Edited by: Maicon Rodrigues Albuquerque, Universidade Federal de Minas Gerais, Brazil
Reviewed by: Hugo Carretero-Dios, University of Granada, Spain; Mariola Laguna, The John Paul II Catholic University of Lublin, Poland
*Correspondence: Lili Tian
This article was submitted to Quantitative Psychology and Measurement, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
We conducted two studies to explore the psychometric properties of the Positivity Scale (P Scale) among Chinese adults and early adolescents, using a sample of 552 adults (Study 1) and a sample of 888 early adolescents (i.e., middle school students) (Study 2). First, item analyses and factor analyses were conducted to investigate the one-factor structure of the P Scale. Second, internal consistency reliability, test-retest reliability, and external evidences of validity were evaluated to examine its reliability and validity. Last, we used multi-group confirmatory factor analysis to test measurement invariance across gender. The two studies both provided evidence for its reliability and validity among Chinese adults and early adolescents. For the test of measurement invariance across gender, full scalar invariance was established among early adolescents; partial scalar invariance was supported among adults. Taken together, the results provided preliminary support in the Chinese context for the P Scale as a valid measure to assess the general disposition toward viewing life and experiences in a positive manner. The potential applications for future research and professional practice are discussed.
Due to the rise of the positive psychology movement, increasing attention has been paid to positive features of individuals' functioning (Seligman and Csikszentmihalyi,
Most notably, research has shown self-esteem, optimism and life satisfaction to all be highly and positively correlated with one another (Diener and Diener,
POS has been corroborated across different cultures (Caprara et al.,
However, given a lack of suitable measures of POS, latent variable models of POS have been estimated with separately developed measures of self-esteem (Rosenberg,
The P Scale was constructed from an initial pool of 36 items. Ultimately, an 8- item scale remained after exploratory and confirmatory factor analyses (CFAs) were interpreted (Caprara et al.,
Although the evidence for the psychometric properties of the P Scale has been promising to date, most studies have been done in developed western countries. In order to achieve a more comprehensive understanding of the stability and applicability of the P Scale in different cultural contexts, research on its psychometric properties in other countries and cultures is necessary. As the most populous eastern country, China is one of the largest developing countries in the world, however, research on POS with the P Scale in China remains scarce. One major reason for the limited studies of the POS in China is likely the lack of psychometrically sound measures suitable for Chinese individuals. Considering the urgent need to lay the groundwork for future cross-culture studies and the relevance of evaluating Chinese people's levels of positivity, this study addresses the psychometric properties of the P Scale in the Chinese context.
Due to differences in the cognitive abilities of adults and adolescents (Baltes,
Specifically, we expected a one-factor structure and good internal consistency for the P scale. This expectation was based on previous studies showing that POS can be assessed by estimating the common factor scores of self-esteem, optimism, and life satisfaction (Caprara et al.,
This study addressed the psychometric properties of the Chinese version of the Positivity Scale (P Scale) with Chinese adults.
A total of 552 adults completed the measures. Participants ranged in age from 18 to 61 years (
The convenience sample was selected from short-term adult education training classes offered at South China Normal University. The study was approved from Human Research Ethics Committee of South China Normal University and the relevant institutional review boards, principals, and training teachers. Participation in the study was voluntary, with written consent required to participate in the study. Six hundred adults consented to participate in the study. Participants were asked to complete the questionnaires on the first and last days of the training, which reflected a 4-week time interval. A series of self-report instruments was administered to the participants by a trained graduate assistant in a regular classroom in the institution. After the participants finished all of the measures, they were debriefed about the purpose of the investigation. Finally, we collected questionnaires from 575 participants and subsequently excluded questionnaires from 23 participants [4.00%; (575–552)/575] because of missing data. Thus, 552 valid respondents were included in the data analyses. Of the 552 participants, 437 participants only completed the first wave of data collection, with 115 participants completing both waves of data collection. Data analyses were performed on the basis of three independent samples (Sample 1, Sample 2, and Sample 3):
The Positivity Scale (P Scale; Caprara et al.,
The P Scale consists of eight items. Items 5, 7, and 8 represent an individual's positive view of herself or himself, item 2 represents an individual's satisfaction with life, items 1 and 4 represent an individual's positive attitude about the future, item 3 represents an individual's confidence in other people, and item 6 is a reverse scored item representing an individual's negative view of the future. Participants answered the eight items on a 5-point scale from 1 =
Descriptive statistics were calculated for the eight items of the P Scale as well as the mean score of the P Scale. Corrected item-total correlations for the eight items of the P Scale were used for the item analysis. The factor structure examination of the P Scale was conducted by EFAs and CFAs. We chose Chi square (χ2), the comparative fit index (CFI), the Tucker-Lewis index (TLI), the standardized root mean square residual (SRMR) and the root mean square error of approximation (RMSEA) to assess model fit. The criteria employed for an acceptable model fit were the following: CFI > 0.90, TLI > 0.90, SRMR ≤ 0.08, RMSEA < 0.10. The criteria for a good model fit were the following: CFI > 0.95, TLI > 0.95, SRMR ≤ 0.04, RMSEA < 0.06 (Bentler,
The means, standard deviations, skewness, kurtosis, corrected item-total correlations, and the correlations of the eight items of the P scale for Sample 1 are shown in Table
Means, standard deviations, skewness, kurtosis, the corrected item-total correlations, the correlations of the eight items of the P scale, and factor loading for Chinese adults (Sample 1,
1.Item1 | – | 3.87 | 0.76 | −0.46 | 0.43 | 0.77 | 0.68 | |||||||
2.Item2 | 0.59 |
– | 3.60 | 0.85 | −0.56 | −0.17 | 0.79 | 0.70 | ||||||
3.Item3 | 0.39 |
0.47 |
– | 3.44 | 0.81 | −0.45 | 0.19 | 0.65 | 0.53 | |||||
4.Item4 | 0.71 |
0.61 |
0.47 |
– | 3.83 | 0.73 | −0.75 | 1.65 | 0.82 | 0.72 | ||||
5.Item5 | 0.52 |
0.59 |
0.44 |
0.63 |
– | 3.78 | 0.73 | −1.15 | 1.74 | 0.78 | 0.68 | |||
6.Item6 | 0.45 |
0.43 |
0.28 |
0.38 |
0.34 |
– | 2.64 | 1.11 | 0.52 | −0.56 | 0.58 | 0.47 | ||
7.Item7 | 0.40 |
0.48 |
0.37 |
0.46 |
0.51 |
0.28 |
– | 3.61 | 0.79 | −0.56 | 0.47 | 0.70 | 0.59 | |
8.Item8 | 0.41 |
0.40 |
0.39 |
0.43 |
0.47 |
0.34 |
0.59 |
– | 3.68 | 0.75 | −0.69 | 0.66 | 0.68 | 0.58 |
POS | 0.76 |
0.79 |
0.65 |
0.79 |
0.76 |
0.65 |
0.69 |
0.68 |
3.56 | 0.59 | −0.31 | 3.56 | ||
1.Item1 | 1 | 3.89 | 0.70 | −0.30 | 0.20 | 0.61 | ||||||||
2.Item2 | 0.41 |
1 | 3.45 | 0.86 | −0.59 | −0.37 | 0.72 | |||||||
3.Item3 | 0.35 |
0.44 |
1 | 3.34 | 0.87 | −0.26 | −0.33 | 0.57 | ||||||
4.Item4 | 0.66 |
0.47 |
0.41 |
1 | 3.81 | 0.70 | −0.92 | 1.69 | 0.70 | |||||
5.Item5 | 0.49 |
0.63 |
0.45 |
0.58 |
1 | 3.73 | 0.80 | −0.90 | 0.73 | 0.80 | ||||
6.Item6 | 0.26 |
0.27 |
0.14 |
0.22 |
0.22 |
1 | 2.63 | 0.97 | 0.53 | −0.26 | 0.33 | |||
7.Item7 | 0.38 |
0.49 |
0.35 |
0.43 |
0.52 |
0.27 |
1 | 3.57 | 0.79 | −0.44 | 0.03 | 0.68 | ||
8.Item8 | 0.47 |
0.39 |
0.37 |
0.50 |
0.47 |
0.25 |
0.56 |
1 | 3.61 | 0.80 | −0.56 | 0.08 | 0.65 | |
POS | 0.70 |
0.74 |
0.64 |
0.75 |
0.78 |
0.50 |
0.72 |
0.71 |
3.50 | 0.56 | −0.17 | 0.26 |
The results of the Bartlett's test (χ2 = 731.98,
Because different factorial structures may emerge for different samples (Kline,
The Cronbach's alpha coefficients (sample 1: α = 0.86; sample 2:
Table
Pearson correlations among the P scale, RSES, SWLS, C-LOT, PA, and NA for Chinese adults
Positivity Scale (P Scale) | ||||||
Self-esteem (RSES) | 0.70 |
|||||
Life satisfaction (SWLS) | 0.62 |
0.46 |
||||
Optimism (C-LOT) | 0.47 |
0.47 |
0.24 |
|||
Positive affect (PA) | 0.49 |
0.47 |
0.30 |
0.26 |
||
Negative affect (NA) | −0.38 |
−0.47 |
−0.29 |
−0.34 |
−0.03 |
Hierarchical multiple regression analyses were applied to estimate the predictive validity of the P scale by assessing the relations between the scores on the P Scale at Time 1 and the scores on the IWB and CES-D at Time 2. Before conducting the regression analyses, we evaluated the multicollinearity of variables via the variance inflation factor (VIF), which is a measure of the degree of multicollinearity in a set of multiple regression variables; values >10 indicate collinearity (Marquardt,
In order to investigate the contribution of the P Scale to the prediction of subsequent subjective well-being and depression scores, the scores on the P Scale at Time 1 were employed as the predictor variable, and the scores on the IWB and CES-D at Time 2 served as the criterion variables. First, correlated demographic variables (including gender and age) were entered as the initial block of control variables, and then, scores on the P Scale were entered as the second block of predictor variables. The categorical variable (Gender) was dummy coded as 0 = males, 1 = females before being entered in the regression analysis. As shown in Table
Hierarchical multiple regression of subjective wellbeing and depression at time 2 on the total score of the P scale at Time 1 (Sample 3,
Step 1 | 0.065 | 0.065 | 3.90 |
(2, 112) | |||
Intercept | 8.99 | 1.26 | |||||
Gender |
1.24 | 0.44 | 0.26 |
||||
Age | 0.02 | 0.04 | 0.04 | ||||
Step 2 | 0.273 | 0.208 | 31.76 |
(1, 111) | |||
Intercept | 3.56 | 1.47 | |||||
Gender |
1.03 | 0.39 | 0.21 | ||||
Age | −0.02 | 0.04 | −0.04 | ||||
T1 P scale | 0.23 | 0.04 | 0.47 |
||||
Step 1 | 0.037 | 0.037 | 2.13 | (2, 112) | |||
Intercept | 19.32 | 4.62 | |||||
Gender |
−3.18 | 1.62 | −0.19 | ||||
Age | −0.16 | 0.16 | −0.10 | ||||
Step 2 | 0.312 | 0.275 | 44.35 |
(1, 111) | |||
Intercept | 41.90 | 5.19 | |||||
Gender |
−2.32 | 1.39 | −0.14 | ||||
Age | 0.01 | 0.14 | 0.004 | ||||
T1 P scale | −0.97 | 0.15 | −0.53 |
Because the result of the maximum likelihood estimator (MLR) can be influenced by the sample size, we conducted a series of stepwise analyses by adopting the robust MLR for the assessment of measurement invariance to determine the extent to which P Scale items reflect comparable meanings across gender. This kind of estimator will correct and adjust the statistics and standard error to improve the accuracy of the result. Correspondently, it will yield the scaled chi-square test statistic (S-Bχ2) as the model indicator (Finney and Distefano,
To test for invariance in the P Scale, baseline CFAs were conducted first for each group (men and women) separately. For men, the indices were S-B
Analysis of measurement invariance across gender in adults (Sample 2,
Man | 29.21 | 19 | 0.950 | 0.926 | 0.075 | 0.055 | |||
Women | 31.04 | 19 | 0.958 | 0.939 | 0.072 | 0.044 | |||
M1:Configural invariance | 60.56 | 38 | 0.955 | 0.934 | 0.074 | 0.049 | |||
M2:Metric invariance | 66.78 | 45 | 0.957 | 0.946 | 0.067 | 0.066 | M1:M2 | 4.54(7) | 0.002 |
M3:Scalar invariance | 104.11 | 52 | 0.894 | 0.885 | 0.112 | 0.112 | M2:M3 | 34.53(7) | −0.063 |
M4:Partial scalar invariance | 71.83 | 49 | 0.955 | 0.948 | 0.065 | 0.068 | M4:M2 | 4.86(4) | −0.002 |
This study aimed to test whether the P Scale also showed good psychometric properties in Chinese early adolescents. The findings of Study 1, which provided support for the psychometric properties of the P Scale with Chinese adults, offered a strong foundation to further examine whether the P Scale is suitable to evaluate the positivity disposition of Chinese early adolescents as well. Therefore, in this study, we investigated the psychometric properties of the P Scale with Chinese middle school students.
The participants were 888 middle school students (440 girls and 448 boys) ranging in age from 11 to 15 years (
The procedures for this investigation were the same as those for the adult participants in Study 1. All students participated voluntarily in the two waves of data collection. Written parent consent and student assent were both required to participate in this study. Although 924 students consented to participate in the study, we collected 902 students' questionnaires. Students [1.5%; (902–888)/902] whose questionnaires had any missing values were excluded from the study. The final sample included 888 valid respondents, in which 694 students completed only the first one wave of data collection, and 194 students completed both waves of data collection (across 6 months). In the current study, data analyses were conducted based on three independent middle school student samples (Sample 1, Sample 2, and Sample 3).
The Rosenberg Self-Esteem Scale (RSES; Rosenberg,
The Brief Multidimensional Students' Life Satisfaction Scale (BMSLSS; Seligson et al.,
The University of California at Los Angeles Loneliness Scale (UCLA; Russell et al.,
Brief Adolescents' Subjective Well-Being in School Scale (BASWBSS; Tian et al.,
The Depression Self-Rating Scale for Children (DSRSC; Linyan et al.,
The data analysis procedures were identical to those of Study 1, including descriptive analyses, analyses of factor structure, reliability and validity, and measurement invariance across gender.
The descriptive analyses for Sample 1 are shown in Table
Means, standard deviations, Skewness, Kurtosis, the corrected item-total correlations, the correlations of the eight items of the P scale, and factor loading for Chinese early adolescents (Sample 1,
1.Item1 | – | - | 4.08 | 0.85 | −0.83 | 0.75 | 0.75 | 0.62 | ||||||
2.Item2 | 0.53 |
– | 4.28 | 0.77 | −1.03 | 1.24 | 0.72 | 0.59 | ||||||
3.Item3 | 0.37 |
0.42 |
– | 3.64 | 0.99 | −0.53 | 0.12 | 0.64 | 0.51 | |||||
4.Item4 | 0.71 |
0.52 |
0.45 |
– | 4.10 | 0.90 | −0.98 | 0.98 | 0.77 | 0.64 | ||||
5.Item5 | 0.40 |
0.42 |
0.37 |
0.52 |
– | 3.84 | 0.91 | −0.70 | 0.44 | 0.76 | 0.63 | |||
6.Item6 | 0.31 |
0.18 |
0.18 |
0.26 |
0.24 |
– | 2.70 | 1.27 | 0.23 | −0.95 | 0.44 | 0.33 | ||
7.Item7 | 0.37 |
0.33 |
0.33 |
0.40 |
0.45 |
0.13 |
– | 3.62 | 0.98 | −0.47 | −0.01 | 0.59 | 0.46 | |
8.Item8 | 0.54 |
0.42 |
0.34 |
0.51 |
0.49 |
0.24 |
0.51 |
– | 3.75 | 0.97 | −0.61 | 0.13 | 0.71 | 0.60 |
POS | 0.77 |
0.67 |
0.63 |
0.78 |
0.72 |
0.52 |
0.64 |
0.74 |
3.75 | 0.64 | −0.47 | 0.29 | ||
1.Item1 | – | 4.08 | 0.91 | −0.99 | 1.11 | 0.76 | ||||||||
2.Item2 | 0.56 |
– | 4.29 | 0.73 | −0.83 | 0.66 | 0.62 | |||||||
3.Item3 | 0.44 |
0.36 |
– | 3.66 | 0.97 | −0.56 | 0.26 | 0.54 | ||||||
4.Item4 | 0.77 |
0.54 |
0.46 |
– | 4.13 | 0.89 | −1.08 | 1.40 | 0.79 | |||||
5.Item5 | 0.45 |
0.38 |
0.35 |
0.54 |
– | 3.85 | 0.91 | −0.69 | 0.46 | 0.67 | ||||
6.Item6 | 0.35 |
0.21 |
0.18 |
0.29 |
0.25 |
– | 2.66 | 1.23 | 0.22 | −0.89 | 0.36 | |||
7.Item7 | 0.42 |
0.35 |
0.35 |
0.46 |
0.48 |
0.18 |
– | 3.66 | 0.99 | −0.54 | 0.03 | 0.64 | ||
8.Item8 | 0.58 |
0.41 |
0.36 |
0.57 |
0.54 |
0.26 |
0.57 |
– | 3.78 | 0.95 | −0.64 | 0.23 | 0.76 | |
POS | 0.81 |
0.65 |
0.62 |
0.82 |
0.71 |
0.53 |
0.68 |
0.77 |
3.77 | 0.66 | −0.65 | 0.88 |
Consistent with the results of Study 1 with adults, the results of the Bartlett's test and the Kaiser-Meyer-Olkin statistic(χ2 = 872.1,
According to the values for skewness and kurtosis as shown in Table
The Cronbach's alpha (sample 1:
For middle school students, scores on the P Scale correlated positively with the scores on the RSES (
Pearson correlations among P scale, RSES, BMSLSS-general, BMSLSS-global, and loneliness among Chinese early adolescent (sample 2,
Positivity (P Scale) | − | |||
Self-esteem (RSES) | 0.60 |
− | ||
General Life Satisfaction (BMSLSS-General) | 0.44 |
0.33 |
− | |
Global Life Satisfaction (BMSLSS-Global) | 0.42 |
0.32 |
0.78 |
− |
Loneliness (ULCA) | −0.60 |
−0.50 |
−0.47 |
−0.42 |
Hierarchical multiple regression analyses were applied to estimate the predictive validity of the P scale by assessing the relations between scores on the P Scale at Time 1 and scores on the BASWBSS and DSRSC at Time 2. According to the value of the variance inflation factor (VIF = 1 < 10), none of the variables in the present study suggested issues of multicollinearity, indicating that hierarchical multiple regression analyses were suitable to perform.
First, correlated demographic variables (including gender and age) were entered as the initial block of control variables. Gender was dummy coded as 0 = boys, 1 = girls. Then, scores on the P Scale were entered at the second step of the models. As shown in Table
Hierarchical multiple regression of subjective wellbeing in school and depression at time 2 on the total score of the P scale at time 1 (Sample 3,
Step 1 | 0.02 | 0.02 | 1.95 | (2, 191) | |||
Intercept | 12.62 | 3.44 | |||||
Gender |
−0.31 | 0.34 | −0.07 | ||||
Age | −0.48 | 0.26 | −0.13 | ||||
Step 2 | 0.196 | 0.176 | 41.72 |
(1, 190) | |||
Intercept | 2.43 | 3.50 | |||||
Gender |
−0.19 | 0.31 | −0.04 | ||||
Age | −0.16 | 0.24 | −0.05 | ||||
T1 P scale | 0.20 | 0.03 | 0.43 |
||||
Step 1 | 0.037 | 0.037 | 3.64 |
(2, 191) | |||
Intercept | −13.09 | 9.33 | |||||
Gender |
1.23 | 0.91 | 0.96 | ||||
Age | 1.74 | 0.71 | 0.17 |
||||
Step 2 | 0.150 | 0.113 | 25.32 |
(1, 190) | |||
Intercept | 9.23 | 9.84 | |||||
Gender |
0.96 | 0.86 | 0.08 | ||||
Age | 1.06 | 0.69 | 0.11 | ||||
T1 P scale | −0.44 | 0.09 | −0.34 |
As in Study 1, three levels of measurement invariance analyses were tested by robust MLR for the assessment of measurement invariance across gender among Chinese early adolescents, and the residuals of items 1 and 4 were correlated in the assessment of measurement invariance as well
Testing the invariance of the P Scale scores across gender, baseline CFA were conducted first for each group (boys and girls) separately. For boys, the model indices were S-B
The result showed that both the CFI difference tests and the Satorra-Bentler different test between the configural invariance model and the metric invariance model was ΔCFI = 0.000; ΔS-B
Analyses of measurement invariance across gender among early adolescents (Sample 2,
Boys | 35.30 | 19 | 0.963 | 0.945 | 0.071 | 0.043 | |||
Girls | 50.05 | 19 | 0.945 | 0.919 | 0.097 | 0.047 | |||
M1:Configural invariance | 67.13 | 38 | 0.966 | 0.949 | 0.067 | 0.042 | |||
M2:Matric invariance | 74.17 | 45 | 0.966 | 0.957 | 0.061 | 0.055 | M1:M2 | 7.04(7) | 0.000 |
M3:Scalar invariance | 83.87 | 52 | 0.962 | 0.959 | 0.060 | 0.059 | M2:M3 | 9.70(7) | −0.004 |
We conducted two studies to provide empirical evidence of the reliability and validity of the P Scale with Chinese adults and early adolescents. The main findings are summarized below.
First, as expected, both studies supported the hypothesized unidimensional model underlying the P Scale with Chinese individuals. The single-factor solution was obtained using both EFA and CFA methods. In addition, the goodness-of-fit indices for the CFA across both groups were adequate, and the results of the CFA procedures in the two studies both suggested that allowing for a correlation between a pair of error terms (the error of items 1 and 4) would increase model fit. This modification was deemed reasonable given that the content of both items appeared related to a positive view of the future. Furthermore, it is important to note that the error correlations of items 1 and 4 have also been encountered in Italian, American, Spanish, Brazilian, Serbian, Polish and Japanese adults (Caprara et al.,
Second, the reliability of the P scale was supported for Chinese individuals. Across both studies, the Cronbach's alpha coefficients for the eight items exceeded 0.80. Both coefficients exceeded the coefficient (α = 0.75) found with Italian adults (Caprara et al.,
Third, we obtained the external evidence of validity of the P scale with Chinese individuals. In both of our studies, we found high positive correlations between P Scale scores and the measures of self-esteem, optimism, and life satisfaction, and moderate negative correlations between P Scale scores and the measures of negative affect and loneliness. The cumulative results also further corroborated the notion that POS can be regarded as a common factor underlying self-esteem, optimism, and life satisfaction, which can be differentiated from other psychological constructs. Furthermore, we provided further evidence that the POS predicts important future outcomes, specifically individuals' reports of subjective well- being and depression for Chinese individuals. Such findings are consistent with the previous finding that individuals with higher P Scale scores were characterized by a pleasant affective state and a low incidence of depressive symptoms (Caprara et al.,
Finally, the measurement invariance findings across gender differed for the two groups. For middle school students, the structural, metric and scalar invariance of the P Scale were all generally supported, which indicated that the psychometric properties of the P Scale are quite similar across gender among middle school students. However, for adults, structural invariance, metric invariance and partial scalar invariance were supported, but full scalar invariance was not supported. Specifically, the female sample demonstrated higher values on the intercepts than the male sample on all of the three non-invariant items (2, 3, 5). This difference suggested that response threshold differences for different genders and the effects of systematic responses biases should be considered in the measurement of POS in the Chinese adult population.
To summarize, our research provides some good support for the psychometric properties of the P Scale when used with Chinese adults and early adolescents. This support is important because the P Scale provides a promising, relatively brief measure that can facilitate research in positive psychology and related areas in China as well as other countries. As mentioned in the introduction, research on POS with the P Scale in China remains scarce. Therefore, the applications of the P Scale would make great contribution to the relative area. First, the use of the P Scale could promote the development of numerous studies related to the antecedents and consequences of individual differences in POS across different age levels, such as health, job success, positive interpersonal relationships and other positive outcomes (Caprara et al.,
Nevertheless, there were some important limitations to the present study. First, considering the length of the survey in relation to the anticipated attention span of middle school students, we did not include an optimism measure as in our examination of the external evidence of validity in Study 2. However, the significant correlations between the scores on the P Scale and the RSES and the BMSLSS provide some evidence that the P Scale could effectively measure middle school students' positivity due to the high correlation between optimism and self-esteem, as well as the significant predictive power of optimism in relation to life satisfaction (Chang,
All the authors substantially contributed to the conception and the design of the work. LT and DZ participated to the acquisition of data. The first author (LT) prepared the draft and the contributing authors (DZ and EH) reviewed it critically and gave important intellectual content. All the authors worked for the final approval of the version to be published. All the authors are accountable for all the aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.