16) (13.75) (24.37) (20.59) Column mean 38.18 37.82 40.07 39.6659.6 59.12 (19.81) (14.89) (24.79) Varying Effects of Text Adjuncts according to Text Structure In the H group, an ANOVA showed that there was a statistically significant difference among the subgroups for th e comparison/contrast text ( p = 0.009). For the collections of descriptions text, the differe nce among the subgroups was not significant


69 Table 4-16. Mean number and percentage of idea units recalled correctly per passage and per text adjunct in L group Control Vocabulary Expanded framework Row mean # % # % # % # % C/D (N = 100) 6.56 6.56 16.40 16.4017.36 17.36 13.44 13.44 (2.35) (8.04) (9.45) (8.71) P/S (N = 142) 15.67 11.03 23.80 16.7629.09 20.49 22.85 16.09 (11.80) (11.67) (16.59) (14.37) C/C (N = 91) 9.33 10.26 18.60 20.4415.55 17.08 14.49 15.93 (3.87) (9.96) (5.77) (7.80) Column mean 10.52 9.28 19.6 17.8720.67 18.31 (6.01) (9.89) (10.60) but neared significance ( p = 0.060). For the problem/solution text, the difference of the subgroups was not significant bu t again neared significance ( p = 0.079). In the L group, an ANOVA showed that there was a statis tically significant difference between the subgroups for the collections of descripti ons and the comparison/contrast texts, p = 0.007 and p = 0.026, respectively. However, there was no significant difference between the subgroups for the ‘problem/solution’ text ( p = 0.112). These results s uggest that the text adjuncts help the lower-level proficiency ESL students to read in an L2 more than the higher-level proficie ncy ESL students. In the H group, the expanded framew ork group recalled the highest mean percentage of idea units per text adjunct fo r all texts together (Table 4-15). An ANOVA showed that the expanded framework group re called significantly more than both the vocabulary and the control groups for all texts together. Therefore, the expanded framework text adjunct was more helpful th an the vocabulary text adjunct. The mean difference between the vocabulary group and the control group was not significant. Individually, the text adjunct effect was only significant for the comparison/contrast text.


70 For this text, there was a significant mean difference between the expanded framework group and the control group (Means: 53.30, 28.00, respectively). In the L group, although the expanded fr amework group recalled the highest mean percentage of idea units per text adjunct for all texts together, the mean difference between the expanded framework group and the vocabulary group was not significant (Table 4-16). An ANOVA showed that the expanded framework group recalled significantly more than did th e control group, as did the voc abulary group, for all texts together. Therefore, I claim that both text adjuncts provide a benefit for the lower proficiency ESL students. The mean diffe rence between the expanded framework group and the vocabulary group was not significant. Furthermore, in the L group, the text adjunct effect was significant for the ‘collections of de scriptions’ and the ‘comparison/contrast’ texts. For the ‘colle ctions of descriptions’ text, there was a significant mean difference be tween the vocabulary group an d the control group (Means: 16.40, 6.56, respectively), and between the expa nded framework group and the control group (Means: 17.36, 6.56, respectivel y). For the ‘comparison/contrast’ text, there was a significant mean difference be tween the vocabulary group an d the control group (Mean: 18.60, 9.33, respectively). Tables 4-17 and 418 are ANOVA tables on the readings in each proficiency group. This section was intended to examine wh ether there is a re lationship between reading proficiency, as determined by the T OAL-3, and L2 reading ability, as measured by recalls. For the reading profic iency, the scores of subtest 6 scores of the TOAL-3 were used.


71 Table 4-17. ANOVAs on the readings in H group df MS F p C/D Between groups 2 1327.83 3.13 0.060 Within groups 28 424.75 Total 30 P/S Between groups 2 1392.92 2.79 0.079 Within groups 28 500.11 Total 30 C/C Between groups 2 1830.64 5.66 0.009 Within groups 28 323.29 Total 30 Table 4-18. ANOVA on the readings in L group df MS F p C/D Between groups 2 339.82 6.04 0.007 Within groups 27 56.27 Total 29 P/S Between groups 2 447.90 2.38 0.112 Within groups 27 188.54 Total 29 C/C Between groups 2 209.59 4.21 0.026 Within groups 27 49.82 Total 29 Relationship between Reading Proficiency and Reading Ability There was a moderately strong correlation between the reading proficiency and the recall score for “The Olympic games” passage that was given to all participants ( r = 0.590, p = 0.000). In other words, if a student had a high reading proficiency score on the TOAL-3, s/he also recalled more of this passage than a student with a low reading proficiency score, and vice versa. Looki ng at the groups indi vidually, a significant correlation was found between th e reading proficiency and th e recall score in the H group ( r = 0.497, p = 0.004). However, there was no such correlation between the reading proficiency and the recall score in the L group.


72 Summary There were no statistically significant differences in scores on the TOAL-3 among the subgroups in each proficiency group. Ther efore, we are able to claim that the subgroups in the H group all have the same hi gher English proficiency and those in the L group all have the same lower English prof iciency. The H group’s expanded framework subgroup performed at the highest level of proficiency for all pa ssages, receiving the highest mean percentage across all passages. For the L group, the expanded framework group performed at the highest le vel of proficiency only for th e collection of descriptions and the problem/solution texts. However, if we look at the column mean in Table 4-16, the expanded framework group received the high est mean percentage across all passages. For the tightly-structured texts, the two types of text adjuncts, (i.e ., expanded framework and vocabulary) both worked well. For exampl e, in the H group, the expanded framework groups received the best score, followed by the vocabulary groups. In the L group, the expanded framework group received the best score for the problem/solution passage, while the vocabulary group received the best score for the comparison/contrast passage. For the loosely-structured text, the control group in H recalled more idea units than the vocabulary group. If we look at the column mean closely, only the expanded framework text adjunct was very helpful for the higher proficiency ESL student s, since there was a significant difference between the mean per centage of the vocabulary group and the expanded framework group (39.66% and 59.12%, respectively). However, both expanded framework and vocabulary text adjuncts we re helpful for the lower proficiency ESL students, and there was not a significant difference between the mean percentage of the vocabulary group and the expanded framework group (17.87% and 18.31%, respectively). Because of this correlation, I will discuss the effect of text adjuncts


73 thoroughly in the discussion chapter. Additi onally, students at both proficiency levels received the best recall score fo r the problem/solution passage. In terms of the relationship between the TOAL-3 scores and recall scores, we do not have consistent results. In other words, there were correlations for the collection of descriptions text and the comparison/contra st text. However, there was no relationship between the problem/solution passages and the TOAL-3 for either proficiency group. This would imply that students’ English prof iciency did not always correlate with their reading ability. Background Information Questionnaire Relationship between FLRAS and Other Factors Some second language researchers have ex amined the notion that ESL students’ background information plays a role in L2 (s econd language) reading. In this section, I examine three factors from the background questionnaire employed in the study: the foreign language reading a nxiety scale (FLRAS), the R eading strategy questionnaire (RSQ), and Length of Residence (LOR). Th e FLRAS was administered to determine whether there is a relationship between a student ’s anxiety level and proficiency level, as measured by the TOAL-3, and a student’s anxiet y level and reading ability, as measured by the recalls. Tables 4-19 and 4-20 give th e average score by item of the FLRAS. A Likert scale was used in the FLRAS questionn aire with 1 being ‘strongly disagree’ and 5 being ‘strongly agree’. Table 4-19. Average score by item in the FLRAS for H group 1 2 3 4 5 6 E 2.9 2.1 2.8 2.1 3.3 3.6 V 2.8 2.9 2.2 2.1 3 4 C 2.82 2.64 2.18 2.73 2.64 3.27 Av. 2.84 2.55 2.39 2.31 2.98 3.62 E: Expanded framework group, V: Vo cabulary group, C: Control group


74 Table 4-20. Average score by item in the FLRAS for L group 1 2 3 4 5 6 E 2.91 2.91 2.82 3.45 2.27 3.36 V 3.5 3.1 2.4 3.3 1.9 3.1 C 3.56 3 2.78 3.22 2.89 3.56 Av. 3.32 3.00 2.67 3.32 2.35 3.34 For items 1 to 5, a higher score signifies th at a student has more anxiety in foreign language reading. The results showed that th e H group students were less anxious than the L group on these five items. The average score of all subgroups in the H group was around 2 ‘strongly disagree’ on items 1 to 5. However, the subgroups in the L group answered 2 ‘strongly disagree’ only on item 3, “I get upset when I don’t understand the grammar in a sentence when reading English,” and item 5, “I feel very uncomfortable when a teacher asks me to read aloud in cl ass.” A moderate negativ e correlation between the TOAL-3 raw score and the average scor e of FLRAS items 1-5 for both groups was found ( r = -0.402, p = 0.001). This means that if a st udent is highly anxious, her/his English proficiency is low, and vice versa. A moderate negative correlation between the FLRAS score and the collection of descri ptions passage for both groups was found ( r = 0.351, p = 0.006). This suggests that if a student is highly anxious, his reading score for this passage will be low, and vice versa. Pearson product correlations indicate that there were relationships between the students’ anxiety levels and th eir reading scores for the H group. Specifically, there was a moderate negative correlation between FLRAS and the ‘problem /solution’ passage in the H group ( r = -0.423, p = 0.018) as well as for the ‘com parison/contrast’ passage ( r = 0.444, p = 0.012).


75 Unlike the H group, there were no signifi cant relationships between FLRAS and the recall scores on the ‘pr oblem/solution’ passage and the ‘comparison/contrast’ passages in the L group. Relationship between Reading Strategy Qu estionnaire (RSQ) and Proficiency Level It is meaningful to compare the read ing strategy tendencies between the two groups. Within the H group, the average score on the top-down reading strategies was 4.03 (agree) and on the bottom-up reading strate gies was 2.52 (disagree). Therefore, topdown reading strategies were used more by the higher proficiency students than bottomup reading strategies. By contrast, within th e L group, the average score on the top-down reading strategies was 3.83 (neither agree nor disagree) and on the bottom-up reading strategies and it was 3.08 (n either agree nor disagree). Two ANOVAs showed that the difference between the top-down reading strate gies and the bottom-up reading strategies in the H group was significant ( p = 0.000), and that the di fference between the two strategies in the L group was also significant ( p = 0.000). Additionally, separate ANOVAs revealed that the difference be tween H group and the L group on the bottomup reading strategies was significant ( p = 0.005) as well as the di fference between the H group and the L group on the top-down read ing strategies was not significant ( p = 0.135). The reading strategies that were queried ranged from the student’s translation practice to focused reading exercises. For exam ple, on item 4, “I usually translate into my native language when I read an English passa ge,” the H group show ed a lower average score than the L group, which means that the H group students admit to translating less into their native languages when they read an English passage than the L group students. An ANOVA showed that the difference be tween the H group and the L group on item 4 was significant ([ F (1, 59) = 4.113, p = 0.047]). There was a significant difference


76 between the two groups on item 6, “I usually pronounce words silently, and when I don’t know how to pronounce a word, I think I won’t get the right meaning.” ([ F (1, 59) = 7.841, p = 0.007]). The average score for this question by the H group was 2.13 and for the L group, it was 2.91. On item 7 “I think it’s helpful if a teacher gives me information about a text before I read ,” both groups answered 4 ‘agree .’ On item 8 “I think reading out loud to myself is helpful,” the average scores for both groups showed they thought reading out loud was helpful (Tables 4-21 and 4-22). Table 4-21. Average score by item in the RSQ for H group Item E V C Av. 1 4.2 4 4.18 4.13 2 4 4.2 3.73 3.98 3 3.9 4 4 3.97 4 2.3 3.1 2.73 2.71 5 2.8 2.4 2.91 2.70 6 2 2.3 2.09 2.13 7 4 4.2 4 4.07 8 3.4 3.8 3.55 3.58 Table 4-22. Average score by item in the RSQ for L group Item E V C Av. 1 3.91 3.6 3.78 3.76 2 4.09 3.8 3.89 3.93 3 4.18 3.9 3.22 3.77 4 3.36 3.4 3.22 3.33 5 2.82 3.6 2.56 2.99 6 2.82 2.7 3.22 2.91 7 3.91 3.9 3.67 3.83 8 4.36 3.5 3.56 3.81 Item 9, “Rate your reading skill in your na tive language,” was intended to compare the students’ reading ability in the L2 with their self-rated reading ability in their native languages. Table 4-23 gives the average scores of item 9 on the RSQ.


77 Correlation between L1 reading ability and L2 reading score on TOAL-3 To test for correlations between students’ self-rated L1 readin g ability and their reading score on the TOAL-3, item 9 on pa rt III of the background information questionnaire and subtest 6 (reading/gram mar) of the TOAL-3 were employed. The results showed that there was no relationshi p between the students’ self-rated first language reading skill and L2 reading profic iency for all groups together. Also, while there was no relationship between the students’ self-rated first langua ge reading skill and reading proficiency for the L group, there was a moderate relationshi p between these two factors for the H group ( r = 0.356, p = 0.049). Table 4-23. Average score of item 9 on the RSQ for both groups Group Subgroup H E 1.7 V 2 C 2.36 Av. 2.02 L E 2 V 2.3 C 2.78 Av. 2.36 Correlation between L1 reading ability and L2 recall score To further examine the correlations between students’ self-rated L1 reading ability and their L2 reading ability, I examined item 9 on part III of the background information questionnaire, and the students’ recall scores The results suggested that there was a relationship between L1 readi ng ability and the recall scor e on “The Olympic games” passage for neither all participants nor the pa rticipants in the H group alone. Furthermore, there were no relationships between L1 r eading ability and the recall scores on other passages for the H group. However, there was a relationship between L1 reading ability and the recall score on “The Olympic games” passage for participants in the L group ( r =


78 -0.424, p = 0.020) and between L1 reading ability and the recall score on the “On being fat in America” passage ( r = -0.366, p = 0.046). Correlation between Length of Residence (LOR) and Other Factors Overall, it was found that there was no re lationship between the students’ anxiety level and LOR. This was true for both th e higher and lower proficiency groups. On the other hand, there was a strong relationshi p between the TOAL-3 score and LOR ( r = 0.625, p = 0.000). However, further analysis re vealed that this relationship was significant only among participants in the higher proficiency group ( r = 0.582, p = 0.001).


79 CHAPTER 5 DISCUSSION The results of this study indicate seve ral key connections between the role of proficiency, the effects of text adjuncts, a nd the interrelationship between L2 learners’ proficiency and reading in an L2. To unders tand the importance of the findings, I will first recapitulate the main results from th e previous chapter as a foundation for the following discussion. To begin with, there was a significant difference between the H group and the L group in terms of the TOAL score. Second, there was no significant difference among the subgroups within each proficiency group on the TOAL score. Third, the expanded framework group’s recall score for the H group was significantly higher than either the vocabular y or the control group for all texts considered together. In the L group, for all texts considered toge ther, the vocabulary group’s and the expanded framework group’s recall score were both si gnificantly higher th an the control group. However, there was no significant differen ce between the expanded framework group and the vocabulary group in the L group. Four th, there was a str ong correlation between the students’ English proficiency and the ‘colle ction of descriptions’ passage. Fifth, there was a moderate negative correlation between the TOAL raw score and the average score of the five FL reading anxiety items (anxi ety score). Finally, there was a moderate negative correlation between the anxiety sc ore and the ‘collecti on of descriptions’ passage for all participants. The most important conclusion that I will dr aw in this discussion is the effects of text adjuncts, the relationsh ip between English proficienc y and second language reading


80 ability, the relationship between FL readi ng anxiety and reading in an L2, and the relationship between students’ r eading strategies and reading ab ility. It also discusses this study’s contribution to previous research in L2 reading. Interrelationship between L2 Proficiency and Text Adjunct Since there was no significant differen ce on the TOAL score among the subgroups in each proficiency group, it was assumed that those subgroups had a similar if not the same proficiency level. However, the two t ypes of text adjuncts provided differential benefits according to proficiency level. Although the expanded framework proved most beneficial for the H group, both the expa nded framework and the vocabulary list benefited the L group equally. As expected, the L group students did not have enough vocabulary knowledge for reading a passage. As such, the vocabulary text adjunct was helpful to the lower level proficiency students. By contrast, the data indicates that it is possible that the H group students did not need additional vocabulary knowledge to read the texts, suggesting that the expanded framew ork is a much more important textual aid. The results of the present st udy support the theory that inefficient readers tend to use low-level text processing skills. As me ntioned previously, supplemental vocabulary information was helpful for low-level stude nts. Although high prof iciency students can use information from outside the text to understand the text when they meet unknown words, low proficiency students may not be ab le to use other information if they do not have sufficient vocabulary knowledge. The interrelationship between a student’s proficiency and the effect of text adjuncts provides an interesting correlation, but it cannot be used to predict successful reading. Within the H group, the control group received the highest pr oficiency score in terms of mean percentage on the TOAL but did not receive the highest recall score.


81 Instead, the expanded framework group received the highest recall score for all texts. Within the L group, the vocabulary group received the highest proficie ncy score in terms of mean percentage on the TOAL but receiv ed the highest recall score only for the comparison/contrast text. This finding suggests that L2 proficiency may help the students understand a text but it does not alwa ys guarantee successful L2 reading. Recall that the subgroups’ differences were only statistically significant for the comparison/contrast text for the H group a nd were statistically significant for the collection of descriptions a nd the comparison/contrast texts in the L group. In other words, although the expanded framework group in the H group benefited most for each text, the results were only meaningful for th e comparison/contrast passage. For this text, the expanded framework group benefited most from text adjuncts, followed by the vocabulary group. These results suggest that text adjuncts facilitate L2 reading depending on the text. For the comparison/contrast text for the H group and for the collection of descriptions and the comparison/contrast text s for the L group, the control groups at each proficiency level received the lowest scores Hence, students who receive supplemental text adjuncts can possibly improve their reading performance. Interaction between Text Type and Text Adjunct within Each Proficiency Level This study supports the claim that a text adjunct enhances reading comprehension. In the H group, the text adjunct effect was significant only for th e comparison/contrast text. It is noted that the text adjunct eff ect neared significance for the collections of descriptions passage. In the L group, the te xt adjunct effect was significant for the ‘collections of descriptions’ and the ‘comparison/contrast’ te xts. Since the text adjunct effect was significant for two texts in th e L group, unlike the H group, these results


82 suggest that the text adjuncts help the low-le vel proficiency ESL student s to read in an L2 more than they help the high-le vel proficiency ESL students. Although the expanded framework group student s recalled significantly more than either the control group students or the vo cabulary group students for all three texts together in the H group, the difference be tween the expanded framework group and the vocabulary group was not significant for all three texts together in the L group. The general explanation that I posit is that both text adjuncts be nefited the lowe r proficiency ESL students. I argue that the expanded fram ework text adjunct was the most helpful to the higher proficiency ESL students when they read the comparison/contra st text. Although the expanded framework text adjunct was the most helpful when the lower proficiency ESL students read the collections of descrip tions (loosely-struc tured type) and the problem/solution (tightly-structured type) pass ages, the vocabulary text adjunct was also very helpful. These results are different from Lee and Riley’s (1990) study. They concluded that an expanded framewor k helped improve their novice readers’ comprehension of a loosely-structured type of passage (description co llection), but not of a tightly-structured type (problem-solution). Since Lee and Riley’s re sults were different from the present study’s results, I discuss se veral possible reasons of the differences. For, example, in Lee and Riley’s st udy, the participants were inex perienced native speakers of English learning French. Since they used cl ass enrollment in order to determine their participants as novice foreign la nguage readers, we really do not know the proficiency of the students, or the effect of their proficie ncy. In my study, the prof iciency of students in the L group is significantly lowe r than that of students in the H group. However, this does


83 not mean that the students in the L group are really low-proficiency students. Similarly, the proficiency of students in the H group is si gnificantly higher than that of students in the L group but this does not mean that the students in the H group are really highproficiency students. Therefore, when th e “lower” level students were given the expanded framework text adjunct before they re ad a text in my study, they may not really be real novice ESL learners. The results from the present study provide many suggestions for researchers/teachers of a second language. Firs t, giving information about a text structure to ESL students was very helpful for both le vels of the ESL students in comprehending the text. Second, vocabulary knowledge was as im portant as the information about a text structure for the lower-proficiency student s. ESL teachers should be aware of this characteristic in particular when they teach L2 reading to low-proficiency students. As several previous researchers have noted, teaching vocabulary items related to the passage is much more important than teaching vocabulary items randomly. In other words, building vocabulary without cons idering a reading passage is almost not useful to a student’s reading ability. This can explain the limitations of short-term memory when an ESL student tries to recall a reading passa ge without understanding what was read. Since there was no main difference be tween the vocabulary subgroup and the expanded framework subgroup for all texts, I cl aim that both text adjuncts were helpful for the lower proficiency student s in reading L2 texts. It is known that low proficiency students tend to use bottom-up reading strategies when reading an L2 text. Hence, it is assumed that the low proficiency students n eed a vocabulary text adjunct to comprehend an L2 text. However, as the results of th e present study showed, bot h the vocabulary list


84 text adjunct and the expanded framework te xt adjunct were helpful for the lowerproficiency ESL students when they read some texts. Yet, when the lower-proficiency students read the comparison/contrast text, only the vocabulary group among the subgroups recalled significantly more than th e control subgroup. This suggests that the vocabulary list text adjunct was helpful for th e low-proficiency student s to read a tightlyorganized L2 text. The purpose of giving a text adjunct before the students read the L2 text is to activate the students’ backgr ound knowledge, and, therefore, the text adjunct helps the students’ L2 reading comprehe nsion. As I discussed before, it is not necessary that the vocabulary list text adjunct activate the botto m-up reading process and that the expanded framework text adjunct activate the top-down reading process. As most recent reading researchers agree, the most successful L2 reading comprehension depends on an interactive process (Fender, 2001; Nassaji, 2002 ). It is assumed that low proficiency students do not have enough vocabulary knowledge to read fluently. It is also assumed that the low proficiency students tend to use bottom-up reading st rategies. The present study results suggest that both te xt adjuncts helped the lowe r proficiency students activate their background knowledge, and, thereby, help ed them comprehend an L2 text. On the other hand, high-level ESL students or native speakers might use other textual or extra-textual information when th ey are not quite sure of the meaning of unknown words. This is related to Hudson’s (1982) notion of ‘behind the eyeball’ knowledge of efficient readers. Therefore, efficient reader s keep using an interaction between bottom-up and top-down reading processes.


85 Among the three passages, the students in both groups recalled most for the ‘problem/solution’ passage. This suggests th at a passage which has a tight structure is easier for ESL students to read than a passage which has a loose struct ure. If we compare the recall scores of the loosel y-structured text with those of the two tightly-structured texts together, students in the H group recalled more of the loos ely-structured text than of the tightly-structured texts, 47.02% and 44.79%, respectively. This suggests that a wellorganized text structure ma y not be important for the high-proficiency students. However, the L group students recalled more of the tightly-structured texts than the loosely-structured text, 16.01% and 13.44%, resp ectively. Because of their relatively low proficiency level, the low-prof iciency students find a tightly-str uctured L2 text easier to comprehend. The results imply that a well-organi zed text structure is more important to low level ESL students. Role of English Proficiency in L2 Reading Another important factor which has been di scussed in L2 reading is L2 proficiency. Taillefer (1996) found that as the reading ta sk became more cognitively difficult, the role of linguistic ability became more important. Al so, several researcher s, such as Clarke (1980) and Kern (1989), claimed that a learne r’s L2 proficiency le vel is an important factor in determining L2 reading comprehens ion. Clarke (1980) argue d that if a student did not achieve some minimal level of profic iency in the L2, (i.e., a threshold), he/she could not comprehend a text successfully or r ead the way native speakers read. It is noted that one can determine whether there is a correlation between English proficiency and reading ability, but one cannot determine cau sation. However in accord with previous research, I claim that my study shows that higher proficiency ES L students’ reading


86 ability was better than the lower proficiency ESL students’ reading ability when they read the same text. The results of the relationshi ps between the students’ proficiency and their reading scores from the present study were not consis tent. Recall that all st udents read a looselystructured text, the ‘collec tions of descriptions’ passage There was a statistically significant relationship between L2 proficiency, as measured by the TOAL-3, and recall of the ‘collection of descriptions’ passage for all participants. Each proficiency group read two tightly-structured texts, which differe d to match their profic iency level. In the H group, there was no significant relationship between students’ proficiency and the “Supertankers” passage. However, there was a relationship between their proficiency and the “Closing down the nuclear power plan ts” passage. In the L group, there was no significant relationship between their profic iency and the “On being fat in America” passage. However, there was a relationship be tween their proficiency and the “Nuclear versus solar energy” passage. These results might be related to the role of the text. Both texts which did not have a corr elation with students’ profic iency in both the H and the L groups were problem/solution passages. This is probably because the text difficulty level was too high for the students, or the stude nts did not have enough background knowledge about the text. These inconsiste nt results could be interpre ted as showing that an L2 learners’ English proficiency itself is not sufficient for successful reading in an L2. Relationship between ESL Learners’ Anxiet y Levels and their Proficiency Levels Individual differences such as anxiety have played a role in L2 reading. Reading in an L2 requires many factors such as t opic knowledge, vocabulary knowledge, and L2 proficiency. Other than these factors, anxiety might be a nother important aspect which


87 some highly anxious L2 learners must cope w ith when reading in an L2. Reading anxiety could prevent ESL readers from reading successfully. The results show that there was a negativ e correlation between the TOAL raw score and the average score of the ‘reading anxiet y’ items. The negative correlation suggests that ESL students’ anxiety impedes their L2 proficiency. While this was only a moderate relationship, it was statistically significant. Since we used a standardized test, we can generalize the results from the pr esent study to the ESL population. Relationship between Students’ Anxiety Level and Reading Ability Several researchers have attempted to examine whether ther e is a relationship between ESL students’ anxiety le vels and their readi ng ability (Saito et al. 1999; Sellers, 2000). The results of this study show that there was a negative relationship between reading anxiety and the collecti on of descriptions passage for all participants. The results confirmed that ESL students’ reading a nxiety played a role in their reading comprehension. As previous studies have shown, anxious students had a difficult time recalling information from a text. In th e H group, there was a relationship between reading anxiety and both the problem/solu tion passage and the comparison/contrast passage. The results imply that highly anxious students tend to have low recall scores in the H group. However, there were no such re lationships between reading anxiety and the reading passages in the L group. This result contradicts previous research, in the sense that lower proficiency students’ reading anxiety level did not really correlate with their reading ability. The L group st udents might have a topic knowledge on the passages, which explains that there was no correlation between their reading anxiety and the recall scores. It is noted that I measured the ES L students’ reading comp rehension ability by using recall protocols, maki ng the results generalizable.

PAGE 100

88 Relationship between ESL Learners’ Re ading Strategies and their Reading Comprehension Ability It has been claimed that there is a re lationship between reading process, (i.e., bottom-up, top-down or interactive) and r eading ability. Many researchers have attempted to show that successful readers us e specific types of reading strategies and unsuccessful readers use other types of r eading strategies. In the present study, the reading strategies on the RSQ were intended to represent students’ preferred reading processes. For example, the item on the reading strategy questionnaire (RSQ), “When reading in English, I skip the words I don’t kno w and continue reading,” was intended to show evidence of a top-down reading process. On the RSQ, items 1, 2, 3 and 7 were topdown reading strategies; items 4, 5 and 6 were bottom-up reading strategies; and item 8 was a reading belief. Results of this study indicat ed that top-down reading strategies were employed more frequently by all students. Within th e H group, the average score on the top-down reading strategies was 4.03 (agree) and on the bottom-up reading strategies was 2.52 (disagree). Therefore, top-down reading stra tegies were preferred more by the higher proficiency students than bottom-up reading st rategies. By contrast, within the L group, the average score on the topdown reading strategies wa s 3.83 (neither agree nor disagree) and on the bottom-up reading st rategies, it was 3.08 (n either agree nor disagree). Separate ANOVAs showed that the difference between the two strategies in the L group was significant, and that the di fference between the H group and the L group on the bottom-up reading strategies was signifi cant. However, the difference between the H group and the L group on the top-down r eading strategies was not significant. Although the average difference between the top-down and bottom-up reading strategies

PAGE 101

89 in the L group was not as big as the H group, I claim that the lower proficiency students also used more top-down strategies than botto m-up strategies when they read in an L2, which challenges the notion th at unsuccessful ESL readers tend to use bottom-up reading strategies more when they read an L2 text. Another implication of my study is that the data permits an investigati on to determine the relationship between self-rated reading skills in students’ na tive languages and their L2 reading ability. The averages of item 9 on the RSQ, “Rate your reading skill in your native language,” for the H group and for the L group were 2.02 and 2.36, respectively, with ‘2’ meaning ‘fairly good.’ The correlation between the self-rated reading skill and the “Olympic games” passage was significant only for the L group. This means th at L1 literacy transfers to L2 literacy especially for low-proficiency students. Sin ce the correlation was si gnificant only for the L group, second language researchers should pay attention to the low-proficiency students’ assessment of their L1 r eading skill for better L2 reading. Relationship between Reading Strategies and Proficiency To understand the relationship between read ing strategies and proficiency it is necessary to look at the aver age score by item. For the H gr oup, Item 1 had the highest average score, followed by Item 7. In other words, the higher-pro ficiency ESL students employed these top-down reading strategies most when they read in an L2. Item 6 had the lowest average score. For the L group, Item 2 had the highest average score, followed by Item 7. The result of my study suggested that even the low-proficiency students used topdown reading strategies more than bottomup reading strategies. ESL researchers and teachers should be aware of this and try not to focus only on bottom-up reading strategies when they instruct reading strategies to low-proficiency students. As for the L group, Item 6 had the lowest average score.

PAGE 102

90 Students in both groups thought that receiving information about a text before they read would be helpful. Since Item 6 had the lowest average score for both groups, the students in this study thought there woul d be no relationship between knowing the pronunciation of a word and gett ing the right meaning. For Item 8, there was an average score of 3.58 and 3.31 for the H group and the L group, respectively. Hence, the students neither agreed nor disagreed about the benef its of reading out loud. These results suggest that the tendency between th e high proficiency students and the low proficiency students on the reading strategies is si milar, which contradicts many of the previous studies. Such a contradiction suggests that fu rther research needs to invest igate how reading strategies and proficiency relate. Correlation between LOR and Other Factors Relationship between LOR a nd Students’ Anxiety Level Several researchers in anxiet y studies have been interest ed in the fact that ESL students’ backgrounds, including their native languages and th e length of residency in a target country, play a role in the students’ anxiety levels. In the present study, there was no relationship between the stud ents’ anxiety level, as meas ured by the ‘reading anxiety,’ and the LOR, for both proficiency groups cons idered together or for each proficiency group individually. The small number of items on L2 reading anxiety might be the reason for the lack of a clear rela tionship between these items. Relationship between LOR and Students’ Proficiency Level There was a strong relationship between the students’ TOAL score and LOR for both groups together. Furthermore, there was a relationship between the students’ TOAL scores and LOR in the H group. This would imply that ESL students’ L2 proficiency depends on how long they have lived in the target language country.

PAGE 103

91 In contrast, there was no re lationship between LOR and the students’ TOAL score in the L group. In terms of LOR, there we re some important differences between the proficiency levels. First, the average LOR for the H group was 26.55 months and that for the L group was 4.67 months. Second, while the lengths of residency varied for the H group, such as 2, 3, or 4 years, one student in the L group who had lived in the U.S. for more than a year. In other words, it is ha rd to examine the relationship between the students’ proficiency and LOR among those who have lived in the targ et country for such a short time. Relationship between LOR and Reading Ability LOR was an important fact or, among many, in determining a student’s reading ability. There was a significant relationship between the students’ LOR and all participants’ recall scores on the ‘Olympic ga mes’ passage. Since successful L2 reading mostly depends on students’ background knowle dge on a text, experience in a target culture is helpful to L2 reading. Interviews with Participants After I finished the experiment with the higher proficiency students, I interviewed them. Regarding the reading comprehension te st, many students said that the recall test was much more difficult than traditional me thods of measuring reading comprehension, such as multiple choice questions or fill-in-the-blank type questions. On the question I asked the participants, “Do you think short te rm memory would affect your recall?” to which they answered that short term memory was not very helpful if they did not understand the passage and tried to recall from the passage. As Nassaji (2003) notes, when L2 reader s have prior knowledge about a passage they read, they should recall that passage better than when they do not have as much prior

PAGE 104

92 knowledge about that passage (p. 460). Seve ral students mentione d the importance of topic familiarity. They said if they were not familiar with the topic of a passage before they read the text, comprehendi ng the text would be difficult fo r them. In fact, there were many graduate students whose majors were en gineering. They told me that the topics such as “Supertankers” or “Nuclear power pl ants” were quite familiar to them, suggesting that the scientific topics gave an advantage to engineering students.

PAGE 105

93 CHAPTER 6 CONCLUSION In this section I will summarize the principal findings from the present study and provide an analysis of the pedagogical and research implications. Additionally, I will discuss the limitations of the present study. Principal Findings To determine a statistical significance be tween the H and the L groups in terms of the TOAL-3 score, a one-way analysis of variance (ANOVA) was performed. As expected, the results were signi ficant. This means that we can claim that the students in the H group are significantly more proficient than those in the L group. In addition, a correlation between the TOAL raw score a nd the groups (Higher and Lower) was found. This means that the Higher group students achieved a higher English proficiency score than the Lower group students and the rela tionship is statistically significant. To look for statistical significance among th e subgroups of each proficiency level, a one-way ANOVA was performed. There was no statistical significance among the subgroups in the H group. Likewise, there was no statistical significance among the subgroups in the L group. Therefore, we assume that subgroups’ proficiency levels within each group are same. For the data generated from the loosely st ructured “The Olympic games” passage that was used for both groups an ANOVA sh owed that there was a statistically significant difference between the groups, (Hig her and Lower). This suggests that the

PAGE 106

94 reading ability of the H group is significantly higher than the reading ability of the L group. In the H group, an ANOVA showed that there was a statistically significant difference between the subgroups, control, vocabulary and expanded framework groups, for the comparison/contrast text. For the collec tions of descriptions text, the difference of the subgroups was not significant but near ed significant. Furthermore, for the problem/solution text, the difference of the subgroups was not significant. In the L group, an ANOVA showed that there was a statistically significant difference between the subgroups, control, vocabulary and expanded framework groups, for the collections of descriptions and the co mparison/contrast texts. However, there was no significant difference between the s ubgroups for the problem/solution text. In order to determine a statistical significan ce of the text types, (the collections of descriptions, problem/solution, or comparison/contrast) in each proficiency group, a oneway ANOVA was performed. There was a statis tically significant difference on the three different text types in the H group. Yet, ther e was a statistical si gnificance between the three different text types in the L group. According to the results, there was no si gnificant difference on the two text types, the loosely-structured text a nd the tightly-structured texts in the H group. However, there was a significant difference between the two text types, the looselystructured text and the tightly-structured texts in the L group. There was a statistically significant re lationship between the TOAL and the loosely-structured text for H. Since both groups read the same passage, e.g., a collection of descriptions passage, a co rrelation test for all participants for the passage was

PAGE 107

95 performed and the correlation was significant. There were no relationships between the students’ English proficiency, as shown by th e TOAL, and the problem/solution passages for both groups. However, there were correl ations for the comparison/contrast passages for both groups. In addition, a correlation betw een the ‘collection of descriptions’ passage raw score and the groups (Higher and Lower) wa s found. This means that there is a very high negative relationship between the two groups and their reading scores. In the H group, an ANOVA showed that there was a statistically significant difference between the subgroups for ‘all texts together.’ This result suggests that the expanded framework group recalled significantly more than the others for all texts. Moreover, in the L group, an ANOVA showed that there was a statistically significant difference between the subgroups for ‘all texts together,’ w ith both the expanded framework group and the vocabulary group recalling significantly more than the control group. The higher proficiency students recalled si gnificantly more of the problem/solution passage than for the comparison/contrast passage. On the other hand, the lower proficiency students recalled significantly mo re of the problem/solution passage than both the comparison/contrast passage and the collection of descriptions passage. This suggests that a passage which has a structure that is easi er for lower proficiency ESL students to read than a passage which has a loose structure. There was a correlation between the reading proficiency as determined by subtest 6 (reading/grammar) of the TOAL, and the recal l score for “The Olympic games” passage among all participants. In other words, if a student has a high readi ng proficiency score, s/he is better at reading in a second language than a student with low reading proficiency

PAGE 108

96 score, and vice versa. A significant co rrelation was found between the reading proficiency and the readi ng score in the H group. The results showed that the H group stude nts were less anxious than the L group on the ‘reading anxiety’ items. A correlation be tween the TOAL raw score and the average reading anxiety items for both groups was f ound. This means that ESL students’ anxiety level might affect their English proficienc y. A correlation between the reading anxiety score and the ‘collection of descriptions’ passage for both groups was found. If a student is highly anxious, her/his read ing score for this passage is lower, and vice versa. Pearson product correlations found that there were relationships between the students’ anxiety levels and th eir reading scores for the H group. Specifically, there was a correlation between the reading anxiety score and the problem/solution passage in the H group. Also, there was a correlation between the reading anxiety score and the comparison/contrast passage in the H group. Within the H group, the average score on the top-down reading strategies was 4.03 (agree) and on the bottom-up read ing strategies was 2.52 (disag ree). Therefore, top-down reading strategies were preferred more by th e higher proficiency students than bottom-up reading strategies. As a cont rast, within the L group, the average score on the top-down reading strategies was 3.83 (neither agree nor disagree) and on the bottom-up reading strategies and it was 3.08 (nei ther agree nor disagree). The results show that there was no relati onship between the students’ self-rated L1 reading skill and reading proficiency, as measured by subtest 6 (reading/grammar) for groups together. There was a relationship be tween the students’ self-rated L1 reading skill and reading proficiency for the H group.

PAGE 109

97 For the correlation tests, item 9 on pa rt III of the background information questionnaire and the students’ reading scor es on “The Olympic Games” passage were employed. There was a relationship between L1 reading skill and the reading scores for the L group. There was a strong relationship between the students’ TOAL score and LOR for both groups. The relationship between the stud ents’ TOAL scores and the LOR was most significant within the H group. Pedagogical Implications The results of this study provide SL teacher s and researchers with insight into some possible practical applications for the clas sroom. Generally, for the higher proficiency students, giving expanded framework info rmation was helpful and for the lower proficient students, giving both expanded framework and vocabulary information was helpful. Of these text adjuncts that assist in L2 reading, the most important for language instructors who mostly deal with lower leve l students is providing accurate vocabulary information. It goes without saying that vocabulary is an essential factor for L2 learners, not only in reading but also in other areas, such as speaking and listening. In many cases, not knowing a word causes L2 learners to lose tr ack of the main idea in spoken or written communication. Also, in order to use a word in speech or in writing, they must know what it means. Therefore, vocabulary is an essential component to all aspects of L2 performance. Regarding vocabulary teaching, the resear cher has some suggestions for L2 classroom teachers. First, when they inst ruct lower level ESL students, the teachers should teach vocabulary important to unde rstanding a passage but unknown to the

PAGE 110

98 students, before they read the passage. After reading, it might be useful to tell them to make sentences using the words as a test Second, the teachers could group all the vocabulary words related to the passage fo r the students so that they build their vocabulary according to a topic. Furthermore, the lower proficiency students recalled significantly more in the problem /solution passage, which has a tight structure than in the collection of descriptions passage, which has a loose structure. Therefore, ESL teachers should examine the rhetorical structure of thei r reading materials, and instruct lower level ESL students with tightly-structu red texts in a reading class. The lower proficiency students’ reading anxiety level was higher than that of the higher proficiency students. However, the di fference was not significant. Recall that for the five items on the FLRAS, a higher score signifies higher anxiet y. L2 teachers should pay attention especially to lower level ESL st udents, because their re latively high anxiety level could hinder their learning in classroom. Research Implications There are several strengths in the present study. First, I measured the participants’ English proficiency using a standardized te st, the TOAL-3. I did not measure only the students’ reading proficiency, for example, as evidence of their overall English proficiency. Since there was a significant difference between the proficiency groups, Higher and Lower, the results from the pr esent study can be generalized to the ESL population. Second, I chose three different t ypes of reading text s to examine the interrelationship between the text types and th e effects of text adjuncts. Third, I used a single scoring template to grade the participants’ recall protocols. This makes the present study’s results easier to generalize.

PAGE 111

99 The study revealed that there was a corre lation between students’ L2 proficiency and their reading ability. Moreover, ther e was a correlation between their reading proficiency in the L2, as measured by the subtest 6 (reading/gram mar) of the TOAL and their reading comprehension of these texts, as measured by recall protocols. This provides evidence that ESL students’ proficienc y should not be neglected in L2 reading. Prior to this study there was al most no research on ESL learners’ L2 proficiency and their reading ability, so this factor shou ld be explored more in future There was a negative correlation between the TOAL raw score and the reading anxiety score in the L group. However, ther e was no correlation between the TOAL raw score and the ‘reading anxiety’ score in the H group. This im plies that reading anxiety and lower proficiency ESL students’ proficie ncy are linked. The results show that there was a negative relationship between the read ing anxiety score and the collection of descriptions passage for all pa rticipants. In the H group, there were relationships between the reading anxiety score and the problem/s olution passage and between the reading anxiety score and the comparison/contrast pa ssage. The results imply that highly anxious students tend to have low recall scores in the H group. However, there were no relationships between the read ing anxiety score and the read ing passages in the L group. In addition, reading in a classroom for a test is different from reading for pleasure. If a student has test anxiety it is possible for him/her to receive a low score on a reading test because of test anxiety. In order to distinguish forei gn language reading anxiety from test anxiety, future resear chers should administer bot h scales to ESL students. Limitations of Current Study Sixty-one students participated in the pres ent study and there we re two proficiency groups. Since the study had a relatively sma ll sample, it was hard to divide the

PAGE 112

100 participants into high, inte rmediate, and low proficie ncy groups. Including an intermediate proficiency group, could have re sulted in having all students from the same language background in any one group, becau se the heterogeneous sample of ESL students contained many Chinese and Korean students. It is recommended that researchers in the future use a larger sample size so that they study the results that include an intermediate proficiency group also. To compare students’ language backgrounds and other factors, such as their proficiency scores in each area, it is ideal to have sufficient numbers of other language backgrounds as we ll. Also, the present study used a ‘balanced random’ sample where I divided the student s into subgroups. This was to avoid the situation of having students from the same language background in one group. If researchers have enough diffe rent language backgrounds, it w ould be better to divide students randomly into subgroups. Another challenge to this study is evaluating the degr ee of proficiency. Although I measured the students’ English proficiency leve l systematically, it is difficult to ascertain the extent of the H group’s high proficienc y, and, likewise, the extent of the L group’s low proficiency students. Since I did not have an intermediate group, some of the students in the present study could be consider ed as an intermediate threshold between the two groups. The texts that I chose might have an effect on the pa rticipants’ performance. I chose the texts according to general matches between the students’ proficiency and the text level; however, the text level might be too high for some students or too low for other students. Furthermore, I did not m easure the students’ background knowledge on the texts they read, which could potentia lly affect anxiety, and proficiency.

PAGE 113

101 The study employed a six-item FLRAS. Th e purpose of administering the FLRAS was to pilot test how the students felt when th ey read in an L2. Since we only had five ‘reading anxiety’ items, it is recommended th at future researchers add more reading anxiety items and examine any relationship between ESL students’ reading and their anxiety. Also, the correlation we found was based only on the students’ answers on the FLRAS. As a part of researching the correlat ions in this study, I have only limited information that would sufficiently explain how one factor affects the other. In other words, I can not claim that the students’ prof iciency affected their reading ability even though there was a relationship between the prof iciency and the reading score. However, I still claim that ESL learners’ proficie ncy cannot be overlooked because it has a relationship with other f actors in SL areas. In that sense, it is important to measure ESL students’ proficiency reliably. Also, when measuring overall proficiency, comprehensiveness should matter. Hence, SL researchers should consider including adequate areas to measure students’ pr oficiency, over and above simply judging proficiency through separate r eading and grammar tests.

PAGE 114

102 APPENDIX A BACKGROUND QUESTIONNAIRE Part I Language Background The questions below are for research pur poses only, and your answers will not be made available to anyone. Please answer the following questions or check the appropriate answers. 1. Name: _______________________________________________ First Last 2. DOB: _________________________ 3. Sex: ____________ Male ____________ Female 4. Native language(s): ______________________________________ 5. Do you speak any language(s) other th an your native language and English? ________________________________________ 6. How long ago did you come to the U.S? ____________________________ 7. How old were you when you first came to the U.S.? ____________________________ 8. What percentage do you speak English in these situations? Your native language ? English Native languages other than English Please specify: ________________________ __________ % at school? ____________ % __________ % at home? ____________ % __________ % with friends? ____________ % __________ % other? ____________ %

PAGE 115

103 9. Places you have lived for more than 6 months: ____________________________ from ___________ to _____________ ____________________________ from ___________ to _____________ ____________________________ from ___________ to _____________ ____________________________ from ___________ to _____________ Part II Directions: Statements 1 through 6 refer to how you feel about learning English. For each statement, please tell me whether you (1) strongly disagree, (2) disagree, (3) neither agree nor disagree, (4) agree, or (5) strongly agree, by circling the appropriate number on the line following each statement. Please give your first reaction to each statement and circle an answer for every statement. 1. When reading English, I often understand the words but I don’t always understand what the author is saying. 1 2 3 4 5 2. When I’m reading, I am ner vous if I don’t know the topic. 1 2 3 4 5 3. I get upset when I don’t understand the grammar in a sentence when reading English. 1 2 3 4 5 4. When reading English, if I don’t understa nd every word, I get nervous and confused. 1 2 3 4 5 5. I feel very uncomfortable when a teach er asks me to read aloud in class. 1 2 3 4 5 6. I like to do group work in English class. 1 2 3 4 5 Part III Directions: Statements 1 through 8 are about reading in English. For each statement, please tell me whether you (1) str ongly disagree, (2) disagr ee, (3) neither agree nor disagree, (4) agree, or (5) strongly ag ree, by circling the appr opriate number on the line following each statement. Please choose the number that best describes your actions to each statement when reading in a forei gn language and circle an answer for every statement. For item 9, please use this s cale (1-very good, 2-fairly good, 3-average, 4below average, 5-bad).

PAGE 116

104 1. When reading in English, I skip the words I don’t know and continue reading. 1 2 3 4 5 2. I read the title and imagine wh at the article might be about. 1 2 3 4 5 3. I start reading and try to figur e out the meaning as I go along. 1 2 3 4 5 4. I usually translate into my native langua ge when I read an English passage. 1 2 3 4 5 5. I read every word and look up the ones I don’t know. 1 2 3 4 5 6. I usually pronounce words silently, and wh en I don’t know how to pronounce a word, I think I won’t get the right meaning. 1 2 3 4 5 7. I think it’s helpful if a teacher gives me information about a text before I read 1 2 3 4 5 8. I think reading out loud to myself is helpful. 1 2 3 4 5 9. Rate your reading skill in your native language. very good fairly good average below average bad

PAGE 117

105 APPENDIX B SUMMARY OF TEXTS For High and Low levels: The Olympic Games (268 words) -source: Oh, J. (1990). Diss. University of Texas at Austin. -a collection of descriptions For High level: Supertankers (240 words) -source: Meyer, Brandt, and Bluth (1 980). Reading Research Quarterly, 16 -problem and solution Closing down nuclear power plants (265 words) -source: Carrell (1992). Language Learning, 42. -comparison/contrast For Low level: On being fat in America (325 words) -source: Connor (1984) TESOL Quarterly, 18. -problem and solution Nuclear energy versus solar energy (254 words) -source: Carrell (1985). TESOL Quarterly, 19. -comparison

PAGE 118

106 The Olympic Games In ancient Greece, athletic festivals were very important, and strongly related to religion. The Olympian athletic festival was held every four years in honor of Zeus. The festival lost its local character, and became first a national event. Then, after the rules against foreign competitors had been removed, it became international. No one knows exactly when the Olympic Games began, but some records date from as early as 776 B.C. The Olympic Games took place in Augus t on the field by the mountain called Olympus. Thousands of people gathered from all parts of Greece. The exact order of events is not certain, but the Olympic Games included events such as gymnastics, horseracing, javelin throwing, and foot races. On the sixth and last day, all the winners were crowned with holy garlands of wild olive. Olympic winners received no prize m oney. However, they were in fact richly rewarded by their provincial authorities. Th e public honor also made worthwhile the tenmonth training the athletes went through. How their results would compare with modern standards, we have no means of telling. After an uninterrupted history of almo st 1,200 years, the Olympic Games were abolished in A.D. 394. The reason was that th e Games were held in honor of Zeus, not the Christian God. It was over 1,500 years befo re there was another such international athletics gathering. The Olympic Games we re revived in 1896, and the first small meeting took place in Athens, Greece. Af ter the 1908 London Olympics, success was reestablished, and nations sent their best repres entatives. In times of peace ever since, the Olympic Games have taken place every four years.

PAGE 119

107 On Being Fat in America Fat people in American society are often discriminated against in their jobs and forced to degrade themselves publicly, sociologists find. Two sociologists, Dr. Ardyth Stimson of New Jersey’s Kean College and Dr. Jack Kamerman, are currently studying fat people and their role in society. According to Dr. Stimson, “We treat peopl e who are fat as handicapped people but we don’t give them the sympathy that we gi ve to other handicapped people. Instead, they’re completely rejected and blamed for their handicap. In addition, they’re expected to participate in what we sociologists call degradation ceremonies. In other words, you’re supposed to stand there and say, “Hee, hee, hee, don’t I look awful? Hee, hee, hee, isn’t it funny I can’t move around?” “Some cities,” Kamerman said, “set overw eight limits for teachers, and if you exceed that limit-25 percent above what the insurance tables define as healthy-you are fired.” He also said that there have been other studies that found fat people do not get promoted as easily and do not advance in a company. Stimson recently completed a study of 40 women, and while none was even remotely medically overweight, she said 39 felt th ey were fat, and it caused some of them trouble in their everyday relationships. “America has become so weight conscious ,” she said, “that 40 percent of all Americans are now considered overweight.” She said there is something wrong in a society when that percentage of people are considered to be abnormal. “The problem is so great,” she said, “if you are overweight, pe ople no longer think of you as a doctor, a lawyer, or a teacher but as that fat person.”

PAGE 120

108 In some instances, the mental pain of fat pe ople is so severe that the effect it has on their lives far surpasses the medical complications that could arise as a result of being fat. If fat men and women were treated as equals, their self-esteem would rise and they would probably lose weight. Supertankers A problem of vital concern is the preven tion of oil spills from supertankers. A typical supertanker carries a half-million tons of oil and is the size of five football fields. A wrecked supertanker spills oil in the o cean; this oil kills animals, birds, and microscopic plant life. For example, when a tanker crashed off the coast of England, more than 200,000 dead seabirds washed ashore. Oil spills also kill microscopic plant life which provides food for sea life and produces 70 percent of the wo rld’s oxygen supply. Most wrecks result from the lack of power and steering equipment to handle emergency situations, such as storms. Supertankers ha ve only one boiler to provide power and one propeller to drive the ship. The solution to the problem is not to im mediately halt the use of tankers on the ocean since about 80 percent of the world’s oi l supply is carried by supertankers. Instead, the solution lies in the training of officers of supertankers, be tter building of tankers, and installing ground control stations to guide tankers near shore. First, officers of supertankers must get top tr aining in how to run and maneuver their ships. Second, tankers should be built with several propeller s for extra control and backup boilers for emergency power. Third, ground control stations should be installed at places where supertankers come close to shore. These sta tions would act like airplane control towers, guiding tankers along busy shipping la nes and through dangerous channels.

PAGE 121

109 Closing Down Nuclear Power Plants There are two different views on the best approach to shutting down nuclear power plants once their usefulness has been se rved. The first approach is immediate dismantlement. Shortly after a plant shutdown, workers would take out highly radioactive fuel parts, clean the internal machinery and then take away all remains of the plant to special nuclear dumps. The second approach to the problem of dismantlement is mothballing. This contrasting approach also calls for the removal of fuel parts and decontamination, but then the plant shell woul d simply wait behind a fence for thirty to a hundred years while its radi ation decays, after which final dismantlement would presumably take place. There are advantages and disadvantages involved with each of the two approaches to the problem of nuclear dismantlement. According to many experts immediate disman tlement is the most effective and least expensive of the approaches. The high cost of maintenance and security as well as the possibility of radioactive leaks are concer ns that weigh against mothballing. Supporters of the mothball approach to dismantlement cite immediate dismantlement as the most dangerous option to undertake. They suggest that immediate disman tlement could release the deadliest amounts of radioactivity onto workers and the public. Extreme precautions would have to be taken and workers would still be exposed to at least four times as much radiation as they would be if dismantlement were delayed th irty years. Another problem with immediate dismantlement is that at th is time there is an absence of permanent disposal sites. Plans have been made to create adequate disposal sites, but it will be years before they are completed.

PAGE 122

110 Nuclear Energy versus Solar Energy Two of the most promising sources of en ergy for the future are nuclear power and solar power. Both sources of energy have num erous advantages and disadvantages which must be considered in the development of energy programs. For several decades, many governments ha ve attached great hopes to nuclear energy as a less expensive alternative to oil and other fossil fuels. However, opposition to nuclear energy has increased in recent years. One of the main concer ns of the critics of nuclear energy is the damage that the releas e of radioactive materi als could inflict upon communities surrounding nuclear power stations Harmful materials could cause serious health problems for individuals and could ha ve disastrous effects on the environment. Another issue concerning nuclear energy involves using nuclear plants for the manufacture of nuclear weapons This is especially a point of concern about countries with unstable governments, where nuclear mate rials could easily be used for nonpeaceful ends by terrorist groups. Solar power is th e other energy source over which there are differing opinions. One of the main advantages of solar energy is that the primary required resource-the sun-is a bundantly available over most pa rts of the earth. Also, if a country’s main source of energy were to come from solar power, it would not have to depend upon other countries in order to main tain its energy supply. The main negative feature of solar power is that the technology is still being developed in this field and is not yet sufficiently reliable or cheap enough to compete with other sources of energy.

PAGE 123

111 APPENDIX C TEXT ADJUNCTS The Olympic Games (High and Low) Vocabulary Group javelin (noun): a long stick with a pointed end, thro wn as a sport garland (noun): a ring of flowers or leaves, wo rn for decoration or in special ceremonies provincial (adj.): rela ting to a province the authorities: the people or organizations that are in charge of a particular place abolish (verb): to officially end a law, system etc. revive (verb): to come back into use or exis tence, or bring something back into use or existence Expanded Framework Group This passage presents a collectio n of descriptions of the Olym pic Games. It explains their origin and details their history. On Being Fat in America (Low) Vocabulary Group discriminate (verb): to treat a person or gr oup differently from another in an unfair way degrade (verb): to treat someone without respect or to make people lose their respect for someone sociology (noun): the scientific study of societies and the be havior of people in groups look awful (verb): to look bad exceed (verb): to go beyond an official or legal limit

PAGE 124

112 promote (verb): to give someone a bette r, more responsible position at work advance (verb): to move forward to a new position surpass (verb): to be better or great er than someone or something else self-esteem (noun): the feeling that you are someone who deserves to be liked, respected, and admired Expanded Framework Group This passage presents a problem and solution of being fat in America. It demonstrates various problems or disadvantages of being fa t in American society and gives a solution. Supertankers (High) Vocabulary Group wrecked (adj.): destroyed, ruined or damaged microscopic (adj.): extremely small crash (verb): to have an accident in which a car, plane etc. hits something and is badly damaged halt (verb): to stop or make something stop backup (noun): someone or something that pr ovides help or support when it is needed Expanded Framework Group This passage presents the problem of oil spills from supertankers and several solutions to the problem. Closing Down Nuclear Power Plants (High) Vocabulary Group dismantle (verb): to gr adually get rid of a sy stem or organization mothball (verb): to close a fact ory or operation, and keep all its equipment or plans for a long time without using them

PAGE 125

113 decay (verb): to be slowly destroyed by natural chemical processes disposal (noun): the act of getting rid of something Expanded Framework Group This passage presents a comparison and contra st of two different approaches to shutting down nuclear power plants. Then, it presents a dvantages and disadvantages of each of the two approaches to the concer n of nuclear dismantlement. Nuclear Energy versus Solar Energy (Low) Vocabulary Group alternative (noun): something you can choose to do or use instead of something else fossil fuel (noun): a fuel such as gas or oil th at has been formed from plants and animals that lived millions of years ago release (verb): to stop holding something inflict (verb): to make someone suffer something unpleasant disastrous (adj.): very bad or ending in failure manufacture (noun): the pr ocess of making goods usually in large numbers Expanded Framework Group This passage presents a comparison of nucl ear energy and solar en ergy. Then, it discusses each energy respectively.

PAGE 126

114 APPENDIX D RECALL -When you have finished reading the passage turn the page and write in English everything that you remember despite whether or not you understand the material or concepts. Do not refer to th e passage during this activity. -If you have any questions about procedures of if you are unclear on any matter, please ask now. -Please try to write as much as you can. -Please try to use complete sentences. If you can’t, you can use the words in the passage or your own words.

PAGE 127

115 APPENDIX E SCORING TEMPLATE The Olympic Games (Idea Units: 100) In ancient Greece, athletic festivals were … important, very and strongly related to religion The Olympian athletic festival was held every four years in honor of Zeus The festival lost its local character, and became… a national event first Then, after the rules…had been removed against foreign competitors it became international No one knows exactly when the Olympic Games began, but some records date from...776 B.C. as early as The Olympic Games took place in August on the field

PAGE 128

116 by the mountain called Olympus Thousands of people gathered from all parts of Greece The exact order…is not certain, of events but the Olympic Games included events such as gymnastics, horse-racing, javelin throwing, and foot races On the sixth (and) last day, all the winners were crowned with holy garlands of wild olive Olympic winners received no prize money However, they were …rewarded in fact richly by their provincial authorities The public honor…made worthwhile also the ten-month training the athletes went through How their results would compare with modern standards, we have no means of telling

PAGE 129

117 After an uninterrupted history of … 1,200 years, almost the Olympic Games were abolished in A.D. 394 The reason was that the Games were held in honor of Zeus, not the Christian God It was over 1,500 years before there was another such international athletics gathering The Olympic Games were revived in 1896, and the first small meeting took place in Athens, Greece After the 1908 London Olympics, success was re-established, and nations sent their best representatives In times of peace ever since, the Olympic Games have taken place every four years On Being Fat in America (Idea Units: 142) Fat people …are …discriminated against in American society often in their jobs and forced to degrade themselves

PAGE 130

118 publicly, sociologists find Two sociologists,…are…studying Dr. Ardyth Stimson of New Jersey’s Kean College and Dr. Jack Kamerman currently fat people and their role in society According to Dr. Stimson, We treat people who are fat as handicapped people but we don’t give them the sympathy that we give to other handicapped people Instead, they ’re… rejected completely and blamed for their handicap In addition, they ’re expected to participate in…degradation ceremonies what we sociologists call In other words, you’re supposed to stand there and say Hee, hee, hee,

PAGE 131

119 don’t I look awful Hee, hee, hee, isn’t it funny I can’t move around Some cities…set Kamerman said overweight limits for teachers and if you exceed that limit 25 percent above what the insurance tables define as healthy you are fired He … said also (that) there have b een other studies that found fat people do not get promoted as easily and do not advance in a company Stimson…completed recently a study of 40 women and while none was…overweight even remotely medically she said 39 felt they were fat, and it caused some of them

PAGE 132

120 trouble in their everyday relationships. America has become…weight conscious so she said (that) 40 percent…are…considered of all Americans now overweight She said there is something wrong in a society when that percentage…are considered of people to be abnormal The problem is ...great so she said if you are overweight people no longer think of you as a doctor a lawyer or a teacher but as that fat person In some instances the mental pain…is…severe of fat people so that the effect…surpasses it has on their lives far the medical complications that could arise as a result of being fat

PAGE 133

121 If fat men…were treated and women as equals their self-esteem would rise and they would … lose probably weight Supertankers (Idea Units: 104) A problem…is the prevention of vital concern of oil spills from supertankers A typical supertanker carries a half-million tons of oil and is the size of five football fields A wrecked supertanker spills oil in the ocean this oil kills animals birds and microscopic plant life For example, when a tanker crashed off the coast of England more than 200,000 dead seabirds washed ashore Oil spills…kill also microscopic plant life

PAGE 134

122 which provides food for sea life and produces 70 percent of the world’s oxygen supply Most wrecks result from the lack of power and steering equipment to handle emergency situations, such as storms Supertankers have only one boiler to provide power and one propeller to drive the ship The solution…is not to…halt to the problem immediately the use of tankers on the ocean since about 80 percent… is carried of the world’s oil supply by supertankers Instead, the solution lies in the training of officers of supertankers, better building of tankers, and

PAGE 135

123 installing ground control stations to guide tankers near shore First, Officers… must get of supertankers top training in how to run and maneuver their ships Second, tankers should be built with several propellers for extra control and backup boilers for emergency power Third, ground control stations should be installed at places and/or where supertankers come close to shore These stations would act like airplane control towers, guiding tankers along busy shipping lanes and through dangerous channels Closing Down Nuclear Power Plants (Idea Units: 98) There are two different views on the best approach to shutting down nuclear power plants once their usefulness has been served

PAGE 136

124 The first approach is immediate dismantlement Shortly after a plant shutdown, workers would take out highly radioactive fuel parts, clean the internal machinery and then take away all remains of the plant to special nuclear dumps The second approach…is mothballing to the problem of dismantlement This contrasting approach…calls also for the removal of fuel parts and decontamination, but then the plant shell would … wait simply behind a fence for thirty to a hundred years while its radiation decays, after which final dismantlement would …take place presumably There are advantages and disadvantages involved with each of the two approaches

PAGE 137

125 to the problem of nuclear dismantlement According to many experts immediate dismantlement is the most effective and least expensive of the approaches The high cost…are concerns of maintenance and security as well as the possibility of radioactive leaks that weigh against mothballing Supporters…cite of the mothball approach to dismantlement immediate dismantlement as the most dangerous option to undertake They suggest (that) immediate dismantlement could release the deadliest amounts of radioactivity onto workers and the public Extreme precautions would have to be taken and workers would … be exposed still to … four times at least as much radiation as they would be if dismantlement were delayed thirty years

PAGE 138

126 Another problem with immediate dismantlement is (that)… there is an absence at this time of permanent disposal sites Plans have been made to create adequate disposal sites but it will be years before they are completed Nuclear Energy versus Solar Energy (Idea Units: 91) Two…are nuclear power of the most promising sources of energy for the future and solar power Both sources…have of energy numerous advantages and disadvantages which must be considered in the development of energy programs For several decades many governments have attached great hopes to nuclear energy as a less expensive alternative to oil and other fossil fuels However, Opposition…has increased to nuclear energy in recent years

PAGE 139

127 One…is the damage of the main concerns of the critics of nuclear energy that the release…could inflict of radioactive materials upon communities surrounding nuclear power stations Harmful materials could cause serious health problems for individuals and could have disastrous effects on the environment Another issue…involves concerning nuclear energy using nuclear plants for the manufacture of nuclear weapons This is…a point especially of concern about countries with unstable governments, where nuclear materials could … be used easily for nonpeaceful ends by terrorist groups Solar power is the other energy source over which there are differing opinions One…is of the main advantages of solar energy (that) the primary require d resource…is…available the sun abundantly

PAGE 140

128 over most parts of the earth Also, if a country’s main source…were to come of energy from solar power it would not have to depend upon other countries (in order) to maintain its energy supply The main negative feature…is of solar power (that) the technology is … being developed still in this field and is not…reliable yet sufficiently or cheap enough to compete with other sources of energy

PAGE 141

129 LIST OF REFERENCES Alderson, J. C., Krahnke, K. J., & Stansfield, C. W. (Eds.). (1987). Reviews of English language proficiency tests Washington, DC: TESOL. Anderson, N. J. (1991). Individual differences in strategy use in s econd language reading and testing. Modern Language Journal 75 460-472. Bailey, K. M. (1983). Competitiveness and anxi ety in adult second language learning: Looking at and through the diary studies. In H. W. Seliger & M. H. Long (Eds.), Classroom oriented research in second language acquisition (pp. 67-102). Rowley, MA: Newbury House. Barnett, M. A. (1988). Reading through cont ext: How real and pe rceived strategy use affects L2 comprehension. Modern Language Journal 72 150-162. Bernhardt, E. (1984). Toward an informati on processing perspectiv e in foreign language reading. Modern Language Journal 68 322-331. Bernhardt, E. (1991). Reading Development in a S econd Language: Theoretical, Empirical & Classroom Perspectives Norwood, New Jersey: Ablex. Block, E. (1986). The comprehension stra tegies of second language readers. TESOL Quarterly 20 463-494. Brantmeier, C. (2005). Effects of reader’s know ledge, text type, and test type on L1 and L2 reading comprehension in Spanish. Modern Language Journal 89 37-53. Carrell, P. L. (1983). Three components of background knowledge in reading comprehension. Language Learning 33 183-207. Carrell, P. L. (1984a). Evidence of a formal schema in second language comprehension. Language Learning 34 87-112. Carrell, P. L. (1984b). The effects of rh etorical organization on ESL readers. TESOL Quarterly 18 441-469. Carrell, P. L. (1984c). Schema theory and ESL reading: Classroom implications and applications. Modern Language Journal 68 332-334. Carrell, P. L. (1985). Facilitating ESL reading by teaching text structure. TESOL Quarterly 19 727-752.

PAGE 142

130 Carrell, P. L. (1989). Metacognitive aw areness and second language reading. Modern Language Journal 73, 121-134. Carrell, P. L. (1991). Second language reading: Reading ability or language proficiency. Applied Linguistics 12 159-179. Carrell, P. L. (1992). Awareness of te xt structure: Effects on recall. Language Learning 42 1-20. Carrell, P. L., & Eisterhold, J. C. (1988) Schema theory and ESL reading pedagogy. In P. L. Carrell, J. Devine & D. E. Eskey (Eds.), Interactive approaches to second language reading (pp. 73-92). Cambridge, MA: Ca mbridge University Press. Carrell, P. L., Pharis, B. G., & Liberto, J, C. (1989). Metacognative strategy training for ESL reading. TESOL Quarterly 23 647-678. Cheng, Y. S., Horwitz, E. K., & Schalle rt, D. L. (1999). La nguage anxiety: Differentiating writing and speaking components. Language Learning 49 417446. Clarke, M. A. (1979). Reading in Spanis h and English: Evidence from adult ESL students. Language Learning 29 121-150. Clarke, M. A. (1980). The short circuit hypot hesis of ESL readingor when language competence interferes with reading performance. Modern Language Journal 64 203-209. Connor, U. (1984). Recall of text: Differences between first and second language readers. TESOL Quarterly 18 239-256. Eskey, D. E. (1986). Theoretical foundations. In F. Dubin, D. E. Eskey and W. Grabe (Eds.), Teaching second language reading for academic purposes (pp. 1-23). Reading, MA: Addison-Wesley Publishing Company. Eskey, D. E. (1988). Holding in the bottom: An interactive appro ach to the language problems of second language readers. In P. L. Carrell, J. Devine & D. E. Eskey (Eds.), Interactive approaches to second language reading (pp. 93-100). Cambridge, MA: Cambridge University Press. Fecteau, M. L. (1999). Firstand second-la nguage reading comprehension of literary texts. Modern Language Journal 83 475-493. Fender, M. (2001). A review of L1 and L2/ESL word integration skills and the nature of L2/ESL word integration development i nvolved in lower-level text processing. Language Learning 51 319-396.

PAGE 143

131 Goodman, K. S. (1971). Psycholinguistic universal s in the reading proces s. In P. Pimsleur & T. Quinn (Eds.), The psychology of second language learning (pp. 135-142). Cambridge: Cambridge University Press. Hammill, D. D., Brown, V. L., Larsen, S. C., & Wiederholt, J. L. (1994). Test of Adolescent and Adult Language (3rd ed.). Aust TX: Pro-Ed. Horwitz, E. K. (1986). Some language acquisiti on principles and their implications for second language teaching. Hispania 69 684-689. Horwitz, E. K. (1987). Surveying student be liefs about language learning. In A. L. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 119-129). London: Prentice-Hall. Horwitz, E. K. (1988). The beliefs about language learning of beginning university foreign language students. Modern Language Journal 72 283-294. Horwitz, E. K., Horwitz, M. B., & Cope, J. A. (1986). Foreign language classroom anxiety. Modern Language Journal 70 125-132. Hudson, T. (1982). The effects of induced schema ta on the “short circu it” in L2 reading: Non-decoding factors in L2 reading performance. Language Learning 32 1-31. Johnson, P. (1982). Effects on reading comprehension of building background knowledge. TESOL Quarterly 16 503-516. Kern, R. G. (1989). Second language readi ng strategy instruction: Its effects on comprehension and word inference ability. Modern Language Journal 73 135149. Kleinmann, H. H. (1977). Avoidance behavi or in adult second language acquisition. Language Learning 27 93-107. Koch, A. S., & Terrell, T. D. (1991). Affec tive reactions of forei gn language students to natural approach activities and teaching techniques. In E. K. Horwitz & D. J. Young (Eds.), Language anxiety: From theory and research to classroom implications (pp. 109-126). Englewood Cliffs, NJ: Prentice Hall. Laufer, B. (1992). Reading in a foreign language: How does L2 lexical knowledge interact with the reader’s general academic ability? Journal of Research in Reading 15 95-103. Laufer, B., & Sim, D. D. (1985). Measuring and explaining the read ing threshold needed for English for academic purposes texts. Foreign Language Annals 18 405-411. Lee, J. F. (1986a). The effects of thr ee components of background knowledge on second language reading. Modern Language Journal 70 350-354.

PAGE 144

132 Lee, J. F. (1986b). On the us e of the recall task to meas ure L2 reading comprehension. Studies in Second Language Acquisition 8 201-212. Lee, J. F. (1999). Clashes in L2 reading: Research versus practice and readers’ misconceptions. In D. J. Young (Ed.), Affect in foreign language and second language learning: A practical guide to creating a low-anxiety classroom atmosphere (pp.49-63). New York: McGraw-Hill. Lee, J. F., & Riley, G. L. (1990). The e ffects of prereading rh etorically-oriented frameworks on the recall of two struct urally different expository texts. Studies in Second Language Acquisition 12 25-41. MacIntyre, P. D. (1999). La nguage anxiety: A review of the research for language teachers. In D. J. Young (Ed.), Affect in foreign la nguage and second language learning: A practical guide to creati ng a low-anxiety classroom atmosphere (pp. 24-45). New York: McGraw-Hill. MacIntyre, P. D., & Gardner, R. C. ( 1989). Anxiety and second-language learning: Toward a theoretical clarification. Language Learning 39 251-275. MacIntyre, P. D., & Gardner, R. C. (1991a). Investigating language class anxiety using the focused essay technique. Modern Language Journal 75 296-304. MacIntyre, P. D., & Gardner, R. C. (1991b) Language anxiety: Its relation to other anxieties and to processing in native and second languages. Language Learning 41 513-534. MacIntyre, P. D., & Gardner, R. C. (1991c). Methods and results in the study of anxiety and language learning: A re view of the literature. Language Learning 41 85-117. MacIntyre, P. D., & Gardner, R. C. (1994a ). The effects of induced anxiety on three stages of cognitive processing in computerized vocabulary learning. Studies in Second Language Acquisition 16 1-17. MacIntyre, P. D. & Gardner, R. C. (1994b). The subtle effects of language anxiety on cognitive processing in the second language. Language Learning 44 283-305. Mathewson, G. C. (1985). Toward a compre hensive model of affect in the reading process. In H. Singer & R. B. Ruddell (Eds.), Theoretical models and processes of reading (3rd ed.), (pp. 841-856). Newark, DE: In ternational Reading Association. Meyer, B. J. F. (1975). Identification of stru cture of prose and its implications for study of reading and memory. Journal of Reading Behavior 7 7-47. Meyer, B. J. F., Brandt, D. M., & Bluth, G. (1980). Use of top-level structure in a text: Key for reading comprehension of ninth-grade students. Reading Research Quarterly 16 72-103.

PAGE 145

133 Meyer, B. J. F., & Freedle, R. O. (1984) Effects of discourse type on recall. American Educational Research Journal 21, 121-143. Naiman, N., Frohlich, M., & Todesco, A. (1975). The good second language learner. TESL Talk 6, 58-76. Nassaji, H. (2002). Schema theory and know ledge-based processes in second language reading comprehension: A need for alternative perspectives. Language Learning 52 439-481. Nassaji, H. (2003). Higher-level and lower-level text processing skills in advanced ESL reading comprehension. Modern Language Journal 87 261-276. Oh, J. (1990). On the relationship between anxiety and reading in English as a Foreign Language among Korean university students in Korea Unpublished doctoral dissertation. University of Texas at Austin. Oller, J. (1979). Research on the measurement of af fective variables: Some remaining questions Paper presented at the Colloquium on Second Language Acquisition and Use Under Different Circumstances 1979 TESOL Convention. Boston, Mass. Oller, J., & Perkins, K. ( 1978). Intelligence and language proficiency as sources of variance in self-reported affective variables. Language Learning 28 85-97. Padron, Y. N., & Waxman, H. C. (1988). The effects of ESL student s’ perceptions of their cognitive strategies on reading achievement. TESOL Quarterly 22 146-150. Park, H. (2001). Foreign language reading anxiety in EFL students Unpublished manuscript. University of Florida at Gainesville. Patten, M. L. (2000). Understanding research methods (2nd ed.). Pyrczak Publishing. Price, M. L. (1991). The subj ective experience of foreign la nguage anxiety: Interviews with anxious students. In E. K. Horwitz & D. J. Young (Eds.), Language anxiety: From theory and research to classroom implications (pp. 101-108). Englewood Cliffs, NJ: Prentice Hall. Pritchard, R. (1990). The effects of cultural schemata of reading processing strategies. Reading Research Quarterly 25 273-295. Ruddell, R. B., & Speaker, R. B. (1985). The in teractive reading process: A model. In H. Singer & R. B Ruddell (Eds.), Theoretical models and processes of reading (3rd ed), (pp. 751-793). Newark, DE: In ternational Reading Association. Rusciolelli, J. (1995). Student response s to reading strategies instruction. Foreign Language Annals 28 262-273.

PAGE 146

134 Saito, Y, Garza, T. J., & Horwitz, E. K. (1999). Foreign Language Reading Anxiety. Modern Language Journal 83 202-218. Scovel, T. (1978). The effect of affect on foreign language learni ng: A review of the anxiety research. Language Learning 28 129-142. Sellers, V. (2000). Anxiety and reading compre hension in Spanish as a foreign language. Foreign Language Annals 33 512-521. Smith, F. (1994). Understanding Reading: A Psycholingu istic Analysis of Reading and Learning to Read (5th ed). Hillsdale, New Jersey: Lawrence Erlbaum. Taglieber, L. K., Johnson, L. L., & Yarb rough, D. B. (1988). Effects of prereading activities on EFL reading by Brazilian college students. TESOL Quarterly 22 455472. Taillefer, G. (1996). L2 reading ability: Furthe r insights into the short-circuit hypothesis. Modern Language Journal 80 461-477. Thomas. M. (1994). Assessment of L2 pr oficiency in second language acquisition research. Language Learning 44 307-336. Tobias, S. (1986). Anxiety and cognitive pr ocessing of instruction. In R. Schwarqer (Ed.), Self-related cognition in anxiety and motivation (pp. 35-54). Hillsdale, NJ: Erlbaum. Urquhart, A. H., & Cyril, W. (1998). Reading in a Second Language: Process, Product and Practice pp. 37-51. London: Addison Wesley Longman. Yano, Y., Long, M. H., & Ross, S. (1994). The e ffects of simplified and elaborated texts on foreign language reading comprehension. Language Learning 44 189-219. Young, D. J. (1991). Creating a low-anxiety classroom enviro nment: What does language anxiety research suggest ? Modern Language Journal 75 426-436. Young, D. J. (1993). Processing strategies of foreign language readers: Authentic and edited input. Foreign Language Annals 26 451-468. Wolf, D. F. (1993). A comparison of assessm ent tasks used to measure FL reading comprehension. Modern Language Journal 77 473-489. Zbornik, J., & Wallbrown, F. H. (1991). The development and validation of a scale to measure reading anxiety, Reading Improvement 28 2-13.

PAGE 147

135 BIOGRAPHICAL SKETCH Heenam Park began her university educati on at the Semyung University in Korea as a student of English Language and Litera ture. She completed her BA in Spring 1995, and was admitted to the master’s program at the Sungkyunkwan University in Korea. She completed her MA in English Language and Lite rature with a concentration in linguistics in Spring 1997. After her master’s degree, she taught English classes at universities in Korea for 2 years. She was admitted to the doctoral program in linguistics at the University of Florida in Fall 1999. During her te rm as a doctoral candidate, she served as a teaching assistant for Introduc tion to Linguistics. Her primar y research interests are in the areas of second-langua ge reading, foreign-language anxiety, second languageacquisition, and Teaching English to Speaker s of Other Languages (TESOL). She was awarded a PhD in linguistics in August 2005.