Journal of Educational Psychology 1979, Vol. 71, No.

1, 64-73
64
Development and Evaluation of a
Learning Strategy Training Program
Donald F. Dansereau, Karen W. Collins, Barbara A. McDonald, Charles D. Holley, John Garland, George
Diekhoff, and Selby H. Evans Texas Christian University
An interactive learning strategy system consisting of primary and support strategies was
developed. In addition, a set of training procedures was created to facilitate the
communication of this system to college-age students. This system was assessed in the
context of a 15-week (2 hours per week) learning skills course. The results of this
assessment indicated that students receiving this training exhibited significantly (p < .05)
greater positive precourse-postcourse changes on comprehension/retention and self-report
measures than did students in the control groups. Further analyses were conducted to
determine which strategy components were most potent and to provide directions for future
modifications.

Previously developed learning skills training programs have improved students' performances on broad
dependent measures such as college grade point averages (Briggs, Tosi, & Morley, 1971; Whitehill, 1972)
and self-report study habit surveys (Bodden, Osterhouse, & Gelso, 1972; Brown, Wehe, Zunker, & Haslam,
1971; Haslam & Brown, 1968; Van Zoost & Jackson, 1974). Al- though these programs seem to benefit
students, they suffer from several drawbacks which may restrict their effectiveness. First, the developers of
these programs have not drawn very heavily on the psychological and educational research literature.
Second, they have typically provided only broad level strategy training. Third, as mentioned previously, they
have usually evaluated these programs with relatively global dependent measures (e.g., grade point average).
Consequently, there has been little evidence on which to base modifications.
On the basis of a review of the literature related to academic learning (Dansereau, Actkinson, Long, &
McDonald, 1974), it was concluded that students would benefit from Comprehension/Retention detailed
training on a strategy system extrapolated from findings in the educational and psychological research
literature. The present research and development effort represents a second step toward this objective. It is
an attempt to expand and modify a modestly successful small-scale program developed by Dansereau et al.
(1975).
The approach to this research has been strongly influenced by the fact that effective interaction with
academic materials requires that the student engage in a complex system of interrelated activities. To assist
the student in this endeavor, a system of mutually supportive strategies has been developed. This system can
be divided into primary strategies, which are used to operate on the material directly, and support strategies,
which are used by the learner to maintain a suitable cognitive climate. Al- though these two sets of strategies
are highly interactive, they are discussed separately in the next two sections. (For a detailed description of
the system and the supporting rationale, see Dansereau, 1978.)

Primary Strategies
The goal has been to identify and develop strategies that will assist the student in reorganizing,
integrating, and elaborating incoming material. To reach this goal an executive strategy and a variety of
alternative substrategies have been developed.
 65
The executive steps, which are communicated by the acronym MURDER, are the following: setting the
mood to study (discussed under support strategies); reading for under- standing (marking important and
difficult ideas); recalling the material without referring to the text; correcting recall, amplifying and storing
the material to digest it; expanding knowledge by self-inquiry; and finally, reviewing mistakes (learning
from tests).
Although varying somewhat in surface structure, the main differences between this executive strategy
and the SQ3R approaches used in previous study skills programs (Robinson, 1946) occurs in the details of
the steps. Typically, training on SQ3R is non- specific; the steps are described and the students are expected
to translate these descriptions into operative substrategies. It appears, however, that a large number of
students have a great deal of difficulty in making this translation. To alleviate this problem detailed
instructions and practice exercises have been designed to communicate alternative substrategies that fall
under the executive framework. As an example of this approach, the recall substrategies (represented by the
R in MURDER) will be described briefly.
Recall and transformation strategies. After an initial reading, the student is instructed to recall the
material he or she has read. This is considered to be the most important phase of the executive
comprehension/retention strategy. Consequently, three substrategies that vary in the degree of
transformation (translation of the text into an alternative symbol system) required on the part of the student
have been developed: paraphrase/imagery, networking, and analysis of key ideas.
1. Paraphrase/imagery. This technique is a simple combination of the paraphrase (the students
intermittently rephrase the incoming material in their own words) and imagery (the students intermittently
form mental pictures of the concepts underlying the incoming material) strategies developed in an earlier
program (see Dansereau et al., 1975). In this version the student is trained on both techniques and is then
instructed to vary the use of the techniques depending on the material being studied.
2. Networking. Unlike the paraphrase/ imagery technique, which requires the student to transform text
material into natural language or pictures, the networking strategy requires material to be transformed into
node-link maps or networks. This technique is designed to assist the student in organizing passage
information based on principles abstracted from relationship- based models of long-term memory (Anderson
& Bower, 1973; Bobrow & Winograd, 1977; Rumelhart, Lindsay, & Norman, 1972). The student is trained
to convert prose into node-link diagrams (networks) using a set of four experimenter-provided links. The
nodes contain paraphrases and images of key ideas, and the links specify the relationships between these key
ideas (concepts). The networking process emphasizes the identification and representation of hierarchies
(type/part), chains (lines of reasoning/temporal orderings/causal sequences), and clusters
(characteristics/definitions/ analogies). Application of this technique results in the production of structured,
two-dimensional maps. These maps provide the student with a spatial organization of the information
contained in the passage.
3. Analysis of Key Ideas. The final comprehension/retention strategy is also derived from network
models of memory (Diekhoff, Note 1). In this structured alternative to networking the student identifies key
ideas or concepts in a body of text, develops systematic definitions and elaborations of the concepts, and
interrelates important pairs of these concepts. The student is aided in these activities by work- sheets that
specify categories of definition and comparison. These categories are iso- morphic to the four links
described earlier (e.g., in defining operant conditioning, one might say that it is a type of learning paradigm,
a part of many behavior modification programs, leads to increases in the target behavior, etc.).
In summary, the comprehension/retention techniques include an executive strategy which guides the
onset of substrategies designed to assist the student in solving com- prehension problems, recalling and
trans- forming learned material, and expanding knowledge.
66
Retrieval/Utilization Strategies
 After comprehending and storing a body of information the student must be able to recall and use the
information under appropriate circumstances (e.g., in taking tests or on the job). To aid the student in this
task an executive strategy with alternative substrategies has been developed. This strategy consists of six
steps: setting the mood; understanding the requirements of the task; recalling the main ideas relevant to the
task requirements (using means-ends analysis and planning); detailing the main ideas with specific
information; expanding the information into an outline; and re- viewing the adequacy of the final response.
These six steps were given the acronym “2nd degree MURDER" in order to facilitate recall of the technique.
Support Strategies
Regardless of the effectiveness of the primary strategies, their impact on learning and performance will
be less than optimal if the internal psychological environment of the student is nonoptimal. Consequently,
the support strategies were designed to assist the student in developing and maintaining a good internal state.
These support strategies include goal-setting and scheduling, concentration management, and monitoring
and diagnosing the dynamics of the learning system. These three classes of strategies are discussed
separately.
Goal-Setting and Scheduling
Students employed a workbook to specify their long- and short-term goals and to schedule their time.
Students were instructed to monitor their progress in achieving their goals and alter their schedules when
appropriate.
Concentration Management
Students employed a technique designed to address attitude (or mood) problems and problems in coping
with distractions. The concentration management technique combined elements of systematic
desensitization (Jacobsen, 1938; Wolpe, 1969), rational behavior therapy (Ellis, 1963; Maultsby, 1971), and
therapies based on positive self-talk (Meichenbaum & Good- man, 1971; Meichenbaum & Turk, Note 2).
The students were first given experiences and strategies designed to assist them in becoming aware of
tension, negative and positive emotions, self-talk, and images they generate in facing a learning task. They
were then encouraged to evaluate the constructiveness of their internal dialogue and were given heuristics
for making appropriate modifications. Worksheets and experimenter-generated sample statements were used
to assist the student in this task. The students were then taught relaxation techniques and instructed to use
these techniques and constructive self-talk to establish an appropriate mood for studying and test- taking.
(These strategies comprise the M in MURDER.) In addition to mood-setting, students were trained to cope
with distractions by using the strategies of relaxation, positive self-talk, and imagery to reestablish an
appropriate learning state.
Monitoring
To be effective, students must be able to detect when their behavior is not sufficient to meet task
demands so that they can make appropriate adjustments. We have not treated monitoring as a separate
component but have embedded monitoring principles in the concentration management component and the
two MURDER strategies. The students were trained to intermittently evaluate their progress and take
corrective actions if necessary.
Research Framework
Due to the complexities of academic learning, a mutually supportive set of interactive strategies is
required to maximize learning potential. In order to examine and capitalize on these interactions, students
must be taught large portions of the strategy system. Unfortunately the time and student motivation required
for training precludes exploring this system in the context of typical short-term experiments.

67
Therefore, the component strategies were put together to form a one-semester (15-week) learning
strategies course. This 2-credit- hour course was offered to Texas Christian University undergraduates
during the 1977 spring semester on a pass/no-credit basis.
 In order to evaluate the effectiveness of the strategy system, two interlocking experiments were created.
In one, the performances of differentially treated sub- groups of the class were compared with each other
and with a no-treatment control group (comprehension/retention controls). The bases of these comparisons
were scores on a series of tests over textbook material that had been studied 7 days earlier. These tests were
given to the class members and the control group prior to the start of the course (precourse test),
approximately halfway through the course (midcourse test), and at the end of the course (postcourse test).
In the second experiment, the performance of the class members on a set of self- report measures was
compared to a separate no-treatment control group (self-report controls) both before and after the course.
The decision to use no-treatment rather than placebo control groups was based on prior research with
learning strategy training. Attempts at equating training time by having students practice their own or less
effective competing methods on the training material have led to suppression of mean performance in
comparison with "untrained" students using their own techniques (e.g., Collins, 1978; Garland, 1977; Long,
Dansereau, McDonald, Collins, & Evans, Note 3). Subjective reports from participants in these placebo
groups indicate that they do not view the training as meaningful and consequently become frustrated and
bored with the task. These reactions apparently carry over to the assessment phase, leading to the reduction
in mean performance. It should also be emphasized that the college-age students participating in these
experiments have had 12-14 years of experience and practice with their own study methods (particularly
with naturally occurring prose) and can therefore be considered no-treatment controls in name only.

Method
Participants
Participants were Texas Christian University undergraduates, heterogeneous with respect to grade level
and majors. Three major groups of these students were employed in this experiment. The learning strategy
class was composed of 38 students. They received 2 semester hours of college credit for successfully
completing this pass/no-credit course.
The comprehension/retention control group consisted of 28 students who were recruited from general
psychology classes. After completing the experiment, they received credit for fulfilling an experimental
participation requirement, a $6 fee, and learning strategy training materials.
Finally, the subjective report control group was composed of 21 students also recruited from general
psychology. Upon completion of the experiment, they received learning strategy training materials,
experimental credit, and a $4 fee.
Interviews with members of the control groups indicated that their prime motivation for participating
was their interest in learning strategy training. Further, a comparison of the profiles of the three groups
indicated that they were very similar in terms of the distributions of majors, grade levels, and sex.
Dependent Measures
Comprehension/retention measures. Multiple- choice and short-answer tests were developed for three
3,000-word passages: one extracted from a textbook on educational psychology (precourse test), one from a
text on ecology (midcourse test), and one from a text on geology (postcourse test). The students were given
1 hour to study each of these passages and then 1 week later were given 45 minutes to take the
corresponding tests.
Self-report measures. These included a 37-item Test Anxiety Scale (a slightly modified version of the
one used by Sarason, 1956), the Survey of Study Habits and Attitudes (Brown & Holtzman, 1966), a 46-
item questionnaire designed to tap concentration difficulties and coping skills, and a 12-item learning
attitude inventory designed to assess students' perceptions of their own academic abilities.
Procedure
For the comprehension/retention experiment the dependent measures were administered to the class
members and the comprehension/retention controls before the course began (precourse test), approximately
halfway through the course (midcourse test), and after the course was completed (postcourse test). For the
final testing the 36 participating class members were subdivided into three groups of 12 each depending on
the comprehension/retention substrategy being employed: paraphrase/imagery, networking, or key ideas.

68
For the self-report experiment the four measures were administered to the class members and the self-
report controls before and after the course.
Class members were given 2 hours training each week for 12 weeks. The two control groups were not
given any training during this time period.
The strategy components described earlier formed the basis of the strategy class training. In general, the
training on primary and support strategies was inter- mixed in order to illuminate the interactions between
the components. Training methods included short lectures, practice exercises, and small group or pair
discussions.
Specific training for the class members, following the precourse test, consisted of 12 hours of training in
goal-setting and scheduling, learning from a test, com- prehension problem solving, and introduction to
MURDER. The class then took the midcourse com- prehension/retention test (n = 25). Following this
measure, class members received the final 12 hours of training. This final strategy training was taught to
three subdivisions of the class: paraphrase/imagery, networking, and key ideas. Each subgroup received not
only the specific recall/transformation strategy but also training on concentration management and retrieval.
Finally, class members participated in the postcourse testing session. Class members were instructed to
apply the recall/transformation strategies they had learned to the material on the comprehension/retention
postcourse test. The students also filled out the self- report measures at this time.

Results and Discussion

Comprehension/Retention
Each of the three comprehension/retention tests (precourse, midcourse, and post- course) contained both
multiple-choice and short-answer subtests. All tests were scored according to predetermined keys without
knowledge of the student's group affiliation. The raw scores were then converted to per- centages of the
maximum possible on each test.
The overall treatment (strategy training) versus control differences will be presented first. The means
and standard deviations of the total scores and changes in total scores for the comprehension/retention tests
are presented in Table 1. (The sample sizes in this and subsequent tables vary depending on the number of
students participating in each testing.) As can be seen in this table there were slight differences in the
precourse test means. This situation also occurred with the other dependent measures. Consequently,

Table 1
Means and Standard Deviations of the Comprehension/Retention Test Scores (Expressed in Terms of
Percent Correct)
 69
all analyses were performed on changes from the precourse test baselines. Since there has been controversy
over the use of change (or difference) scores in the past, a brief justification for this approach is presented.
In a detailed analysis of this issue, Overall and Woodward (1975) concluded the following:
Numerous writers have emphasized the unreliability of difference scores, which results from summation
of measurement errors (Bereiter, 1963; Lord, 1963; Webster & Bereiter, 1963). While this may be a problem
for certain types of correlational studies, it is not a cause for concern in the use of simple difference scores to
measure treatment-induced change in experimental research. (p. 85)
This position was reasserted by Overall and Woodward in 1976 and supported by Huck and McLean
(1975). Further, according to the criteria suggested by Huck and McLean, the analysis of difference scores is
a more conservative procedure than analysis of co- variance given the conditions of the present study.
Analysis of the mean changes in total scores indicated that in comparison with the
comprehension/retention control group, the strategy class members exhibited significantly greater positive
change from the precourse to the midcourse test, t(51) = 1.85, p< .05, and from the precourse to the post-
course test, t(57) = 1.73, p < .05. The negative change scores exhibited by the control group indicate that the
midcourse test was more difficult than the postcourse test, which in turn was more difficult than the
precourse test (see Table 1). This ordering was supported by subjective ratings of comprehensibility elicited
from the subjects at each testing.
The means and standard deviations of the short-answer and multiple-choice scores for the
comprehension/retention tests are also presented in Table 1. To clarify the effects of training, separate
analyses were conducted for each measure. Analyses of the mean changes in short-answer scores indicated
that the class members exhibited a significantly greater positive change from the precourse to the postcourse
test than did the comprehension/retention controls, t(57) = 1.69, p <.05. Although the class members' mean
short-answer change from the precourse to the midcourse test was more positive than that of the controls, the
difference did not reach significance, t(51) = 1.13, p = .13. As with the total scores, the negative changes
exhibited by the controls indicate that the midcourse short-answer subtest was more difficult than the
postcourse test, which was in turn more difficult than the precourse subtest (see Table 1).
Compared with the comprehension/retention controls, the strategy class members exhibited significantly
greater positive multiple-choice changes from the precourse test to the midcourse test, t(51) = 1.78, p < .05.
Class members showed more positive mean multiple-choice change from the pre- course test to the
postcourse test than did the controls, but the difference in change scores did not reach significance, t(57) =
833, p.21. Examination of the control group's change scores indicates that the midcourse and postcourse
multiple-choice subtests were approximately equal in difficulty and both were more difficult than the
precourse subtest (see Table 1).
The amounts (expressed in percentages) by which strategy class members outscored the control group on
each of the comprehension/retention tests are presented in Table 2. (Only participants who took all three
tests are included in the data display.) Examination of this table indicates that the training appeared to have a
greater influence on the short-answer subtest than on the multiple-choice subtest. This result is encouraging,
since short-answer tests are much less likely to be influenced by guessing and differential test-taking
strategies than multiple-choice tests.

Table 2
Percentage Amounts by Which Strategy Class Members Outscored the Control Group on Each of the
Comprehension/Retention Testsa

Note. Class member n = 24; control group n = 23.

a Mean differences between the two groups expressed as per- centages of the control group's mean scores.

70
The data in Table 2 also suggest that the major effects of the strategy training occurred prior to the
midcourse test. (The percentages by which the class members outscored the controls are at or near their
peaks for each of the measures at the time of the midcourse test.) Although this is a tenable hypothesis, other
factors exist that could have influenced these results. First, as has been stated earlier, the midcourse test
appeared to be more difficult than the postcourse test. It is possible that the strategy training has a greater
impact on more difficult material, thus producing the results exhibited in Table 2. Second, during the
training subsequent to the midcourse test and prior to the postcourse test the class members were subdivided
into three groups and each group was trained on a different recall/transformation technique: para-
phrase/imagery, key ideas, and networking. If one or more of these techniques were not as effective (at the
time of testing) as the more global strategy (i.e., MURDER with free recitation) taught prior to the
midcourse test, then the data in Table 2 would not be unexpected. This possibility is explored in the
following paragraphs.
Between the midcourse and postcourse tests the strategy class members were divided into three groups
based on their stated preference for a particular comprehension/ retention substrategy. Each group received
approximately 4 hours of training on one of the following strategies: paraphrase/im- agery, networking, or
key ideas. At the time of postcourse assessment, students were instructed to use the substrategies they had
received. Comparisons between these groups indicated that there were no significant differences between the
change scores from the precourse to postcourse tests (for each group, n = 12). However, it should be noted
that there were substantial differences in the percentage amounts by which these groups outscored the
control group. For example, on the short-answer subtests the networking group performed 1.16% below the
controls on the precourse test and 42.26% above the controls on the postcourse test. For the other two groups
the differential was substantially less (5.39% above on the pre- course and 16.36% above on the postcourse
test for paraphrase/imagery and 11.95% above on the precourse and 29.42% above on the postcourse test for
the key ideas group).
Table 3 provides data for the strategy subgroups on all three tests. Again, the percentage amounts by
which these groups outscored the control group are illustrated. (Only the data from those participants who
took all three tests are included in Table 3; this reduces the ns from 12 to 8 for each group.) Examination of
Table 3 indicates that the networking group increased its ad- vantage over the controls from midcourse to
postcourse, while the Key Ideas group maintained approximately the same ad- vantage, and the
paraphrase/imagery group decreased its advantage.
Table 3
Percentage Amounts by Which the Strategy Subgroups Outscored the Control Group on Each of the
Comprehension/Retention Tests

Note. Each substrategy (paraphrase/imagery, key ideas, networking) occurred between the midcourse test
and the postcourse test.
a Mean differences between the strategy subgroups and the control group expressed as percentages of the
control group's mean
scores.
b For each subgroup, n = 8; for the control group, n = 23.
71
Therefore it is reasonable to speculate that the peak in class-control differences at the midcourse
examination may be due to the lack of impact of the key ideas training and the ap- parent negative impact of
the paraphrase/ imagery training. The only group that seemed to benefit from the substrategy training was
the networking group. Consequently, if all class members were taught networking, then one might expect
that the class-control differences would have been greater on the postcourse test than on the midcourse test.
(It is also possible that the utilization of equally difficult midcourse and postcourse tests would influence the
results in a similar way.)
The reasons for the negative impact of the paraphrase/imagery training on postcourse test performance
are not readily apparent, especially in light of prior work with versions of this technique (e.g., Dansereau et
al., 1975). There are two possible explanations. First, evaluators of paraphrasing and imaging have typically
used shorter materials (generally passages of 1,000 words or less); in this study the passages were
approximately 3,000 words in length. Because these techniques do not emphasize organization of the
material, they may be relatively ineffective with longer passages. Second, students chose which type of
training they would receive. The paraphrase/imagery technique was undoubtedly perceived as the easiest
technique to learn and implement. Consequently, it may have attracted relatively unmotivated students.
Their performance on the postcourse test may be more reflective of motivational deficits than strategy
deficits.
Self-Report Measures
The self-report measures were scored ac- cording to predetermined keys, and a total score was created
for each individual on each test. The precourse, postcourse, and change (from pre- to postcourse
administration) score means and standard deviations for the four self-report measures are presented in Table
4. The different sample sizes reflected in Table 4 are a consequence of the fact that uncompleted self-report
inventories were not scored. Because there were initial mean differences between the two groups of
participants, all statistical analyses were performed on the mean changes from pre- course to postcourse
administrations.
The t tests comparing the precourse- postcourse change scores of class members with those of the
controls reached signifi- cance on three of the measures: Survey of Study Habits and Attitudes, t (39) = 2.57,
p < .01; Test Anxiety Scale, t(45) = 3.57, p < 72 .01; and the learning attitude inventory, t(48) = 3.57, p
< .01

Table 4
Precourse, Postcourse, and Change Scores for the Four Self-Report Measures

a Lower score indicates lower anxiety.

72
The t test approached significance on the concentration questionnaire, t (44) = 1.55, p <.06. In all cases the
class members reported greater positive changes on academically-related dimensions than did the self-report
controls.
 These results may, however, have been confounded by either or both of the following factors. First,
because the class members were generally lower than the controls on the precourse measures, the significant
effects may have been due to regression toward the mean. The power of this type of explanation is
substantially diminished in the present case because previous administrations have shown that these four
measures are generally very reliable. In addition, on the Survey of Study Habits and Attitudes, the class
members scored below controls on the precourse test and above this group on the postcourse test. These
results would not be expected if the only factor operating were regression toward the mean. Another
potential explanation for the results is that the class members may have been yea-saying on the postcourse
measures. The fact that the class members did not show significant changes on some of the items on these
scales reduces the possibility that the group's responses to the postcourse measures were artifactual.
Although the confounding factors cited above cannot be completely discounted, it seems very likely that the
learning strategy training had a positive in- fluence on the academic behavior and attitudes reflected in the
four self-report measures.

Summary and Conclusions

The comprehension/retention and self- report results, coupled with the positive feedback arising from the
students' course evaluations, indicate that the strategy system improved the students' learning behaviors and
attitudes. The learning strategy class members showed significantly greater positive precourse-postcourse
changes on the comprehension/retention tests than did the no-treatment controls. Further, the training
appeared to have a greater impact on the short-answer portions of the tests than on the multiple-choice
portions.
Examination of the performance of differentially treated subgroups of the class indicated that the
networking recall/trans- formation strategy produced somewhat greater positive changes from the midcourse
test to the postcourse test than did the paraphrase/imagery and key ideas strategies. One possible explanation
is that the organizational aspects of networking (i.e., the production of two-dimensional maps) were
responsible for the mean differences in performance.
In addition to the formal assessments, participants rated the perceived value of each strategy component
included in the program. Although all components were rated positively, networking, concentration
management, and the MURDER executive routine received the highest ratings. In subsequent studies these
components will be evaluated independently, modified, and then recombined.

Reference Notes
1. Diekhoff, G. M. The node acquisition and integration technique: A node-link based teaching/ learning
strategy. Paper presented at the annual meeting of the American Educational Research As- sociation,
New York, April 1977.
2. Meichenbaum, D., & Turk, D. The cognitive-behavioral management of anxiety, anger, and pain. Paper
presented at the Seventh Banff International Conference on Behavioral Modification, Banff, Canada,
1975.
3. Long, G. L., Dansereau, D. F., McDonald, B. A., Collins, K., & Evans, S. H. The development, as-
sessment, and modification of an effective learning strategy training program. Paper presented at the
annual meeting of the American Educational Re- search Association, San Francisco, April 1976.

References
Anderson, J. R., & Bower, G. H. Human associative memory. Washington, D.C.: Winston, 1973. Bereiter,
C. Some persisting dilemmas in measurement of change. In C. W. Harris (Ed.), Problems in mea- suring
change. Madison: University of Wisconsin Press, 1963.
Bobrow, D. G., & Winograd, T. An overview of KRL, a knowledge representation language. Journal of
Cognitive Science, 1977, 1, 3-46.
Bodden, J. L., Osterhouse, R., & Gelso, C.J. The value of a study skills inventory for feedback and criterion
purposes in an educational skills course. Journal of Educational Research, 1972. 65, 309–311.Survey of
Study New York: The

73
Briggs, R. D., Tosi, D. J., & Morley, R. M. Study habit modification and its effect on academic
performance: A behavioral approach. Journal of Educational Research, 1971, 64, 347-350. Brown, W.
F., & Holtzman, W. H. Habits and Attitudes, Form C. Psychological Corporation, 1966. Brown, W. F.,
Wehe, N. O., Zunker, V. G., & Haslam, W. L. Effectiveness of student-to-student counseling on the
academic adjustment of potential college dropouts. Journal of Educational Psychology, 1971, 62, 285-
289.
Collins, K. W. Control of affective responses during academic tasks. Unpublished master's thesis, Texas
Christian University, 1978.
Dansereau, D. F. The development of a learning strategy curriculum. In H. O'Neil (Ed.), Learning strategies.
New York: Academic Press, 1978. Dansereau, D. F., Actkinson, T. R., Long, G. L., & McDonald, B. A.
Learning strategies: A review and synthesis of the current literature AFHRL-TR- 74-70, Contract
F41609-74-V-0013). Lowry Air
Force Base, Colorado, 1974. (NTIS No. AD-A007 722)
Dansereau, D. F., Long, G. L., McDonald, B. A., Act- kinson, T. R., Ellis, A. M., Collins, K. W., Williams,
S., & Evans, S. H. Effective learning strategy training program: Development and assessment (AFHRL-
TR-75-41, Contract F41609-74 C-0013). Lowry Air Force Base, Colorado, 1975. (NTIS No. AD-A014-
722)
Ellis, A. Reason and emotion in psychotherapy. New York: Lyle Stuart, 1963.
Garland, J. C. The development and assessment of an imagery based learning strategy program to improve
the retention of prose material. Unpublished master's thesis, Texas Christian University, 1977. Haslam,
W. L., & Brown, W. F. Effectiveness of study-skills instruction for high school sophomores. Journal of
Educational Psychology, 1968, 59, 223-226.
Huck, S. W., & McLean, R. A. Using a repeated mea- sures ANOVA to analyze the data from a pretest-
posttest design: A potentially confusing task.
73
Psychological Bulletin, 1975, 82, 511-518. Jacobsen, E. Progessive relaxation. Chicago: Uni- versity of
Chicago Press, 1938.
Lord, F. M. Elementary models for measuring change. In C. W. Harris (Ed.), Problems in measuring change.
Madison: University of Wisconsin Press, 1963. Maultsby, M. Handbook of rational self-counseling.
Madison, Wisc.: Association for Rational Thinking, 1971.
Meichenbaum, D. H., & Goodman, J. Training im- pulsive children to talk to themselves: A means of self-
control. Journal of Abnormal Psychology, 1971, 77, 115-126.
Overall, J. E., & Woodward, J. A. Unreliability of dif- ference scores: A paradox for measurement of
change. Psychological Bulletin, 1975, 82, 85-86. Overall, J. E., & Woodward, J. A. Reassertion of the
paradoxical power of tests of significance based on unreliable difference scores. Psychological Bulletin,
1976, 83, 776-777.
Robinson, F. P. Effective study. New York: Harper, 1946.
Rumelhart, D. E., Lindsay, P. H., & Norman, D. A. A process model for long-term memory. In E. Tulving &
W. Donaldson (Eds.), Organization of memory. New York: Academic Press, 1972.
Sarason, I. Effect of anxiety, motivational instructions, and failure on serial learning. Journal of Experi-
mental Psychology, 1956, 51, 253–260.
Van Zoost, B. L., & Jackson, B. T. Effects of self- monitoring and self-administered reinforcement on study
behaviors. Journal of Educational Research, 1974, 67, 216-218.
Webster, H., & Bereiter, C. The reliability of changes measured by mental test scores. In C. W. Harris (Ed.),
Problems in measuring change. Madison: University of Wisconsin Press, 1963.
Whitehill, R. P. The development of effective learning skills programs. Journal of Educational Research,
1972, 65, 281-285.
Wolpe, J. The practice of behavioral therapy. New York: Pergamon, 1969.
Received April 10, 1978 ■