Sunteți pe pagina 1din 4

Clayton's "Premarital Intercourse: A Substantive Test of the Contingent Consistency Model" Revisited Author(s): David W.

Grimes and Roger W. Libby Reviewed work(s): Source: Journal of Marriage and Family, Vol. 35, No. 1 (Feb., 1973), pp. 9-11 Published by: National Council on Family Relations Stable URL: http://www.jstor.org/stable/351090 . Accessed: 10/04/2012 05:34
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

National Council on Family Relations is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marriage and Family.

http://www.jstor.org

Letters

to

the

Editor

CLAYTON'S"PREMARITAL INTERCOURSE: SUBSTANTIVE A TEST OF THECONTINGENT CONSISTENCY MODEL"REVISITED In a recent article (Journal, May, 1972) Richard R. Clayton attempted to apply the contingent consistency perspective to premarital intercourse in order to delineate the relationshipbetween attitudes and behavior.He simultaneouslyassessedsubjects'beliefs toward premarital permissiveness, religiosity, perception of reference group premaritalpermissiveintercourse. ness, and self reports of premarital Overall, a higher degree of attitude-behavior consistency was found than has been reported in past studies (cf., Wicker,1969). Noting that contingent factors (i.e., referencegrouppermissiveness) had little uniform effect as predictors of behavior, Clayton suggested the contingent consistency approachrequiresfurther clarification. However, there are certain difficulties in the Clayton study that call into question whether he actually had "a substantivetest of the contingent consistencymodel." Clayton divided his contingent consistency model into three major sections: attitudes, contingent factors, and behavior.Attitudes and contingent factors were designatedas independent variablesto predict premarital intercourse. We will comment first on his treatment of attitudes. Clayton asserts "there are at least two levels at which [attitudes] ... operate" (p. 274). General attitudes "deal with content which is somewhat abstract and sometimes rather far removed from any kind of behavioralorientation" (p.274). Behavior-specificattitudes, on the other hand, are similar in content to substantive behavior toward a particular attitude object-in the present case, premarital intercourse. attiClayton's index of "behavior-specific" tudes consisted of five items from the Reiss (1967) PremaritalPermissivenessscales. However, in spite of Clayton's conceptualization only two of the five items utilized are actually towardpremarital intercourse. behavior-specific The other three items deal with kissing and petting-a far cry from coitus. Thus, Clayton's personal attitudinal sexual permissiveness (P.A.P.) scale turns out to be a general scale of premarital permissiveness of doubtful unidimensionality with regardto the attitude object of premarital intercourse. February1973

The five P.A.P. items constituted six Guttman-scale types which were used to dichotomize respondents into low and high attitude categories. The three "less difficult" scale types, i.e., (0) no endorsement;(1) kissing-no affection; and (2) petting-engaged, were classified as low permissivenessand the three "more difficult" scale types, i.e., (3) petting-no affecand tion; (4) full sexual relations-engaged; (5) full sexual relations-strong affection, were classified as high permissiveness. Thus, subjects who endorsed items up to and including "petting-no affection" and rejected the two items dealingwith coitus were classifiedas high on permissiveness. The implications of this "cookbook" grouping procedure on the attitude-behaviorrelationshipare as follows: Those subjects who expressed agreement toward the kissing and the two petting items and disagreement toward both the coital items were consistent if (wrongly) classifiedas behaviorally they reported previous engagement in coitus. Further, those subjects who similarlyagreedto the kissing and the two petting statements while rejecting both coital items, were (again wrongly) classified as behaviorallyinconsistent if they reported no previoussexual intercourse. Clearly, the logical division of lows and highs on the P.A.P. scale should have been between those who agreed to either one or both of the intercourse items and those who did not, i.e., between scale types 3 and 4, not 2 and 3. Thus Clayton in effect treated congruenciesbetween attitudes and behavior as inconsistencies and incongruenciesas consistencies.These criticisms suggest that any interpretation of Clayton's results must be viewed with considerable caution. In addition to questioningthe P.A.P. scale as a valid measure of behavior-specificattitudes, we disagreewith its status as a valid indicatorof personal attitudes. Thus, item three reads: "I believe that petting is acceptable for the male (female) before marriagewhen he (she) is not particularlyaffectionate toward his (her) partner." To make it an item of personalattitudes, the reader should rereadthe statement, substituting "me," "I am," and "my" for "male (female)," "he (she) is," and "his (her)," respectively. Clearly, such a rephrasing describes a personal standard which Clayton intended to assess. The original P.A.P. items elicit beliefs about the acceptance of same-sex social standardsto which a respondentmay or may not havepersonallysubscribed. 9

JOURNAL OF MARRIAGEAND THE FAMILY

Turning now to Clayton's indices of general attitude, we are somewhat confused as to why he designated religiosity as an attitude toward premarital sex. True, religious attitudes have been empiricallyrelated to permissiveness (religiosity has been related to many variables),but it is difficult to conceive of religiosity as a general attitude toward coitus, premarital or otherwise. Moreover, we seriously doubt its relevance as an independent variablein testing the contingent consistency postulate. In a prior paper reporting on the same study (1971), Clayton's purpose was to investigatethe effect of religiosity on premaritalpermissiveness attitudes. Therefore,the more recent paper appears to be a post hoc reanalysisof data intended for other purposes, rather than a study designedto test the relationship between verbal attitudes and overt behavior. The second section of Clayton's research purports to deal with contingent factors. However, they are nothing more than a of rephrasing the five-itemP.A.P. scale. Whereas each P.A.P. item begins with the phrase, "I believe . . .," the two contingenI variables, normative reference group (N.R.G.) and comparison reference group (C.R.G.), begin with, "They believe that.. ." and "Most people on campus believe that...," respectively.We find it rather arbitraryto consider P.A.P. items attitudes and N.R.G. and C.R.G. measuresas. as contingent factors. The N.R.G. and C.R.G. items just as clearly elicit cognitions or beliefs as the P.A.P. items. To the present authors all three scales seem to elicit a perceived norm concerning same-sex category. And like the P.A.P. scale, tle two referencegroup scales are limited to same-sex groups, whereas the predicted behavior involves a heterosexual ph-enomenon. Unfortunately, Clayton failed to provide opposite-sex reference group information. Moreover, the same criticisms applied to the P.A.P. scale concerningits generalcontent and the procedureof dividingsubjectsinto low and high categories also apply to the Guttmantype N.R.G. and C.R.G. scales. A further biasing limitation of Clayton's operationalization of both reference group indices involves their restriction to the campus normativesystem (cf., Clayton, 1971). Considering that approximately 47 per cent of the subjects were freshmen who had been on campus less than one semester, it is doubtful that the campus normativesystem provided an appropriatebase for selection of salient contingent factors (cf, Linn, 1965). The relatively low relationshipbetween reference group data and reportedcoitus may have been due in part 10

to the fact that this large percentage of new arrivals were not familiar with the prevalent campus attitudes and sanctions governing sexual permissiveness.In addition, if respondents had been provided the opportunity, many of them might have designatedtheir most important referencegroup as off-campus(e.g., family, friends at home, etc.). Clayton claims "there is the logical necessity of assumingthat the attitudinaland contingent factor variables are antecedent to the behavior under study" (p. 274, italics added). However, it is surprisingto find his researchdesignis the very antithesis of this "logical necessity." Clayton'sbehaviorcriterionwas reportedcoitus during the calendaryear prior to the measurement of attitudes and contingent factors. Thus, it would appear that Clayton has actually inadvertently tested the effect of previous behavioron currentlyheld attitudes and contingent factors. Clayton maintains that "behavior should always be very similar to the content of the behavior-specificattitudes and could refer to either overt behavior or an attitudinal disposition" (p. 274). We have already documented the dissimilar content of the P.A.P. and the substantive behavior indices. The concept "overt behavior" as utilized in the contingent consistency literatureimplies a commitment to engage in behavior or actual engagement in behavior. The concept clearly omits mere retrospectiveverbal self reports of one's prior behavior(Wicker, 1969). By referringto behavior as an attitudinal disposition, Clayton is actually designatingbehavioras a component of attitudes, an assumption which the contingent consistency postulate seriouslychallenges(Warner andDeFleur, 1969). Although Clayton did attempt to measure attitudes toward premarital intercourse, he completely neglected to recognize that attitudes and behaviordiffer in.situational aspects (cf. Figa-Talamanca,1972). All of his indices were situationally similar as they were assessed in the same anonymoussetting. Since behavior, attitudes, and contingent factors were simultaneously assessed, the issues of reactivity, dissonance reduction, social desirability, demand characteristics, etc., are problematic. On the basis of his interpretation of the contingent consistency perspective, Clayton concludes "it was expected that reported behavior would be the result of additive relationships..." (p. 274, italics added). However, studies which have operated under this perspectivestrongly suggest that the realkey to predicting behavior lies in discovering interFebruary 1973

JOURNAL OF MARRIAGEAND THE FAMILY

relaconsistency in the attitude-behavior tionship."AmericanSociologicalReview37 (December): Forthcoming. Clayton,RichardR. 1971 "Religiosity and premaritalsexual permisand siveness:elaborationof the relationship debate." Sociological Analysis 32 (Summer):81-96. a 1972 "Premarital sexualintercourse: substantive test of the contingent consistencymodel." Journal of Marriageand the Family 34 (May):273-281. Coleman,JamesS. 1964 Introduction to MathematicalSociology. New York:McGraw-Hill. Irene Fig'a-Talamanca, 1972 "Inconsistencies attitude and behaviorin of family-planningstudies." Journal of Marriageand the Family34 (May):336-344. Goodman,Leo A. 1972 "A modified multiple regressionapproach to the analysisof dichotomousvariables." American Sociological Review 37 (February):28-46. S. Linn,Lawrence 1965 "Verbal attitudes and overt behavior: a study of racial discrimination." Social Forces43 (March):353-364. Reiss,Ira Context of PremaritalSexual 1967 The Social. Permissiveness.New York:Holt, Rinehart andWinston. Warner, Lyle G. and MelvinL. DeFleur 1969 "Attitudeas an interactional concept: social constraintand social distanceas intervening variables between attitudes and action." American Sociological Review 34 (April): 153-169. AllenW. Wicker, 1969 "Attitudesversusaction: the relationship of DAVID W. GRIMES verbal and overt behavioral responses to ROGERW. LIBBY attitude objects." The Journal of Social State University Issues25 (Autumn):41-78. Washington

REFERENCES action effects of attitude and situational variables ratherthan their separateadditive effects Acock, AlanC. andMelvinL. DeFleur 1972 "A configurational approachto contingent (Acock and DeFleur, 1972; Wicker,1969).

Consistent with Clayton's hypothesis of additivity was his utilization of Coleman's (1964) additivetechniqueof multivariateanalysis for dichotomous variables.He also applied Coleman'stest for interactioneffects. It was on the basis of comparing the results of these statistical tests that Clayton claims "rather strong support" for an additivemodel. Unfortunately, as Coleman (1964) and Goodman (1972) caution, measuring and interpreting interactioneffects amongindependentvariables using Coleman'stechnique remainsan unsolved problem. Further, not all forms of interaction are handled by the Coleman model. These considerations plus Clayton's failure to meet necessary statistical assumptions and procedures, i.e., homoscedasticity, equal marginals, unweighted estimates, etc., cast doubt that an additiveinterpretationis appropriate. Since attitudes and behaviorwere assessedat the same time, in the same setting, and in the same manner, one would expect an even higher consistency than Clayton reported.The reasons for lack of a higher correspondencebetween attitudes and behaviorwere documentedin our comments on serious conceptual and methodological problems. In carryingout a secondary analysis of his data which were obviously not appropriate to test a contingent consistency model, Clayton cannot be credited with a substantive test of the attitude-behaviorrelationship. Hopefully, future studies will develop a closer fit between theory and researchdesign.

February 1973

JOURNAL OF MARRIAGEAND THE FAMILY

11

S-ar putea să vă placă și