10 pages

Humor and Competence in School-Aged Children Author(s): Ann S. Masten Source: Child Development, Vol. 57, No. 2 (Apr., 1986), pp. 461-473 Published by: Wiley on behalf of the Society for Research in Child Development Stable URL: http://www.jstor.org/stable/1130601 . Accessed: 18/11/2014 10:00

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 [email protected].

.

Wiley and Society for Research in Child Development are collaborating with JSTOR to digitize, preserve and extend access to Child Development.

http://www.jstor.org

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Humor and Competence in School-aged Children

Ann S. Masten

University of Minnesota

MASTEN, ANN S. Humor and Competence in School-aged Children. CHILD DEVELOPMENT, 1986, 57, 461-473. Multiple aspects of humor were evaluated in children between the ages of 10 and 14 and related to several areas of competence manifested at school. Humor measures assessed appreciation (including mirth, subjective ratings, and response sets), comprehension, and production, while competence measures included teacher ratings of classroom behavior, peer reputation, and achieve- ment. Humor was related to competence in several ways consistent with previous theory and research: (1) through the manifestation of intellectual ability both in humor behaviors and in compe- tent functioning; (2) through the role of mastery motivation enhancing both types of functioning; and (3) through peer relations, resulting from the effects of humor on peer acceptance or the effects of peer relations on humor behaviors. Ideas for further research relating humor to social competence, social cognition, and mastery motivation are discussed.

The appreciation and creation of humor have been credited with many adaptive func- tions (Goldstein & McGhee, 1972; Ziv, 1984). Social functions include the enhancement of interpersonal relations and group cohe- siveness as well as the expression of sexual or aggressive feelings. Physiological functions include arousal reduction or relaxation pro- duced by laughter. Defensive functions in- clude the avoidance or alleviation of fears and anxiety. Cognitive functions ascribed to humor center on the process of "pleasure in mastery," referring to enjoyment derived from successfully meeting the intellectual challenge involved in understanding or creat- ing humor (McGhee, 1974). This concept was based on White's (1959) theory of competence (or mastery) motivation, although Freud and Piaget each observed the process (Keith- Spiegel, 1972). Humor is thought to satisfy the competence motive and thereby to en- courage subsequent mastery efforts.

Given the multiple adaptive functions at- tributed to humor, it is not surprising that a good "sense of humor" is often theoretically associated with adjustment in terms of matu-

rity, health, or effective coping (Martin & Lef- court, 1983). These theoretical perspectives suggest that humor would be positively re- lated to competence, a primary characteristic of adjustment referring to the capacity for effective functioning in the environment (White, 1959; Yarrow et al., 1983). Up to now, however, this relation rarely has been exam- ined empirically, except with respect to intel- lectual ability as measured by IQ, a relatively narrow approach to competence (McClelland, 1973).

Research on children's humor apprecia- tion, comprehension, and production pro- vides evidence of the cognitive, social, defen- sive, and concomitant physiological functions of humor (McGhee & Chapman, 1980). First, children prefer humor they can understand and appear to derive considerable enjoyment from figuring out jokes. It has also been dem- onstrated, although not consistently, that op- timal cognitive challenges can enhance hu- mor appreciation (McGhee, 1976; Zigler, Levine, & Gould, 1966, 1967). Second, stud- ies of infants and older children illustrate the social and physiological functions of humor.

This paper is based in part on a doctoral dissertation submitted to the University of Minnesota. The research reported is part of the ongoing Project Competence studies of stress resistance in children. The author is grateful to Norman Garmezy, adviser and project director, and also to Auke Tellegen, David Pellegrini, Richard Weinberg, Alan Sroufe, Kevin Larkin, Robert Cudeck, Sherry Mulrooney, and Anne Chodzko for their assistance. The author also wishes to thank Tom Wilson and Universal Press Syndicate for permitting "Ziggy" to participate in this research. The investiga- tion was partially supported by the National Institute of Mental Health (MH-33222) and the William T. Grant Foundation, by an Eva O. Miller Fellowship, and by the University of Minnesota Com- puter Center. Requests for reprints should be sent to Ann S. Masten, Department of Psychology, N414 Elliott Hall, University of Minnesota, 75 East River Road, Minneapolis, MN 55455.

[Child Development, 1986, 57, 461-473. ? 1986 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/86/5702-0007$01.00]

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

462 Child Development

Both in infant-mother games and in the in- teractions of older children watching cartoon films, humorous laughter and smiling appear to facilitate arousal modulation, thereby en- hancing the quality of the transactions (Chap- man, Smith, & Foot, 1980; Pien & Rothbart, 1980; Sroufe & Waters, 1976). The shared positive affect appears to be reciprocally rein- forcing, which would encourage further social engagement. Third, observers of preschoolers and older children have noted the use of humor for coping with fear, anxiety, and cur- rent concerns (Gesell, Ilg, & Ames, 1946; Murphy & Moriarty, 1976; Ransohoff, 1975; Wolfenstein, 1954).

Often these functions are intertwined. For example, sexual jokes in the group of ado- lescent girls observed by Ransohoff (1975) served both to relieve anxieties about current developmental concerns and to enhance group cohesion. Similarly, the sharing ofjokes by elementary school peers provides the plea- sure of getting the joke along with the fun of sharing it with friends. Humor appreciation and creation thus can be multiply reinforced and reinforcing in a social context and would be expected to enhance competence, particu- larly social competence.

In addition to the evidence that humor may serve multiple functions in children, de- velopmental research also suggests another mechanism by which humor and adjustment become linked. Much of the recent research on children's humor has focused on how it is mediated by cognitive development. Data clearly show that appreciation and creation of humor depend to some degree on the level of factual knowledge, the development of sym- bolic, logical, and abstract reasoning abilities, language development, and the linguistic knowledge of the child (McGhee & Chap- man, 1980). Since competence is also medi- ated by cognitive development, some relation would be expected between humor abilities and competence as a function of the devel- opment of general cognitive abilities. Zigler et al. (1966) have suggested that humor is as- sociated with adjustment because both are mediated to some extent by cognitive func- tioning: "The conceptual demands of a joke or a cartoon require a variety of cognitive functions not unlike those demanded by a multitude of life's problems, the adequate solutions of which characterize the adjusted individual" (p. 509).

Thus, research on children's humor as well as various theoretical perspectives sug- gest that humor should be positively related to social, emotional, and cognitive compe-

tence. A number of studies have considered the relation of humor to intellectual compe- tence (Brodzinsky & Rightmyer, 1980). How- ever, the results are inconsistent, quite possi- bly because humor behaviors have been measured inconsistently.

Studies relating children's humor abili- ties to social competence are scarce. In his studies of humor, Ziv (1984) found that ado- lescent humorists were more popular than their less humorous peers and were also viewed by peers as leaders. Humor and intel- lectual ability were positively related, partic- ularly for creative humor and for males. In another study of eighth-grade class clowns identified sociometrically by their peers, Damico and Purkey (1978) found that "clowns" were rated by their teachers as lead- ers who were more cheerful, attention seek- ing, and unruly than nonclowning classmates.

The purpose of this study was to examine the relation between humor and competence in children, more specifically between three types of humor behavior and several signifi- cant aspects of effective functioning in school- aged children. One index of competence, peer reputation, was assessed by means of a class play instrument. Behavioral competence in the classroom was assessed by teacher ratings, and academic competence was as- sessed by an individual achievement test. In- tellectual competence was assessed by an IQ measure so that the broader measures of so- cial and academic competence could be com- pared with this traditional index of intellec- tual ability.

Humor assessment included two mea- sures of humor appreciation, the funniness ratings of cartoons by the children and the mirth visibly expressed by them. In addition, neutral cartoons (humor removed) were used to measure possible response sets in the rat- ings of cartoons by the children, an important but neglected technique introduced by Brod- zinsky (1975, 1977). These neutral cartoons allowed both appreciation scores to be ad- justed for indiscriminate responding. Com- prehension was measured by judges' ratings of the children's explanations of the cartoon humor. Finally, creative humor ability was measured by the quality of humorous cap- tions or ideas children were able to produce on demand.

Given the cognitive and motivational re- quirements of the humor comprehension task and the production of humor on demand, both these aspects of humor were expected to be positively related to general intellectual com-

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 463

petence as measured by IQ and to academic performance and effective classroom behav- ior. Appreciation was also expected to be re- lated to IQ since appreciation (when cor- rected for baseline responsiveness) should depend to some degree on comprehension. In addition, the pervasive association in the liter- ature between laughter or smiling and mas- tery suggested that appreciation might also be positively associated with academic compe- tence, perhaps beyond the role of comprehen- sion. Finally, since both humor appreciation and production appear to be socially facilitat- ing and highly valued characteristics, they were expected to relate to social competence as assessed by peer reputation.

Given the expected relations among IQ, other forms of competence, and humor, the question arose as to whether humor would be related to social and academic competence beyond whatever variance was shared with IQ. It was important to address this question in order to determine whether humor and competence are related only because they are both cognitively mediated. On the premise that humor and competence-particularly so- cial competence-involve motivational and social qualities that go beyond individual dif- ferences in intellectual ability, it was pre- dicted that some unique relations would be found between humor and competence be- yond the contributions of IQ.

Method

Subjects Participants in this study included 93

children (51 girls and 42 boys, ages 10-14) in grades 5 (N = 32), 6 (N = 23), 7 (N = 22), and 8 (N = 16). Twenty-five percent of the chil- dren were of ethnic minority backgrounds, in- cluding black, American Indian, Hispanic, and Asian students. These children were part of a series of investigations dealing with com- petence and stress resistance in school-aged children (Garmezy, Masten, & Tellegen, 1984; Garmezy & Tellegen, 1984). Partici- pants in these studies were originally drawn from all the third- to sixth-grade students of two centralized urban schools housed in the same building complex with a substantial pro- portion of lower- and middle-class families and single parents. At the time of this study, the participants had been involved in various Project Competence studies for approxi- mately 2 years, during which time a number of children had gone on to secondary school or had moved to new school districts. A large proportion of these children and their families were participating in interview studies that

provided information regarding socioeco- nomic status (SES). The Duncan Socioeco- nomic Index (Hauser & Featherman, 1977) was used to estimate SES levels. There was a considerable range of SES scores (11-92) on this 100-point scale. The mean SES level (43) falls among such occupations as craft workers, highly skilled labor, and clerical workers.

Measures Humor stimuli.-Cartoons were chosen

in order to include a broad range of nonverbal humor and on the assumption that humor with a visual element has great appeal for children. "Ziggy" cartoons created by Tom Wilson (Universal Press Syndicate) were se- lected for the pool of humor stimuli because they were available in abundance, because they varied widely in difficulty and content, and because they featured one central charac- ter of ambiguous age and occupation who of- ten faced situations that would be familiar to children. Limiting the pool to one type of car- toon was intended to promote the formation of a humor set that might facilitate the diffi- cult task of producing humor on demand, us- ing cartoon drawings of the same type as stimuli.

Cartoons from a large collection (Wilson, 1977) were initially screened on the basis of wordiness (cartoons with more than a dozen words were eliminated), difficulty, appeal, and suitability for children. Adult judges (graduate students in psychology and three elementary school principals) were instructed to rate the cartoons on appropriateness as well as "humorousness, engagingness, and inter- est" for children ranging from age 8 to age 14, without regard to difficulty. The final pool of 96 cartoons was rated again by 163 fourth to eighth graders for funniness and by five judges (graduate students) for difficulty level. The following humor measures were devel- oped from these cartoons.

Humor appreciation.-Two sets of car- toons were selected for appreciation assess- ment. Set A, chosen to maximize the effects of individual differences not due to comprehen- sion or group differences, consisted of 14 ap- pealing, easy cartoons that had shown no sex or grade effects in the pilot results. Set B, cho- sen to maximize the likelihood of obtaining appreciation differences related to difficulty and comprehension, was selected on the basis of multiple criteria: high interjudge agree- ment on difficulty level, the absence of sex effects, significant grade effects in pilot re- sults, and varied content. Four cartoons were selected from each of four difficulty levels. Finally, sex neutral (nonfunny) cartoons were

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

464 Child Development

created by removing the intended humor. Three neutral cartoons were randomly added to both set A and set B.

The cartoons were presented individu- ally in random order to each child to assess funniness on a five-point scale, ranging from 1 (not funny at all) to 5 (very, very funny). Five simple facial line drawings illustrated each scale point, ranging from a neutral, unsmiling face at level 1 to an open-mouth laughing face at level 5. Simultaneously and unknown to the child, the administrator rated the affect ("mirth") expressed by the child in response to the cartoon on a four-point scale: 1 = no or negative response; 2 = slight smile (silent, lips closed or almost closed); 3 = full smile (open mouth smile, usually silent or with one expiration of breath at the outset or one noise); and 4 = laughter (vocalized smile, with "haha" breathing pattern and vocaliza- tion). Previous reports of mirth scoring sug- gested that a simple behaviorally defined scale was the most reliable (e.g., Brodzinsky, 1975, 1977; Zigler et al., 1966). Pilot testing indicated interrater agreement for this scale of 77% for perfect agreement and 96% for agreement within one mirth level. The -inter- rater reliability correlation for mirth scores summed over 30 cartoon items (15 subjects) was .97. In a pilot study with 33 elementary school children, test-retest reliability for alter- nate forms of a 48-item group-administered appreciation test given 2 weeks apart was .78.

Funniness ratings and mirth scores were summed for sets A and B and the neutral car- toons. Subsequently, the ratings and mirth scores were adjusted for baseline respon- siveness. When a single score was needed for analysis, a regressed difference score (Cohen & Cohen, 1975) was derived by the formula C = ZA - rABZB, where ZA is the standardized response to the regular cartoons, ZB is the standardized response to the neutral cartoons, and rAB is the correlation of the A and B scores. Otherwise, the baseline appreciation scores were used as covariates.

Humor comprehension.-After the ap- preciation and production tasks were given, each child was asked to explain the humor in the set B cartoons, and the explanations were noted verbatim. Subsequently, each response by each subject was judged independently by presenting them separately (on index cards) to each judge in random order. Judgments were based on comprehension criteria for each cartoon developed by Masten (1982). Com- prehension was rated by two judges on a three-point scale: 3 = full comprehension; 2 = partial comprehension; 1 = no com-

prehension. Subsequent to independent rat- ings, the two judges discussed disagreements to reach a consensus. Scores were summed to obtain a global comprehension score. Inter- rater reliability was .96 for the global score.

Humor production.-Two tasks, Cap- tions and Titles, were developed to assess the creative aspects of humor. Captions consisted of nine cartoons in which part of the humor, usually a verbal punch line, was removed. The child was instructed to add something to make the cartoon funnier. Titles included six unaltered, nonverbal cartoons previously rated by judges as easy to understand. The child was asked to make up a funny title for the cartoon. In both tasks the cartoons were selected to have a strong "pull" for humor, although the Captions task did not require a verbal "punch line" response; children could and did suggest nonverbal changes in the car- toon drawings. If unable to produce a re- sponse, the child was encouraged to return to an item later. For Captions only, if no attempt at an item was made, a single prompt was provided: for example, "What could the bird be saying to Ziggy that would be funny here?" (In the final scoring no penalty was deducted for cued responses since this pro- duced a negligible effect on scores.)

Subsequently, each response was inde- pendently scored by two judges who first scored each response for appropriateness (in- appropriate or bizarre responses were scored 0) and then scored all appropriate responses using a five-point scale with two anchor points: 1 (not at all funny) and 5 (exception- ally funny, witty). These independent judg- ments were averaged to obtain quality scores. Unknown to the judges, the original captions for the Captions task cartoons were randomly included in the deck of responses for each item. For eight of the nine items, a number of children scored as high or higher than the car- toonist. Interrater reliabilities were .63 for each task and .75 for a 15-item composite.

Intelligence.-General intellectual abil- ity was estimated by two subtests of the Wechsler Intelligence Scale for Children- Revised (WISC-R), Vocabulary and Block De- sign. This two-test short form has shown the highest correlations (r = .88) with the Full Scale IQ score (Silverstein, 1975). The mean score for the two scale scores summed (N = 90) was 20.2 (SD = 5.3).

Competence.-Academic achievement was assessed by the Peabody Individual Achievement Test (PIAT) (Dunn & Mark- wardt, 1970). Behavioral competence in the

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 465

classroom was measured by teacher ratings on the Devereux Elementary School Behavior Rating Scale (DESB) (Spivack & Swift, 1967), which has shown evidence of reli- ability and validity for differentiating mal- adjusted and poorly achieving children in the classroom (Kendall, Pellegrini, & Urbain, 1981). Four scores were computed: Dis- ruptive-Oppositional (based on 18 items), Poor Comprehension-Disattention (14 items), Cooperative-Initiating (eight items), and Performance Anxiety (four items). These scores had shown factorial validity, stability, and high internal consistency in a Project Competence study of 648 children (Finkel- man & Ferrarese, 1981). Interpersonal com- petence was assessed with a peer assessment instrument, the Revised Class Play (Masten, Morison, & Pellegrini, 1985). The class play method has shown good evidence of concur- rent and predictive validity regarding adjust- ment (e.g., Bower, 1969; Cowen, Pederson, Babigian, Izzo, & Trost, 1973; Lambert, Hartsough, & Zimmerman, 1976). The Re- vised Class Play measures three reliable dimensions of peer reputation: sociability- leadership, aggressive-disruptive, sensitive- isolated.

Procedure Humor session.-Children were tested

individually at their schools. The humor bat- tery was administered in one session of about 30-min duration by a female administrator (the investigator, who was uninformed about the status of the children on all SES, compe- tence, and IQ variables). Every attempt was made to maintain a warm and pleasant con- text for all the tasks, without emphasizing time or evaluation, with the goal of fostering an atmosphere conducive to humor apprecia- tion and particularly to creating humor on de- mand.

Appreciation tasks were administered first. After several examples, set A cartoons (including three neutral cartoons) were pre- sented one at a time followed by set B car- toons (also including three neutral cartoons). Each set was presented in random order ex- cept that the first cartoon was not one of the difficult or neutral cartoons. The children were instructed that some cartoons would be funnier than others and that some would not be funny at all, and the rating scale was fully explained. Following the appreciation tasks, the two humor-production tasks (Captions and then Titles) were administered. Each task included a practice example. Comprehension was administered last since this task, by na- ture, seems more evaluative or academic and

might have stifled appreciation or creativity if it had preceded the others. The children were asked to explain what was funny (or supposed to be funny) about each of the set B cartoons, which were presented in the same order as in the appreciation ratings.

Other measures.-Procedures for the competence and other measures have been described in detail elsewhere (Garmezy & Tellegen, 1984). The PIAT was individually administered, and standardized scores were computed according to the manual. The Re- vised Class Play was group administered to each classroom, and scores were standardized within classroom and sex (to adjust for un- equal sex and classroom size distributions). Teachers completed the DESB on each mem- ber of their classes, and scores were standard- ized within class (to adjust for variations in teacher rating tendencies).

The humor tests were administered in the fall. The PIAT had been administered shortly before by another administrator. The peer and teacher assessments reported here were collected during the late spring of the same calendar year (but the previous school year) as the humor measures.

Results

Reliability Internal consistencies (alpha) for the ap-

preciation and comprehension measures were satisfactory, .91 for funniness ratings (30-item composite), .93 for mirth, and .71 for com- prehension. Alpha was more moderate for humor-production scores: .55 for the Captions task, .51 for Titles, and .68 for a 15-item com- posite of the two. Recall that interrater reli- ability was also lower for judgments of cre- ative humor quality. Both producing humor and judging its quality may be more difficult and less consistent than other aspects of humor behavior.

Since the neutral cartoons were altered with the goal of removing the joke, the mean responses to these cartoons should be lower than the means for the "funny" cartoons. All neutral versus "funny" cartoon mean com- parisons for ratings and mirth indicated that the means for neutral cartoons were signifi- cantly lower (one-tailed t tests, all p < .001). For example, the mean funniness rating and mirth scores were 3.28 and 1.97, respectively, for funny cartoons and 1.63 and 1.23, respec- tively, for the neutral cartoons.

Intercorrelations among the Humor Scores The correlations of the same appreciation

scores across cartoon sets A and B were quite

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

466 Child Development

TABLE 1

INTERCORRELATIONS AMONG HUMOR SCORES

1 2 3 4 5 6 7

1. Production .......... 2. Comprehension..... .53***

Funniness Ratings

3. Adjusted ..........

.09 .15 4. Unadjusted ......... -.13 - .03 .79*** 5. Control ............. -.32*** -.24** .00 .61**

Mirth

6. Adjusted ............ .27** .35*** .47*** .29** -.13 7. Unadjusted ......... .27** .40*** .48*** .33*** - .07 .90*** 8. Control ............. .06 .20* .13 .17* .12 .00 .45***

NOTE.-N = 93. For humor appreciation (ratings and mirth), "adjusted" scores are corrected for response to neutral (control) cartoons.

* p < .05. ** p < .01. *** p < .001.

high-.78 and .81 for funniness ratings and mirth, respectively. Given the better reliabil- ity of the composite scores, these scores were combined. The two humor-production tasks (Captions and Titles) were also combined to form a single, more reliable scale. A number of the items on each task had correlations with items on the other task as high or higher than intratask correlations.

Intercorrelations among the humor scores are shown in Table 1. Humor produc- tion was most strongly related to comprehen- sion, and the two aspects of appreciation showed a similar but lower relation. Fun- niness ratings and mirth were more strongly related to each other after each was adjusted for responsiveness to control stimuli.

Relations of Humor to Sex, Age, SES, and IQ

To test for the importance of sex, age, SES, and IQ differences in humor-task perfor- mance, a series of hierarchical regression analyses (Cohen & Cohen, 1975) were con- ducted, each treating humor as the dependent variable. The four subject variables were treated as independent measures and entered in hierarchical fashion in the following order: (1) sex, (2) age, (3) SES, and (4) IQ. Sex and age were not affected by other variables; they were therefore entered first. Socioeconomic status preceded IQ in order better to separate the unique variance explained by IQ after shared SES factors were already accounted for. Two interaction effects of interest, sex x age and sex x IQ, were tested last since the component variables must first be partialed to

determine if an interaction adds significant unique variance to the multiple R2. The hier- archical analysis consisted of a series of simul- taneous regression equations, beginning with one independent variable, sex. The next re- gression equation included sex and age, the next sex, age, and SES, and so on.

Results presented in Table 2 indicated that boys expressed more mirth than girls (X = 2.15 for boys and 1.83 for girls). When the frequencies of each type of mirth response were analyzed by sex (one-way ANOVAs for sex differences), results indicated that girls produced more neutral or negative (score of 1) responses, F(1,91) = 5.60, p < .05, that boys expressed more laughter, F(1,91) = 5.53, p < .05, and that there were no sex differences in the other two scores (smiling).

Mirth was also positively associated with higher SES level and IQ. In order to explore the role of IQ and comprehension in these relations, several additional analyses were conducted. When results were considered for only set A (easy) cartoons (when comprehen- sion as a factor should be reduced), SES and IQ still showed very similar relations to mirth. When IQ was treated as the dependent vari- able in a separate regression analysis, with humor comprehension and corrected mirth scores as "predictors," comprehension alone increased R2 by .30 (p < .001) and mirth con- tributed an additional 6% to R2 (p < .05) be- yond that accounted for by comprehension. Similarly, when mirth (corrected) was treated as the dependent variable, IQ alone as a pre- dictor accounted for 15% of the variance in

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 467

TABLE 2

HIERARCHICAL MULTIPLE REGRESSION ANALYSES OF INDIVIDUAL DIFFERENCES IN HUMOR RESPONSE: R2 CHANGE DUE TO SEX, AGE, IQ, AND INTERACTION EFFECTS

DEPENDENT VARIABLES

Funniness INDEPENDENT VARIABLES Ratings Mirth Comprehension Production

Control .............. ........... .41*** .20*** Sex ............................00.

.0.8**- .00 .01

Age ............................ .03*- .00 .11"**+ .05* + SES ............................ .00 .08**+ .02 .00 IQ

............................. .00 .07**+ .27*** + .25***+

Sex x age ....................... .00 .00 .02 .00 Sex x IQ .................... .00 .00 .01 .00

Overall R2 ...................... .45*** .42*** .43*** .31***

NOTE.-Each column represents a separate regression analysis for the humor score heading the column. Each regression was carried out hierarchically with each of the six (or seven) steps added in sequence. The resulting increment in R2 due to the addition of each step is indicated in the table along with the significance of the increment (F test) and the direction of the effect. Sex was coded 0 = boy, 1 = girl; thus a negative effect indicates that boys scored higher. N = 87.

* p < .05. ** p < .01. *** p < .001.

mirth (p < .001). When SES was added as a predictor, SES provided a significant increase in R2 of .06 (p < .05) over and above the IQ effect, despite its substantial overlap with IQ variance. Thus the relation of mirth with SES was not due only to IQ variance, nor was that of mirth with IQ entirely due to varying com- prehension levels.

Funniness ratings declined slightly but significantly with age. Both humor compre- hension and production were greater with in- creasing age and intellectual ability.

Relation of Humor to Competence Before the humor measures were ana-

lyzed in relation to the competence measures, each humor score was standardized within grade level so the scores were more compara- ble to the competence scores, all of which were standardized by classroom or age (see Procedure above).

The correlations between the humor measures and the competence measures are presented in Table 3. Socioeconomic status and IQ scores are also included. Given the sex difference in the mirth scores, these cor- relations are presented separately by sex. It should also be noted that the sample size is smaller for the peer and teacher ratings. Com- parable data were not available for children who had moved to different schools, includ- ing the oldest grade cohort, who had moved on to seventh grade at the time of these as- sessments.

Generally, humor showed a positive rela- tion to competence, although it showed little relation to disruptiveness as judged by peers or teachers. Humor production and com- prehension had somewhat stronger relations to the IQ and achievement measures than did humor appreciation.

In order to discover whether mirth was related to competence beyond the possible role of comprehension, eight regression anal- yses were conducted with each competence indicator as the criterion and two humor pre- dictors. Comprehension was entered first as a control variable, followed by mirth. This hier- archical analysis revealed that, after com- prehension was taken into account, mirth ac- counted for very little unique variance, except in the case of cooperative-initiating behavior, which was significantly related to mirth after comprehension was partialed out.

Overall, the results in Table 3 raised the question of how general intellectual ability as traditionally represented by the IQ score might be affecting the correlations between humor and competence. It was important to consider to what extent the various aspects of humor were uniquely associated with compe- tence, above and beyond any relation shared with the individual differences in sex, SES, and, especially, IQ. Another question con- cerned the relative strength of association be- tween different aspects of humor and com- petence, particularly when other individual differences were controlled.

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

z 0

V~C) 0 00 m (7)Okfl

114 1 C~~ o Iq I4 I Co C4) q

p 4

z 0

z t0 - t- 10 0 (7 CM C)

z

e' arLnco ) c~ca' (C r3 c3 a

GO

z O

0 * *0

r zI I : I

000-*

m CA

4) o

- 0 04CC1

z0 AC

0 a~OO~4-0 -,-o .0 > I1

w4-1

0 Q

.d In

0 -0

0 0C. ,-( 0

rA -e r A -d- 0 CL C-)

=C 0Z .Z.

I 3

L rj; LrH

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 469

To address these questions, a series of regression analyses were carried out in which each competence measure served as a depen- dent variable. As a first test of the association between the eight "dependent" and seven "independent" measures, the first canonical correlation was calculated. The resulting cor- relation of .88 (p < .001) suggested that non- random relations existed between the two sets of variables. Next, hierarchical regression was conducted for each aspect of competence with the following steps (independent vari- ables): (1) sex; (2) SES; (3) IQ; and (4-7) four humor scores, ordered by contribution to the multiple correlation. The four humor vari- ables were entered in order of contribution because there was no compelling rationale for preselecting the order. Thus the "strongest" humor variable was allowed to take prece- dence. Results of these analyses are pre- sented in Table 4. Then, in order to compare the predictive strength of humor versus IQ, a second set of regressions was conducted mov- ing IQ to step 7 so that all humor variables were entered before IQ. Salient results from this set of analyses will be described below.

As expected, general intellectual ability, as measured by the IQ score, appeared to play a substantial role in the assessed areas of com- petence, whether teachers, peers, or objective tests were the source of measurement. In sev- eral instances, however, humor showed a unique relation to competence above its shared variance with IQ.

Humor production and appreciation rat- ings were significantly related to sensitive- isolated peer reputation even after IQ was ac- counted for. When, in the second set of analyses, humor preceded IQ in the analysis, humor production alone accounted for as much variance as IQ did at step 3, an incre- ment in R2 of .24 (p < .001). Thus IQ and humor production related quite strongly to this aspect of peer reputation, together ac- counting for about 32% of the variance in this score, about half of which is shared variance plus a significant unique contribution of each type of measure with respect to the other. Set- ting aside IQ, humor production and appreci- ation (ratings) accounted together for 33% of the variance in the isolated peer reputation score. For positive reputation, these two as- pects of humor can account for 16% of the variance if they precede IQ, leaving IQ (as step 7) accounting for 5% unique variance.

Humor comprehension made a small contribution to the prediction of achievement even after the overwhelming effect of IQ was accounted for. When comprehension was en-

tered before IQ (in step 3), it increased R2 by .30 (p < .001), and IQ subsequently added .20. Thus the IQ score showed a stronger rela- tion to achievement.

Affective expressiveness in response to humor, indexed by the mirth score, had a significant, positive relation to performance anxiety. Interestingly, this occurred only when IQ was controlled. Thus while simple correlations showed a positive relation be- tween IQ and mirth, a negative relation be- tween IQ and performance anxiety, and very little relation between mirth and performance anxiety, one would expect that, for children with the same level of intellectual ability, mirth would be positively related to perfor- mance anxiety.

As can be seen in Table 4, humor scores did not make a unique contribution to the prediction of the teacher ratings of coopera- tive-initiating behavior and poor comprehen- sion-disattention when IQ was entered first in step 3. When humor preceded IQ, however, humor production, the strongest predictor of the humor scores, increased R2 by .15 (p < .01) for cooperative-initiating behavior (a positive relation) and by .14 (p < .01) for poor comprehension-disattention (a negative rela- tion). However, most of these effects were shared with IQ. When the humor variables preceded IQ in the second set of regressions, humor production was the strongest predictor of the humor variables in five of the eight analyses.

Discussion

Results from this study support the hy- pothesis, derived from the developmental lit- erature as well as from theories of humor, that humor and competence are positively related. Better humor production, comprehension, and greater mirth were associated with aca- demic and social competence. Children who expressed these humor abilities were viewed by their teachers as more effectively engaged in the classroom and more attentive, coopera- tive, responsive, and productive. Their peers viewed them as more popular, gregarious, and happy and as leaders with good ideas for things to do. Neither humor nor IQ was re- lated to disruptive-aggressive behavior as judged by teachers or peers.

As expected, intellectual ability was moderately related to humor, except in the case of humor-appreciation ratings. Mirth, comprehension, and production were signifi- cantly related to IQ even when the effects of sex, age, and SES were controlled. The rela-

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

--4 ?CCl 0 0 0 I +

in, 4- 0C C,,cr I3;

vv v i~ 0 o~.9i 1*

a, c" c'1 oo o cr 2

Cn 00

U 09 n

Hjc +

Z 0j7 A 4-j

(Ux

Cis 0

CIO t - M,0 c

00OD0 --

H

oac~ol1 l Q

COC.

P. . It 14

bO CID 0 O Oq 4-0

z P-0 0

C H

0-4 cl,

zn P

* *

zc- Cq 10 CIO 00 ---4 ? rc

cn- 009 999 9 C'J3

0 0 Cq coo 0 *

z (L)"O

0X

0 C.)C4-0

CO

O tn a II (L) t:4 4-JO -

e~ 8~

E b(L) -4 W54 E

M., m -Jl~ ,,, , 4-J C3 a 000 ~o 00 L1 z >14-JQ

z

Pd r. C4

r. cis bj c I I q z V. Q rl) rn r i"

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 471

tively strong relation of these aspects of humor to IQ raised the question of whether humor was related to competence in this study primarily because both are cognitively mediated rather than because humor and competence influence each other. This ques- tion was addressed by regression analyses in which IQ variance could be statistically con- trolled.

Results suggest that humor is related to achievement and the teacher's view of effec- tive classroom participation primarily because all these abilities are related to general intel- lectual abilities. However, humor was not ex- pected to have a functional relationship with academic competence. Rather, it was as- sumed that common cognitive characteristics might underlie the development of humor as well as successful adaptation in the class- room. Humor comprehension may have had a small but unique relation to achievement (be- yond IQ) because the task resembled a verbal achievement test.

Although humor is not likely to influence academic competence directly, in the case of peer relations humor may well influence or be affected by peer reputation. In the case of positive peer reputation, results do not sup- port a functional relation between humor and peer reputation but suggest again that under- lying general cognitive abilities account for the relation. In an academic setting, positive peer regard and leadership may be more re- lated to cognitive abilities than would be the case in other peer groups outside the context of school.

Results for the sensitive-isolated dimen- sion of peer reputation, on the other hand, are consistent with the possibility of a functional relation between a lack of humor behaviors and social isolation. Reputation as withdrawn, quiet, shy, or unhappy appears to be substan- tially related to humor production and appre- ciation judgments even when IQ is con- trolled, suggesting that these relations are not due entirely to cognitive mediation. Humor production and appreciation may reflect an "engaged" quality, involving the child's so- cial awareness and responsiveness.

One cannot determine from these results whether humor may be subdued in such chil- dren because of their social isolation, whether their lack of humor contributes to their social reputations, or both. Peer rejection could af- fect a child's mood and outlook in a way that would dampen humor appreciation and cre- ation. One way to address this issue would be the research strategy employed recently by

Coie, Dodge, and their colleagues (Coie & Kupersmidt, 1983; Dodge, 1983) to study the development of peer status over time in newly formed groups. For example, one could ascertain whether humor measured before group formation is predictive of peer status or reputation that emerges over time and also observe humor behavior during group in- teraction. Greater understanding of the rela- tion between humor and peer reputation might also be obtained from studies of the social facilitation of humor appreciation (e.g., Chapman et al., 1980). One would expect children of varying peer reputations to re- spond differently to experimental manipula- tions of the social context for humor.

Finally, it would be interesting to exam- ine the relations between social cognition and humor. Pellegrini (1985) has shown that social problem-solving abilities and interpersonal awareness are related to peer reputation in middle childhood. His results, together with those of this study, suggest that humor may be related to peer reputation in part because of underlying social-cognitive abilities. The ap- preciation and creation of most humor re- quires social "know-how," in terms either of getting or creating the joke or of sharing the humor with others appropriately. Hence the level of social awareness, knowledge, and sensitivity would be expected to mediate humor behaviors as well as other social activi- ties that contribute to peer reputation.

Mirth, interestingly, did not relate to so- cial competence once IQ was controlled. The affective expression of appreciation appears to have more in common with general intellec- tual ability than does funniness as perceived and rated by the child. Mirth was associated with IQ even for easy cartoons or when com- prehension as a factor was partialed out. More intelligent children may be more sensitive to subtle elements of cartoon humor not cap- tured by simple ratings of difficulty or com- prehension level. The humor of the cartoons could be enhanced for them by their greater social understanding, for example, or by an appreciation for irony or other subtle sources of humor. Another possibility is that brighter children tend to be more motivated and re- sponsive, particularly in "performance" situa- tions. They may have had more mastery expe- riences or more mastery motivation and consequently have come to express more pleasure in cognitive tasks.

Two other findings are consistent with the possibility that there is a motivational fac- tor being tapped by mirth. First, with IQ con- trolled, mirth was substantially associated

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

472 Child Development

with greater performance anxiety. Second, even after comprehension was partialed out, mirth continued to show a positive relation with cooperative-initiating behaviors in the classroom (such as offering to do things for the teacher and spontaneously contributing to classroom discussions).

Future studies might direct more atten- tion to the exact nature of the relation be- tween mastery motivation and humor. For ex- ample, one might compare patterns of affec- tive response accompanying different tasks such as humor appreciation, humor produc- tion, and nonhumorous cognitive tasks (e.g., anagrams, as used by Harter, 1974). Careful observations might lead to the differentiation of cognitive mastery from other sources of smiling and laughter. Mastery responding could then be more clearly examined with re- spect to competence.

In conclusion, the results suggest that humor may be related to competence in three ways: (1) through the manifestation of intel- lectual ability both in competent functioning at school and in humor comprehension, ap- preciation, and production; (2) through the role of mastery motivation in enhancing humor responsiveness and skills as well as competence at school; and (3) through peer relations, either by humor behaviors influenc- ing peer reputation or being influenced by peer relations or, most likely, by both.

In an academic context, many of the rela- tions between humor and competence appear to be mediated by cognitive abilities rather than the result of reciprocal influences among humor behaviors and different aspects of competence. Nevertheless, the exceptions suggest important areas for further research, particularly in regard to peer relations and mastery motivation. Moreover, results for dif- ferent humor behaviors (such as clowning or spontaneous joking) and for competencies ex- hibited and measured outside the school con- text may provide greater evidence of relations resulting from the hypothesized functions of humor or from social-emotional development.

There appear to be multiple pathways by which humor and competence become linked in the course of development: humor and competence may influence each other over time; they may both satisfy the competence motive; and both appear to be mediated by social, emotional, and cognitive development. These processes need to be examined more thoroughly across age. It will be important to consider the relations of humor to compe- tence as manifested in diverse social activi-

ties, family relationships, or friendships and in terms of emotional adjustment or psycho- logical well-being. Such research may pro- vide insight into the development of social competence and mastery as well as humor it- self. It would also provide an important foun- dation for the empirical investigation of the notion that humor is an adaptive coping strat- egy-a modifier of stress for both children and adults.

References

Bower, E. M. (1969). Early identification of emo- tionally handicapped children in school (2d ed.). Springfield, IL: Thomas.

Brodzinsky, D. M. (1975). The role of conceptual tempo and stimulus characteristics in chil- dren's humor development. Developmental Psychology, 11, 843-850.

Brodzinsky, D. M. (1977). Children's comprehen- sion and appreciation of verbal jokes in relation to conceptual tempo. Child Development, 48, 960-967.

Brodzinsky, D. M., & Rightmyer, J. (1980). Individ- ual differences in children's humour develop- ment. In P. E. McGhee & A. J. Chapman (Eds.), Children's humour (pp. 181-212). Chichester: Wiley.

Chapman, A. J., Smith, J. R., & Foot, H. C. (1980). Humour, laughter, and social interaction. In P. E. McGhee & A. J. Chapman (Eds.), Children's humour (pp. 141-179). Chichester: Wiley.

Coie, J. D., & Kupersmidt, J. B. (1983). A behav- ioral analysis of emerging social status in boys' groups. Child Development, 54, 1400-1416.

Cohen, J., & Cohen, P. (1975). Applied multiple regression/correlation analysis for the behav- ioral sciences. Hillsdale, NJ: Erlbaum.

Cowen, E. L., Pederson, A., Babigian, H., Izzo, L. D., & Trost, M. A. (1973). Long-term follow- up of early detected vulnerable children. Journal of Consulting and Clinical Psychol- ogy, 41, 438-446.

Damico, S. B., & Purkey, W. W. (1978). Class clowns: A study of middle school students. American Educational Research Journal, 15, 391-398.

Dodge, K. A. (1983). Behavioral antecedents of peer social status. Child Development, 54, 1386-1399.

Dunn, L. M., & Markwardt, F. C. (1970). Peabody Individual Achievement Test. Circle Pines, MN: American Guidance Service.

Finkelman, D., & Ferrarese, M. (1981). The mea- surement of classroom behavioral competence: The Devereux teacher ratings (Tech. Rep. No. 3). Minneapolis: University of Minnesota, Proj- ect Competence.

Garmezy, N., Masten, A. S., & Tellegen, A. (1984).

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions

Ann S. Masten 473

The study of stress and competence in chil- dren: A building block for developmental psychopathology. Child Development, 55, 97- 111.

Garmezy, N., & Tellegen, A. (1984). Studies of stress-resistant children: Methods, variables and preliminary findings. In F. Morrison, D. Keating, & C. Lord (Eds.), Applied develop- mental psychology (Vol. 1, pp. 231-287). New York: Academic Press.

Gesell, A., Ilg, F. L., & Ames, L. B. (1946). The child from five to ten. New York: Harpers.

Goldstein, J. H., & McGhee, P. E. (Eds.). (1972). The psychology of humor: Theoretical per- spectives and empirical issues. New York: Aca- demic Press.

Harter, S. (1974). Pleasure derived by children from cognitive challenge and mastery. Child Devel- opment, 45, 661-669.

Hauser, R. M., & Featherman, D. L. (1977). The process of stratification: Trends and analysis. New York: Academic Press.

Keith-Spiegel, P. (1972). Early conceptions of humor: Varieties and issues. In J. H. Goldstein & P. E. McGhee (Eds.), The psychology of humor: Theoretical perspectives and empirical issues (pp. 3-39). New York: Academic Press.

Kendall, P. C., Pellegrini, D. S., & Urbain, E. S. (1981). Approaches to assessment for cogni- tive-behavioral interventions with children. In P. C. Kendall & S. D. Hollon, Assessment strategies for cognitive-behavioral interven- tions (pp. 227-285). New York: Academic Press.

Lambert, N. M., Hartsough, C. S., & Zimmerman, I. L. (1976). The comparative predictive effi- ciency of intellectual and nonintellectual com- ponents of high school functioning. American Journal of Orthopsychiatry, 46, 109-122.

Martin, R. A., & Lefcourt, H. M. (1983). Sense of humor as a moderator of the relation between stressors and moods. Journal of Personality and Social Psychology, 45, 1313-1324.

Masten, A. S. (1982). Humor and creative thinking in stress-resistant children. Dissertation Ab- stracts International, 43, 3737B. (University Microfilms No. 83-08,093)

Masten, A. S., Morison, P., & Pellegrini, D. S. (1985). A revised class play method of peer as- sessment. Developmental Psychology, 21, 523- 533.

McClelland, D. (1973). Testing for competence rather than for "intelligence." American Psy- chologist, 28, 1-14.

McGhee, P. E. (1974). Cognitive mastery and chil-

dren's humor. Psychological Bulletin, 81, 721- 730.

McGhee, P. E. (1976). Children's appreciation of humor: A test of the cognitive congruency principle. Child Development, 47, 420-426.

McGhee, P. E., & Chapman, A. J. (Eds.). (1980). Children's humour. Chichester: Wiley.

Murphy, L. B., & Moriarty, A. E. (1976). Vulnera- bility, coping, and growth: From infancy to adolescence. New Haven, CT: Yale University Press.

Pellegrini, D. S. (1985). Social cognition and com- petence in middle childhood. Child Develop- ment, 56, 253-264.

Pien, D., & Rothbart, M. K. (1980). Incongruity humor, play, and self-regulation of arousal in young children. In P. E. McGhee & A. J. Chap- man (Eds.), Children's humour (pp. 1-26). Chichester: Wiley.

Ransohoff, R. (1975). Some observations on humor and laughter in young adolescent girls. Journa, of Youth and Adolescence, 4, 155-170.

Silverstein, A. B. (1975). Validity of WISC-R short forms. Journal of Clinical Psychology, 31, 696- 697.

Spivack, G., & Swift, M. (1967). Devereux Elemen- tary School Behavior Rating Scale manual. Devon: Devereux.

Sroufe, L. A., & Waters, E. (1976). The ontogenesis of smiling and laughter: A perspective on the organization of development in infancy. Psy- chological Review, 83, 173-189.

White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Review, 66, 297-333.

Wilson, T. (1977) The Ziggy treasury. Kansas City: Andrews & McMeel.

Wolfenstein, M. (1954). Children's humor: A psy- chological analysis. Glencoe, IL: Free Press.

Yarrow, L. J., McQuiston, S., MacTurk, R. H., McCarthy, M. E., Klein, R. P., & Vietze, P. M. (1983). Assessment of mastery motivation dur- ing the first year of life: Contemporaneous and cross-age relationships. Development Psychol- ogy, 19, 159-171.

Zigler, E., Levine, J., & Gould, L. (1966). Cognitive processes in the development of children's ap- preciation of humor. Child Development, 37, 508-518.

Zigler, E., Levine, J., & Gould, L. (1967). Cognitive challenge as a factor in children's humor appre- ciation. Journal of Personality and Social Psy- chology, 6, 332-336.

Ziv, A. (1984). Personality and sense of humor. New York: Springer.

This content downloaded from 63.138.247.2 on Tue, 18 Nov 2014 10:00:43 AM All use subject to JSTOR Terms and Conditions