In a previous article, humor preference was found to change systematically with age but not with gender. Here, the pattern of change is confirmed and evidence of linkage between humor preference and Big 5 personality scores was found.

Introduction:

In a previous article⁽¹⁾ humor types defined in terms of motivational scales and cognitive divides were found to support predicted change of preference with age, but no gender preference was found. Here, links between humor preference on one hand and age and Big 5 personality scales on the other are sought. The internet questionnaire form contains 40 Jokes and either 20 or 50 items IPIP Big 5 personality items. The questionnaire is offered on line in multiple languages but responses in languages other than English over several months were too small to consider for separate analysis; the responses were commutated into one.

The Hypothesis:

The predictions of age related humor preference are explained in detail elsewhere⁽²⁾, here is a sketch of the essential concepts. Self preservation is identified as the motivation behind Emotional type humor, which either reassures self-existence or falsifies perceived threat to directly harm the subject. Threats to self preservation are generally external and come from the environment, which generally does not change very much over the age of the individual, therefore it is not unreasonable to expect that Emotional humor should have relatively constant preference; no age related change is hypothesized. The motivation behind Feeding humor is protecting the resources needed for growth, which is highest at childhood and declining with age, therefore preference of Feeding humor is predicted to start high then gradually fall with age. SocioSexual humor is predicted to increase and reach its highest at adulthood then falls with age. Parenting humor is predicted to gradually increase with age, reflecting increasing relevance of care for supporting the offspring. The motivational types above are also expressed in a cognitive divide, either externalizing or internalizing, for example Emotional humor saves the self from immediate harm either by escape (fear=internalizing) or by falsifying the source of threat (aggression=externalizing), so that the total count of humor types between motivational and cognitive scales is eight. The predictions of link between personality scores and humor preference are based on the observation of similarity of scale interpretations between personality and cognitive divides; the dependence on cognition is not surprising since it is viewed as a stable measure with respect to age after maturity⁽⁴⁾. Here, Big 5 IPIP items are included in the survey, the scores are correlated with cognition scores, or differences between externalizing and internalizing humor scores, the aim is to be able to predict the four (cognitive) scores of Big 5 scales from cognitive scales and the fifth Stability score from motivational scales. As for gender differences in humor preference, no predictions offered at this stage.

The Survey:

The IPIP translations of the 50-item Big 5 scales were used in the beginning. The 20-item �mini� version of Donnellan et al.⁽³⁾ was introduced later in order to attract more participation through simplification. However, perceptible increase was observed. Furthermore, an interim analysis showed no significant correlation using the mini version, so its use was abandoned in favour of the 50-item version. Noting Costa and McCrae�s finding⁽⁴⁾ that B5 scales were significantly more stable for ages over 30, age groups under 26 were removed then Big 5/Humor preference correlation values computed for older age groups.

The Analysis:

The analysis was performed using R Software . Appendix A contains survey details such as line statements of humor (jokes) and their categorization, raw responses and explanations. The Cronbach alpha coefficient for combined motivational scores is computed: 0.82. The same when computed using the robust estimator⁽⁵⁾ is 0.952, and for the scales of Emotion, Feeding, SocioSex and Parenting are respectively: 0.78, 0.88, 0.87 and 0.89. In preparation for the analysis, and in order to make raw data easily available for later research, subject contact information was removed and the responses were saved in a form ready to be copied and pasted in .csv file in Appendix B. Humor scores were standardized (Z-scored) for each subject and the scores averaged separately for sub-totals of genders and cognitive divides, so there are four sub-groups making up each humor motivational type. The tables of mean values and standard deviations for the four motivational types and cognitive divides are shown in Appendix C. In order to remove age group score bias (it was evident that subjects increased their average scores with age) the totaled scores were scaled for the four motivational types. Appendix D contains the scored preferences of each gender. The histogram of humor preference scores for combined genders is presented Figure 1 and numerical values in Table 1 below. Linear regression and other statistics can be read directly from the R run in Appendix E, which is offered as is with little annotation. The analysis for Big 5/Humor preference link is straight forward; scores from Big 5 scales are correlated with the motivational scales and cognitive divides. Subjects are grouped by personality assessment item totals; either 20 items �mini� or 50 items, tables are in Appendix F. This grouping will allow the assessment of the impact of lowering the item count on correlation values. In addition to individual scale correlation, total motivational scores are correlated with each Big 5 scale. Appendix F contains 20 and 50 items correlation tables as well as total motivation with the Big 5 scales.

Fig. 1 Humor preference changes with age, 479 participants.

	Under 15 #25	16 to 25 #187	26 to 35 #94	36 to 45 #79	46 to 55 #43	Over 55 #51
Emotion	0.236	0.00	-0.336	0.935	-0.553	0.095
Feeding	1.332	1.004	-0.067	-0.480	-0.573	-1.406
SocioSex	-0.732	0.053	1.225	-1.087	0.232	0.46
Parenting	-0.845	-1.057	-0.821	0.632	0.894	1.041

Table 1: Scaled mean values in standard deviations of humor preference scores.

The Results:

Fig. 1 histogram shows the data trends visually, which clearly supports the evidence that Emotional humor preference remains relatively constant over age, Feeding starts high and falls gradually with age, SocioSexual rises until adulthood at around 35 then gradually falls, here we find a freak value in the 36-45 age group, but its significance is tempered by a higher standard deviation of the cell (see Appendix C), notably for female motivational scores in Table C8, which indicates that there may be outliers in this age and gender group. The trends were closely similar for both genders (see Appendix D), standard deviation does not fluctuate by large values, Cronbach�s alpha values as above and linear regression parameter values in Appendix E support confidence in the results. The correlation coefficient values between Big 5 and motivation/cognition scales are noteworthy, for 20-item mini Big 5 scales there is hardly any significant value greater than 0.1, contrasting with 50-item scales where higher correlation values begin to show as expected with cognitive but not motivational scales, except as expected with the Stability scale. And as the age groups of the lower than 26 are removed we find the same coefficient values increase systematically and new coefficients begin to emerge (see tables F1/F2/F4/F5).

	Emotion	Feeding	SocioSex	Parenting
Extroversion	-0.11	0.15	-0.13	0.07
Agreeableness	-0.01	0.17	-0.02	-0.03
Conscientiousness	-0.02	0.03	0.02	0.15
Stability	0.24	-0.01	0.03	0.28
Imagination	-0.02	-0.19	-0.29	0.18

Table 2: 50-Item Big 5/Cognitive Correlation values for over 26 age groups, 264 participants.

	Emotion	Feeding	SocioSex	Parenting
Extroversion	-0.02	-0.07	-0.06	-0.03
Agreeableness	-0.04	-0.04	-0.02	-0.05
Conscientiousness	0.07	0.1	-0.01	0.04
Stability	0.09	0.11	0.09	0.15
Imagination	-0.01	-0.06	-0.01	-0.06

Table 3: 50-Item Big 5/Motivational correlation values for over 26 age groups.

The Conclusions:

The survey results confirm the strong link between humor preference and the age of the subjects. On the other hand, the link between Big 5 personality assessment and humor preference becomes clearer and conforms with expectation, yet not clear enough at this stage to define scale relationships. In order to achieve higher definition of scale relationships the data provided some direction for further research: To use the full 100-item Big 5 assessment and to focus on over 26-year old age groups.

References:

1. Kadri, FL : Humor Preference Changes with Age, http://www.artificialpsychology.com/TechNotes/humor_preference_age.htm , 2008.

2. Kadri, FLT and Duncan, IJH: A New Nonlinear Model of Mechanisms of Motivation, Behavioural Processes, Vol. 33, No. 3, pp. 273-288, 1995.

3. Donnellan, MB, Oswald, FL, Baird, BM and Lucas, RE : The Mini-IPIP Scales: Tiny-yet-effective measures of the Big Five Factors of Personality, Psychological Assessment 18(2): pp. 192-203, June 2006.

4. Costa, PT and McCrae, RR: Stability and Change in Personality Assessment: The Revised NEO Personality Inventory in the Year 2000. Journal of Personality Assessment, 68(1), pp. 86-94, 1997.

5. Christmann, A and Van Aelst, S : Robust Estimation of Cronbach�s Alpha, Journal of Multivariate Analysis, 97, pp. 1660-1674, 2006.

* Parts of this report were included in a presentation at the Department of Psychology, University of Zurich on January 9, 2009.

About the author: Faisal L. Kadri is an independent researcher not affiliated with any educational institution. His research interest since 1986 is in applying the mathematical tools of nonlinear systems engineering to modeling motivational mechanisms in animals and humans. For more information please visit: www.artificialpsychology.com 

Copyright � 2009 Faisal L. Kadri, all rights reserved. Reproduction of this article in Internet media is permitted on condition that all links are maintained. Use in college essays welcome.