Bond UniversityePublications@bond

Humanities & Social Sciences papers Faculty of Humanities and Social Sciences

10-1-2002

Relationship of humour to health: A psychometricinvestigationGregory J. BoyleBond University, Gregory_Boyle@bond.edu.au

Jeanne M. Joss-ReidBond University

Follow this and additional works at: http://epublications.bond.edu.au/hss_pubs

Part of the Psychology Commons

This Journal Article is brought to you by the Faculty of Humanities and Social Sciences at ePublications@bond. It has been accepted for inclusion inHumanities & Social Sciences papers by an authorized administrator of ePublications@bond. For more information, please contact Bond University'sRepository Coordinator.

Recommended CitationGregory J. Boyle and Jeanne M. Joss-Reid. (2002) "Relationship of humour to health: Apsychometric investigation" ,, .

http://epublications.bond.edu.au/hss_pubs/27

Relationship of humour to health: A psychometric investigation

Gregory J. Boyle1,2

and Jeanne M. Joss-Reid1

1Department of Psychology, Bond University, Australia

2Department of Psychiatry, University of Queensland, Australia

The effects of humour on health were investigated using a sample of 504 individuals comprising three groups (community group, university students, and respondents with a medical condition). Hypotheses were:

1. that after controlling for other variables, humour would be significantly associated with health:

2. that individuals with a greater sense of humour would report significantly higher levels of good health as compared with those with less humour; and

3. that the assessment of the factor structure of the Multidimensional Sense of Humour Scale (MSHS) would support its construct validity in the Australian context.

The present findings supported the view that a sense of humour is associated with health, and also provided support for the validity and reliability of the MSHS instrument.

There is considerable evidence that a good sense of humour plays a positive role in physical health and well-being (Berk, 2003; Carroll, 1990; Carroll & Schmidt, 1992; Dean, 1997; Johnston, 1990; Kamei, Kumano, & Masumura, 1997; Lefcourt, Davidson-Katz, & Kueneman, 1990; MacHovec, 1991; McClelland & Cheriff, 1997; McGhee, 1999; Martin, 2001; Martin & Dobbin, 1988; Porterfield, 1987; Provine, 2000; Richman, 1995; Rotton, 1992; Solomon, 1996). Nevertheless, although humour is a multidimensional construct (Svebak, 1996; Thorson & Powell, 1993b), many studies have used unidimensional instruments to measure humour, thereby limiting their findings with respect to the role of a sense of humour on health. For example, Thorson and Powell (1991) conducted principal components analyses on several frequently used humour instruments and concluded that use of any of the existing instruments alone would provide an inadequate measure of the sense of humour construct. Thorson and Powell (1993a) subsequently developed the Multidimensional Sense of Humour Scale (MSHS). This scale assesses four dimensions of sense of humour: humour production, humour appreciation, the use of humour as a coping mechanism, and attitudes towards humour. Likewise, the World Health Organization (see Bowling, 1997) has suggested that health is a multidimensional construct comprising physical health, psychological health and social health, the dimensions of which interact with one another (Brannon & Feist, 1997). The RAND 36-Item Health Survey (Hays, Sherbourne, & Mazel, 1993), the instrument used in the current study, assesses eight dimensions of health and includes those suggested by the WHO and by Fitzpatrick et al. (1992). Health is affected by a number of demographic variables and lifestyle correlates (Argyle, 1997). Variables such as morbidity (Stewart et al., 1989), age (Brannon & Feist, 1997), social relationships (Berkman & Syme, 1979; House, Robbins, & Metzner, 1982; Penninx et al., 1997), occupational status (Kessler, Turner, & House, 1998), smoking (Rosenbaum, Sterling, & Weinkam, 1998; Surgeon General, 1990), drinking (Klatsky & Friedman, 1995; Power, Rodgers, & Hope, 1998) and exercise (Fontane, 1996) might either detract from or enhance the effects on health of a sense of humour. The aim of the present study was to overcome the limitations of previous studies by using a large, diverse, community-based sample to investigate the relationship between humour and health (many studies have utilized small samples of university students) ( Julious, George, & Campbell, 1995) to investigate the relationship between humour and health, and thus have limited the generalizability (Martin, Kuiper, Olinger, & Dance, 1993; Thorson & Powell, 1996). It was hypothesized (H1) that after

controlling for other variables, humour would contribute significantly to health. Secondly, if individuals do differ in humour levels, then those who have a greater sense of humour might also be healthier. Therefore, it was hypothesized (H2) that individuals high in humour, as compared with those low in humour (as measured by the MSHS), would report significantly better health. Thirdly, an item-factor analysis was undertaken to investigate the purported multidimensionality of Thorson and Powell’s (1993a) MSHS instrument. Individuals low in humour levels might be similar to what Eysenck (1991) termed ‘the disease-prone personality’ (p. 54); in other words, individuals low in sense of humour may be more vulnerable to ill health. If Fry’s (1994) assertion that humour does perform a protective function (like some sort of psychological immune system) is correct, then the establishment of an empirical relationship could help justify the promotion of humour in various health-related environments such as hospitals, hospices and counselling rooms.

Instruments The questionnaires comprised eight demographic questions, 36 questions from the short-form version of the RAND 36-Item Health Survey (Hays et al., 1993)—measuring health-related quality of life (HRQL)—and 24 questions from the MSHS (Thorson & Powell, 1993a). RAND 36

The RAND 36 comprises 36 items representing eight dimensions of HRQL: physical functioning (10 items), role limitations due to physical health (four items), role limitations due to emotional problems (three items), energy/fatigue (four items), emotional well-being (five items), social functioning (two items), pain (two items), general health (five items) and health change (1 item). Scores for all dimensions are measured on a scale of 0 to 100, a higher score indicating a better state of well-being. The RAND 36 is identical to the Medical Outcomes Study 36-Item Short-Form Health Survey (MOS SF-36; Ware & Sherbourne, 1992). Therefore, tests of the reliability and validity of the SF-36 are applicable to the RAND 36 and vice versa (Bowling, 1997; Hays et al., 1993).

Population norms for the SF-36 in Australia have been developed by the Australian Bureau of Statistics (1995). According to Anderson, Aaronson, and Wilkin (1993), the validity of the subscales of the SF-36 has largely been established against clinically defined criterion groups and the proven validity of the full version. McHorney, Ware, Rogers, Raczek, and Lu (1992) reported that the SF-36 is able to discriminate between relatively healthy individuals and those with a chronic medical condition. Hays, Wells, Sherbourne, Rogers, and Spritzer (1995) compared the functioning and well-being of patients with depression to patients with chronic physical illnesses, and reported that the SF-36 was able to discriminate well between the two groups.

Van der Zee, Sanderman, and Heyink (1996) assessed the concurrent validity of the RAND 36 subscales by correlating them with measures of physical health such as the Groninger Activity Restriction Scale (GARS) and the General Health Questionnaire (GHQ). Van der Zee et al. reported adequate correlations of .65 and .76, respectively. The construct validity of the SF-36 was assessed by Lewin-Epstein, Sagiv-Schifter, Shabtai, and Shmueli (1998) using confirmatory factor analysis; the eight-factor structure of the scale was supported, giving a goodness-of-fit (GFI) index of .81, and a root mean square error of approximation (RMSEA) of .08.

In regard to item homogeneity, Hays et al. (1993), Lewin-Epstein et al. (1998), and Van der Zee et al. (1996) reported Cronbach alpha coefficients ranging from .71 to .93 for the RAND 36 (in the present study, alpha coefficients across the three groups ranged from .70 to .91). However, very high alpha coefficients may result from item redundancy rather than internal consistency (see Boyle, 1991). The test–retest reliability of the RAND 36 was assessed by Van der Zee et al. Correlations after a two-month interval ranged from .58 for social functioning to .82 for physical functioning. MSHS

Sense of humour was assessed using the MSHS, which comprises 24 items that assess four separate dimensions of humour: humour generation (11 items), use of humour as a coping mechanism (seven items), appreciation of humour (two items) and attitudes towards humour and humorous persons (four items). The items are scored on a Likert-type scale ranging from 0 = strongly

disagree to 4 = strongly agree. Scores range from 0 to 96, with higher scores suggesting a greater sense of humour.

Thorson and Powell (1993a), as well as Thorson, Brdar, and Powell (1997) used Kaiser’s crude ‘Little Jiffy’ factor analytic approach in constructing the MSHS, thereby producing less than optimal factor solutions. They extracted principal components followed by orthogonal rotation. However, principal components analysis combined with orthogonal rotation leads to inflated factor loadings, and a less than optimal factor solution (Boyle, 1988; Boyle, Stankov, & Cattell, 1995; Cattell, 1978; Child, 1990; Gorsuch, 1983, 1990; McDonald, 1985). Moreover, the dimensions of humour are not likely to be independent of one another.

Cann, Holt, and Calhoun (1999) assessed the relationship between the MSHS and several other measures of sense of humour: the SHQ (Svebak, 1996), the SHRQ (Martin & Lefcourt, 1984) and the CHS (Martin & Lefcourt, 1983). Cann et al. reported that only two of the subscales of the MSHS (humour production and coping humour) showed a positive relationship with the other humour measures.

Consequently, the present study not only investigates further the relationship between humour and health, but also aims to ascertain the factor structure of the MSHS instrument, using methodologically sound factor analytic procedures.

Method Participants

Preliminary analyses began with 504 participants from the Brisbane and Gold Coast areas in Queensland. The sample was divided into three groups and analysed separately to avoid possible confounding of results. Altogether, there were 306 females and 198 males, ages ranging from 18 to 65 years plus, the majority of whom were married, employed, light drinkers, non-smokers, and moderate exercisers.

Participants who agreed voluntarily to take part in the study were given a package consisting of the demographic questions together with the two questionnaires, as well as an explanatory letter outlining the nature of the study, and a reply paid envelope. The first group consisted of 300 respondents who were not students and had no medical condition (community group). The second group consisted of 103 non-students with a medical condition (medical group). The third group consisted of 101 students with and without a medical condition (student group). Table 1 provides a description of the community group, medical group, and student group participants by gender, age, marital status, occupational status, alcohol consumption, tobacco usage, exercise habits and medical condition. Methodology

To answer the first hypothesis (H1) of whether humour contributes to HRQL, hierarchical multiple regression analyses were conducted with the score on each dimension of the RAND 36 as the dependent variable. The independent variables of age, marital status, occupational status, alcohol consumption, tobacco usage, exercise habits and medical condition (for the student group only) were entered into the regression equation together as Block 1. Humour scores from MSHS were subsequently entered into the regression equation as Block 2.

The second hypothesis (H2), concerning the difference in health between those high and low in sense of humour, was answered using a one-way, between groups MANOVA. A conservative median split was used to divide the high and low humour individual. These two groups comprised the two levels of the independent humour variable. The eight subscale scores of the RAND 36 were the dependent variables.

The dimensionality of the MSHS instrument was also assessed using a maximum likelihood (ML) factor analytic procedure, together with direct oblimin simple structure rotation. As found in previous research (Thorson et al., 1997; Thorson & Powell, 1993a, 1993b, 1997), the scree test indicated that at least four factors should be extracted, thereby attesting to the multidimensionality of the humour construct.

Table 1. Sample description Community

Group (n =300)

Medical Group

(n=103) Students (n=101)

Gender Female 178 63 65 Male 122 40 36

Age 18–24 17 8 80 25-34 92 12 17 35–44 58 18 1 45-54 53 20 3 55-64 34 24 — 65+ 11 21 —

Marital status Partner 220 77 23 No Partner 80 26 78

Occupational status Employed 249 62 — Unemployed 21 11 — Student — — 101 Retired 27 29 —

Alcohol consumption 0–2 drinks per day 208 75 58 3–5 drinks per day or non-drinker 85 27 39 6+ drinks per day 7 1 4

Tobacco usage Never smoked 188 52 63 Former smoker 49 31 14 Less than 20 cigarettes per day or pipes/cigars 42 14 21 Equal to or more than 20 cigarettes per day 21 6 3

Exercise habits No exercise 63 33 28 Exercise once per week 49 14 16 Exercise 2–3 days per week 87 28 23 Exercise 4–5 days per week 55 21 20 Exercise once per day 46 13 14

Medical condition — 103 21

Results Cronbach alpha coefficients obtained for the MSHS instrument in the present study were .92 for the normal group, .93 for the medical group and .91 for the student group. Test–retest reliability of the MSHS was assessed using a sample of 36 undergraduate students who completed the test twice within a four-week interval. A test–retest coefficient of .83 was obtained, suggesting satisfactory reliability of the MSHS instrument.

Preliminary comparisons of means were made across the three groups. In each comparison, a one-way, between-groups MANOVA was used with respect to the eight subscales of health, and an independent groups t test was used to compare means with respect to the variable humour (collapsed

across the four MSHS factors which were measured as a single variable). There was no significant gender difference with respect to health, and none with respect to humour; therefore, males and females were analysed together (cf. Abel, 1998). A comparison of means between respondents with a medical condition (n = 124) and the rest of the sample (n = 380) revealed a significant multivariate difference on the health measure, F(8,495) = 13.23, p < .001. Likewise, a comparison of means between students (n = 101) and the rest of the sample (n = 403) revealed a significant multivariate difference on the health measure, F(8,495) = 7.39, p < .001. Six of the eight subscales of the health scale were responsible for this difference in both instances.

After removing all respondents with a medical condition from the sample, a further comparison of means between students (n = 80) and the rest of the sample (n = 300) revealed a significant multivariate difference on the health measure, F(8,371) = 6.47, p < .001. Again, six of the eight subscales of the health scale were responsible for this difference. However, there was a significant difference on humour, t(99) = 2.9, p < .01. The 21 students with a medical condition were then compared to the 103 non-students with a medical condition. There was a significant multivariate difference on health, F(8,115) = 3.72, p < .01. Two of the eight subscales of the health scale were responsible for this difference. There was also a significant difference with respect to humour, t(122) =2.83, p < .01. Therefore, the 21 students with a medical condition were removed from the medical group, but were retained within the student group. Analyses proceeded with the three distinct groups in order to prevent possible confounding of results. Analysis of the community group (n = 300) Correlations between demographic variables, humour and transformed dependent variables (health subscales) were calculated. Using two-tailed tests, significant, positive correlations were identified between humour and four of the dependent variables (health subscales): general health (r = .12, p <.05); role limitations (physical) (r = .11, p < .05); energy/fatigue (r = .19, p < .01); and emotional wellness (r = .15, p < .01). Multiple regression analyses The independent variables (demographic variables [age, marital status, employment, unemployment, retired, alcohol consumption, tobacco usage, exercise habits] and humour) were regressed on each of the eight transformed dependent health subscales. After the demographic variables had been entered (Block 1), humour (Block 2) significantly contributed variance to three of the eight transformed dependent variables.

The regression analysis on role limitations (physical) accounted for 7.4% of the variance, F(9,290) = 2.57, p < .01. Block 1 contributed 6.1% of the variance, Fchange(8,291) = 2.35, p < .01. Humour added 1.3% of the variance, Fchange (1,290) = 4.09, p < .05. The regression on energy/fatigue accounted for 22.2% of the variance, F(9,290) = 8.15, p < .001. Block 1 contributed 16.4% of the variance, Fchange (8.291) = 7.12, p < .001. Humour contributed a further 3.8% of the variance, Fchange (1,290) = 13.89, p< .001. The regression on emotional wellness accounted for 12.6% of the variance, F(9,290) = 4.65, p< .001. Block 1 contributed 9.4% of the variance, Fchange (8,291) = 3.78, p < .001. Humour added 3.2% of the variance, Fchange (1,290) = 10.57, p < .01. MANOVA A one-way, between-groups MANOVA revealed that there was no significant multivariate difference between the low humour group and the high humour group across the transformed dependent variables (health subscales). Examining univariate effects following a non-significant multivariate effect is justified, given the rationale by Huberty and Morris (1989). Univariate F tests revealed that four of the eight transformed dependent variables differed significantly for the high and low humour groups, respectively: general health (M = 4.56, SD = .15 and M = 4.14, SD = .15; F = 3.87, p < .05); role limitations (physical) (M = 1.49, SD = .10 and M = 1.71, SD = .10; F = 4.73, p < .03); energy/fatigue (M = 6.38, SD = .13 and M = 5.93, SD = .13; F = 5.66, p < .02); and emotional wellness (M = 5.05, SD = .13 and M = 4.72, SD = .12; F = 3.43, p < .05). Analysis of the medical group (n=103) Although the multiple regression analysis on pain was non-significant overall, humour contributed a small, but significant 4.2% of the predictive variance, Fchange (1,93) = 4.56, p < .05.

A one-way, between groups MANOVA yielded a significant multivariate difference between the low

and the high humour groups across the transformed dependent variate, F(8,94) = 1.26, p < .05, despite the reduced sample size as compared with the community group; the observed power level was .83. Univariate F tests revealed that only one of the eight transformed variables, emotional wellness, differed significantly across the high and low humour groups, F(101) = 7.68, p = .007; observed power level = .78. The mean for the high humour group (n = 56) was 4.42, (SD = .25), and the mean for the low humour group (n = 47) was 5.37 (SD = .23), for emotional wellness. Analysis of the student group (n =101) Using two-tailed tests, significant, positive correlations were identified between humour and medical condition (r = .28, p < .01), and humour and emotional wellness (r = .28, p < .01). The multiple regression analysis for emotional wellness predicted 19.3% of the variance, F(7,93) = 3.19, p < .01. Block 1 predicted 15.3% of the variance, Fchange (6,94) = 2.84, p < .05. Humour added 4% to the predictive variance, Fchange (6,94) = 4.63, p < .05.

A one-way, between groups MANOVA yielded a significant multivariate difference in mean scores between the low humour group and the high humour group across the transformed dependent variable, F(8,92) = 3.58, p < .01, the observed power level being .98. Univariate F tests revealed that three of the eight transformed dependent variables were significantly different for the high and low humour groups, respectively: energy/ fatigue (M = 7.30, SD = .20 and M = 6.43, SD = .16; F = 11.24, p< .001); emotional wellness (M = 6.40, SD = .23 and M = 5.15, SD = .19; F = 17.79, p < .0001); and social functioning (M = 2.89, SD = .21 and M = 2.34, SD = .17; F = 4.08, p < .046). Factor analysis Factor analysis of the 24-item MSHS was undertaken using all respondents (N = 504). A maximum likelihood (ML) factoring procedure with direct oblimin rotation to simple structure was used to obtain a four-factor simple structure solution that accounted for 49.75% of the variance. A criterion of four factors was used for extraction as the MSHS purportedly measures four factors: humour production, humour appreciation, attitudes towards humour, and coping humour. Also, the scree test (see Fig. 1) indicated at least four factors.

Figure 1. Eigenvalue Plot for Scree Test Criterion. Table 2 presents the obtained factor pattern matrix together with the eigenvalues, and percentage

of variance accounted for by each factor. For ease of interpretation, only factor loadings that attained the 6.32 cut-off suggested by Comrey and Lee (1992, p. 243) are shown. Factor 1 (35.53% of the unrotated variance) was defined by items pertaining to humour production. Factor 2 (8.63% of the unrotated variance) consisted of items related to attitudes towards humour. Factor 3 (2.62% of the

unrotated variance) consisted of items concerned with the production of humour for coping in difficult situations. Factor 4 (2.96% of the unrotated variance) appeared to be related to personal coping. No factor loaded on Item Q14, ‘I appreciated those who generate humour’. (See the Appendix for the full list of items.)

Table 3 presents the inter-correlation matrix between the derived factors. The highest factor correlation was between Factors 1 and 3 (r = .63), reflecting the fact that both factors were concerned with humour production. Table 3. Correlation matrix for MSHS factors

Factor 1 2 3 4

1

2 .29

3 .63 .29

4 -.32 -.47 -29

Discussion H1 predicted that humour would contribute unique variance to health. H1 was not well supported since humour contributed variance to only three of the eight subscales of the health scale for the community group. However, the results were different for both the medical group and the student group, in that humour contributed variance to just one subscale of the health scale in each group; humour contributed to pain in the medical group, and to emotional wellness in the student group. H2 predicted that individuals with a greater sense of humour would report higher levels of good health compared to individuals with a lower sense of humour. Despite the use of a conservative median split, H2 was, nevertheless, also supported to some extent. Although no significant difference existed between the low and high humour individuals in the community group, members of both the medical group and the student group were significantly different in this respect. It was also predicted that the multidimensionality of the MSHS would be supported with the Australian sample. Although the factor analysis produced a four-factor structure, the structure differed from that proposed by the test authors (who had employed less than optimal factor analytic procedures).

Multiple regression analyses were used to ascertain how much of the variance in health could be attributed to humour once the variance contributed by age, marital status, occupational status, alcohol consumption, tobacco usage, exercise habits, and medical condition (students only) had been accounted for. Humour contributed unique variance to three of the health subscales for the community group: role limitations (physical), energy/fatigue and emotional wellness. However, the results for the medical group were different. After controlling for the demographic variables listed above, humour contributed just over 4% of the variance to one of the health subscales, namely pain. As scores decreased, indicating that pain became worse, humour scores increased. The relationship between humour and pain for the medical group was significantly different from the other two groups (p < .05). This finding suggests that humour may have been used as a coping mechanism to relieve pain. This is interesting in view of the fact that it was Cousins’ (1989) experience using humour as a natural analgesic, that accelerated the research into humour’s effects on health. One study that investigated the effects of humour on pain, using individuals with a medical condition, also found that humour was used to ameliorate pain (Rotton & Shats, 1996).

The results for the student group were also different from those of the community group. After controlling for the variance contributed to health by age, marital status, alcohol consumption, tobacco usage, exercise habits and medical condition, humour contributed 4% of the variance to one health subscale, namely emotional wellness. It appears that students’ sense of humour is related to just one variable, emotional wellness, although they reported lower levels of health across five of the eight health subscales when compared to the community group. Deaner and McConatha (1993) suggested students are subject to higher levels of anxiety and depression due to the pressure of study and exams.

Although it had already been ascertained that the 21 students with a medical condition were not answering the health questionnaire differently from the students without a medical condition, the variable ‘medical condition’ contributed to the subscale social functioning. Specifically, students with a medical condition (such as asthma) scored significantly lower on social functioning.

A conservative median split was used to divide the three groups into low and high humour groups, in order to test the hypothesis that persons in high humour groups would report higher levels of good health than those in low humour groups. Use of a median split, rather than comparing the highest and lowest 25–30% of individuals with respect to their MSHS humour scores, enabled a conservative test of the hypotheses, so that any resultant significant effects could be regarded as being real effects. In the community group, univariate F tests revealed that for three of the eight subscales— general health, role limitations (physical) and energy/fatigue—humour differed significantly. The high sense of humour group obtained a lower mean for role limitations (physical) but higher means for general health and energy/fatigue.

There was a significant difference between the high and low humour groups for respondents with a

medical condition. Univariate F tests revealed this difference to be due to emotional wellness (p <.007). Respondents with lower health scores had higher humour scores. This finding was contrary to H2, although it was in accordance with findings reported by Anderson and Arnoult (1989) and Cann et al. (1999) that lower levels of emotional wellness (mental health) are associated with higher levels of humour. A possible explanation of humour increasing in the presence of emotional unwellness is that it is being used as a coping mechanism, as also reported by Martin (1996).

There was also a significant difference between the high and low humour groups within the student group. Univariate F tests revealed this difference to be due to three health subscales. Emotional wellness was the most significant variable separating the high and low sense of humour groups, although unlike the medical group, higher humour was associated with higher scores on emotional wellness. This finding was in accordance with those of Deaner and McConatha (1993), Kuiper and Martin (1993), and Martin and Lefcourt (1983), although it contradicts the findings of Anderson and Arnoult (1989), Cann et al. (1999) and Martin (1996).

The relationship between humour and emotional wellness was apparent for both the medical and student groups. However, the direction of the relationship appeared to depend on the group in question. That is, lower scores for emotional wellness were associated with higher scores for humour for the medical group, but higher scores for emotional wellness were associated with higher scores for humour for the student group. These findings suggest that to some extent humour’s contribution to health depends on whether members of the sample tested have a medical condition, and/or whether they are students or non-students. Viewed in this light, it seems logical to suggest that the inconsistencies of past research could have been due to the fact that researchers have failed to control for possible differences among respondents. In addition, most of the research into the effects of humour on health has been conducted using student samples, and little consideration has been given to whether students’ responses are different from those of the general community or whether the students had a medical condition.

Thorson and Powell’s (1993a) reported factor structure of the MSHS was generally supported in this study, although there were some differences. The first factor extracted (35.53% of the unrotated variance) was clearly associated with humour production. All nine items that formed the first factor were the same items as those contained in Thorson and Powell’s first factor, labelled ‘humour generation’. The second factor (8.63% of the unrotated variance) loaded on six items, four of which originally belonged to the fourth factor in Thorson and Powell’s structure, which was called ‘attitudes towards humour’. The other two items were composed of one coping humour item and one humour appreciation item. For example, Item Q13, ‘humour is a lousy coping mechanism’, belonged to Thorson and Powell’s second factor, labelled ‘coping humour’, whereas Q10, ‘I like a good joke’, was originally related to Thorson and Powell’s third factor, a factor concerned with the appreciation of humour. Thorson and Powell’s second factor was clearly related to coping humour, whereas the second factor for the Australian sample appeared to be most closely related to attitudes towards humour. Interestingly, the same difference existed between the Croatian sample and the American sample in the Thorson et al. (1997) study. That is, Factor 2 loaded on attitudes towards humour, not coping items.

The third factor appeared to represent coping humour and to producing humour in order to cope with difficult situations. However, the four items forming this factor were items associated with coping humour and humour production according to Thorson and Powell’s (1993a) structure. For example, Item Q2, ‘Uses of wit or humour help me master difficult situations’, and Item Q6, ‘I can use wit to help adapt to many situations’, both belonged to Thorson and Powell’s second factor, labelled ‘coping humour’, whereas Item Q3, ‘I’m confident that I can make other people laugh’, and Item Q7, ‘I can ease a tense situation by saying something funny’, are items related to Thorson and Powell’s first factor, namely ‘humour production’.

The fourth factor also appeared to be related to the use of humour to cope with life on a more personal basis. There were four items loaded by this factor, three of which were related to coping humour, and one to humour appreciation. For example, Item Q19, ‘Coping by using humour is an elegant way of adapting’, Item Q16, ‘Humour helps me cope’, and Item Q20, ‘Trying to master situations through uses of humour is really dumb’, clustered under Thorson and Powell’s second

factor, namely ‘coping humour’, whereas Item Q22, ‘Uses of humour help put me at ease’, was related to Thorson and Powell’s third factor, labelled ‘humour appreciation’.

Conclusions The present investigation into the effects of humour on health found overall that a better sense of humour is not particularly related to higher levels of health. H1 was not well supported, and the various oddities of the negative relationship of humour to well-being and pain as found in the data of the medical group did not lend support to the view that humour and health are positively interrelated. Moreover, it appears that the results obtained are dependent on the sample used, or more specifically on whether the individuals in the sample are free from actual health problems. The results from the present study suggest that individuals who are basically healthy appear to use and view humour differently from those who have a medical condition. Although individuals who are unwell or in pain may turn to humour as a coping strategy, it does not necessarily follow that humour is consistently related to health, and the relationship, when it does occur, is sometimes only a weak one. The findings from the present study also suggested that students may differ from non-students with respect to health. Therefore, care should be taken with regard to generalizing from the results of studies that have used only student samples.

Factor analysis of the MSHS with the combined sample suggested that humour is used as a personal coping mechanism in difficult situations. It appears that this may be even more evident among those who have health problems. In addition, although the factor pattern obtained in the present study (using methodologically sound factor analytic procedures) was somewhat different from that previously reported by Thorson and Powell (1993a) and Thorson et al. (1997), the multi-dimensionality of the MSHS instrument was nevertheless supported.

References Abel, M. H. (1998). Interaction of humour and gender in moderating relationships between stress and

outcomes. Journal of Psychology, 132, 267–237. Anderson, C. A., & Arnoult, L. H. (1989). An examination of perceived control, humor, irrational beliefs,

and positive stress as moderators of the relation between negative stress and health. Basic and Applied Social Psychology, 10, 101–117.

Anderson, R. T., Aaronson, N. K., & Wilkin, D. (1993). Critical review of the international assessments of health-related quality of life. Quality of Life Research, 2, 369–395.

Argyle, M. (1997). Is happiness a cause of health? Psychology and Health, 12, 769–781. Australian Bureau of Statistics. (1995). National Health Survey: SF-36 population norms Australia.

Canberra: Australian Bureau of Statistics. Berk, R. A. (2003). Humor as an instructional defibrillator: Evidence-based techniques in teaching and

assessment. Sterling, VA: Stylus Publishing. Berkman, L. F., & Syme, S. L. (1979). Social networks, host resistance, and mortality: A nine-year

follow-up study of Alameda county residents. American Journal of Epidemiology, 109, 186– 204. Bowling, A. (1997). Measuring health: A review of quality of life measurement scales. (2nd ed.).

Buckingham: Open University Press. Boyle, G. J. (1988). Elucidation of motivation structure by dynamic calculus. In J. R. Nesselroade & R.

B. Cattell, (Eds.), Handbook of multivariate experimental psychology (2nd ed., pp. 737– 787). New York: Plenum.

Boyle, G. J. (1991). Does item homogeneity indicate internal consistency or item redundancy in psychometric scales? Personality and Individual Differences, 12, 291–294.

Boyle, G. J., Stankov, L., & Cattell, R. B. (1995). Measurement and statistical models in the study of personality and intelligence. In D. H. Saklofske & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 417–446). New York: Plenum.

Brannon, L., & Feist, J. (1997). Health psychology: An introduction to behavior and health. PacificGrove, CA: Brookes Cole.

Cann, A., Holt, K., & Calhoun, L. G. (1999). The roles of humor and sense of humor in responses to stressors. Humor: International Journal of Humor Research, 12, 177–193.

Carroll, J. L. (1990). The relationship between humour appreciation and perceived physical health. Psychology—A Journal of Human Behavior, 27, 34–47.

Carroll, J. L., & Shmidt, J. L. Jr. (1992). Correlations between humorous coping styles and health. Psychological Reports, 70, 402.

Cattell, R. B. (1978). The scientific use of factor analysis in behavioral and life sciences. New York: Plenum.

Chapman, A. J., & Foot, H. C. (Eds.) (1996). Humor and laughter: Theory, research, and applications.New Brunswick, NJ: Transaction Publishers.

Child, D. (1990). The essentials of factor analysis (2nd ed.). London: Cassell. Comrey, A. L., & Lee, H. B. (1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum. Cousins, N. (1989, October). Minding your health. Psychology Today, 22–25. Dean, R. A. (1997). Humour and laughter in palliative care. Journal of Palliative Care, 13, 34–39. Deaner, S. L., & McConatha, J. T. (1993). The relation of humor to depression and personality.

Psychological Reports, 72, 755–763. Eysenck, J. J. (1991). Smoking, personality, and stress. New York: Springer-Verlag. Fitzpatrick, R., Fletcher, A., Gore, S., Jones, D., Spiegelhalter, D., & Cox, D. (1992). Quality of life

measures in health care. I: Applications in assessment. British Medical Journal, 305, 1074– 1077. Fontane, P. E. (1996). Exercise, fitness, and feeling well (on-line), 39(2), p. 288. Available from:

http://proquest.umi.com/pqdweb? (1998, Dec. 22). Fry, W. F. (1994). The biology of humor. Humor: International Journal of Humor Research, 7, 111–

125. Gorsuch, R. L. (1983). Factor analysis (Rev. 2nd ed.). Hillsdale, NJ: Erlbaum. Gorsuch, R. L. (1990). Common factor analysis versus component analysis: Some well and little

known facts. Multivariate Behavioral Research, 25, 33–39. Hays, R. D., Sherbourne, C. D., & Mazel, R. M. (1993). The RAND 36-Item Health Survey 1.0. Health

Economics, 2, 217–227. Hays, R. D., Wells, K. B., Sherbourne, C. D., Rogers, W., & Spritzer, K. (1995). Functioning and

wellbeing outcomes of patients with depression compared with chronic general medical illnesses. Archives of General Psychiatry, 52, 11–19.

House, J. S., Robbins, C., & Metzner, H. L. (1982). The association of social relationships and activities with mortality: prospective evidence from the Tecumseh community health study. American Journal of Epidemiology, 116, 123–140.

Huberty, C. J., & Morris, J. D. (1989). Multivariate analysis versus univariate analyses. Psychological Bulletin, 105, 302–308.

Johnston, R. A. (1990). Humor: A preventive health strategy. International Journal for the Advancement of Counselling, 13, 257–265.

Julious, S. A., George, S., & Campbell, M. J. (1995). Sample sizes for studies using the Short Form 36 (SF-36). Journal of Epidemiology and Community Health, 49, 642–644.

Kamei, T., Kumano, H., & Masumura, S. (1997). Changes of immunoregulatory cells associated with psychological stress and humor. Perceptual and Motor Skills, 84, 1296–1298.

Kessler, R. C., Turner, J. B., & House, J. S. (1998). Effects of unemployment on health in a community survey: Main, modifying, and mediating effects. Journal of Social Issues, 44, 69–85.

Klatsky, A. L., & Friedman, G. D. (1995). Alcohol and longevity. American Journal of Public Health, 85,16–18.

Kuiper, N. A., & Martin, R. A. (1993). Humour and self-concept. International Journal of Humour Research, 6, 251–270.

Lefcourt, H. M., Davidson-Katz, K., & Kueneman, K. (1990). Humor and immune-system functioning. Humor: International Journal of Humor Research, 3, 305–321.

Lewin-Epstein, N., Sagiv-Schifter, T., Shabtai, E. L., & Shmeuli, A. (1998). Validation of the 36-item Short-Form Health Survey (Hebrew version) in the adult population of Israel. Medical Care, 36, 1361–1370.

MacHovec, F. (1991). Humour in therapy. Psychotherapy in Private Practice, 9, 25–33. Martin, R. A. (1996). The Situational Humor Response Questionnaire (SHRQ) and Coping Humor

Scale (CHS): A decade of research findings. Humor: International Journal of Humor Research, 9,251–273.

Martin, R. A. (2001). Humor, laughter, and physical health: Methodological issues and research findings. Psychological Bulletin, 127, 504–519.

Martin, R. A., & Dobbin, J. P. (1988). Sense of humour, hassles, and immunoglobulin A: Evidence for a stress-moderating effect of humour. International Journal of Psychiatry in Medicine, 18, 93–105.

Martin, R. A., Kuiper, N. A., Olinger, J., & Dance, K. A. (1993). Humor, coping with stress, self-concept, and psychological well-being. Humor: International Journal of Humor Research, 6, 89–104.

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.

Martin, R. A., & Lefcourt, H. M. (1984). Situational humor response questionnaire: Quantitative measure of sense of humour. Journal of Personality and Social Psychology, 47, 145–155.

McClelland, D. C., & Cheriff, A. D. (1997). The immunoenhancing effects of humour on secretory IgA and resistance to respiratory infections. Psychology and Health, 12, 329–344.

McDonald, R. P. (1985). Factor analysis and related methods. Hillsdale, NJ: Erlbaum. McGhee, P. E. (1999). Health, healing, and the amuse system: Humor as survival training (3rd ed).

Dubuque, IA: Kendall/Hunt. McHorney, C. A., Ware, J. E., Rogers, W., Raczek, A. E., & Lu, J. F. R. (1992). The validity and

relative precision of MOS short-, and long-form health status scales and Darmouth COOP charts. Medical Care, 30, 253–265.

Penninx, B. W. J. H., van Tilburg, T., Kriegsman, D. M. W., Deeg, D. J. H., Boeke, A. J. P., & van Eijk, J. T. M. (1997). Effects of social support and personal coping resources on mortality in older age: The longitudinal aging study Amsterdam. American Journal of Epidemiology, 146, 510–519.

Porterfield, A. L. (1987). Does sense of humor moderate the impact of life stress on psychological and physical well-being? Journal of Research in Personality, 21, 306–317.

Power, C., Rodgers, B., & Hope, S. (1998). U-shaped relation for alcohol consumption and health in early adulthood and implications for mortality. Lancet, 352(9131), 877–879.

Provine, R. R. (2000). Laughter: A scientific investigation. New York: Viking Penguin. Richman, J. (1995). The lifesaving function of humour with the depressed and suicidal

elderly.Gerontologist, 35, 271–273. Rosenbaum, W. L., Sterling, T. D., & Weinkam, J. J. (1998). Use of multiple surveys to estimate

mortality among never, current, and former smokers: Changes over a 20-year interval. American Journal of Public Health, 88, 1664–1668.

Rotton, J. (1992). Trait humour and longevity: Do comics have the last laugh? Health Psychology, 22, 262–266.

Rotton, J., & Shats, M. (1996). Effects of state humor, expectancies, and choice on postsurgical mood and self-medication: A field experiment. Journal of Applied Social Psychology, 26, 1775–1794.

Solomon, J. C. (1996). Humor and aging well. American Behavioral Scientist, 39, 249. Stewart, A. L., Greenfield, S., Hays, R. D., Wells, K., Rogers, W. H., Berry, S. D., McGlynn, E. A., &

Ware, J. E. (1989). Functional status and well-being of patients with chronic conditions: Results from the medical outcomes study. Journal of the American Medical Association, 262, 907– 913.

Surgeon General’s 1990 report on the health benefits of smoking cessation. Morbidity and Mortality Weekly Report, 39, 1–15.

Svebak, S. (1996). The development of the sense of humor questionnaire: From SHQ to SHQ-6. Humor: International Journal of Humor Research, 9, 341–361.

Thorson, J. A., Brdar, I., & Powell, F. C. (1997). Factor-analytic study of sense of humor in Croatia and the USA. Psychological Reports, 81, 971–977.

Thorson, J. A., & Powell, F. C. (1991). Measurement of a sense of humor. Psychological Reports, 69,691–702.

Thorson, J. A., & Powell, F. C. (1993a). Development and validation of a multidimensional sense of humor scale. Journal of Clinical Psychology, 49, 13–23.

Thorson, J. A., & Powell, F. C. (1993b). Sense of humor and dimensions of personality. Journal of Clinical Psychology, 49, 799–809.

Thorson, J. A., & Powell, E. C. (1996). Women, aging, and sense of humour. Humor: International Journal of Humor Research, 9(2), 169–186.

Thorson, J. A., & Powell, E. C. (1997). Psychological health and sense of humor. Journal of Clinical Psychology, 53, 605–619.

Van der Zee, K. I., Sanderman, R., & Heyink, J. (1996). A comparison of two multidimensional measures of health status: The Nottingham Health Profile and the RAND 36-Item Health Survey 1.0. Quality of Life Research, 5, 165–174.

Ware, J. E., & Sherbourne, C. D. (1992). The MOS 36-item Short-Form Health Survey (SF-36). Medical Care, 30, 473–481.

Submitted 24 August 2001; revised version submitted 21 October 2002

Appendix: Items in the Multidimensional Sense of Humour Scale

Item no.

Q1 Sometimes I think up jokes or funny stories. Q2 Uses of wit or humour help me master difficult situations. Q3 I’m confident that I can make other people laugh. Q4 I dislike comics. Q5 Other people tell me that I say funny things. Q6 I can use wit to help adapt to many situations. Q7 I can ease a tense situation by saying something funny. Q8 People who tell jokes are a pain in the neck. Q9 I can often crack people up with the things I say. Q10 I like a good joke. Q11 Calling somebody a ‘comedian’ is a real insult. Q12 I can say things in such a way as to make people laugh. Q13 Humour is a lousy coping mechanism. Q14 I appreciate those who generate humour. Q15 People look to me to say amusing things. Q16 Humour helps me cope.Q17 I’m uncomfortable when everyone is cracking jokes.Q18 I’m regarded as something of a wit by my friends.Q19 Coping by using humour is an elegant way of adapting.Q20 Trying to master situations through uses of humour is really dumb.Q21 I can actually have some control over a group by my uses of humour.Q22 Uses of humour help to put me at ease.Q23 I can use humour to entertain my friends.Q24 My clever sayings amuse others.