Iranian and Swedish adolescents: differences in personality traits and well-being

Participants and procedure

The data was collected at a high school south of Sweden (N = 109, 38 boys and 71 girls, mean age 16.69 years SD .91 years) and two high schools in Tehran and Zanjan, Iran (N = 122, 72 boys and 50 girls, mean age 15.23 years SD 1.26 years). The sampling procedure of schools was based on convenience and included the entire schools; although they had no particular interest or knowledge about our research interest beforehand. Teachers and parents were informed about the nature of the study and that participation was voluntary and that pupils had the right to withdraw at any moment. The school nurse from each school was contacted by the researchers and informed about the study in case any of the students needed counseling. Participants were informed that the study examined how pupils think about their lives in different situations. They were ensured anonymity and informed that participation was voluntary; they had consent from their teachers to participate. The study was conducted in the participants’ own classrooms in groups of 20 to 30 pupils; the questionnaires were distributed on paper. The entire procedure, including debriefing, took approximately 30 min. The Ethics Committee of Gothenburg University approved this research protocol and written informed consent was obtained from all the participants.

Questionnaires

The Big Five Inventory (Benet-Martínez & John, 1998) is a 44-item (5-point Likert scale: 1 = strongly disagree to 5 = strongly agree) instrument that enables efficient assessment of the five personality dimensions: Neuroticism (e.g., I see myself as a person “who worries a lot”), Extraversion (e.g., I see myself as a person “who is talkative”), Openness (e.g., I see myself as a person who “is original, has new ideas”), Agreeableness (e.g., I see myself as a person who “has a forgiving nature”), and Conscientiousness (e.g., I see myself as a person who “does things efficiently”). This self-report measure has been empirically demonstrated to be apt for cross-language and cross-cultural research (Benet-Martínez & John, 1998; Allik & McCrae, 2004). For the Swedish sample the instrument was translated from English to Swedish and then backtranslated by Swedish native speakers; no significant discrepancies were found. Cronbach’s α varied between .84 and .92 among traits. The Iranian version has been used in previous studies (e.g., Joshanloo & Afshari, 2009; Joshanloo & Nosratabadi, 2009). For the Iranian version Cronbach’s α varied between .65 and .72 in the present study.

The Satisfaction With Life Scale (Diener et al., 1985) assesses the cognitive component of subjective well-being (i.e., life satisfaction) and consists of 5 items (e.g., “In most of my ways my life is close to my ideal”) that require a response on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). Both the Swedish and the Iranian versions of this instrument have previously been used in these cultures (e.g., Garcia, 2011c; Garcia, 2012a; Garcia, 2012b; Garcia, Rosenberg & Siddiqui, 2011; Garcia & Moradi, 2012; Garcia & Moradi, 2013; Jokar, Samani & Sahragard, 2007; Joshanloo & Ghaedi, 2009). In the current study the instrument had a Cronbach’s α = .83 in the Swedish sample and .87 in the Iranian sample.

The Positive Affect and Negative Affect Schedule (Watson, Clark & Tellegen, 1988) assesses the affective components of subjective well-being by requiring participants to indicate on 5-point Likert scale to what extent (1 = very slightly, 5 = extremely) they generally experienced 20 different adjectives (10 positive affect and 10 negative affect) within the last few weeks. The positive affect scale includes adjectives such as strong, proud, and interested; and the negative affect scale includes adjectives such as afraid, ashamed, and nervous. The Swedish version has been used in previous studies (e.g., Garcia et al., 2012a; Garcia et al., 2012b; Garcia, Kerekes & Archer, 2012; Garcia & Erlandsson, 2011; Garcia et al., 2010; Nima, Archer & Garcia, 2012; Nima et al., 2013; Schütz et al., 2013) and demonstrated good internal consistency in the present study (positive affect Cronbach’s α = .86, negative affect Cronbach’s α = .78). The Iranian version was translated from English to Farsi and backtranslated by Farsi native speakers; no significant discrepancies were found. The Iranian version demonstrated low but acceptable internal consistence; Cronbach’s α = .56 for positive affect and .57 for negative affect.

The Scales of Psychological Well-Being (the short version; Clarke et al., 2001) comprises 18 items, 3 items for each of the 6 dimensions, using a 6-point Likert scale (1 = strongly disagree, 6 = strongly agree). These dimensions are: (i) positive relations with others (e.g., “People would describe me as a giving person, willing to share my time with others”), (ii) environmental mastery (e.g., “I am quite good at managing the responsibilities of my daily life”), (iii) self-acceptance (e.g., “I like most aspects of my personality”), (iv) autonomy (e.g., “I have confidence in my own opinions, even if they are contrary to the general consensus”), (v) personal growth (e.g., “For me, life has been a continuous process of learning, changing, and growth”), and (vi) purpose in life (“Some people wander aimlessly through life, but I am not one of them”). The Swedish version has been used in previous studies (e.g., Garcia, 2011a; Garcia & Siddiqui, 2009b; Nima et al., 2013) and in the current study the total psychological well-being score (i.e., the sum of the 18 items) demonstrated a Cronbach’s α of .72. The Iranian version has also been used in several studies (e.g., Garcia & Moradi, 2013; Joshanloo & Ghaedi, 2009) and in the current study the total psychological well-being score demonstrated a Cronbach’s α of .62.

Differences between Swedish and Iranian adolescents

Table 1 shows mean scores and standard deviations as well as indicating significant differences for all variables in both samples. Independent factors for all MANOVAs were nationality and gender. For the first MANOVA the dependent factors were the Big Five personality traits, for the second, the three components of subjective well-being, and for the third, the six dimensions of psychological well-being and its total score.

The first MANOVA showed that nationality (F(5, 192) = 15.27, p < .001, Wilks’ Lambda = .72, Observed Power = 1.00) and gender (F(5, 192) = 6.05, p < .001, Wilks’ Lambda = .86, Observed Power = 1.00) had an effect on the personality traits. The interaction of nationality and gender was also significant (F(5, 192) = 2.91, p < .01, Wilks’ Lambda = .93, Observed Power = .84). Iranian adolescents scored significantly higher in Agreeableness (F(1, 196) = 6.41, p < .01, Observed Power = .71), Openness (F(1, 196) = 18.89, p < .001, Observed Power = 1.00) and Conscientiousness (F(1, 196) = 12.17, p < .001, Observed Power = .94). Overall, the national samples did not differ significantly in Extraversion and Neuroticism; however, Swedish boys scored significantly lowest and Swedish girls scored significantly higher than boys. In Neuroticism (F(1, 196) = 9.03, p < .001, Observed Power = .99). See Table 1 for more details.

The second MANOVA showed that nationality (F(3, 201) = 12.23, p < .001, Wilks’ Lambda = .85, Observed Power = 1.00) had a significant effect on the subjective well-being measures. Although the effect of gender was not significant (p = .20), the interaction of nationality and gender was significant (F(3, 201) = 20.48, p < .001, Wilks’ Lambda = .77, Observed Power = 1.00). Swedish adolescents reported higher positive affect (F(1, 203) = 12.02, p < .001, Observed Power = .93) and lower negative affect (F(1, 203) = 26.29, p < .001, Observed Power = 1.00) than Iranian adolescents. Of all the groups, Iranian boys reported lowest positive affect and highest negative affect (see Table 1). No difference in life satisfaction was found between cultures and the interaction of nationality and gender had no effect on life satisfaction either.

The third MANOVA showed that nationality (F(6, 191) = 14.24, p < .001, Wilks’ Lambda = .69, Observed Power = 1.00) and gender (F(1, 191) = 6.21, p < .001, Wilks’ Lambda = .84, Observed Power = 1.00) had a significant effect on the psychological well-being measures. The interaction of nationality and gender was also significant (F(6, 191) = 4.37, p < .001, Wilks’ Lambda = .88, Observed Power = 1.00). Swedish adolescents scored higher on positive relations (F(1, 196) = 11.79, p < .001, Observed Power = .93), personal growth (F(1, 196) = 38.35, p < .001, Observed Power = 1.00), purpose in life (F(1, 196) = 5.87, p < .001, Observed Power = .67), and the psychological well-being total score (F(1, 196) = 4.28, p < .05, Observed Power = .54). Iranian adolescents scored higher on autonomy (F(1, 196) = 28.32, p < .001, Observed Power = 1.00). Moreover, boys as a group scored higher on environmental mastery (F(1, 196) = 15.36, p < .001, Observed Power = .97). The interaction of nationality and gender was significant for the psychological well-being total score (F(1, 196) = 5.65, p < .01, Observed Power = .66) and self-acceptance (F(1, 196) = 15.83, p < .001, Observed Power = .98). Specifically, Swedish boys reported higher scores on both psychological well-being total score and self-acceptance. Iranian girls reported higher self-acceptance than Iranian boys (see Table 1).

Relationships between personality and well-being within samples

Swedish Adolescents: Table 2 displays the correlations between personality and well-being variables for the Swedish sample. Multiple Regression Analysis (MRA) showed that Extraversion was positively related to life satisfaction; Neuroticism was negatively related to life satisfaction and positive affect and positively related to negative affect (see Table 3). Neuroticism, Extraversion, and Conscientiousness were associated to the psychological well-being total score. Among the six psychological well-being dimensions, Neuroticism was negatively associated and Extraversion was positively associated to positive relations (see also Agreeableness which was positively associated to this psychological well-being dimension), environmental mastery (see also Conscientiousness which was positively associated to this psychological well-being dimension), self-acceptance, and autonomy. Personal growth was positively associated to the traits of Conscientiousness and Openness, while purpose in life was solely positively associated to Conscientiousness (see Table 2).

Each of the six dimensions of psychological well-being were related to at least one of the subjective well-being measures: life satisfaction was positively related to self-acceptance, positive affect was positively related to environmental mastery and self-acceptance but negatively related to positive relations with others, and negative affect was negatively related to environmental mastery. See Table 2 for details.

Iranian adolescents: Table 4 displays the correlations between personality and well-being variables for the Iranian sample. In accord with the Swedish sample and our hypothesis, the MRA for the Iranian data showed that Neuroticism was negatively related to life satisfaction and positively related to negative affect. In contrast to the results among Swedes, Neuroticism was not associated to positive affect and Extraversion was not related to any subjective well-being measure (see Table 5). Instead, Openness was positively related to life satisfaction and positive affect. The only psychological well-being dimensions related to personality traits among Iranian adolescents were personal growth (positively related to Extraversion) and purpose in life (positively related to Openness). Also, in contrast to the results among Swedish adolescents, only self-acceptance was positively related to life satisfaction and positive affect. Negative affect was not associated to any of the six psychological well-being dimensions. See Table 5 for details.

Mediation analyses between personality traits, subjective well-being and psychological well-being

We conducted mediation analyses applying procedures recommended by Baron & Kenny (1986) to test if the effect of personality on subjective well-being was mediated by psychological well-being. For this specific analysis we created a subjective well-being total score by first standardizing positive affect, negative affect, and life satisfaction, and then subtracting the standardized negative affect score from the standardized positive affect score and finally summarizing life satisfaction (i.e., (zpositive affect − znegative affect) + zlife satisfaction). Personality variables were set to be the predictors, the psychological well-being total score as the mediator and subjective well-being total score as the outcome variable. Importantly, as recommended by Baron & Kenny (1986), only personality traits that significantly predicted the outcome variable (i.e., subjective well-being) and the mediator (i.e., psychological well-being) were used in the analyses. For the Swedish sample, Extraversion and Neuroticism met these criteria. However, although Openness and Neuroticism predicted subjective well-being in the Iranian sample, neither of the traits predicted psychological well-being in this sample. Hence, mediation analyses could only be conducted for the Swedish sample.

A series of equations were conducted for analyzing the associations between personality (predictor variable), psychological well-being (mediator), and subjective well-being (outcome variable) within the Swedish dataset. Firstly, psychological well-being was regressed on Extraversion and Neuroticism. Both Extraversion (positively: β = .28, t = 3.30, p < .001) and Neuroticism (negatively: β = −.53, t = −6.17, p < .001) were related to psychological well-being. Secondly, applying hierarchal regression, subjective well-being was regressed on Extraversion and Neuroticism in the first step and to psychological well-being in the second step. In the first step Extraversion (β = .26, t = 2.82, p < .01) and Neuroticism (β = −.51, t = −5.40, p < .001) were significantly associated with subjective well-being. In the second step, the effect of Extraversion on subjective well-being was no longer significant (β = −.18, t = 1.90, p = .06) while Neuroticism was still associated with subjective well-being (β = −.33, t = −3.01, p = .004). Moreover, in the second step of the regression, psychological well-being was associated with subjective well-being (β = .33, t = 2.94, p = .004) and even increasing the prediction of the model (ΔR² = .06, F(1, 75) = 8.62, p = .004). Psychological well-being yielded a Sobel Z-value = 1.80 (p = .07) for the relationship Extraversion-psychological well-being-subjective well-being and a Sobel Z-value = 2.08 (p < .05) for the relationship Neuroticism-psychological well-being-subjective well-being (see Fig. 1). In other words, these reductions in beta weights suggest that psychological well-being serves to partially mediate the relationship between Extraversion and subjective well-being and the relationship between Neuroticism and subjective well-being.

Limitations

This study was based on self-reported data and correlational methods that ought to be replicated and supported with other methods such as experimental studies. Further, considering the rather small size of both samples, cautiousness in generalizing the results is warranted as well as acknowledging the need for replicating the core purposes of the study. Nevertheless, with the exception of the differences between the samples in purpose in life and psychological well-being total score, the Observed Power was between .84–1.00—an Observed Power of .80 is generally considered acceptable (Cohen, 1988). It is also important to note that the use of current well-being measures can be seen to mirror Western cultures’ conceptions of well-being and might not necessarily be equally suited for Eastern Muslim samples (Joshanloo, 2012). Perhaps this explains the substantial difference in reliability of used scales between the two samples. To address this in future research might involve a re-conceptualization of well-being and the development of new culturally sensitive scales (Joshanloo, 2012). Nonetheless, these results make up an initial platform to build further research upon.

Unfortunately the Iranian dataset did not allow for the mediation analyses because the statistical criteria for such analyses put forward by Baron & Kenny (1986) were not met. Perhaps this is related to the cultural differences in the conceptions of well-being discussed above. Nevertheless, within the Swedish sample, mediation analyses revealed that psychological well-being mediated the relationship between both Neuroticism and Extraversion, to subjective well-being. Hence, this lends empirical support that psychological well-being promote subjective well-being as theorized by Ryan & Deci (2001), among others.