Acquisition and retention of academic material learned while tutoring, being tutored, or studying independently

MISSING IMAGE

Material Information

Title:
Acquisition and retention of academic material learned while tutoring, being tutored, or studying independently
Physical Description:
vii, 144 leaves : ill. ; 28 cm.
Language:
English
Creator:
Monroe, Craig Michael
Publication Date:

Subjects

Subjects / Keywords:
Tutors and tutoring   ( lcsh )
Peer-group tutoring of students   ( lcsh )
Education -- Experimental methods   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1984.
Bibliography:
Includes bibliographical references (leaves 139-143).
Statement of Responsibility:
by Craig Michael Monroe.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000468618
notis - ACN3304
oclc - 11627593
System ID:
AA00003421:00001

Full Text











ACQUISITION AND RETENTION OF ACADEMIC MATERIAL
LEARNED WHILE TUTORING, BEING TUTORED, OR
STUDYING INDEPENDENTLY













By

CRAIG MICHAEL MONROE


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




UNIVERSITY OF FLORIDA
1984






























This manuscript is dedicated
to my mother Anne, who so
dearly wanted to see that
it was completed.














ACKNOWLEDGMENTS


I extend my thanks to Drs. James Johnston, Henry

Pennypacker, and Edward Malagodi, who generously gave of their

time and expertise in helping me solve the day-to-day prob-

lems one encounters when conducting research.

I wish to thank Kathryn Saunders and Dianne Trifiletti

for carefully implementing the experimental procedures in

their classrooms, Special thanks are extended to Kathryn for

being supportive while I tried to keep three experiments

afloat.

I wish to thank Barbara Bader, Charles Hamid, and Gregory

LaForme, whose recollections of their own dissertation

ordeals provided some of the inspiration and moral support

necessary to complete this manuscript. I am especially

grateful to Gregory LaForme for his helpful editing of

earlier drafts and persistent friendship and encouragement.

Special appreciation is extended to Paula Beck, whose

love and support in so many areas helped me over many of

the rough spots encountered when writing this manuscript.


iii


















TABLE OF CONTENTS


PAGE


ACKNOWLEDGMENTS . .


ABSTRACT . .


GENERAL INTRODUCTION. .


EXPERIMENT I . .


Method . .
Results and Discussion .


EXPERIMENT II . .


Introduction .
Method . .
Results and Discussion .


EXPERIMENTS I AND II: DISCUSSION


EXPERIMENT III . .


Introduction .
Method . .
Results and Discussion .


GENERAL DISCUSSION .


REFERENCE NOTES . .


REFERENCES . .


BIOGRAPHICAL SKETCH .


. . iii


. . v


. . 1


. . 12


. . 12
. . 19


. . 53


. . 53
. . 53
. . 58


. . 94


. . 103


. . 103
. 105
. . 111


. . 131


. . 139


. . 140


. . 144












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

ACQUISITION AND RETENTION OF ACADEMIC MATERIAL
LEARNED WHILE TUTORING, BEING TUTORED, OR
STUDYING INDEPENDENTLY

By

Craig Michael Monroe

April, 1984

Chairman: James M. Johnston, Ph.D.
Major Department: Psychology

Three experiments were conducted to examine (1) whether

children could effectively tutor others without instruction

about how to tutor, (2) the relative benefits of serving as

a tutor, being tutored, or studying independently, and (3)

whether retention of academic material is differentially

affected as a function of the manner in which the material

was learned.

In the first two experiments, the experimental design

allowed a simultaneous comparison of each subject's spelling

performance when the subject tutored another subject, was

tutored by another subject, or studied independently in one

experimental phase; during the other phase, tutor-tutee roles

were reversed and daily probes were conducted instead of







allowing the subjects to study independently. Retention

probes were conducted after 17 to 28 days in which the

subjects had no instructional contact with the spelling

words. These experiments found that (1) all subjects were

effective tutors without instruction about how to tutor,

(2) while the subjects learned most rapidly when being tutored,

the results comparing learning when tutoring or studying inde-

pendently were highly individualized, (3) certain subject

pairings resulted in more rapid learning for both tutor and

tutee than other pairings, and (4) no systematic differential

effects of the condition in effect during acquisition upon

retention was observed.

The third experiment was designed to examine the feasi-

bility of using an unstructured tutoring arrangement with

second-grade students. Two subjects were selected based

upon their poor performance on subtraction facts. Each

served in the following conditions: tutor, tutee, inde-

pendent study. When neither child improved during any

experimental condition, they were both taught a specific

tutoring procedure to use when tutoring, This structuring

of the tutorial process did not improve either subject's

performance. Some improvement was observed when the

subjects were made to correct their errors. Different

experimental procedures may have resulted in different

findings.







The results of these experiments suggest that teachers

need to analyze the effects of tutorial arrangements at the

level of the individual child.


vii














GENERAL INTRODUCTION


John Dewey (1916) posited that children learn by doing.

Hall, Delquadri, and Jasper (note 1) have stated that chil-

dren have very few opportunities for responding to academic

subject matter. As an example, Hall et al. (note 1) cite

the results of Fox (1974) who observed that first-grade

students in a school averaged no more than 20 seconds of

reading instruction per day. Additionally, Hall et al.

(note 1) state that students at a university-affiliated

elementary school spent less than five seconds per day prac-

ticing basic arithmetic facts.

As educational budgets are being decreased nation-

wide, lower teacher-student ratios result with the effect

of limiting the time available for teachers to provide

practice opportunities for the individual student. One

potentially economical and effective means to increase

learning opportunities for a student would be to use other

students as tutors. This would allow teachers to delegate

routine tasks to tutors and concentrate on planning and

more complex teaching tasks (Worell & Nelson, 1974).





2

Students have taught one another academic material for

many years. The earliest report of peer tutoring being used

was in early American rural schools where older students

were used to help younger students memorize their lessons

(Lovitt, 1977). Tutoring achieved widespread usage during

the 1960's spurred by major government programs such as Youth

Tutoring Youth and Mobilization For Youth. The focus of

these government-financed programs was on the social bene-

fits gained by older underprivileged tutors. Benefits such

as improved attitudes (Dreyer, 1973; Fleming, 1969), a sense

of personal satisfaction (Haun, 1975; Rossi, 1969), and the

development of self-esteem (Frager & Stern, 1970; Lippitt &

Lippitt, 1970; Mohan, 1971; and Squires, 1971) were reported.

Few investigations reported on the academic achievements of

either tutor or tutee.

As is the case with many educational practices, broadly

based empirical underpinnings are notably lacking for the

use of tutoring (Devin-Sheehan, Feldman, & Allen, 1976;

Lovitt, 1977; Norris, note 2). While it has been said that

the tutor and/or tutee benefit academically from participa-

tion in tutoring (Cloward, 1969; Hassinger & Via, 1969; and

Landrum & Martin, 1970), such general statements seem

premature given the paucity of data in this area. Reports








have been historically devoid of data, advocating tutoring

based upon anecdotal observation (Fleming, 1969; Frager &

Stern, 1970; and Mohan, 1971) or using group experimental

designs which do not allow statements to be made about the

functional relations between tutoring and academic improve-

ment by the individual student (Cloward, 1969; Gartner,

Kohler, & Riessman, 1971; and Rosner, 1970).

At this point in the study of tutoring questions about

whether a tutor can effectively teach another student or can

learn the academic material while tutoring seem to miss the

point; the available research provides ambiguous answers to

both questions. Tutoring arrangements have resulted in

tutors and tutees learning and failing to learn academic

material. Questions must be raised regarding the costs and

benefits of various tutoring arrangements accruing to the

tutor, the tutee, and the teacher. Quite simply is tutoring

a worthwhile investment relative to other educational prac-

tices?

Studies examining tutoring have extended along a con-

tinuum from using untrained tutors in an unstructured tutor-

ing situation (Harris, Sherman, Henderson, & Harris, 1972;

and Harris & Sherman, 1973) to the opposite extreme where

tutors are trained to follow a specific tutoring procedure








with frequent monitoring by the teacher to maintain adherence

to those procedures (Davis, 1972; Dineen, Clark, & Risley,

1977; Drass & Jones, 1971; Hassinger & Via, 1969; Johnson &

Bailey, 1974; and Willis, Crowder, & Morris, 1972). Con-

siderable teacher time was invested in controlling tutor-

tutee interactions in investigations incorporating structured

tutoring.

For example, Johnson and Bailey (1974) conducted three

30-minute tutor training sessions prior to the investigation.

Similarly, Willis et al. (1972) spent five hours training

their tutors in behavioral principles related to teaching

reading, and teams of tutors competed to win a trophy for

the highest "efficiency" (i.e., adherence to tutoring pro-

cedures) rating on an ongoing basis. This required regular

systematic observation of the tutoring sessions by the school

counselor. Davis (1972) trained tutors to praise and

prompt tutees for correct and incorrect responses, respec-

tively. Tutoring sessions were tape-recorded and tutors

were reinforced based upon how well they conducted these

sessions. Two raters each spent 15 hours per week evaluating

the taped sessions. During the Dineen et al. (1977) investi-

gation, the teacher was required to attend to the tutoring

process approximately 45 times each hour. The average






5

teacher may not have this much time available for controlling

the tutoring process.

However, the results of two investigations point out

that structuring the tutoring process may adversely effect

academic gains made by tutors. Duff and Swick (1974) had

third and fourth graders who read below grade level tutor

first and second graders using the Sullivan Programmed Read-

ing materials. Lessons were planned with a teacher and

lasted 30 to 40 minutes per day for six weeks. Same age

peers were chosen as controls for tutors and tutees; no

information was provided about the activities of the control

groups. Tutees made statistically significant gains rela-

tive to controls on the Metropolitan Achievement Test.

Tutors showed a trend to make greater gains than controls

but the differences were not statistically significant.

Similarly, the eighth-grade tutors in the Willis and Crowder

(1974) investigation made smaller gains on the California

Achievement Test than control group subjects. In discussing

their findings the authors noted that the (undescribed)

tutoring procedures used in their study were very structured

relative to other studies which allowed more flexible

arrangements between tutor and tutee.

Several studies have demonstrated that unstructured

tutoring can have beneficial results for both tutor and








tutee. Harris et al. (1972) compared the effects of a 10-

minute unstructured tutoring session and a 10-minute inde-

pendent study session upon spelling performance. On Monday

of each week students took two 20-word protests. During

the week they worked on assignments in class related to

these words. At the end of the week students were either

allowed 10 minutes to help each other learn one 20-word

list in groups of two to four children or were allowed 10

minutes (on other weeks) to study one 20-word list alone.

Two 20-word posttests covering that week's words were then

administered. During five weeks of this investigation,

performance gains on tutored words were compared to per-

formance gains on untutored words. The within week gains

for the tutored set of words ranged from 14 percent to

25 percent higher than the gains for the untutored words.

During the other five weeks, the within week gains for the

set of words studied independently ranged from 3 percent

to 9 percent higher than the gains for the untutored words.

The comparison of tutored to untutored words was then

replicated in four other classrooms for one week each.

Within week pre- to posttest gains on the tutored words

compared to the performances on the untutored sets of words

ranged from 16 percent to 23 percent greater. The authors







noted that because of the unstructured nature of the tutoring

groups, they were unable to assess the differential effects

of a child serving as a tutor or a tutee. It should be noted

that the gains noted on tutored words or words studied inde-

pendently result from interactions with the effects of the

lessons in class related to these words. Harris et al.

(1972) present data on retention of tutored, independently

studied and comparison words lists at intervals of two to

five weeks following administration of the posttest. Both

words studied independently and tutored words showed greater

declines from posttest to retention test than did comparison

word lists. Data were presented in group formats only,

In a similar study Harris and Sherman (1973) compared

unstructured tutoring with independent study with fourth

and fifth graders in math. Consistent with the earlier

findings of Harris et al. (1972), students performed better

on math problems when tutored than when working alone. Once

again, the data presentation formats chosen did not reveal

the effects of these conditions upon individual children

nor could differential effects upon tutor or tutee per se

be ascertained.

In two similar reports Hassinger and Via (1969) and

Landrum and Martin (1970) found larger than expected gains








(simply as a function of time passage) for both tutors and

tutees on standardized reading tests following six weeks

of unstructured tutoring. No data were presented for compari-

son in either report of comparable students who did not par-

ticipate in tutoring. In both studies gains for tutors

were larger than gains for tutees.

Several other studies using group experimental designs

have demonstrated the efficacy of unstructured tutoring

arrangements (e.g., Cloward, 1969; Gartner et al., 1971);

and Rosner, 1970) as well. However, group designs do not

reveal the functional relations between tutoring procedures

and behavior change.

Justification of spending a student's time in a tutoring

role must include not only a question of "Does that student

learn as a function of tutoring?" but "Is this educational

gain at least equal to the potential gain possible from

other pedagological arrangements?" Myers, Travers, and

Sanford (1965) had 192 fourth through sixth graders learn

20 German words for 15 to 20 minutes each day for three days

in one of four conditions: studying independently, being

taught by a peer, tutoring a peer, or being a tutor for one-

half of the session and being tutored during the second half.

Of interest is that students who tutored others performed

lower than students in all other conditions both in terms of







immediate learning and retention at three- to nine-week

intervals. The observation that tutors would have learned

the subject matter more readily if they had studied inde-

pendently calls their participation as tutors into question.

The results of the Duff and Swick (1974) investigation dis-

cussed earlier, which did not find significantly greater

gains for tutors than for control subjects, lend support.

If an individual student does not benefit academically in

the process of tutoring, another teaching approach is called

for with that individual.

Dineen et al. (1977) present the first investigation to

examine the relative benefit of serving as a tutor or being

tutored for the individual child. Children's spelling

performance under tutor and tutee conditions was contrasted

with their performance on words with which they have no con-

tact except for pre- and posttesting. Prior to the collec-

tion of data all three students were trained to follow a

highly structured tutoring procedure and the teacher praised

or fined the children during the tutoring sessions. Follow-

ing each tutoring session, the tutee took a spelling test

over the words just studied. The children's performances on

the word list on which they tutored and the no-contact

comparison were assessed via pre- and posttesting. This








may present an interpretive problem as any effect of daily

testing upon performance would only be noted for the word

list on which a child served as a tutee. One of the three

children learned the most serving as a tutee, one gained

the most serving as a tutor, and the third child gained

equally serving as tutor and tutee. The mean percent change

for all three children was a 1 percent loss on the control

words. Dineen et al. conclude that "The process of tutoring

is not necessarily a waste of the tutor's time" (1977,

p. 236). This conclusion is not universally supported in

the tutoring literature (Duff & Swick, 1974; and Myers et al.,

1965). Further, a classroom teacher must arrange optimal

learning conditions for an individual child and cannot

assert that tutoring or being tutored is an effective edu-

cational technique when contrasted with performance on a

"no contact" control word list; these performances need

comparison with some alternative educational approach to

learning. Dineen et al. (1977) acknowledge that the highly

structured tutoring procedure and frequent teacher inter-

ventions maintained high levels of on-task behavior during

tutoring sessions which influenced both tutor and tutee

gains. This study was conducted in a classroom where two

teachers were available for 12 children. Few classrooms

have such a rich teacher-student ratio, and it is








questionable whether such a highly controlled tutoring pro-

cedure could be carried out by a single teacher.

The current series of experiments were conducted to

examine whether children can be effective tutors without

prior training or supervision during tutoring sessions. In

reviewing the tutoring literature, no studies were found

that examined whether individual tutors learned academic

material while serving as a tutor when the tutoring process

was unstructured. The current experiments examined this ques-

tion and compared the relative effects of serving as a tutor,

being tutored, and studying independently upon the acquisi-

tion of academic material. Because the age of the subjects

might influence the effectiveness of a tutoring arrangement

(Devin-Sheehan et al., 1976; and Wright, 1967), the three

experiments were conducted with subjects of three different

age brackets.

Harris et al. (1972) present the only data regarding

retention of academic material learned within a tutoring

paradigm and they were unable to ascertain differential

effects upon retention of having served as a tutor or a

tutee. Experiments I and II examined whether retention of

academic material is differentially effected as a function

of the manner in which the material was learned.














EXPERIMENT I


Method

Subjects and Setting

In a class of six children, three boys served. This

class was situated in a large room with two other classes

of six children each. Within each of the three classes each

child had an individual desk and chair which were clustered

around their own teacher. The classes were each separated

from the other classes by approximately two to three meters

of vacant space which served as passageways for movement

throughout the room. Each of these other classes was super-

vised by its own teacher and did not participate in this

experiment. All phases of this investigation were conducted

while the other children were present and engaged in other

learning tasks.

All children who participated in this study were aca-

demically deficient but intellectually unimpaired. Chuck

was 12 years of age and had an IQ of 98; Don was 13 years

of age and had an IQ of 109; and Alan was 13 years of age

and had an IQ of 113. All three subjects had taken daily









timed probes on their other academic material for approxi-

mately seven months prior to the start of this study. The

subjects did not exhibit any habitual or significant inap-

propriate classroom behaviors prior to or during this

investigation. Each subject was selected based on his

teacher's casual observations of deficient spelling perform-

ance compared to other children the same age. Pseudonyms

were provided for each subject.

Word Selection and List Construction

Spelling words used in this investigation were taken

from Levels 4, 5, and 6 of Basic Goals in Spelling (Kott-

meyer & Ware, 1964). Pretesting was conducted over a seven-

day period and 400 words were presented. After the teacher

presented each word verbally and used it in a sentence,

the subjects were asked to spell that word. The next word

was presented when all subjects were finished spelling the

previous item. Each day's pretest was corrected prior to

the subject's lunch period, and each subject was told the

number of words he had spelled correctly. The subjects

were given a small piece of candy for every 10 words they

had spelled correctly. They were not informed which words

were correct or incorrect.








Lists of words incorrectly spelled (a word was con-

sidered misspelled if it contained one or more letters) in

sequence incorrect; see below) were compiled for each

subject. From these lists a list was developed consisting

of only those words that all three boys had misspelled.

This list was used to randomly generate six lists of 15

words (without repetitions) used during probes. Each word

was printed on a 9.6 cm by 12.1 cm index card for use by

the subjects during the various experimental conditions.

Procedure

Experimental conditions.

A) Tutor: One subject was asked to teach another

boy his spelling words. The tutor was given a

stack of 15 index cards which contained the

words. These cards were returned daily following

the 10-minute tutorial session. No instruc-

tions were provided regarding how to tutor.

B) Tutee: A boy was tutored by another for 10

minutes.

C) Independent Study: A child was given a stack of

index cards and asked to study his spelling

words while seated at his desk alone for 10

minutes.








D) Control: The child has no instructional contact

with these words.

Experimental design. The boys were exposed to two

phases in this experiment. In Phase I each served as a

tutor for another boy, was tutored on a second word list,

and independently studied a third word list. Experimental

sessions were only held when all three boys were in attend-

ance. Following the completion of Phase I, tutor-tutee

pairings were changed and new word lists were substituted

for those previously used.

During Phase II each boy served as a tutor, a tutee,

and had no instructional contact with a third word list.

Figure 1 summarizes these arrangements. In each phase

each word list served a different function for each boy.

For example, Word List 1 served as the tutor word list for

Art, as the tutee word list for Chuck, and as the inde-

pendent study word list for Don.

Measurement units. Spelling involves sequencing letters

correctly in relation to one another (Liberty, note 3; and

White & Haring, 1976). This study examined spelling per-

formance in terms of letters in sequence correct (LISC)

and letters in sequence incorrect (LISI). Examples of these

responses are































FIGURE 1

Arrangement of Experimental Conditions in Experiment I.
Arrow directions indicate tutor-tutee interactions (e.g.,
in Phase I Art tutors Chuck on word list 1).

















PHASE I


ART
LIST 3 LIST I

DON < CHUCK
LIST 2


INDEPENDENT STUDY:
DON, LIST I
ART, LIST 2
CHUCK, LIST 3


PHASE IL


ART
LIST 4 LIST 6


DON CHUCK
LIST 5


CONTROL
DON, LIST 6
ART, LIST 5
CHUCK, LIST 4








Word Subject's Spelling Scoring


A A AA
cat c a t 4 LISC


cat ca 2 LISC, 1 LISI
v

cat b a 3 LISI
V V V


cat c a b 2 LISC, 2 LISI
vv

AA A
cat c a t t 3 LISC, 2 LISI
v v


Any word which was omitted by the speller was scored as being

incorrect (e.g., if cat were omitted 4 LISI would be scored).

Combining these units with the time spent completing

the probe words allowed this study to not only measure spell-

ing accuracy but also to measure spelling fluency. Data

on words correctly spelled were collected as well.

Acquisition probe sessions. At the completion of each

day's spelling period, the teacher presented each subject

individually with a series of three probes containing 15

words each. The teacher read each word to the subject, who

wrote it on his paper. The subject determined the rate of

word presentation by raising his hand when he wanted to hear

the next word. The teacher used a stopwatch to measure the

time taken by the subject to spell the words in each list.

Timing began when the first word was verbalized by the





19

teacher, and timing stopped when the subject raised his hand

indicating that he was finished writing the final word of

that list. At the completion of the probe session, each

subject returned to his desk and other assignments.

Retention probe sessions. Twice during Phase II of

this experiment retention probe sessions were conducted.

Each of these sessions occurred following the completion of

each subject's acquisition probe session on the word lists

used in Phase II. During these retention probe sessions,

the teacher presented each subject individually with a

series of three probes using the Phase I word lists. The

same procedures were used as during acquisition probe

sessions with the exception that the subjects had not

studied these word lists on the day of the probe adminis-

tration.

Contingencies. Probe performances were corrected prior

to the following school day. Feedback was provided by tell-

ing each boy the sum total of letters-in-sequence-correct/

minute from the previous day's three probes. This measure-

ment unit had been explained to the students prior to the

beginning of data collection. Points exchangeable for ac-

tivities (e.g., early recess, class trip, ride in the

teacher's car) were provided based on their previous day's

performance. In addition, each day their letters-in-





20

sequence-correct/minute were plotted on a graph posted on the

classroom wall. No feedback was given regarding specific

words spelled correctly or incorrectly in order to avoid con-

founding the effects of the conditions with specific feedback.

Interobserver Agreement

Each day, the primary observer scored the spelling per-

formances on mylar sheets. Twice during Phase I, and once

during Phase II, a second observer independently scored that

day's spelling performances on separate mylar sheets. Agree-

ment was scored when both observers placed a ^ or v in the

same place indicating a letters-in-sequence-correct or a

letters-in-sequence-incorrect, respectively. A disagreement

was scored when either observer made a mark that the other

had not, or when the observers disagreed as to whether two

letters-in-sequence were correct or incorrect. During the

three checks made in the course of this study, the observers

agreed 100 percent of the time.

Results and Discussion

Acquisition

Figures 2 through 10 present the acquisition and reten-

tion data for each subject in terms of LISC/minute and LISI/

minute. The charting conventions used in these figures are

those of the standard behavior chart (Pennypacker, Koenig, &

Lindsley, 1972). A computer program fitted celebration lines































FIGURE 2

The open circles represent Art's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Art's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.










100
90
80
70
60
50
40

30


20




10
9
8
7
6
5
4

3


15
SUCCESSIVE


25 35
CALENDAR DAYS


45






























FIGURE 3

The open circles represent Art's letters-in-sequence-
correct/minute on the daily probes when tutoring. The
open squares represent Art's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.











I00
90
80
70
60
50
40

30


20




10
9
8
7
6
5
4

3


2





.9
.8
.7
.6
.5


15
SUCCESSIVE


25
CALENDAR


35
DAYS


45
































FIGURE 4

The open circles represent Art's studying independelty
during Phase I and on the control words during Phase II.
The open squares represent Art's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probe of the word
list used in Phase I.











100
90 -
80
70 X I.16 '--
50 -0-- 1.07

40
ART
30


Z
D- 20 ------0-


1.03

z 10
D 9
ca 8
0 8 \ 1
2u 7- \


WZ 5

--1





8 \
\r





.6 -

.5 I I I I
5 15 25 35 45
SUCCESSIVE CALENDAR DAYS































FIGURE 5

The open circles represent Don's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Don's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.










100 .
9 0 0 b
80 o / 9
70 -
60 -" X X 1.24
50 x.38
40 -

30 DON
30 -



z
I- 20



w + 3.03 1
z 10 -204
w 9- o o
8
W 7 -7
S 6
z I5
U)'1
w 4\
1- \0



2





.9


.6 -
.5I I I I
5 15 25 35 45
SUCCESSIVE CALENDAR DAYS
































FIGURE 6

The open circles represent Don's letters-in-sequence-
correct/minute on the daily probes when tutoring. The
open squares represent Don's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.









TUTORS ART
(List 3)


SX 1.36
f-
*I"


100
90
80
70
60
50
40

30


20




10
9
8
7
6
5
4

3


2





.9
.8
.7
.6
.5


0


TUTORS CHUCK
(List 5)



x 1.14 ,-

---D-N


DON


+ 1.32


1T
1r


SUCCESSIVE CALENDAR


35
DAYS


45


-1.47
































FIGURE 7

The open circles represent Don's letters-in-sequence-
correct/minute on the daily probes when studying
independently during Phase I and on the control words
during Phase II. The open squares represent Don's
letters-in-sequence-incorrect/minute. The closed
circles and squares represent letters-in-sequence-
correct/minute and incorrect/minute on the retention
probes of the word list used in Phase I.









INDEPENDENT STUDY
(List I)


X1.19


- oo


100
90
80
70
60
50
40

30


20




10
9
8
7
6
5
4

3


2




I


CONTROL
(List 6)



X 1.07
-. .- -- o
---~ -- r-V


DON


- I.IU
I-- "


Y)


I II I I


15 25
SUCCESSIVE CALENDAR


S1.45


35
DAYS
































FIGURE 8

The open circles represent Chuck's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Chuck's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.







TUTORED BY ART
(List I)


100
90
80
70
60
50
40
30

20


-0


--~-o --cf. -

-1.01


TUTORED BY DON


TUTORED BY DON
(List 5)

o-Jo- xI.-06 -oo
y0" x 1.06 O


0


,, [
K


- 1.47 ,
0-


N


CHUCK






.1


5 15 25 35 45
SUCCESSIVE CALENDAR DAYS


x 1.11


.41-































FIGURE 9

The open circles represent Chuck's letters-in-sequence-
correct/minute on the daily probes when tutoring. The
open squares represent Chuck's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.








TUTORS DON
(List 2)
x 1.19


100
90
80
70
60
50
40

30


20


- 1.14


TUTORS ART
(List 6)

x 1.00



^--J^-^-'(-^-^
P- .- -------o-~-



x 1.00








CHUCK














1"


5 15 25 35 45
SUCCESSIVE CALENDAR DAYS


-V--































FIGURE 10

The open circles represent Chuck's letters-in-sequence-
correct/minute on the daily probes when studying
independently during Phase I and on the control words
during Phase II. The open squares represent Chuck's
letters-in-sequence-incorrect/minute. The closed
circles and squares represent letters-in-sequence-
correct/minute and incorrect/minute on the retention
probes of the word list used in Phase I.








100
90
80
70
60


o -o .0


CONTROL
(List 4)


I -R-


50 ,-" x I.16 --- o-
40 o o

30 -


20 0 x 1.03
X|.|0 0
LLJ


SI10
9
w 8
0 7
z
UL 6
o 5
,,w CHUCK
U) 4
z
U) 3
cr-
UJ

H 2





.9-
.8 -
.7- '

.6 -
.5
.4

.3 i- i I i-I
.3
5 15 25 35 45
SUCCESSIVE CALENDAR DAYS


T


"a;



tP-
P








to the data display in Figures 2 through 10 using the least

squares method. Overall procedural effectiveness is pre-

sented in Table 1 which prescribes each subject's improvement

index (i.e., celebration of daily accuracy ratios [LISCt LISI]).

Figures 11 through 13 present the acquisition and retention

data for each subject in terms of words correct.

Each of the subjects improved the least on the control

word list. However, while Don and Chuck were unaffected by

daily probes in this condition, Art's improvement index of

1.10 points out that daily assessments may make a contribu-

tion to learning independent of other experimental proce-

dures. This observation is supported by the fact that Art

learned to spell three words from the control word list

correctly during Phase II, in spite of having no instruc-

tional contact with these words.

Being tutored resulted in the most learning by Don

as his two highest number of words correct (14 words and

15 words) and two highest improvement indices (4.13 and

2.52) resulted during the tutee conditions. Chuck learned

to spell the most words correctly (10 words) and his highest

improvement index of 1.57 resulted during Phase II's tutee

condition. Art improved the most (2.63 improvement index

and 15 words correctly spelled) when allowed to study alone,

but his next two highest improvement indices (2.49 and 2.02)



















rO o o o
CD NO ICM ID O













o) '- 1l F-- o
- '- N -









N i '- o 0 ,D -








0)

F-
I-

z


LO

0 0 U IU I-
I I-I- 5 Z

H- F- F- 0


0
4-
*4J
.r4










-,-
,CQ







o
-H
0









0













4
a)

0


*I-I












0)































FIGURE 11

Words spelled correctly in the various conditions by Art.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
tutor words; the open circles represent independent study
words; the open stars represent control words. Closed
squares, triangles, and circles represent performance
on the retention probes from the Phase I word lists.










0






0 A


1


ART


a


5 15 25 35 45


SUCCESSIVE CALENDAR DAYS


6-

5

4

3


A I


*

J































FIGURE 12

Words spelled correctly in the various conditions by Don.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
tutor words; the open circles represent independent study
words; the open stars represent control words. Closed
squares, triangles, and circles represent performance on
the retention probes from the Phase I word lists.












I
0 L~1


DON


0


j


0


*
*r


I 4


i


5 15 25 35 45


SUCCESSIVE CALENDAR DAYS


*r


0- ^































FIGURE 13

Words spelled correctly in the various conditions by Chuck.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
independent study words; the open stars represent control
words. Closed squares, triangles, and circles represent
performance on the retention probes from the Phase I
word lists.












CHUCK


*


0 r r


5 15 25 35 45


SUCCESSIVE CALENDAR DAYS


5

4








O-
2



0


ql r








and next two highest number of words spelled correctly on

the final day of acquisition (14 words and 14 words)

occurred when he was tutored. With the exception of the

negligible gains made by Chris when tutored by Art (1.12

improvement index and 2 words spelled correctly), these

data suggest that each subject benefited from being tutored,

and that each was effective in teaching previously unknown

spelling words without being instructed in how to provide

tutoring.

Whether a student's involvement as a tutor is justified

can only be answered on an individual basis. The question

becomes does the student learn as much or more when tutor-

ing compared to how much he learns when his time spent

tutoring is spent in some alternative educational activity?

Having Art serve as a tutor for Chuck seems ill-advised

as his improvement index of 1.23 and words correctly spelled

(10 words) were much lower than his gains when allowed to

study independently (2.63 improvement index and 14 words

correct). The results for Don and Chuck provide a mixed

answer to the question. When their improvement indices in

the tutoring conditions are contrasted with the indices

obtained in the independent study condition, both subjects

learned more studying independently than they did in one

of the tutoring conditions. In the other tutoring








condition, each boy learned more quickly than when studying

alone.

Certainly tutor-tutee pairings of these three boys

resulted in better performance when compared to other pair-

ings. This arrangement of design elements allowed each boy

to serve as a tutor and be tutored by each of the other two

boys. When the improvement indices for each dyad's learning

under tutee and tutor conditions in the two experimental

phases are-averaged, the pairing of Chuck and Art resulted in

the poorest overall performance (x = 1.29, refer to Table 2).

Chuck tutoring Don seemed to be a good pairing as Chuck's

index score of 1.36 was his second highest improvement index,

and Chuck's highest index score of 1.57 was obtained when

Don was tutoring him. However, Don's improvement index of

1.49 in this condition was his poorest performance (with

the exception of performance in the control condition); his

continued participation as a tutor in this pairing may not

be advised. As mentioned previously, Art did not learn

very well when he served as a tutor for Chuck. These

results suggest that while a teacher can allow children to

tutor one another in an unstructured arrangement, tutor-

tutee pairings which are empirically based seem advisable

in order to insure maximum learning by each student.













TABLE 2

Improvement Indices of Each Tutor-Tutee Dyad


Chuck tutors Don

Don tutors Chuck






Art tutors Chuck

Chuck tutors Art






Don tutors Art

Art tutors Don


Don

4.13

1.49




Chuck

1.12

1.01




Art

2.49

1.56


Chuck

1.36

1.57




Art

1.23

2.02




Don

1.99

2.52


x = 1.90








x = 1.29








x = 2.10








Retention

Table 3 presents the individual retention data for the

word lists used in Phase I. The number of words correctly

spelled on each retention probe was divided by the number of

words correctly spelled during the final acquisition probe

in Phase I. The number of words correctly spelled are pre-

sented in Figures 11 through 13.

The only consistent feature of the retention data is

that each subject in each condition (tutor, tutee, and inde-

pendent study) spelled more words correctly during the reten-

tion probes than he had during the first acquisition probe.

No consistent relationship emerged to indicate that any one

of the three experimental conditions was more effective than

the other conditions in promoting the retention of what was

learned. Similarly, no consistent relationship emerged

indicating that spelling performance deteriorated more from

the first to the second retention probe as a function of

experimental condition.

While data were not gathered regarding what teaching

techniques were used in the tutoring process, on several

occasions the tutors were noted to be providing practice

probes with the tutor verbally presenting the words as

quickly as the tutee spelled them. At no point in this












TABLE 3

Words Correct (WC) on Each of the Retention Probes (RPI and
RPII) Divided by Words Correct on the Final Acquisition
Probe (FA) in Phase I for Each Word List Expressed as a
Percentage.


Art


Tutor

Tutee


Independent Study


Don


Tutor

Tutee


Independent Study



Chuck Tutor

Tutee


Independent Study


RPI WC
FA WC


90%

86%

93%


63%

79%

75%


83%


100%

200%


RPII WC
FA WC

100%

71%

93%



75%

86%


88%



100%

68%

100%






52

experiment were the boys observed to exchange tutor-tutee

roles. During this investigation the teacher was able to

work with the three other children in the class who did not

participate in the experiment. Supervision of the tutoring

sessions was unnecessary as periodic observation indicated

that the boys stayed on-task during these sessions.















EXPERIMENT II


Introduction

The objective of Experiment II was to serve as a

replication of Experiment I. The major distinction between

these two investigations was that the subjects were two

years younger.

Method

Most of the procedures used in Experiment I were

repeated in Experiment II. This section only notes pro-

cedural differences.

Subjects and Setting

In a class of six children, three boys served. This

class was located in a large room shared with one other

class of six children. Within each of the classes each

child had an individual desk and chair which were clustered

around their own teacher's desk. The classes were separated

from each other by approximately two to three meters of

vacant space which served as a passageway. The other class

in the room was supervised by its own teacher and did not

participate in this experiment. All phases of this








investigation were conducted while the other children were

present and engaged in other learning tasks.

All children who participated in this study were

academically deficient but intellectually unimpaired. Kurt

was 10 years of age and had an IQ of 104; Dave was 11 years

of age and had an IQ of 112; and Rob was 11 years of age

and had an IQ of 118. All three subjects had taken daily

timed probes on other academic material for approximately

seven months prior to the start of this study. The subjects

did not exhibit any habitual or significant inappropriate

classroom behaviors prior to or during this investigation.

Each subject was selected based on his teacher's casual

observations of deficient spelling performance compared to

other children the same age. Pseudonyms were provided for

each subject.

Word Selection and List Construction

Spelling words used in this study were taken from

Levels 4 and 5 of Basic Goals in Spelling (Kottmeyer &

Ware, 1964). Pretesting procedures were the same as those

used in Experiment I. Pretesting was conducted over a six-

day period and 320 words were presented. All words used in

this study were randomly chosen from words that all three

boys had misspelled during protesting.








Feedback

Each day feedback on the boys' sum total of letters-

in-sequence-correct per minute from the previous day's

three probes was written by the teacher on a paper which

hung by each boy's desk. Points based on the previous

day's sum total of letters-in-sequence-correct/minute were

exchangeable for activities in this experiment as well.

Experimental Design

The boys were exposed to two experimental phases in

this experiment. The experimental design is identical to

that used in Experiment I with the exception that the con-

trol condition occurred during Phase I and the independent

study condition occurred during Phase II. Figure 14 sum-

marizes these arrangements.

Interobserver Agreement

Once during Phase I and once during Phase II a second

observer scored that day's spelling performances on a

different mylar sheet than was used by the primary observer.

The check conducted during Phase I resulted in three dis-

agreements between the observers; in Phase II one disagree-

ment was noted. Both agreement checks yielded agreements

above 99 percent. Where these disagreements were noted,

the words were rescored. This rescoring occurred prior to

plotting the data.






























FIGURE 14

Arrangement of Experimental Conditions in Experiment II.
Arrow directions indicate tutor-tutee interactions (e.g.,
in Phase I Dave tutors Kurt on word list 1).
















PHASE I


DAVE
LIST 3 LIST I

ROB KURT
LIST 2


CONTROL:
ROB, LIST I
DAVE, LIST 2
KURT, LIST 3


PHASE IT


DAVE
LIST 6 LIST 4

ROB > KURT
LIST 5


INDEPENDENT STUDY:
ROB, LIST 4
DAVE, LIST 5
KURT, LIST 6








Results and Discussion

Acquisition

Figures 15 through 23 present the acquisition and

retention data for each subject in terms of LISC/minute

and LISI/minute. Overall procedural effectiveness is pre-

sented in Table 4 which describes each subject's improvement

index. Figures 24 through 26 present the acquisition and

retention data for each subject in terms of words correct.

All of the subjects improved the least on the control

word list, but all had improvement index scores greater than

1.00 indicating that daily testing even without specific

feedback about errors resulted in improved spelling. This

observation is supported by both Rob's (Figure 24) and

Dave's (Figure 25) data on words correct which shows that

both subjects were able to spell more words from the con-

trol word list correctly at the end of Phase I than they

had at the beginning of this phase.

Overall, being tutored resulted in the largest im-

provement in spelling performance. Dave's highest improve-

ment indices of 4.35 and 2.73 were achieved when he served

as a tutee. His data on words correct reveal that he

spelled more words correctly on the tutee word list than

on the other word lists in both phases. Kurt achieved his































FIGURE 15

The open circles represent Rob's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Rob's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.









TUTORED BY KURT
(List 2) o


- x 1.17


-4.17


0


\j
\
\Oc
\0
\


TUTORED BY DAVE (List6)-
x 1.10 -
? O--o ,
o .. --- "


-1.20
% 1



1.20


ROB


*1


I.


15 25 35
SUCCESSIVE CALENDAR DAYS


45


100
90
80
70
60
50
40

30


20


10
9
8
7
6
5
4

3


1


L






























FIGURE 16

The open circles represent Rob's letters-in-sequence-
correct/minute on the daily probes when tutoring. The
open squares represent Rob's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.








- TUTORS DAVE
(List 3)
0

x

0


+1.32


.4-


TUTORS KURT
(List 5)


x --I
X 1. 11


+ 1.32









ROB


ROB


;%












n


5 15 25 35 45
SUCCESSIVE CALENDAR DAYS


100
90
80
70
60
50
40

30


20


'I- -


--%


- --
-































FIGURE 17

The open circles represent Rob's letters-in-sequence-
correct/minute on the daily probes on the control words
during Phase I and when studying independently during
Phase II. The open squares represent Rob's letters-
in-sequence-incorrect/minute. The closed circles and
squares represent letters-in-sequence-correct/minute
and incorrect/minute on the retention probes of the
word list used in Phase I.







CONTROL


CONTROL
(List I)


-x .0 3 o
x 1.03


~0 -


-.0
+1.10


INDEPENDENT STUDY
(List 4)


100
90
80
70
60
50
40

30


20


x .0---
x 1.04


- 1.18 -






ROB










T


15 25 35
SUCCESSIVE CALENDAR DAYS


45


6


I i I I I I































FIGURE 18

The open circles represent Dave's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Dave's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.








TUTORED BY ROB
(List 3)


100
90
80
70
60
50
40


x 1.37


oe


+ 3.23


\0


TUTORED
(List


BY KURT
4)


x 1.19
y-o"


- 2.27\
1


DAVE


15 25 35
SUCCESSIVE CALENDAR DAYS


45


o0 /


30


20


o~'
//






























FIGURE 19

The open circles represent Dave's letters-in-sequence-
correct/minute on the daily probes when tutoring. The
open squares represent Dave's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.





68

100
90 TUTORS KURT TUTORS ROB
80 (List I) (List 6)
70 -
60 -
50 X 1.23 -"15
40- o ,"
3O 9-
30 -

DAVE
20
20 -



\ +1.85
S10 1.61
9- \ -

5 \ \-
w 8
o 7
.6 6 -



z
U) 3
uJ

S2
_J





.9
.8-
.7
.6-
.5-

.4

.3 I I
5 15 25 35 45
SUCCESSIVE CALENDAR DAYS






























FIGURE 20

The open circles represent Dave's letters-in-sequence-
correct/minute on the daily probes on the control words
during Phase I and when studying independently during
Phase II. The open squares represent Dave's letters-
in-sequence-incorrect/minute. The closed circles and
squares represent letters-in-sequence-correct/minute
and incorrect/minute on the retention probes of the
word list used in Phase I.








100
90 CONTROL INDEPENDENT STUD
80 (List 2) (List 5)
70
60
50 ~ P X 1.01
40 X 1.06

30- --I 30
30 I





o o
20 3







LrJ 8
u 7
z
LU 6
0 5
LUi
U) 4
z
S3 DAVE
U--

1-
L- 2





.9-
.8-
.7-
.6 --
.5 -

.4

.3 I I I
5 15 25 35 45
SUCCESSIVE CALENDAR DAYS































FIGURE 21

The open circles represent Kurt's letters-in-sequence-
correct/minute on the daily probes when being tutored.
The open squares represent Kurt's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.









'8R
80
70
60
50
40

30


20


- 1.35


nI


TUTORED BY ROB
(List 5)


TUTORED BY DAVE
(List 1)







x 1.19 .-


o0

011


u

. 1.52 %\ \

\\


KURT


- I


- _I,


15 25 35
SUCCESSIVE CALENDAR DAYS


x 1.20,.- "

,- -


45
































FIGURE 22

The open circles represent Kurt's letters-in-sequence-
correct/minute on the daily probes when Tutoring. The
open squares represent Kurt's letters-in-sequence-
incorrect/minute. The closed circles and squares
represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.







100
90
80
70
60
50
40

30

20


0


-1.06


TUTORS DAVE
(List 4)


TUTORS ROB
(List 2)





x 1.27,-o
o/- -0


o %-



1.03


-fi


KURT


I~


5 15 25 35 45
SUCCESSIVE CALENDAR DAYS


x 1.17































FIGURE 23

The open circles represent Kurt's letters-in-sequence-
correct/minute on the daily probes on the control words
during Phase I and when studying independently during
Phase II. The open squares represent Kurt's letters-
in-sequence-incorrect/minute. The closed circles and
squares represent letters-in-sequence-correct/minute and
incorrect/minute on the retention probes of the word
list used in Phase I.








100
0 CONTROL INDEPENDE
80 (List 3) (List
70
60
50
40
xl.16





I.12
Sx 1.01
zI
- 10
S9
8-
o 7
Lz 6
D
a 5
.IJ
( 4 KUR
z
u) 3
cr
L.J
I-
S2
-J




.9
.8
.7
.6
.5-

.4-

.3 I I I L
5 15 25 35
SUCCESSIVE CALENDAR DAYS
















(D N -
SCNM ( C M










D N C U r0 0 -
0 o N- N 0










10 N- 0 N M
U' ) 07) rO CN









I-



I- I-- -
0




O w
O-








?~ ? ?


a
o

-4


4-,
u
rd

C4






0


a)
tJ
0
rt
U
a)





a)





Q)































FIGURE 24

Words spelled correctly in the various conditions by Rob.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
tutor words; the open circles represent independent study
words; the open stars represent control words. Closed
squares, triangles, and stars represent performance on
the retention probes from the Phase I word lists.











ROB


14 o

13 0 o

12

II

-0J


9

0 8- A a


4- AT
J 7 -
6 o I


S5 -



3


2-

2-

0-


5 15 25 35 45


SUCCESSIVE CALENDAR DAYS






























FIGURE 25

Words spelled correctly in the various conditions by Dave.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
tutor words; the open circles represent independent study
words; the open stars represent control words. Closed
squares, triangles, and stars represent performance on
the retention probes from the Phase I word lists.












DAVE
14

13

12

II 0




0
cr 9-
0
0 8

-j
-J 7
w
0-
Cl

a:



4-

3 t
rI








2 0 Y




0 ov


5 15 25 35 45


SUCCESSIVE CALENDAR DAYS






























FIGURE 26

Words spelled correctly in the various conditions by Kurt.
The open squares represent performance on tutee words
during the daily probes; the open triangles represent
tutor words; the open circles represent independent study
words; the open stars represent control words. Closed
squares, tiangles, and stars represent performance on the
retention probes from the Phase I word lists.












14 -

13 KURT

12

I I

>O
0 -
ui6
w
-J 9
o



-j 7


U) T
6



4 o



2
4- ff r1




5 5 25 35 4
5 15 25 35 45


SUCCESSIVE CALENDAR DAYS


-









two highest improvement indices (1.82 and 1.61) and spelled

the most words correctly in both phases on word lists on

which he was tutored as well. Similarly, Rob achieved his

highest improvement index of 4.90 and spelled all 15 words

correctly when being tutored by Kurt. He was able to spell

13 words correctly when being tutored by Dave, although his

performance on the final two tutee acquisition probes in

Phase II slipped slightly. These data show that each sub-

ject benefited from being tutored, and that each was an

effective tutor without instruction on how to perform this

task.

When serving as tutors both Rob and Dave learned to

spell more words correctly and had higher improvement

indices than when allowed to study independently. When

serving as a tutor, Kurt learned at a similar rate of

improvement (his improvement index during Phase I was 1.36

and during Phase II was 1.21) and knew how to spell a

similar number of words (seven in both phases) as when he

studied independently (improvement index of 1.31 and six

words spelled correctly). In this instance the improvement

index data are deceptive as the values from both tutoring

conditions and the independent study condition are either

equal to or just slightly greater than Kurt's improvement








index on the control word list (1.21). But Kurt clearly

learned how to spell several words correctly from the

tutoring and independent study word lists, but did not learn

to spell any words from the control list.

When the improvement indices for each dyad's learning

under tutor and tutee conditions in the two experimental

phases are averaged, the pairing of Dave and Kurt resulted

in the poorest overall performance (x=1.80, refer to

Table 5). This is not to say that some alteration of the

tutorial arrangement, such as providing teacher supervision

over their interactions, might not improve the learning

achieved by this dyad. If a teacher wanted to continue

using an unstructured and unsupervised tutoring arrangement,

different pairings for each boy may be advised. The same

conclusion may be drawn when Dave tutors Rob. Dave was not

an effective tutor in this situation as Rob's improvement

index of 1.31 is barely higher than his independent study

and control performance of 1.22 and 1.14, respectively.

Dave's data on words spelled correctly support this

conclusion.

Retention

Table 6 presents the individual retention data for the

word lists used in Phase I. The number of words correctly

















TABLE 5

Improvement Indices of Each Tutor-Tutee Pairing


Rob tutors Dave

Dave tutors Rob






Dave tutors Kurt

Kurt tutors Dave






Rob tutors Kurt

Kurt tutors Rob


Dave

4.35

2.12




Kurt

1.61

1.21



Kurt

1.82

1.36


Rob

1.55

1.31




Dave

1.98

2.73



Rob

1.47

4.90


x = 2.08








x = 1.80








x = 2.05











TABLE 6

Words Correct (WC) on Each of the Retention Probes (RPI and
RPII) Divided by Words Correct on the Final Acquisition
Probe (FA) in Phase I for Each Word List Expressed as a
Percentage.


RPI WC
FA WC


Rob Tutor

Tutee

Control


67%

80%

113%


RPII WC
FA WC

44%

80%

100%


Dave Tutor

Tutee

Control



Kurt Tutor

Tutee


Control


46%

14%

100%



57%

63%

100%


31%

21%

67%




57%

38%

100%









spelled on each retention probe was divided by the number

of words correctly spelled during the final acquisition probe

in Phase I.

All three subjects' data display a common configura-

tion in that each retained less on words learned while

tutoring or being tutored than on control words. This is

similar to the results of Harris et al. (1972), who found

that performances on words studied independently and words

used during unstructured tutoring showed greater declines

from posttest to retention test than did performances on

comparison (control) word lists. Harris et al. (1972)

present their data in group format, and three of the four

classes for which data are presented show this relationship.

It would seem plausible that this relationship might be

explained by performances on no contact control words

remaining stable from pretest to posttest (acquisition)

and remaining stable when retested three to five weeks

later. Put another way, the subjects in Harris et al.

(1972) were making the same errors on the control words

on retests as they had when pretested. However, two of the

three classes had substantial pretest to posttest gains in

mean classroom percentage on the control word lists.










Because data on individual performances are not presented

by Harris et al. (1972), it is difficult to posit a plausible

explanation of their retention data and the authors offer

none of their own.

Returning to the data from this investigation, Kurt

did not learn how to spell any control words during acqui-

sition (Figure 26) and, thus, had no knowledge to retain.

Since Kurt did learn approximately half of the words on the

lists he tutored with and was tutored on, his lower number

of words spelled correctly can be attributed to simply for-

getting what he had learned due to no contact with these

words between the final acquisition probes during Phase I

and the retention probes 22 and 28 days later. Dave's

data on words spelled correctly (Figure 25) may be explained

in a similar fashion. While Dave had learned how to spell

13 words and 14 words correctly on the lists used when he

tutored and was tutored, respectively, perhaps he had for-

gotten much of what he had learned by the time the retention

probes were given. However, Dave learned how to spell at

least two and possibly three control words (all subjects

had misspelled all of the words used in this experiment

during protesting) during Phase I. Dave was able to spell










three words and two words correctly on the retention probes,

suggesting that he may have retained a greater percentage

of what he had acquired on the control words relating to

what was retained on the words used when he served as tutor

and tutee.

It could be argued that the upward trend in Dave's

control words correctly spelled data does not represent

acquisition at all, but rather is not representative of what

would have been revealed had daily probes been continued.

Perhaps the number of words he spelled correctly may have

varied between zero words and four words without a discern-

ible trend. Examining Dave's performance on the control

word list in terms of LISC and LISI (Figure 20), it appears

that his performance did improve as Phase I progressed,

but his performance on the retention probes is similar in

terms of accuracy to that observed on the first day of

Phase I. Essentially, Dave did improve his spelling of the

control words during Phase I, but has forgotten what he

learned by the time he took the retention probes. In this

case, words spelled correctly may have been too coarse of

a measurement unit to reveal the decay in Dave's performance

on the control word list.










Rob appears to have retained the least on the words

he used when tutoring. During the retention probes, he

was able to spell more words correctly from the control

word list than from the tutor word list. During the

retention probes, Rob was able to spell as many control

words correctly as he has during the final acquisition

probe. The high degree of retention on the control word

list appears to be unique, particularly as both Rob's words

correct data (Figure 24) and LISC and LISI (Figure 17)

show that his spelling did improve as Phase I progressed,

i.e., he had acquired some knowledge which could be for-

gotten.

The fact that Rob performed better on the tutee word

list than the tutor word list during the retention probes

may be due to the fact that he had learned the tutee word

list to a higher level of mastery during acquisition than

the words on the tutor list. Eaton and Swenson (1973) found

that having a subject perform at a criterion level for two

days during acquisition enhances retention more than if the

subject only performed at that criterion level for one day.

The fact different levels of performance were achieved both

between conditions for each subject and between subjects






92

limits the conclusions which may be drawn from the retention

data.

However, it is clear from the retention data that

without instructional contact following acquisition perform-

ance deteriorates. Information about the rate of deteriora-

tion of the subjects' spelling performance is limited some-

what because only two retention probes were conducted.

Increasing the number of retention probes in each condition

would have allowed a regression curve to be fitted to the

retention data and may have revealed relationships which

cannot be seen with only two data points.

In spite of the limitations upon interpretation imposed

by having only two retention data points for each condition

and the fact that Phase I (acquisition) ended when the

subjects were performing at different performance levels,

the current results do provide a reminder for those who

use tutoring procedures. While the data did not reveal any

differential effects upon retention as a function of the

experimental condition in effect during acquisition, it was

clear that the subjects performed less accurately on the

retention probes in the tutor and tutee conditions than

they had during the final acquisition probe during Phase I.

Lest all the knowledge gained during acquisition be lost,

some educational intervention (e.g., periodic probes,






93



tutoring sessions, or the use of the material in another

context) seems indicated.