Probing the Dimensions of the Need For Cognition

Kenneth R. Lord, Mercer University
Sanjay Putrevu, Bryant College
EXTENDED ABSTRACT - Cacioppo and Petty (1982) developed a 34-item Need for Cognition Scale that quickly became the definitive tool for measuring this construct. Later, Cacioppo, Petty, Feinstein and Jarvis (1996) offered an abbreviated (18-item) version of the scale. This paper reports the results of two studies that evaluate the factor structure of the instrument in its two forms and examine the relationship of the underlying dimensions to consumer response to advertisements.
[ to cite ]:
Kenneth R. Lord and Sanjay Putrevu (2003) ,"Probing the Dimensions of the Need For Cognition", in NA - Advances in Consumer Research Volume 30, eds. Punam Anand Keller and Dennis W. Rook, Valdosta, GA : Association for Consumer Research, Pages: 103-104.

Advances in Consumer Research Volume 30, 2003     Pages 103-104

PROBING THE DIMENSIONS OF THE NEED FOR COGNITION

Kenneth R. Lord, Mercer University

Sanjay Putrevu, Bryant College

EXTENDED ABSTRACT -

Cacioppo and Petty (1982) developed a 34-item Need for Cognition Scale that quickly became the definitive tool for measuring this construct. Later, Cacioppo, Petty, Feinstein and Jarvis (1996) offered an abbreviated (18-item) version of the scale. This paper reports the results of two studies that evaluate the factor structure of the instrument in its two forms and examine the relationship of the underlying dimensions to consumer response to advertisements.

STUDY 1

Completing the original 34-item scale, presented as part of a larger "Lifestyles Study," were 195 undergraduate marketing students from a midwestern university in the US. Three weeks later, the same individuals read allegedly unrelated editorial and advertising materials and responded to measures of ad recall, recognition and amount read.

The initial factor analytic procedures were similar to those undertaken by Cacioppo and Petty. Specifically, principal-components analysis of the Need for Cognition Scale yielded nine eigenvalues greater than one. Final communality estimates ranged from .04 to .56; none were sufficiently near 1.0 to justify an assumption of minimal unique variance. Therefore, principal-components analysis was abandoned in favor of the common-factor model.

A principal-axis solution produced four eigenvalues greater than one. To account for 95 percent of the variance, it would be necessary to retain seven factors. Depending on the degree of conservatism employed, use of the scree procedure could point to the last marked discontinuity as occurring after the second, fourth, fifth or ninth eigenvalue. Based on this evidence and the value of accounting for a large proportion of variance, it seems appropriate to reject a single-factor solution for this data set.

Given the inferential advantages of maximum-likelihood factor analysis, that procedure was employed for remaining analyses. The chi-square value remains significant until the test of the ninth factor. However, the Tucker and Lewis (1973) rho statistic, which is less prone to overfactoring, reaches .95 with eight factorsBan indicator that an eight-factor solution will provide good fit to the data.

Since the eight "need for cognition" factors may be expected to correlate with one another, a Harris-Kaiser oblique rotation was applied, with Varimax as the prerotation method. The first factor emerged with high loadings on six items associated with the enjoyment of cognitive tasks. Five variables characterized the second factor, which assesses confidence in cognitive ability. The third latent variable, with four items as its primary contributors, highlights a preference for complexity. Factor 4 is associated with four items centering on the need for purposeful cognitive effort. The three primary contributors to the fifth factor address a desire for understanding. Dominant in the sixth latent variable are two items measuring tolerance for cognitive responsibility. The essence of the seventh factor is captured in a single variable that measures trust in the consequences of cognitive effort. The two items with high loadings on Factor 8 capture intellectual self-perception. Interfactor correlations range from .01 to .27.

The relatively large number of factors required to provide a good fit to the data, together with the low correlations between those latent variables, suggests that need for cognition, as measured by the 34-item scale, is multidimensional. To determine whether the dimensions of need for cognition vary in their relationship to advertising processing, factor scores were correlated with ad recall, amount read and argument recall. The second factor (confidence in cognitive ability) achieved uniformly higher correlations with the ad-processing measures than did the total scale. This provides tentative evidence that when measuring need for cognition it may be useful to single out the specific latent factors embedded in the scale.

STUDY 2

The purpose of Study 2 was to analyze the dimensionality of the Need for Cognition Scale in its shorter version. Data were collected from a sample of 244 undergraduate students attending an Australian university. Each participant read a ten-page booklet containing two test ads (one simple and the other complex) for fictitious automobiles and filler materials. Subjects completed the 18-item version of the Need for Cognition Scale and measures of attitude toward the ad, brand attitude and purchase intention.

Maximum-likelihood factor analysis yielded four eigenvalues greater than one. Sixteen factors would be required to explain 95 percent of the variance. The scree procedure shows the last substantive discontinuity after the fifth factor. Similarly, chi-square drops to insignificance with the addition of a fifth factor and that is the point at which the Tucker-Lewis rho statistic reaches .95. Based on these criteria, a five-factor solution (with Promax oblique rotation) was adopted.

The first factor captures abstract thinking. The loadings for the second factor link it to preference for depth of thinking. Factor 3 relates to pragmatic thinking. The fourth factor represents a preference for complexity. The final dimension is a psychological-discomfort factor. The correlations between the factors range from .12 to .43.

Again ad response varies across the factors. All five factors significantly affect attitude toward the ad in both the simple and complex conditions, a result that is largely duplicated for brand attitude in the complex-ad condition. However, the influence of need for cognition on brand attitude in the simple-ad condition derives from the preference-for-complexity, psychological-discomfort and pragmatic-thinking factors. The same three factors account for need for cognition’s effect on the intention to purchase in the complex condition, with little effect of the factors in the simple condition. A difference between the effects in the simple and complex conditions is the direction of the parameters: they are consistently negative in the simple condition and positive for the complex ad.

DISCUSSION

The Need for Cognition Scale’s usefulness may be enhanced by the consideration of its factor structure. Both studies provided evidence that the construct (as measured by either form of the scale) is multidimensional. The observation that different factors account for need for cognition’s effect on brand attitude and purchase intention in simple- and complex-ad conditions establishes the value of isolating and exploring the effects of the factors underlying the need-for-cognition construct.

REFERENCES

Cacioppo, John T. and Richard E. Petty (1982), "The Need for Cognition," Journal of Personality and Social Psychology, 42 (1), 116-131.

Cacioppo, John T., Richard E. Petty, Jeffrey A. Feinstein and W. Blair G. Jarvis (1996), "Dispositional Differences in Cognitive Motivation: The Life and Times of Individuals Varying in Need for Cognition," Psychological Bulletin, 119 (2), 197-253.

Haugtvedt, Curtis, Richard E. Petty and John T. Cacioppo (1992), "Need for Cognition and Advertising: Understanding the Role of Personality Variables in Consumer Behavior," Journal of Consumer Psychology, 1 (3), 239-260.

Haugtvedt, Curtis, Richard E. Petty, John T. Cacioppo and Theresa Steidley (1988), "Personality and Ad Effectiveness: Exploring the Utility of Need for Cognition," in Advances in Consumer Research, Vol. 15, ed. Michael T. Houston, Provo, UT: Association for Consumer Research, 209-212.

Lord, Kenneth R. and Sanjay Putrevu (1997), "Advertising-evoked Images: The Effects of Word Association Set Size, Need for Cognition, and Style of Processing," in Proceedings of the Annual Conference of the Administrative Sciences Association of Canada, Marketing Division, Vol. 18, No. 3, H. F. MacKenzie, ed., St. John’s, Newfoundland: Administrative Sciences Association of Canada, 69-77.

Tucker, L. R. and C. Lewis (1973), "A Reliability Coefficient for Maximum Likelihood Factor Analysis," Psychometrika, 38, 1-10.

Zhang, Yong (1996), "Responses to Humorous Advertising: The Moderating Effect of Need for Cognition," Journal of Advertising, 25 (Spring), 15-32.

Zhang, Yong and Richard Buda (1999), "Moderating Effects of Need for Cognition on Responses to Positively versus Negatively Framed Advertising Messages," Journal of Advertising, 28 (Summer), 1-15.

----------------------------------------