development self-report inventory for assessing learning · 2016. 5. 18. · 413 development of a...

413

Development of a Self-Report Inventory for AssessingIndividual Differences in Learning ProcessesRonald Ray Schmeck, Fred Ribich, and Nerella RamanaiahSouthern Illinois University

Five studies are presented—all related to the de-velopment and application of a self-report inventoryfor measuring individual differences in learningprocesses. Factor analysis of items derived by trans-lating laboratory learning processes into the contextof academic study yielded four scales: Synthesis-Analysis, Study Methods, Fact Retention, and Elab-orative Processing. There were no sex differences,and the scales demonstrated acceptable reliabilities.The Synthesis-Analysis and Elaborative Processingscales both assess aspects of information processing(including depth of processing), but Synthesis-Analysis assesses organizational processes, whileElaborative Processing deals with active, elaborativeapproaches to encoding. These two scales werepositively related to performance under incidentallearning instructions in both a lecture-learning andtraditional verbal-learning study. Study Methodsassessed adherence to systematic, traditional studytechniques. This scale was positively related to per-formance in the intentional condition of the verbal

learning study. The Fact Retention scale assessedthe propensity to retain detailed, factual informa-tion. It was positively related to performance in theincidental condition of the verbal-learning but notthe lecture-learning study. Future research and ap-plications are discussed.

Recently, several researchers in the area ofhuman learning and memory have emphasizedthe importance of processes such as encoding

(Melton & Martin, 1972), organization (Tulving& Donaldson, 1972), imagery (Paivio, 1971 ), anddepth of processing (Craik & Tulving, 1975).The present paper describes the development ofan instrument for measuring individual dif-

ferences in such learning processes.Since Cronbach’s (1957) plea for research

combining correlational and experimental ap-proaches, there has been a growing interest inthe study of individual differences in learning(e.g., Gagn6, 1967). Much of this research hasbeen done in the applied areas of teaching andcounseling. With the exception of a few studies(e.g., Biggs, 1970; Goldman & Warren, 1973),many of the applied studies have mainly em-ployed personality, attitudinal, and cognitivestyle measures (e.g., Cowell & Entwistle, 1971;Cropley & Field, 1969). After reviewing the re-search in this area, Tallmadge and Shearer

(1969, 1971) have concluded that learning styleassessed from a behavioral-process orientation ismore likely to be useful than one based purelyon a personality- or cognitive-style orientation.

In laboratory research, experimenters havefrequently employed performance measures ofindividual differences. For example, in order tostudy individual differences in organizationalprocesses, Earhard (1974) and Thompson, Ham-lin, and Roenker (1972) essentially ran subjectsonce to establish individual differences and then

ran them again in the same or similar laboratory

Downloaded from the Digital Conservancy at the University of Minnesota, http://purl.umn.edu/93227. May be reproduced with no cost by students and faculty for academic use. Non-academic reproduction

requires payment of royalties through the Copyright Clearance Center, http://www.copyright.com/

414

situation to determine the impact of the indi-vidual differences on further learning. Unfor-tunately, such measures are time-consumingand laborious, and they may be subject to the in-fluence of common method variance, since themeasure of individual differences and the ex-

perimental measures are based on the samemethod of measurement.

Following the recommendation of Tallmadgeand Shearer (1969, 1971), the present authorsdeveloped a self-report inventory usingbehaviorally oriented statements to assess im-portant learning processes in the academic set-ting. The primary goal in the preparation of theitem pool was to represent the findings andtheoretical processes of information processing,especially as they have been applied to the ver-bal and prose learning areas. Essentially, theauthors consulted advanced information-proc-essing and verbal- and prose-learning texts inorder to list the major processes which have beenuncovered by research or advocated by varioustheoretical points of view. Each of the authorsthen attempted to write items to assess theseprocesses by phrasing behavioral descriptions interms of the environment and activities of the

typical college student. In addition, the authorsprepared a variety of items which were related toacademic activities but which had no obvious

preconceived relationship to any established

learning processes. These items were included inthe hope that some of them would cluster withthe information-processing items or form

separate clusters revealing unanticipated learn-ing processes. These latter items were concernedwith behaviors which students engage in while

attending lectures and films, taking notes, writ-ing papers, making plans, interacting with otherstudents, and generally preparing for academicexaminations.

In order to illustrate how the item pool wasdeveloped, it will be useful to consider a few ofthe well-known processes which served as start-

ing points and present some sample items sug-gested by these processes. Basic to the informa-tion processing area are the concepts of encod-ing, storage, and retrieval. The concept of en-

coding has been studied extensively in the areaof verbal learning and memory (e.g., Melton &

Martin, 1972). Encoding refers to the processwhereby the learner transforms new informationinto a form which can more readily be related tothe old information which is already stored inmemory. Relevant inventory items may take theform: &dquo;I learn new words and ideas by associat-ing them with words and ideas which I alreadyknow&dquo;; or &dquo;I learn new ideas by relating them tosimilar ideas.&dquo; Storage refers to the holding ofinformation in memory and could be measured

by inventory items such as: &dquo;I memorizematerial as given in the text&dquo;; &dquo;my memory isactually pretty poor&dquo;; and &dquo;I have trouble

remembering definitions.&dquo; Retrieval refers to

the process whereby information is broughtforth from memory, and Eysenck (1974) pro-vides evidence of individual differences in infor-mation retrieval. Inventory items related to re-trieval might take the form, &dquo;When I studysomething, I devise a system for recalling it la-ter.&dquo;Another information-processing variable,

which has been stressed in recent verbal-learn-

ing and memory studies, concerns the changesin the organization of information which occurbetween the time it is originally presented to asubject and the time that he/she is asked to re-call it (Tulvig & Donaldson, 1972). Such organi-zational processes clearly improve recall. In ver-bal learning the terms clustering (e.g, Bousfield& Bousfield, 1966) and subjective organization(e.g., Tulving, 1962) refer to systematic changesin word order which occur between the timewhen the subject studies a list of words and thetime when his/her recall is tested. In this con-

text, Earhard (1974) and Thompson, Hamlin,and Roenker (1972) reported that there are

individual differences in the tendency to or-

ganize word lists and that such differences arerelated to amount and content of recall. Similar

organizational processes have been studied inrelation to textbook and lecture learning (e.g.,Frase, 1969; Shultz & DiVesta, 1972). Inventoryitems concerned with organizational processesmight take the form, &dquo;I have trouble organizing



415

the information that I remember&dquo; or &dquo;I general-ly prepare a set of notes integrating the informa-tion from all sources in a course.&dquo;

Another aspect of information processingmight be referred to as elaborative processing.There are several ways in which an individualcan elaborate upon given information. One in-volves the use of imagery, i.e., visually imagingsituations which exemplify the given informa-tion. Paivio (1971) reports substantial improve-ment in memory produced by the use of ima-gery. He has also reported some successful at-tempts to measure individual differences in theuse of imagery. This process suggests inventoryitems such as: &dquo;I learn new words and ideas byvisualizing a situation in which they occur.&dquo;

Elaborative methods of processing informationalso involve the depth to which information is

processed. Craik and Tulving (1975) suggest thatinformation processing can vary along a con-tinuum from &dquo;shallow,&dquo; involving attention tosuperficial aspects of the material, to &dquo;deep,&dquo; inwhich the learner delves into the subtle nuances,

meaning, and personal relevance of new infor-mation by thinking about it. In the verbal-learn-ing area, it has been found that &dquo;deeper&dquo; proc-essing improves recall (Craik & Tulving, 1975).Similar findings have been obtained in regard totextbook learning(e.g., Anderson and Kulhavy,1972). Inventory items related to shallow proc-essing might take the form, &dquo;I often memorizematerial that I don’t understand.&dquo; At a deeperlevel, we might have: &dquo;I learn new ideas by ex-pressing them in my own words&dquo; or &dquo;While

learning new concepts, their practical applica-tions often come to mind.&dquo;

This paper presents five separate studies con-cerned with the development of a self-report in-ventory of learning processes. The first studyreports the actual development of scales assess-ing various learning processes. Since some of theitems in the scales developed in the first studyhad to be reversed in order to balance the num-

ber of true-keyed and false-keyed items, thesecond study tested the effect of item reversalson the stability of the item responses. The nextstudy was concerned with sex differences, inter-

correlations, internal consistency, and test-retestreliability of the various scales. The last twostudies were designed to provide preliminary in-dications of the validity of the scales. One studyexamined the relationships between the scalesand performance on an unannounced objectiveexamination in a lecture-learning setting. Thefinal study examined similar relationships usinga standard laboratory verbal learning paradigm.In addition, this study attempted to assess therelative independence of the learning processscales as compared with selected measures ofpersonality and cognitive style.

SCALE DEVELOPMENT

Study 1

The aim of the first study was to identify clus-ters of items which could be used to developscales with adequate reliability. Responses to theinitial item pool were factor analyzed and, asrecommended by Berdie and Campbell (1968)and Wang and Stanley (1970), scales and scoringkeys were constructed by giving unit weight tothe salient items.

Method

Subjects. The subjects were 503 undergrad-uate students at Southern Illinois University.The sample was composed of 237 males and 266females, with 87 freshmen, 199 sophomores, 142juniors, and 75 seniors representing various

academic majors. Participation in the study wasvoluntary. -

Procedure. The preliminary form of the in-ventory consisted of 121 true-false items describ-

ing learning behaviors which might be utilizedwithin an academic setting. The students wereinstructed to respond to the statements (onstandard IBM answer sheets) by considering themanner in which they &dquo;generally&dquo; learn ratherthan how they learn within any particular courseor academic discipline.

In general, the analysis followed the guide-lines recently suggested for questionnaire de-velopment by Cattell (1974) and Vaughan



416



417



418

(1973). Intercorrelations among 121 items basedon the responses of the 503 students were first

subjected to a principal components solution,and a scree test (Cattell, 1966) was used to deter-mine the number of factors. The correlation ma-trix was then factor analyzed using the principalfactor method with squared multiple correla-tions as estimates of communalities, and the

previously determined number of factors wererotated to the Varimax criterion (Kaiser, 1958).Since individual scale reliabilities depend uponthe number of items within a scale, only thosefactors were retained which had five or moreitems with loadings exceeding .25 and havingminimal overlap with other factors.

Results and Discussion

Although the scree test yielded 15 factors, in-spection of the rotated factor matrix indicatedthat only four of them met the criteria for reten-tion. The four factors encompassed 62 of the ori-ginal 121 items. Table 1 presents the results ofthis analysis, giving the items which compriseeach factor and their loading on that factor.Factor I is marked by items which stress

evaluation, organization, discrimination, and

extrapolation and is referred to as the Svnthesis-Ana~vsis factor. Factor II represents the ubiqui-tous study methods dimension of learningindicating the use of systematic, traditional

study techniques. It will henceforth be labeledthe Studv Methods factor. Factor III, which is

called Fact Retention, has substantial loadingsfor items indicating a preference for factual in-formation and retention of details. Factor IV is

comprised of items which stress visualizing,summarizing, relating, encoding, and applyinginformation; it is called the Elaborative Process-ing factor.

Study 2

An inspection of the scoring keys for each ofthe four scales revealed that within each scale

there was an imbalance of true-keyed and false-keyed items. Thus, in order to reduce the in-fluence of response set, several items were

selected within each scale which could be con-

veniently reworded in the reverse direction. Atotal of 17 items was selected in this fashion, and

Study 2 was designed to determine the effect ofitem reversal on students’ responses to the in-

ventory by examining item-stability coefficients(Rorer & Goldberg, 1965).

Method

The effect of item reversals was studied bycomparing the stability of the responses of stu-dents who first responded to the original itemand then responded to the reversed item with thestability of the responses of students who twiceresponded to the original item. An experimentalgroup of students (N = 95) completed the orig-inal version of the inventory. This was followedtwo weeks later by the form containing the 17 re-versed items. A control group (N = 120) filledout the original version of the inventory twiceover the same time period. The reversed itemsare presented in Table 2.

Following Rorer and Goldberg (1965), stabili-ty coefficients were computed for the items thatwere reversed and those that were not reversed.

Stability coefficients indicate the percentage ofsubjects who respond consistently to a particularitem on two different occasions. In the case of

the control group, stability was defined as thenumber of subjects who responded &dquo;true&dquo; on

both the first and second administration plusthe number of responding &dquo;false&dquo; on both occa-

sions, divided by the sample size (N = 120). Inthe case of the experimental group, stability fornonreversed items was computed in the samemanner as the control group. But, the stabilityof reversed items for this group was defined as

the number of subjects who respond &dquo;true&dquo; to

the given item on the first occasion and &dquo;false&dquo;

on the second, plus the number who respond&dquo;false&dquo; on the first and &dquo;true&dquo; on the second,divided by the sample size (N = 95). The controlgroup’s mean item-stability coefficient for re-

versed items was compared to that of the experi-mental group. A similar comparison was madefor nonreversed items.



419

Table 2. Reversed wording of items incorporated in the final formof the ILP.


For the experimental group which respondedfirst to the original item and then to its reversedform, the mean and standard deviation of thestability coefficients for the 17 reversed itemswere 72.6 and 6.4 (values ranging from 63.2 to82.9), respectively. The mean and standard de-viation of the stability coefficients for the controlgroup which twice responded to the 17 items intheir original form were 78.1 and 4.6 (valuesranging from 69.2 to 84.1). Since there was nosignificant difference between the mean stability

coefficients of the experimental and control

groups, it was concluded that the reversal of the17 items had not altered their original meaning.It may also be noted that the mean and standarddeviation of the stability coefficients for the non-reversed items were 78.9 and 5.9 for the control

group and 79.2 and 6.4 for the experimentalgroup. The 62-item inventory administered tothe experimental group, including the 17 re-worded items shown in Table 2, is hereinafter re-ferred to as the Inventory of Learning Processes(ILP).



420

Table 3. Means and standard deviations of the four ILP scales for

male and female samples.

Study 3

The third study was aimed at investigating sexdifferences, internal consistency and test-retestreliabilities, as well as scale intercorrelations forthe final form of the ILP. The first step involvedthe examination of sex differences in terms ofindividual scales and average profiles. Next, theintercorrelations of the four scales were deter-mined. The subsequent determination of relia-bilities involved the calculation of both internal

consistency and temporal stability. In order todetermine temporal stability, a subsample ofstudents was retested following a two-week in-terval.

Method

The final form of the ILP was administered to434 volunteer undergraduate students (232males and 202 females) from a number of dif-ferent academic majors, including freshmenthrough seniors. For estimating test-retest relia-bility, a subsample of 100 students was asked toreturn two weeks later for another testing ses-sion ; 95 of the 100 returned for this session inwhich the ILP was re-administered. Multivariate

analysis of variance was used for testing profiledifferences between males and females, whileunivariate F-tests were used at the individualscale level.


Table 3 presents the means and standard de-viations of males and females on the four ILP

scales. Multivariate analysis of variance revealedno significant sex differences for the averageprofiles. Univariate analysis indicated no signifi-cant differences at the individual scale level.

Hence, the data for males and females werecombined for subsequent analyses. Scale inter-correlations are presented in Table 4, where itcan be seen that the scales are relatively inde-pendent. Internal consistencies (KR -21 ) are alsoreported on the diagonal in Table 4. The test-re-test reliabilities are also presented in the lastcolumn of Table 4. These internal-consistencyand test-retest reliability values are in the

acceptable range, and the remaining studies re-ported below are concerned with the applicationof the ILP within the research setting.

EXPERIMENTAL VALIDATION STUDIES

Study 4

This study investigated the relationship be-tween ILP scales and performance in a lecture-learning setting, using a single videotaped lec-ture followed by an objective examination. SincePlenderleith and Postman (1957) have suggestedthat individual differences emerge more stronglyin incidental (unintentional) learning, the exa-



421

Table 4. Intercorrelations, internal consistency, and test-retestreliabilities of the four ILP scales.

Note: Intercorrelations and internal consistencies were based on a

sample of 434 subjects, and test-retest reliabilities werebased on a sample of 95. Internal consistencies are listedon the diagonal. Decimal points have been omitted.

mination in the present study was unannounced.In addition, in order to investigate the degree towhich the Synthesis-Analysis and ElaborativeProcessing scales of the ILP deal with high-levelcognitive skills, half of the examination was

composed of comprehension (high-level) ques-tions and half was composed of knowledge (low-level) questions written according to the guide-lines of Bloom’s (1956) taxonomy.

Since the incidental learning situation pre-cludes study behavior, it was hypothesized thatthere would be no relationship between the

Study Methods scale and performance on theexamination. On the other hand, if the Synthe-sis-Analysis and Elaborative Processing scalesassess the &dquo;depth&dquo; to which subjects habituallyprocess information, then according to Craikand Tulving’s (1975) &dquo;levels of processing&dquo; no-tion, scores on these scales should be related toexamination performance. Furthermore, if syn-thesis-analysis and elaborative processing re-

quire high-level cognitive skills, then individualswho score high on these two scales should alsobe more proficient at answering high-level cog-nitive examination questions. Thus, it was hy-

pothesized that the Synthesis-Analysis andElaborative Processing scales would be positive-ly related to both the comprehension and knowl-edge examination subscores, but the correlationwould be greater with the comprehension sub-score than with the knowledge subscore.

Method

The subjects were 32 undergraduate volun-teers from an introductory psychology course.Subjects were run as a group and observed a 20-minute videotaped lecture on the principles ofinstrumental conditioning under instructions tolist the concepts covered by the lecturer. To pro-vide a rationale for the task, subjects were toldthat the tape was being screened before beingaired to the public, in order to be certain that itdid not contain too much information. The

videotaped presentation was in color and con-tained numerous diagrams and illustrations ofthe principles covered. After watching the video-tape, subjects were administered an unan-

nounced 30-item, multiple-choice test composedof 15 items written according to Bloom’s (1956)



422

guidelines to assess low-level cognitive activity(knowledge) and 15 items written to assess

high-level cognitive activity (comprehension).Scoring of the objective test yielded a total scoreand two subscores---one of knowledge and onefor comprehension questions. Following the test,all subjects were administered the ILP.


The four learning process scales were corre-lated with the three objective test scores and theresults are presented in Table 5. As expected,there was no significant correlation between theStudy Methods scale and examination perfor-mance. Also, the results indicate that the totalobjective test score was significantly related toboth the Synthesis-Analysis and the ElaborativeProcessing scales. This supports the hypothesisthat individuals who score high on these twoscales habitually process information in depthand thus, in accordance with Craik and Tul-

ving’s (1975) notion, retain more of the informa-tion, even when they are not advised of an up-coming examination.The means and standard deviations for the

two subtests were 5.0 and 2.0 for the knowledge

items and 4.3 and 1.7 for the comprehensionitems. Although the correlations between theSynthesis-Analysis and Elaborative Processingscales and the comprehension subscores are

slightly higher than those for the knowledge sub-score (as seen in Table 5), the difference betweenthe correlation coefficients is not significant. Itshould be noted, however, that the correlationbetween the comprehension and knowledge sub-scores themselves was .46. Thus, the absence ofa significant differential correlation with the twoILP scales may be accounted for by the methodvariance shared by the questions entering intothe two subscores as well as the fact that knowl-

edge acquisition is a prerequisite to answeringcomprehension questions. Further research em-ploying independent assessments of knowledgeand comprehension levels of learning will benecessary to establish whether the Svnthesis-Ana(vsis and Elaborative Processing scales doindeed assess high-level cognitive activities,which bring about more than knowledgeacquisition.

Study 5The previous study indicated significant rela-

tionships between the Synthesis-Analvsis and

Table 5. Correlations of the four ILP scales with scores on an

objective examination given under incidental learning--

conditions (n = 32).

Note: Decimal points have been omitted.*p <.05

**p <.01



423

Elaborative Processing scales and the test per-formance within an incidental learning situa-tion. The present study investigated learningunder both incidental and intentional learninginstructions, in order to look for differential re-

lationships between the ILP scales and learningunder the two conditions. In addition, the learn-ing task employed in this last study was moretypical of laboratory research, involving the

presentation of a list of words followed by reten-tion tests.

Since subjects under conditions of intentionallearning are specifically instructed to study andretain the words, a positive relationship betweenretention and the ILP scales of study methodsand fact retention was expected. However, sinceinstructions to learn might cause study habitsand memorization styles to take precedence overthe more subtle information processing strate-gies of synthesis-analysis and elaborative proc-essing, the authors anticipated that the two lat-ter ILP scales would not be related to retentionin the intentional condition. This analysis is sup-ported by Craik and Tulving (1975), who reportthat incidental learning subjects who are in-

structed to engage in activities which requirethat information be processed in depth oftenperform as well as, or better than, subjects in in-tentional conditions who are simply instructedto learn. This is taken to be an indication thatthe study strategies aroused by instructions tolearn in the intentional condition may actuallybe less effective than the deep processing pro-duced by appropriate instructions in the

incidental condition.As in the previous study, a positive relation-

ship between word-list retention and the Svn-thesis-Ana(vsis and Elaborative Processingscales should be evident in the incidental condi-

tion if these scales do, in fact, assess deep proc-essing strategies which are not specific to thelearning task. The Studv Methods and Fact Re-tention scales should be unrelated to learning inthe incidental condition, since studying and ac-tive attempts at retention are not required.

Recently, researchers (e.g., Paivio & Csapo,1973) have shown that memory for imageable

material (e.g., pictures, concrete words) is betterthan that for nonimageable material (e.g., ab-stract words). In order to investigate whetherthere are differential relationships betweenindividual difference measures and retention ofthese two types of material, the word list used inthe present study was comprised of half concreteand half abstract nouns. It was expected that thetwo scales which relate to the organization andencoding of information (i.e., $vnthesis-Ana(vsisand Elaborative Processing) would bear a

stronger relationship to retention of concretewords as opposed to abstract words.

Traditionally, the two measures of retentionof verbal-learning research (i.e, recall and recog-nition) are assumed to reflect different theore-tical processes. Both measures are assumed toreflect storage processes, but recall measures areassumed to reflect search and retrieval processesas well. Consequently, the present study in-cluded both measures in order to provide an op-portunity to observe differential relationshipswith the ILP scales. For example, if the ILPscales demonstrate relationships with the recallmeasures while demonstrating no relationshipswith the recognition measures, it could be takenas an indication that the scales are mainly tap-ping search and retrieval processes.

In order to investigate whether the ILP scalesare relatively independent of personality andcognitive-style measures, one personality andtwo cognitive-style measures were also adminis-tered in the present study. Recent researchwhich has demonstrated superior learning by ex-troverted subjects (Eysenck, 1976) served as therationale for including the Pittsburg Social Ex-traversion-Introversion Scale (Bendig, 1962). Arecent proposal by Schwartz (1975) that field-in-dependent subjects rely more heavily upon inter-nal organizing schema prompted the inclusionof the Hidden Figures Test (French, Ekstrom, &

Price, 1963), with the expectation that it maypositively correlate with the ILP Svhthesis-Analvsis and Elaborative Processing scales. As asecond cognitive style measure, the CategoryWidth Scale (Pettigrew, 1958) was also included,since it has been suggested (Schwartz, 1975) that



424

narrow categories focus on details and mighttherefore score high on the ILP Fact Retentionscale.

Method

Materials. Social Extraversion-IntroversionScale (SEIS): This scale, developed by Bendig(1962), consists of 30 true-false, self-reportitems; high scores indicate social extraversion.Bendig reports high internal consistency; no es-timates of test-retest reliability were available.Hidden Figures Test (HFT~: The HFT was

the standard version supplied by the Educa-tional Testing Service, (French, Ekstrom &

Price, 1963). It consists of 32 items-16 complexdesigns presented in each of two 10-minute

(timed) halves. In each of the complex patterns,subjects are to locate one of five possible geo-metric designs. The test was administered usingthe standard procedure. The individual scoreswere comprised of the number of figures cor-rectly identified minus one-fourth the numbermisidentified (to correct for guessing). A highscore is indicative of field independence. Inter-nal consistency and test-retest reliabilities are

reported to be in the acceptable range (Boersma,1968).Category Width Scale (CWS): This scale

consists of 40 multiple-choice items which asksubjects to estimate the extremes of a number ofdiverse categories, from the length of whales toannual rainfall in Washington. The response al-ternatives are weighted according to categorybreadth such that high scores indicate broadcategory width. High scorers have been charac-terized as having broad equivalence ranges andbeing risk-takers (Pettigrew, 1958).The word list for the verbal-learning task con-

sisted of 15 concrete and 15 abstract nouns

chosen from those published by Paivio, Yuille,and Madigan (1968). These 30 words were tape-recorded in random order at the rate of one

word spoken every 3.5 seconds. A recognitiontest for the word list was constructed by listingon a sheet of paper, in random order, the 30

original nouns and 30 new distractornouns-the latter being another set of 15 ab-stract and 15 concrete words. The ILP was alsoadministered (in the form described elsewhere inthe present paper).

Subjects and procedure. The subjects were62 undergraduate students enrolled in an intro-ductory psychology course at Southern IllinoisUniversity; they received additional course

credit for participation. The study was con-ducted in two sessions. In the first, all subjectswere administered the ILP, SEIS, CWS, andHFT in that order. During the second session,one week later, subjects were randomly assignedto either an intentional or incidental learningcondition, depending upon the instruction sheetthey received. Due to a clerical error, 32 subjectswere assigned to the incidental group while 30were in the intentional condition.The incidental orienting task employed in the

present study was typical of those used by Craikand Tulving (1975) to bring about deep levels ofprocessing. Subjects in this condition were in-structed to rate the connotative meaning of eachword presented on the dimension strong-weak.A rating sheet containing 30 seven-point ratingscales with the word strong typed at one end andweak at the other was provided. In the instruc-tions, the word &dquo;daffodil&dquo; was given as an exam-ple of a word which would likely be rated at theweak end of the scale, while &dquo;horse&dquo; was givenas an example of a word connoting strong mean-ing. For subjects in the incidental condition, nomention was made of the subsequent memorytests for the list of words. Subjects in the inten-tional condition were instructed to learn the listof words in preparation for a subsequent condi-tion.

Following a single presentation of the wordlist, all subjects were given a blank sheet of pa-per and instructed to write down, in any order,all the words they could remember from the pre-sented list. This free-recall test was scored for

the number of concrete, abstract, and totalwords correctly recalled. The recognition test

immediately followed free recall. Subjects were



425

Table 6. Means and standard deviations of word list recall and

recognition scores for incidental and intentional groups.

instructed to circle those words which they rec-

ognized as having been presented in the wordlist. The number of concrete and abstract words

correctly identified (hits) was separately re-

corded ; a total corrected recognition score wasobtained by subtracting the number of &dquo;false

alarms&dquo; (circled distractors) from the total num-ber of hits.


The means and standard deviations of the

various word-list retention measures for the two

learning conditions are presented in Table 6. In

regard to the recall measures, unweighted-means analysis of variance indicated no signifi-cant effects of word type (concrete vs. abstract),condition (intentional vs. incidental) or their in-teraction. However, analysis of recognition hitsyielded a significant main effect for word type (F(1, 60) = 42.8, p < .001), indicating a higher hitrate for concrete words and for learning condi-tion (F(1, 60) = 19.8, p < .001), with the inci-dental group outperforming the intentional sub-jects. The interaction was not significant. Theincidental group also had a significantly highertotal corrected recognition score (t (60) = 4.5 p <

.001) with the overall false-alarm rates not dif-fering for the two groups (mean = 5.2 forincidental and 5.6 for intentional). Thus, withrespect to the effects of imageability and level ofprocessing, results of the present study are inagreement with previous research.While a comparison of group means on the

individual difference measures indicated no sig-nificant differences between the incidental andintentional conditions, the pattern of correla-tions with the retention scores, presented in Ta-ble 7, did differ for the two groups. As can beseen, the predicted differential relationships be-tween intentional and incidental learning andthe ILP scales emerged. Specifically, the StudyMethods scale obtained significant correlationswith recall only in the intentional condition,while in the incidental condition the Synthesis-Analysis and Elaborative Processing scales werethe only scales significantly correlated with re-call. None of the ILP scales correlated with the

recognition scores, with the exception of one sig-nificant relationship between Synthesis-Analysisand the total corrected recognition score in theincidental condition (r = .44). This latter rela-tionship is due to the fact that there was a sig-



427

nificant negative correlation between Synthesis-Analysis and false-alarm making (r = - .35).Thus, high scorers on the Synthesis-Analysisscale were not necessarily better in recognizingcorrect words, but they were better in their abili-ty to reject incorrect words. The overall tendencyfor the ILP to relate to recall measures but not

recognition measures seems to indicate that theILP scales are tapping search and retrieval proc-esses more than storage processes. Note also thatabstract word retention is not related to any ofthe scales, while concrete word measures doshow some relationships. This supports the no-tion that concrete words are more amenable tothe encoding and processing strategies utilizedby individuals in such a learning task (Paivio,1976) and measured by the scales of the ILP.

Contrary to expectation, the Fact Retentionscale showed a stronger relationship to recall inthe incidental condition than in the intentional.This outcome suggests that the Fact Retentionscale might not be so much a measure of the ten-dency to actively memorize facts as it is a meas-ure of general memory for details. Of the threeadditional scales employed, only the HFT ob-tained a significant correlation with any of theretention measures. We can think of no reason-

able explanation for the HFT-concrete hits cor-relation in the incidental condition and the re-

sult is likely to be spurious.The correlations of the ILP scales with the

SEIS, CWS, and HFT are also given in Table 7,as are the intercorrelations of the latter three

scales. The independence of the ILP as a meas-ure of learning processes is evident. The three

additional measures are also quite unrelated toeach other. It is interesting to note that the cor-relation between Fact Retention and CategoryWidth is in the expected direction, with narrowcategorizers indicating better memory for de-tails.

In general, then, the pattern of correlationsbetween the ILP scales and the verbal-learningmeasures supports the notion that instructionsto learn elicit well-practiced study strategieswhich restrain the learner from engaging in in-

formation-processing strategies and which leadto the same quality or better performance whenthere is no pressure to learn (i.e., in incidentallearning). Such results provide valuable infor-mation both about the nature of the processesassessed by the ILP as well as their function inthe acquisition of knowledge.

SUMMARY AND CONCLUSIONS

Based on recent laboratory research in humanlearning and memory, an instrument was de-veloped to assess individual differences in learn-ing processes within the academic setting, utiliz-ing behaviorally oriented statements. Using thefactor analytic approach, the following fourscales were obtained: Synthesis-Analysis, StudyMethods, Fact Retention, and Elaborative Proc-essing. These scales have acceptable levels of in-ternal consistency and test-retest reliability.There were no sex differences on any of thescales. While future research will necessarily in-volve further explication of the meaning andutility of each learning-process scale, some basicinterpretations are suggested by the availabledata.The Svnthesis-Analysis scale appears to assess

the ability of students to glean organizationfrom a unit of material as well as the ability toreorganize it. In addition to its relation to the or-ganizational processes typically studied in theverbal- and prose-learning areas, the scale isalso related to Bloom’s (1956) three highest cate-gories of cognitive activity (i.e., synthesis, analy-sis and evaluation). Although the label Syn-thesis-Analysis emphasizes only two of Bloom’s(1956) high-level cognitive activities, it should benoted that the significant negative correlationbetween this scale and &dquo;false alarms&dquo; on the re-

cognition test of Study 5 suggests that the scalealso assesses the evaluation process, with highscorers seeming to evaluate the content of theirmemory more carefully before making choiceson a recognition test.The extent to which a student rigidly adheres

to study techniques and behaviors which have



428

traditionally been regarded as conducive to

learning in academic settings is assessed by theStudy Methods scale. A high scorer on this scalewould likely be labeled a &dquo;good student&dquo; in thetraditional sense. The finding from Study 5 thatthis scale related to performance only under in-tentional learning instructions implies that thestudy techniques assessed by the scale are ap-plied only when the individual is specifically in-structed to learn the material.A student’s ability to store or retain detailed

factual information is assessed by the Fact Re-tention scale. Since Study 5 indicated a signifi-cant positive relationship between scores on theFact Retention scale and recall in the incidental

learning condition, the scale may be assessing amemory capacity which influences learning evenin the absence of &dquo;studying&dquo; in the traditionalsense. Although a similar relationship was notobtained under the incidental conditions of

Study 4, the difference between the conceptuallecture material used in that study and the ver-bal list used in Study 5 makes it difficult to in-

terpret this discrepancy in findings at the pres-ent time.

Finally, the Elaborative Processing scaleseems to assess the lengths to which a studentwill go in order to encode new information. Theelaborative techniques assessed by the scale in-clude interrelating new and old information, us-ing visual imagery, rephrasing in one’s own

words, and thinking of practical applications.The Synthesis-Analysis and Elaborative Process-ing scales taken together appear to assess thetendency to take an active rather than a passiverole in the processing of new information. Like-wise, both scales seem to assess the habitual useof &dquo;deep&dquo; rather than &dquo;shallow&dquo; information-

processing strategies (Craik & Tulving, 1975), anassumption supported by the finding in Studies4 and 5 that high scorers do well on incidental-learning tasks. Thus, as one would expect, theresults of Study 3 show a positive correlation be-tween these two scales, and Studies 4 and 5 indi-cate that the two scales tend to relate or not re-

late to performance as a pair.However, there is some evidence regarding the

differential validity of the Synthesis-Analysisand Elaborative Processing scales. On a prose-learning task (intentional paradigm), the secondauthor (Ribich, 1976) reported significant posi-tive relationships between the Synthesis-Analy-sis scale and note-taking efficiency, objectiveexamination performance, and free recall ofidea units, while obtaining no relationships be-tween those prose-learning variables and theElaborative Processing scale. In addition,Ribich noted significant relationships betweenscores on the American College Testing Pro-gram (A CT) college entrance exam and the Syn-thesis-Analysis and Fact Retention scales butnot on the Elaborative Processing or StudyMethods scales. Thus, although both the

Synthesis-Analysis and Elaborative Processingscales deal with similar aspects of informationprocessing, it is not likely that they are measur-ing precisely the same processes.

Study 5 of the present report indicated thatthe Study Methods scale was significantly re-lated to performance only under intentional

learning instructions, while Synthesis-Analysis,Fact Retention, and Elaborative Processingevidenced such a relationship under incidental-learning instructions. A similar pattern of re-sults was obtained in Study 4; performance onan unexpected exam following a videotaped lec-ture was only related to the Synthesis-Analysisand Elaborative Processing scales. Such resultsare consistent with the findings of Craik andTulvig (1975) who note that in verbal-learningstudies, incidental learners will often perform aswell as (or even better than) intentional learnersif the instructions given to the incidental groupencourage deep processing of the material. Theimplication is that the well-practiced studyhabits which are elicited when subjects are sim-ply instructed to learn might, on occasion, dis-place more effective information-processingstrategies such as those assessed by the ILP’sSynthesis-Analysis and Elaborative Processingscales or those elicited by &dquo;deep processing&dquo; in-structions in an incidental learning task. Takenas a whole, this analysis suggests that students(and teachers for that matter) are not always



429

aware of which learning behaviors are most con-ducive to learning. As a matter of fact, some ofthe supposedly &dquo;effective study habits&dquo; which

they employ may turn out to be less effectivethan simply listening to or reading the informa-tion in a thoughtful, elaborative manner.

Tangential support for this analysis comes fromRibich’s (1976) finding that ACT scores show asignificant positive relationship with scores onthe Svnthesis-Analvsis scale but are not relatedto scores on the Studv Methods scale.McKeachie (1974) notes that very few innova-

tive techniques produce clearly superior learningwhen studied in the context of empiricalresearch. He also voices the opinion that more ofthese studies should permit the observation ofattribute-treatment interactions. It is frequentlythe case that half of the studies concerned with

an instructional technique report that it is effec-tive and half report that it is not. In fact, most ofthese techniques are probably effective with cer-tain students under certain conditions. For

example, Duchastel and Merrill (1973) reportthat the widely acclaimed procedure of provid-ing students with behavioral objectives some-times aids learning and sometimes has no efFect,but they suggest that some of the confusionwould be eliminated if researchers attempted torelate the performance measures to the charac-teristics of the individuals within the subjectsample. For example, they refer to one study(Etter, 1969) that obtained no main effect of be-havioral objectives but did find that males withhigh socioeconomic status profited more thanany other group from the availability of objec-tives. A major problem is that the attribute

measures employed in such studies are eithervery gross (e.g., socioeconomic status) or elsethey consist of personality and cognitive-stylemeasures which have not been specifically de-signed to assess learning processes. Since theILP was specifically designed for the latter pur-pose, future research will be aimed at determin-

ing whether any of the more publicized instruc-tional techniques interact with ILP scale scores.Furthermore, Glaser (1972) suggests that thereis a need in the educational field for new meas-

ures which deal with skills that are trainable andwhich are conceptualized in terms of processes.Since the Synthesis-Analvsis and ElaborativeProcessing scales of the ILP seem to provideoperational definitions of two &dquo;skills&dquo; of the

type referred to by Glaser, the present authorswill attempt to devise training procedures whichwill increase a student’s ability to employ themto his/her advantage in the academic setting.

In addition, Underwood (1975) suggests thatthe validation of a theoretical learning processshould begin by determining whether individ-uals differ reliably in the extent to which

they demonstrate the process. The present in-strument was designed to measure individualdifferences in the processes which served as

guides while preparing items, but further re-search is required to determine whether the ILPdimensions and the processes suggested bylaboratory research do in fact correspond. Forexample, the theoretical organizational proc-esses of clustering and subjective organizationseem to be assessed by the Synthesis-Analvsisscale, but further research is needed to deter-mine whether highs scorers on this scale do, infact, demonstrate more clustering and subjectiveorganization as measured by standard labora-tory techniques. Similar studies should be con-cerned with the relationship between encodingstrategies (e.g., imagery usage) and the Elab-orative Processing scale and between depth ofprocessing and both the Svnthesis-Atialvsis andElaborative Processing scales.

ReferencesAnderson, R. C., & Kulhavy, R. W. Imagery and

prose learning. Journal of Educational Psy-chology, 1972, 63, 242-243.

Bendig, A. W. The Pittsburgh scales of social extra-version-introversion and emotionality. The Jour-nal of Psychology, 1962, 53, 199-209.

Berdie, R. F., & Campbell, D. P. Measurement of in-terest. In D. K. Whitla (Ed.), Handbook of meas-urement and assessment in behavioral sciences.

Reading, MA.: Addison-Wesley, 1968.Biggs, J. B. Faculty patterns in study behavior.

Australian Journal of Psychology, 1970, 22,161-174.



430

Bloom, B. S. (Ed.). Taxonomy of educational objec-tives. Handbook: Cognitive domain. New York:David McKay, 1956.

Boersma, F. J. Test-retest reliability of the CF-1 Hid-den Figures Test. Educational and PsychologicalMeasurement, 1968, 28, 555-559.

Bousfield, A. K., & Bousfield, W. A. Measurement ofclustering and of constancies in repeated free re-call. Psychological Reports, 1966, 19, 935-942.

Cattell, R. B. The scree test for the number of factors.Multivariate Behavioral Research, 1966, 1,245-276.

Cattell, R. B. How good is the modern questionnaire?General principles for evaluation. Journal of Per-sonality Assessment, 1974, 38, 115-129.

Cowell, M. D., & Entwistle, N. J. The relationship be-tween personality, study attitudes, and academicperformance in a technical college. British Journalof Educational Psychology, 1971, 41, 85-90.

Craik, F. I. M., & Tulving, E. Depth of processingand retention of words in episodic memory. Jour-nal of Experimental Psychology: General, 1975,104, 268-294.

Cronbach, L. J. The two disciplines of scientific psy-chology. American Psychologist, 1957, 12,671-684.

Cropley, A. J., & Field, T. W. Achievement in scienceand intellectual style. Journal of Applied Psy-chology, 1969, 53, 132-135.

Duchastel, P. C., & Merrill, P. F. The effects of be-havioral objectives on learning: A review of em-pirical studies. Review of Educational Research,1973, 43, 53-68.

Earhard, M. Free recall transfer and individual dif-ferences in subjective organization. Journal of Ex-perimental Psychology, 1974, 193, 1169-1174.

Etter, D. C. G. Adult learner characteristics and in-structional objectives. Unpublished doctoral dis-sertation, University of California, 1969.

Eysenck, M. W. Individual differences in speed of re-trieval from semantic memory. Journal of Re-search in Personality, 1974, 8, 307-323.

Eysenck, M. W. Extraversion, verbal learning, andmemory. Psychological Bulletin, 1976, 83, 75-90.

Frase, L. T. Paragraph organization of writtenmaterials: The influence of conceptual clusteringupon the level and organization of recall. Journalof Educational Psychology, 1969, 60, 394-401.

French, J. W., Ekstrom, R. B., & Price, L. A. Manual

for Kit of Reference Tests for Cognitive Factors.Princeton, N.J.: Educational Testig Service, 1963.

Gagné, R. M. Learning and individual differences.Columbus, OH: Charles E. Merrill PublishingCo., 1967.

Glaser, R. Individual and learning: The new apti-tudes. Educational Research, 1972, 13, 5-13.

Goldman, R. D., & Warren, R. Discriminant analysisof study strategies connected with college gradesuccess in different major fields. Journal of Edu-cational Measurement, 1973, 10, 39-47.

Kaiser, H. F. The varimax criterion for analytic rota-tion in factor analysis. Psychometrika, 1958, 23,187-200.

McKeachie, W. J. Instructional psychology. AnnualReview of Psychology, 1974,25. 161-193.

Melton, A. W., & Martin, E. (Eds.). Coding processesin human memory. Washington, DC: Winston &

Sons, 1972Paivio, A., Imagery and verbal processes. New York:

Holt, Rinehart and Winston, 1971.Paivio, A. Imagery in recall and recognition. In J.

Brown (Ed.), Recall and Recognition, New York:John Wiley and Sons, 1976.

Paivio, A., & Csapo, K. Picture superiority in free re-call : Imagery or dual coding. Cognitive Psy-chology, 1973, 5, 176-206.

Paivio, A., Yuille, J. C., & Madigan, S. Concreteness,imagery, and meaningfulness values for 925nouns. Journal of Experimental PsychologyMonograph Supplement, 1968, 76, No. 1, Part 2,1-25.

Pettigrew, T. F. The measurement and correlates ofcategory width as a cognitive variable. Journal ofPersonality, 1958, 26, 532-544.

Plenderleith, M., & Postman, L. Individual dif-ferences in intentional and incidental learning.British Journal of Psychology, 1957, 48, 241-248.

Ribich, F. Performance on word list and prose learn-ing tasks as a function of learning style. Un-

published Master’s thesis, Southern Illinois Uni-versity, 1976.

Rorer, L. G., & Goldberg, L. R. Acquiescence in theMMPI? Educational and Psychological Measure-ment, 1965, 25, 801-817.

Schultz, C. B., & DiVesta, F. J. Effects of passage or-ganization and note taking on the selection ofclustering strategies and on recall of textualmaterials. Journal of Educational Psychology,1972, 63, 244-252.

Schwartz, S. Individual differences in cognition:Some relationships between personality andmemory. Journal of Research in Personality, 1975,9, 217-225.

Tallmadge, G. K., & Shearer, J. W. Relationshipsamong learning styles, instructional methods, andthe nature of learning experiences. Journal ofEducational Psychology, 1969, 60, 222-230.

Tallmadge, G. K., & Shearer, J. W. Interactive rela-tionships among learner characteristics, types oflearning, instructional methods and subject mat-ter variables. Journal of Educational Psychology,1971, 62, 31-38.



431

Thompson, C. P., Hamlin, V. J., & Roenker, D. L. Acomment on the role of clustering in free recall.Journal of Experimental Psychology, 1972, 94,108-109.

Tulving, E., Subjective organization in free recall of"unrelated" words. Psychological Review, 1962,69, 344-354.

Tulving, E. & Donaldson, W. (Eds.). Organization ofmemory, New York: Academic Press, 1972.

Underwood, B. J. Individual differences as a cruciblein theory construction. American Psychologist,1975, 30, 128-134.

Vaughan, D. S. The relative methodological sound-ness of several major personality factor analyses.Journal of Behavioral Science, 1973, 1, 305-313.

Wang, M. D., & Stanley, J. C. Differential weighting:A review of methods in empirical studies. Reviewof Educational Research, 1970, 40, 663-705.

Author’s Address

Ronald Ray Schmeck, Department of Psychology,Southern Illinois University, Carbondale, IL62901



development self-report inventory for assessing learning · 2016. 5. 18. · 413 development of a...

Documents