Download Curriculum, Instruction, and Learning Outcomes: A Historical View and more Lecture notes Experimental Psychology in PDF only on Docsity!
DOCUMENT RESUME
ED 067 363 SP 005 782
AUTHOR TITLE Posner, George J.
PUB DATE Education:Apr 72 Its^ Components^ &^ Constructs.
NOTE 27p.; Paper presented at the Annual Meeting of the
American 1972 Educational Research Assn., Chicago, April
EDRS PRICE MF-$0.65 HC-$3.
DESCRIPTORS Curriculum; *Education; Educational Principles;
Instruction;^ Educational^ Research;Models; ResearchEducational Methodology^ Theories;
ABSTRACT
purpose of specifying^ This^ paper major^ presentseducational^ a^ conceptual components^ model and theirfor^ the
relationship.', The initial stage of building this theoretical
curriculume^ framework^ for'education instruction,^ includes and learning^ the^ clarification outcomes. The^ of next^ three section^ terms:
deals nature with of constructs,'constructs generalembodied types by each of validation,of the above and components, the importance the
of one of these types, namely, construct validity. Construct validity
is achieved explicated learning i.) the)-contextoutcomes. Suggestions of one educational are given component: for the the
investigation The need for clarificationof construct validityof educational for other components educational for thecomponents. purpose
of clarified, theory building empirical in researchcurriculum can and lead instruction to a theoretical is stressed. framework Once
eachidentifying of the components.the^ relationships A 43-item^ between bibliography^ various^ constructsis included.^ embodied^ by
(Author/MJM)
EDUCATION: ITSCOMPONENTSANDCONSTRUCTS*
CornellUniversity^ GeorgeJ.Porier
PaperpresentedattheAnnualConventionofthe AmericanEducationalResearchAssociation,Chicago,
April,
*Originaltitle: ModelforCurriculumandInstructionAMethodologicalResearch
I
U.S.DEPARTMENTOFHEALTH, EDUCATION8WELFARE THISDOCUMENTHASBEEN^ OFFICEOFEDUCATIONREPRO- DUCEDEXACTLYASRECEIVED THEPERSONORORGANIZATIONFROMORIG- INATINGIT.POINTSOFVIEW IONSSTATEDDONOTNECESSARILYOROPIN- REPRESENTOFFICIALOFFICE CATIONPOSITIONORPOLICY.OFEOU-
DefiningCurriculurn Johnson(1967)recognizedthatthemajorproblemplaguingthe fieldofcurriculumandinstructionwasoneofdefinitionofterms. Definitionsofcurriculumvaryfromwritertowriter,althoughmost definitionscanbecategorizedintooneofseveralpredominanttypes. AlthoughJohnsonisnotexplicitaboutthesetypes,onecaninferfrom hisdescriptionofdefinitionsofcurriculumthatthefollowingparame- tersdeterminethetypeofdefinition: (^) (a)process-versusproduct- orientation,(b)temporalrelationshipofcurriculumtoinstruction, and(c)programmaticversusdescriptiveimport. Processversusproduct.Theprocess-productdistinctionpertains towhethercurriculumreferstotheeventsorresultsofinstruction.The "process"viewisexemplifiedbydefinitionsofcurriculumintermsof "learningactivities"(AlbertyandAlberty,1962)or"alloftheexperi- encesofchildrenforwhichtheschoolacceptsresponsibility" (^) :lOgan, 1966,p.4).Definitionsofcurriculumintermsof"intendedlearnings" (GoodladandRichter,1966)or"whateverachildlearnsundertheguidance anddirectionoftheschool"(Wagner,1958,p.328)areexamplesof product-orienteddefinitions. Temporalrelationships. Thetemporalrelationshipbetween curriculumandinstructiondescribesthecurriculumaspriorto,concur- rentwith,orareportofinstruction. Definitionssuchas" ... the designofasocialgroupfortheeducationalexperiencesofthOrchildren inschool"(Beauchamp,1964,p.15)or" ... thosethingswewishchildren tolearn"(SaylorandAlexander,1966,p.3)areexamplesofdefinitions thatimplycurriculumtobeanticipatoryofinstruction,i.6.,priorto
3
instruction. On the other hand, definitions such as " ... the means of instruction ... " (Krug, 1957, p. 3) or " .., what a teacher uses when he teaches children" (Wilhelms, 1967), suggest that curriculum is a part of instruction. Definitions such as "whatever ... achild learns under the guidance and direction of the school (^) ... " (Wagner, 1958, p. 328) imply that curriculum is a report of what was learned. Descriptive versus programmatic. Some definitions promote a particular doctrine and are thus programmatic. Phrases such as "experi- ences ... having a maximum of lifelikeness for the fearnee (Rugg, 1926, p. 18) and " ... set up in the school for the purpose of disciplining children and youth in group ways of thinking" (Smith, Stanley, and Shores, 1957, p. 3) are highly programmatic. Others, such as that of Goodlad and Richter cited above make no statement prescribing the nature or content of the curriculum and are, therefore, non-programmatic or descriptive. Programmatic definitions take a position about what the curriculum should be, whereas non-programmatic definitions describe what the curriculum is. A proposed type of definition. For the purposes of explanatory theory and basic research a definition of curriculum is preferred that (a) is product-oriented, (b) places curriculum prior to instruction, and (c) is descriptive. This preference is justified by considering (a) the dis- tinction between curriculum and instruction, (b) the role of curriculum with respect to instruction, and (c) the state of the field of curriculum and instruction. Johnson (1967) pointed out that definitions involving experience rather than learning outcomes cloud the distinction between curriculum and instruction. Process definitions involve interactions since no experience
years. Melvin(1931)discussedtheconfusionbetweenthefieldsof "curriculumandmethod." Heemphasizedthatcurriculumshouldbedefined asdesiredlearnings(1931,p.730). "Inatheoreticalsensethecurricu- lumshouldlistthoselearningsindependentofthemethodoftheteacher whoistohelpthepupilattainthem"(1931,p.731). Asindicatedabove, Johnson(1967)alsodistinguishedbetweencurriculumandinstruction. tO:' definedcurriculumas"astructuredseriesofintendedlearningoutcomes" (1967,p.130),implyingthatcurriculumisanticipatoryto,anddistin- guishedfrom,themeansusedtoachievetheoutcomes. Thelatterhe viewedasthedomainofinstruction,whichinvolvesthemeansusedto convertintendedoutcomesintoachievedlearningoutcomes. Heregarded thisdistinctionbetweencurriculumandinstructiontobecrucialin research. Severalotherwritersinthefieldareingeneralagreement withJohnson'sandMelvin'sconceptionofcurriculum. Goodladand Richter(1966,p.11-12)definedcurriculumas" ...a setofintended learnings." Gagng'sdefinitionofcurriculumas" asequenceof contentunits ... "(1967,p.22)withcontentdefinedas"descriptions oftheexpectedcapabilitiesofstudentsinspecifieddomainsofhuman activity"(1967,p.21)issimilartoJohnson'sdefinition. Macdonald definedcurriculumasthesystemof"plannedactionsforinstruction," andinstructionas"thesystemforputtingtheplanintoaction"(1965, p.5). Macdonald,therefore,alsorecognizedtheneedtodistinguish plansfromactualexperiences,butfailedtodistinguishthecurriculum fromtheinstructionalplan.
Conceptions of curriculum similar to Johnson's definition can be found in both experimental and descriptive studies. In an experimental study Hutchinson (1963) attempted to determine how learning and thinking processes are affected by different teaching techniques when subject matter is kept constant. Various^ aspects^ of^ the^ teaching^ process^ have been studied independently of the subject matter taught by researchers analyzing classroom discourse (see, for example, Bel lack and Davitz, 1968; , Amidon^ and^ Flanders,^ 1963;^ Smith^ and^ Meux,^ 1968;^ Withall^ ,^ 1949).^ These studies imply that by conceptually separating subject matter from teaching methods, a more sophisticated approach to curricular and instructional research results. Relationship between curriculum, instruction, and achieved learning outcomes. By accepting Johnson's definition of curriculum as a "structured series of intended learning outcomes," one can conceive of the relationship between curriculum, instruction, and achieved learning outcomes as depicted in Figure 1.
Curriculum is shown in Figure 1 to anticipate instruction; it represents the intended learnings without specifying the instructional means of attaining them (Johnson, 1967). Both^ curriculum^ (C)^ and^ numerous administrative factors ("frame factors") act as inputs to or influences on the instruction (I). This influence is symbolized by a dotted line. Instruction, on the other hand, contributes to the production of the achieved learning outcomes (0). This^ process^ of^ production^ is^ symbolized by a solid line. Instruction Instruction is composed of the means by which the intended learning outcomes are achieved. Since it is guided by the curriculum, instruction is an intentional activity (Johnson, 1969). Gaga (1965) specified eight functions instruction: (1) presenting the stimulus, (2) directing attention and learner activity, (3) providing a model for terminal performance, (4) furnishing external prompts, (5) guiding the direction of thinking, (6) inducing transfer of knowledge, (7) assessing learning attainments, and (8) providing feedback. Functions one and three can be summarized as the "display" functions: they provide the stimulus^ component^ of^ instruction.^ Functions^ two,^ four, five, six, seven, and eight, on the other hand, can be summarized as the "control" functions: they^ provide^ the^ shaping-of-response^ component of instruction. For^ the^ purpose^ of^ this^ study^ variables^ related^ to^ one^ or more of the two instructional functions, i.e., display and control, are considered instructional variables.
Achieved Learning Outcomes According to Figure 1, the purpose of instruction is to enable pupils to achieve the intended learning outcomes. That is, the purpose of instruction is to promote learning, learning occurring when a particu- lar pupil 's capability of behaving in a particular situation changes (Gagne, 1965). Learning distinguished from measures of learning. Learning and behavior, although often related, are not the same (Smith, 1961). Learning can occur without a measurable change in gross behavior, al- though behavioral indicators of learning are needed to certify its occurrence. Measures of learning can only be considered indices of learning and not the learning itself. This distinction is discussed later in this paper in a more generic sense with regard to the distinc- tion between a construct and a particular measure of that construct. Learning distinguished from instruction. As^ Smith^ (1961 , p.
- pointed out, learning is distinct from instruction, learning often occurring either without instruction or in spite of instruction. The former case is indicated IA Figure 1 by "non-instructional experiences"; the latter case is indicated by "achieved unintended learning outcomes," occurring either as a concomitant result of instruction or from non- instructional experiences. Multiple Outcomes of Instruction Psychologists and educationists have pointed out that anticipated learning outcomes are often accompanied by unanticipated results (Figure 1). Cronbach (1964) argued for the consideration of multiple learning outcomes,
Lastly, when alternative courses in a particular area are being compared, it might be desirable to find out how well the graduate of one course can understand issues in another, i.e., how much overlap of learning outcomes exists (Cronbach, 1964, p. 244). Without this in- formation proper decision-making is difficult. Thus, research investigating multiple outcomes of instruction could aid in the answering of several questions: (a)^ What^ are^ the achieved learning outcomes of a particular combination of a curriculum and an instructional treatment? (b) How do different combinations com- pare in their effect on both a particular outcome and also across various outcomes? (c) Does a particular combination of a curriculum and an instructional treatment discriminate; between two or more outcomes? If all the outcomes considered are desirable, then discrimination between the outcomes, whether they be intended or unintended, is neither necessary nor desirable. If,^ however,^ some^ outcomes^ are^ desirable^ and^ some^ are^ not desirable, then discrimination is necessary. Research into the relative effectiveness of curriculum and instruction in the differential achievement of intended and unintentional outcomes is certainly not new. For^ several^ years^ researchers^ in^ psychology have been interested in the relative effectiveness of various independent variables on intentional versus unintentional or "incidental" learning. Using incidental versus intentional learning as the dependent variable, many independent variables, both curricular and instructional in nature, have been investigated, such as syntactical versus unsyntactical material (Epstein and Arlinsky, 1965); time (Bahrick, 1957; Miller and Lasko, 1964);
contingent reinforcement (Dixon and Moulton, 1967); rate of presentation and instructions given (Dornbush and Winnick, 1967); and amount of induced attention (Schneider and Kintz, 1967). Although^ not^ unanimous,^ most^ of these researchers report differential achievement of incidental compared with intentional learning outcomes as a result of varying different curricular or-instructional variables. Constructs and Construct Validity
Researchers in curriculum and instruction continually deal with constructs embodied by each of the previously discussed educational components. The^ following^ sections^ are^ concerned^ with^ the^ nature^ of constructs, general types of validation, and the importance of one of these types, namely, construct validity. Construct^ validity^ is^ then explicated in the context of one educational component, i.e., the achieved learning outcomes. Finally,^ suggestions^ are^ given^ for^ the investigation of construct validity for other educational components. Constructs The term "construct" has been used in various ways by various writers. For^ the^ purpose^ of^ this^ paper^ construct^ is^ defined^ as^ any postulated attribute of people or of situations assumed to be reflected in some objective indicator. This^ definition,^ although^ more^ general,^ is consistent with Cronbach and Meehl's definition. They defined a constrat as "some postulated attribute of people assumed to be reflectedin tint performance" (1955, p. 283). The^ proposed^ definition^ is^ also^ consistent with Kantor's description of constructs "as products derive() from [observers] interbehaving with events" (1963, p. 181) hoilfilargenau's (1950) 13
attributable to the^ / assessment method. In order to accomplish this sort of separation, the same construct must be assessed by two or more diver- gent methods (^) (Cronbach^ / and Meehl, 1955; Campbell and Fiske, 1959; Webb et al, 1966; Garner, 1954; Garner, Hake, and Eriksen, 1956). (^) Once these independent assessments have been obtained they can be compared; if the tests are purported to assess the same construct, then test scroes should correlate highly (Cronbach and Meehl, 1950; Campbell and Fiske, 1959). Campbell and Fiske (1959) suggested that tests representing a common assessment to hnique but different constructs also be compared. Cor- relations be (^) een measures of supposedly different constructs should be low even when the assessment method is the same. This latter aspect of construct alidity, Campbell and Fiske (1959) termed "discriminant validity";itke correlation between different measures of the same construct they termed^ I "convergent validity." (^) Several writers have suggested matrices of intercorrelations for assessing the various aspects of construct validity (Cronbach and Meehl, 1955; Campbell and Fiske, 1959). Achieved learning Outcomes as Constructs 1 and, the measures outcome is viewe
Achieved learning outcomes may be viewed as a class of constructs efore, when assessing t'em, the construct validity of their must be considered. To determine construct validity a "multi- ulti-method" approach is needed in which any measure of an outcome as an [outcome - method] unit:
One task of the researcher is to differentiate between the achiev d learning outcomes and their methods of assessment. (^) The
differentiation task may be visualized by considering the variance associated with an outcome assessed by a particular method as a Euler circle.
variance^ [outcome- method] unit
Assume, for example, that one is studying any two achieved learning outcomes101 and 02, e.g., spelling ability and vocabulary know- ledge. Assume also that one has two methods for assessing each outcome, M 1 and M2, e.g., multiple-choice and completion tests. The result is four [outcome-method] units, which are called outcome measures (Cronbach and Meehl, 1955). Assume further that each outcome measure ([outcome-method of assessment] unit) has a hypothetical variance of one, represented as follows.
Intercorrelating these four[outcome-method]units yields the following relationships:
If the achievement of a particular learning outcome by a group of subjects (Ss) is assessed by different methods, the results can be considered to consist of part "outcome variance" and part "method variance." Measures of purportedly similar achieved learning outcomes should share a substantial amount of variance regardless of the assess- ment method used; that is, they should possess convergent validity. Information about the discriminant validity of outcome measures is also useful. That is, do purportedly different learning outcomes, even when-assessed by similar methods, possess considerable common variance? Empirical differences are not necessarily expected even when obviously different outcomes are compared. For example, individuals who score highest on a measure of achievement might also score highest on an interest measure. Or, individuals who score highest on an arithmetic measure might also score highest on a reading measure. Yet even though the scores on these outcomes might correlate highly, the measures represent obviously different outcomes. Whether they are different or similar, how- ever, the degree of correlation is important information for the researcher, as it reveals much about the extent to which different individuals achieve different outcomes. The empirical construct validity of outcome measures can be determined by the development and use of correlational matrices similar to those of Campbell and Fiske (1959) (See Figure 2).
M1 (^) ri 1 m
0
Ill (^) r 1 112 r 2 r 5 02 M1^ r^3 r^6 r^8 M2 r4 r 7 r9 r
Figure 2. A generalized paradigm of a simple correlation matrix useful for construct validation of measures of outcomes, 01 and 02, as assessed by methods, Ml and M2. Correlations ri, r5, r8 and rio are reliability cells; r2 and r9 are con- vergent validity cells; r3 and r7 are discriminant validity cells; r4 and r6 are hetero-outcome hetero-method cells.
19
