Examination of the Validity of the Chinese Version of the Children’s Hope Scale Among Chinese Adolescents

Abstract

Given possible developmental and/or cultural differences in the meaning and levels of hope among children, we evaluated the validity, measurement invariance, and latent mean differences for the Chinese version of the Children’s Hope Scale (CHS) across gender. Our sample comprised 2045 Chinese adolescents (46.4% men), with a mean age of 12.94 years. Confirmatory factor analyses (CFAs) supported one-factor and two-factor models. Cronbach’s alphas, mean inter-item correlations, and test–retest coefficients supported their reliability for both models. Also, multigroup CFA for both models indicated measurement invariance across gender. Analyses of latent means revealed no significant gender differences for the CHS total or Agency factor. However, men scored higher than women on the Pathways factor. Such differences suggest consideration of gender when assessing and promoting hope in Chinese adolescents.

Keywords

Children’s Hope Scale measurement invariance gender difference adolescents

Introduction

Hope is critical to individuals’ pursuit and achievement of goals. Within the context of studies of hope among children and adolescents, (Snyder et al., 1991) theory of hope has been the most widely employed. Snyder defined hope as a positive cognitive-motivational construct, incorporating two components. The motivational component involves agency, which refers to the ability to initiate and sustain motivation to achieve desired goals. The cognitive component involves pathways thinking, which refers to the perceived cognitive capability to generate multiple routes or pathways to achieve desired goals. These two components are theorized to reinforce each other during goal pursuit and attainment.

Adolescence is a key period of mental and physical development (Silverberg, 1990). It is also a critical period for protecting adolescents from psychosocial problems (Gray, Taffe, Keating, 2012). During adolescence, Chinese adolescents appear to bear particularly heavy burdens in the major life domain of schooling, more than Western adolescents (Zhao, Zhu, & Ma, 2009). Thus, the levels and importance of hope for adolescents may vary across development or cultures, although cross-age and cross-cultural studies remain sparse. Nevertheless, hope has been shown to be crucial for adolescents’ positive development in studies of adolescents from various cultures. Hope appears to facilitate the establishment of positive qualities in adolescents, such as readily exploring their environments, forming positive relations with others, and making appropriate commitments to future adult roles (Callina, Johnson, Buckingham, & Lerner, 2014). Furthermore, high hope adolescents typically display better academic achievement, physical health behaviors, and mental health outcomes, such as greater life satisfaction and optimism and fewer internalizing and externalizing problems (Merkaš & Brajša-Žganec, 2011).

Snyder and his colleagues developed the Children’s Hope Scale (CHS) to evaluate individual differences in hope among children of ages 8–16 years (Snyder et al., 1997). The CHS is a six-item dispositional, self-report measure addressing the two dimensions of Agency and Pathways. Because the two dimensions of hope are conceptualized to be synergistic, Snyder, Cheavens, and Michael (2005) recommended that the CHS total score as well as the separate scores for the Agency and Pathways subscales be interpreted for a comprehensive understanding of hope. Previous studies have provided promising evidence for the reliability and validity of the CHS in different countries, such as China (Zhao & Sun, 2011), America (Shadlow, Boles, Roberts, & Winston, 2015), and South Africa (Savahl, Casas, & Adams, 2016). In various studies, the two-factor structure proposed by Snyder et al., 1997 has been supported (Jovanović, 2013; Zhao & Sun, 2011).Despite the support for the two-factor model of the CHS, a one-factor model has also been used (Ciarrochi, Heaven, & Davies, 2007). For instance, Pulido-Martos, Jiménez-Moral, Lopez-Zafra, and Ruiz, 2014 recommended the use of the total score, given that the separate subscale scores did not fit their sample data well. Edwards, Ong, and Lopez, 2007 and Guse, Bruin, and Kok (2016) also recommended the use of the total CHS score, interpreted as the overall level of hope in youth.

Beyond studies of its internal structure, research has been conducted with the CHS to evaluate its external relations with other variables. Evidence for validity has been offered through a meaningful array of correlations with other measures. For instance, hope, as measured by the CHS, has been associated with positive individual and social–environmental characteristics (Merkaš & Brajša-Žganec, 2011). Moreover, studies have shown that higher scores on the CHS protect adolescents from negative outcomes, such as depressive symptoms (Wong & Lim, 2009; Jovanović, 2013). Furthermore, researchers have addressed convergent and discriminant relations of the Agency and Pathways subscales. For example, Jovanović (2013) found that scores on Agency subscale were more closely linked to scores on optimism and self-esteem measures, whereas scores on Pathways subscale were more closely linked to scores on self-efficacy and resilience measures, and both CHS subscales showed strong correlations with measures of positive affect. Subsequently, a deeper understanding of adolescents’ hope is of considerable importance, particularly for adolescents in the context of schooling in China.

Extant research on hope, using the CHS, has also addressed gender differences in hope among adolescents. The existing studies have typically, but not always, demonstrated lack of mean differences across gender in hope in children (Snyder et al., 2005; Snyder, Lopez, Shorey, Rand, & Feldman, 2003) and adolescents (Jovanović, 2013). Because adolescents may experience differing gender-based expectations and stereotypes related to their specific environments and developmental levels; hope reports may be quantitatively different for boys and girls in some circumstances (Hendricks-Ferguson, 2006). Indeed, previous research has suggested that boys and girls may display different levels of hope across the differing stages of development. For example, Bernardo (2015) found that girls reported higher hope than boys in seventh grade, whereas girls reported lower hope than boys in 10th grade. Regarding China in particular, Chinese men are exposed to higher expectations than women for achieving their goals (Wang, Yuan, & Yang, 2013). Thus, gender differences in adolescents’ hope in China seem plausible.

Given the gender differences in hope in previous studies, it seems important to consider the possibility that there are gender differences in how the dimensions of hope are experienced and reported during adolescence in particular contexts (Bernardo, 2015; Hendricks-Ferguson, 2006). Furthermore, because the majority of research on the structural validation and invariance of the CHS has been conducted with Western adolescents, it is unclear to what extent these findings can be generalized to adolescents in China, given the substantial cultural differences between China (i.e., more collectivistic) and Western countries (i.e., more individualistic) (Daphna, Coon, & Markus, 2002). Prior to further study of the nature and origins of hope differences among youth from differing cultures, the nature and magnitude of the relations between gender and hope in Chinese mainland adolescents remain unknown. Thus, evidence is needed regarding the measurement invariance across gender for hope scales, such as the CHS, in Chinese mainland adolescents.

As an important statistical tool in establishing the validity of making comparisons across groups, such as gender groups, tests of measurement invariance facilitate the determination of whether observed differences reflect true differences in the relevant construct or whether the differences reflect possible gender-based measures’ biases, for example, biases related to the interpretation of the meaning of items across genders (Cheung & Rensvold, 2002). Thus, we employed multigroup confirmatory factor analysis (MGCFA), which is frequently used to assess measurement invariance across groups or time (French & Finch, 2008). Also, latent mean differences across gender may be examined if at least two items per factor exhibit metric and scalar invariance (Steenkamp & Baumgartner, 1998). Latent means are better indicators of true differences than observed means because they are not associated with measurement error (Brown, 2006).

With a sample of Chinese adolescents, we thus addressed the psychometric properties of the CHS through examinations of its (a) factor structure using confirmatory factor analyses techniques, (b) measurement invariance for the CHS total scale and its two subscales across gender, and (c) latent mean differences for the one-factor model and the two-factor model between Chinese men and women adolescents. Also, we evaluated its internal consistency, test–retest reliability, and relations with four theoretically related measures. This study thus extended beyond previous research by addressing the psychometric properties of the CHS with adolescents from a non-Western culture, including testing for gender differences in levels of hope.

Method

All procedures in our study were in accordance with the ethical standards of Hunan Normal University Institutional Review Board and 1964 Helsinki declaration, as well as its later amendments or comparable ethical standards. Informed parental consent and student assent were obtained from all participants.

Participant

The sample comprised 2045 participants who were recruited using cluster sampling. A total of 949 (46.4%) participants were men and 1096 (53.6%) were women. The participants ranged in age from 11 to 16 years (M = 12.94; SD = 1.12 years), and grade levels ranged from 7 to 9 (M = 8.10; SD = 0.86). Most participants (95.6%) were Han, the predominant nationality in China. All participants spoke the Mandarin language. There was no significant gender difference in age (t = −1.591, p >.05).

Procedure

Participants were from three cities representing different levels of economic development in the Hunan province of China. Four classes in each grade (7–9) were selected randomly from three middle schools in each city. Prior to participation, 2208 participants were informed about the purpose of the study, and consent forms were sent to their parents. A total of 2190 students returned consent and assent forms. Participants took about 30 minutes to complete study questionnaires supervised by trained research assistants during the school day in the students’ regular classrooms without the attendance of teachers. Our final sample size was 2045, with a 93.3% effective response rate. A total of 598 participants were randomly selected to complete the CHS twice across a 1-month interval to examine its test–retest reliability.

All measures were administered in the students’ primary language of Mandarin. The translation of the CHS for this study was originally conducted by a native Chinese speaker with a strong knowledge of English. Then, a native English speaker who was fluent in Chinese performed a back translation from Chinese to English. Discrepancies between the Chinese and English translators were resolved by their agreement. The Chinese version was judged to be an accurate translation of the original English version. Both the Chinese and English items were also evaluated by other authors to ensure equivalence in meaning.

Measurements

Children’s Hope Scale (CHS)

The CHS is a six-item self-report scale. Students rated items using a Likert-type format from 1 (none of the time) to 6 (all of the time), yielding a total score from 6 to 36. Higher scores denoted higher hope. In the original study, Snyder et al. (1997) reported that internal consistency coefficients ranged from .74 to .81, with a 1-month test–retest coefficient of .71 for the total score. Also, Zhao and Sun (2011) assessed the two-factor model with Chinese adolescents, reporting alpha coefficients of .74, .55, and .59 for the total, Agency, and Pathway scores, respectively, with a test–retest reliability coefficient of .71 for the total score. In the current study, both one-factor and two-factor models of the CHS were examined.

Brief Multidimensional Students’ Life Satisfaction Scale (BMSLSS; (Seligson, Huebner, & Valois, 2003) and Chinese Validation by Zeng, Ling, Huebner, He, & Lei, 2017.

The BMSLSS is a 6-item self-report measure assessing students’ life satisfaction, with response options ranging from 1 (terrible) to 7 (delighted). Total scores ranged from 6 to 42; higher scores reflected higher general life satisfaction. Cronbach’s alpha for the BMSLSS among Chinese adolescents was .77 in a previous study (Tian, Zhang, & Huebner, 2015) and .85 in the current study.

Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1989, Chinese Version, (Ling, Huebner, Liu, Liu, Zhang, & Xiao, 2015).

The RSES consists of 10 items designed to assess positive global self-evaluations. Response options ranged from 1 (strongly disagree) to 4 (strongly agree), with total scores ranging from 10 to 40. Past research with Chinese adolescents yielded an alpha coefficient of .63 (Chan, 2012). For the present study, the alpha was .65.

Multidimensional Scale of Perceived Social Support (MSPSS); (Zimet, Dahlem, Zimet, & Farley, 1988), Chinese Version, Ling et al., 2015).

The MSPSS is a 12-item self-report measure of support from family, friends, and others. Participants responded to a 7-point scale ranging from very strongly disagree (1) to very strongly agree (7). Higher total scores indicated greater perceived overall social support (Ling et al., 2015). Evidence of its reliability and validity has been provided for Chinese adolescents (Zhao, Wang, & Kong, 2014). In our study, its alpha coefficient was .91.

The Revised Life Orientation Test (LOT-R; Scheier, Carver, & Bridges, 1994, Chinese Version, Ling et al., 2015).

The LOT-R was designed to measure participants’ dispositional optimism. Respondents answered 10 items each on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). Vassar and Bradley (2010) reported a Cronbach’s alpha coefficient of .75 with Chinese adolescents. Its alpha coefficient was .64 in the current study.

Data Analyses

The SPSS statistical 17.0 version was used to analyze t-tests, test–retest reliabilities, and internal consistency reliabilities of the CHS. Given the very low rate of missing data (.2%) in this study, we used mean replacement based on data split by group.

Confirmatory factor analyses and MGCFA were computed with AMOS 17.0 software by using maximum likelihood (ML) as the estimation method. The one-factor and two-factor structures of the CHS were examined through CFA in the complete sample and in both gender subsamples. In this study, we utilized multiple, complementary fit indexes to evaluate model fit: the comparative fit index (CFI), Tucker–Lewis Index (TLI, a non-normal fitting index), standardized root mean square residual (SRMR), and root-mean-square error of approximation (RMSEA). Hu and Bentler (1999) tentatively denoted CFI ≥.90 as an “adequate” fit and values greater than .95 as a “good” fit. They also cautiously designated fit indices for RMSEA and SRMR ≤.06 and .08, respectively, as “good” and values less than .08 and .10 as “mediocre” (Hu & Bentler, 1999).

Multigroup confirmatory factor analysis procedures were used subsequently to examine whether the CHS demonstrated measurement invariance across gender. Four nested models were estimated by default: configural model, metric model, scalar model, and strict model (Vandenberg & Lance, 2000). Configural invariance (Model 1) was the baseline model that relaxed all equality constraints. Metric invariance (Model 2) tested the invariance of factor loadings among groups by placing equality constraints on these parameters. Scalar invariance (Model 3) not only tested the invariance of factor loadings but also the equality of intercepts across groups and covariance groups. The strict invariance (Model 4) was tested with factor loadings, intercepts of variables, and error variances all being set to be equal across gender (Liu, 2008). The change in CFI (△CFI ≤ .010) and the change in RMSEA (△RMSEA ≤ .015) were considered as evidences of equivalence between different consecutive models (Cheung & Rensvold, 2002).

Finally, latent mean differences across gender were tested. In the structural means model, the male group served as the reference group with its latent mean being constrained to 0, while the female group was estimated freely. The critical ratio (CR) was chosen as the index to evaluate whether the latent means were different across groups (Tsaousis & Kazi, 2013). If the CR was higher than 1.96 or lower than −1.96, the estimate of equality was rejected. Therefore, a higher score for the latent mean was regarded as a positive value for the comparison group relative to the reference group, whereas a negative value indicated a lower score for the latent mean for the comparison group.

Results

Internal Structure

To assess internal structure of the CHS, two competing models were examined: a one-factor model and the two-factor model. The fit indexes for both theoretically plausible models for the three samples using CFA are summarized in Table 1. Models were fit for the total sample, and separately for boys and girls. The CFI, TLI, and SRMR for both models were in the acceptable ranges for all the samples. To further compare the fit of competing models, the X² difference test was also conducted. No significant difference was found between the one-factor and two-factor models (X² DIFFTEST = .002, df = 1, p >.05). Therefore, both the one- and two-factor models served as the baseline model for the subsequent tests of measurement invariance. The RMSEAs for both models were slightly higher than the acceptable range for the female sample. Additionally, the standardized factor loadings for the six items were all statistically significant (see Figure 1 and Figure 2).

Table 1.

Model Fit Statistics for the CFA Models.

	Model	X ² (df)	CFI	TLI	RMSEA	SRMR
Total sample/men/women (n = 2045/949/1096)
CFA	One-factor model	108.598(9)/34.481(9)/90.886(9)	.973/.984/.961	.954/.973/.935	.074/.055/.091	.031/.025/.039
CFA	Two-factor model	108.596(8)/34.396(8)/90.747(8)	.972/.983/.960	.948/.968/.926	.078/.059/.097	.031/.025/.038

Note. X² = chi-square; df = degrees of freedom; CFI = comparative fit index; TLI = Tucker–Lewis Index; RMSEA = root-mean-square error of approximation; SRMR = standardized root mean square residual.

Figure 1.

Standardized parameter estimates of Agency factor and Pathways factor of CHS (full sample/men/women). Note. CHS = Children’s Hope Scale.

Figure 2.

Standardized factor loadings of CHS full scale (full sample/men/women). Note. CHS = Children’s Hope Scale.

External Relations

Evidence of the validity of the CHS was also provided through comparisons of scores between the low-level and high-level hope groups on the BMSLSS, REES, MSPSS, and LOT. Four hundred seven adolescents who scored one standard deviation lower than the mean on the CHS total scores were classified as the low-level hope group, and 391 adolescents who scored one standard deviation higher than the mean were classified as the high-level hope group. Independent sample t-tests showed that the differences between the low-level hope group and the high-level group on the BMSLSS, REES, MSPSS, and LOT were all statistically significant, with medium to large effect sizes ranging from .73 to 1.70 (Cohen, 1988) (Table 2). As expected, youth with higher CHS scores showed higher scores on measures of life satisfaction, self-esteem, social support, and optimism.

Table 2.

Comparison of BMSLSS, REES, MSPSS, and LOT between Low- and High-Level Group.

Scale	Low-level hope n = 407	High-level hope n=391	Mean difference	Cohen’s d
BMSLSS	23.71 ± 6.60	28.13 ± 5.37	−4.43	−.73
REES	19.89 ± 4.42	23.76 ± 3.59	−3.86	−.96
MSPSS	39.65 ± 12.31	59.58 ± 11.01	−19.93	−1.70
LOT	26.93 ± 4.97	33.28 ± 5.50	−6.36	−1.21

Note. MSPSS = Multidimensional Scale of Perceived Social Support; LOT-R = The Revised Life Orientation Test; BMSLSS = Brief Multidimensional Students’ Life Satisfaction Scale; RSES = Rosenberg Self-Esteem Scale.

Cronbach’s Alpha and Mean Inter-Item Correlation

Table 3 shows Cronbach’s alphas and mean inter-item correlations for the CHS and its two subscales for the total sample and male and female groups.

Table 3.

Cronbach’s Alpha and Mean Inter-Item Correlation.

	Hope		Agency		Pathways
	α	MIC	α	MIC	α	MIC
Total	.811	.420	.666	.407	.671	.408
Male	.800	.402	.652	.389	.665	.401
Female	.820	.436	.680	.425	.676	.413

Note. α = Cronbach’s alpha; MIC = mean inter-item correlation.

Temporal Stability

Test–retest reliability provides an estimate of temporal stability, which is particularly important for measures purporting to address psychological constructs that incorporate state- and trait-like properties, like hope. The 1-month test–retest correlations for the CHS total, Agency, and Pathways scores were all statistically significant (p < .001), with r(584) = .739, r(584) = .710, and r(584) = .709, respectively.

Measurement Invariance Across Gender

To evaluate the measurement invariance of the one-factor and two-factor models of the CHS across gender, MGCFA was examined. As mentioned, the X² statistic is sensitive to sample size. Thus, △CFI ≤ .010 and △RMSEA ≤ .015 were employed. As seen in Table 4, configural invariance (model 1) was tested, which involved an initial analysis with no constraints estimated simultaneously across gender. The factor loadings, intercepts, and error variances for this model were all allowed to vary. Metric invariance (model 2), in which factor loadings were specified to be equal across gender, was supported (△CFI = .001/.001 and △RMSEA = −.007/−.007). This finding indicated that the CHS has the same structure and meaning across gender groups (with intercepts and error variances estimated freely). Scalar invariance (model 3), in which both factor loadings and intercept values were specified to be equal across gender groups, was supported (△CFI = −.002/−.002 and △RMSEA = −.008/−.004), which was within the cutoff score of △CFI ≤.010 and △RMSEA ≤.015. Next, we examined strict invariance (model 4) for the CHS, which also supported △CFI = −.009/−.010 and △RMSEA = .002/.001 for one- and two-factor models, respectively.

Table 4.

Multigroup Analysis Across Gender.

Two-factor model/one-factor model	X²(df)	CFI	TLI	RMSEA	SRMR	△X²(df)	△CFI	△RMSEA
Model 1	125.139(16)/125.364(18)	.970/.971	.944/.951	.058/.054	.0384/.0248	—	—	—
Model 2	126.460(20)/126.825(23)	.971/.972	.956/.963	.051/.047	.0395/.0244	1.321(4)/1.461(5)	.001/.001	−.007/−.007
Model 3	141.033(29)/141.033(30)	.969/.970	.968/.970	.043/.043	.0394/.0246	14.573(9)/14.208(7)	−.002/−.002	−.008/−.004
Model 4	180.705(35)/180.707(36)	.960/.960	.966/.967	.045/.044	.0433/.0280	39.672(6)/39.674(6)	−.009/−.010	.002/.001

Note. CFI = comparative fit index; TLI = Tucker–Lewis Index; RMSEA = root-mean-square error of approximation; SRMR = standardized root mean square residual. △, change in the model; Model 1, configural invariance; Model 2, metric invariance; Model 3, scalar invariance; Model 4, strict invariance.

Latent Mean Differences

Given that invariance was achieved for the configural, metric, scalar, and strict models, the latent factor mean differences across gender were computed, along with Cohen’s d effect size indices (Cohen, 1988). For the one-factor model, the latent factor CHS means were not significantly different between boys and girls (CR = −1.899, d = .08). For the two-factor model, the latent factor CHS means were not significantly different between boys and girls for the Agency subscale (CR = −.874, d = .04). However, the latent means were significantly different between boys and girls for the Pathways subscale, with boys reporting a higher mean level than girls (CR = −2.429, d = .11).

Discussion

Based on Snyder et al. (1997) and Zhao and Sun (2011), we examined one-factor and two-factor models of the CHS in a large sample of Chinese adolescents, subsequently assessing its measurement invariance and latent mean differences across gender. We further examined its internal consistency, test–retest reliability, and relations with theoretically related variables.

First, consistent with previous studies (Edwards et al., 2007; Jovanović, 2013), our CFA results for one- and two-factor models for the CHS all exceeded relative standards. Nevertheless, the RMSEAs for both models were slightly higher than the acceptable range for our female sample, which was also consistent with previous research (Dixson, 2017). According to Browne and Cudeck (1992), .05 < SRMR and RMSEA < .10 are also considered acceptable.

In this study, the fit indices for the one-factor model were slightly better than the two-factor model, but not statistically different given the chi-square difference test. Thus, both models revealed an acceptable fit to the data among Chinese adolescents, providing preliminary evidence for the use of the CHS with Chinese adolescents. Additionally, although Item 2 (“I can think of many ways to get the things in life that are most important to me”) showed a lower factor loading than the other items in both models, the mean inter-item correlations were acceptable. Such a low factor loading for Item 2 may be a result of translation of context-specific concerns. In more collectivistic cultures (e.g., China), individuals tend to be socialized to be part of interdependent groups with common fates, goals, and values (Daphna, Coon, & Markus, 2002). Thus, adolescents more likely rely on their parents or others to find ways to obtain the things in life that are most important to them. Thus, it does not appear warranted to revise or discard Item 2. In addition, because Zhao and Sun (2011) argued persuasively that the retention of Item two is essential to the construct validity of the Pathways subscale, we retained Item two in both models in subsequent analyses.

The finding of support for both models suggested that the total score as well as the Pathways and Agency scores are meaningful and interpretable with Chinese adolescents (cf. Snyder et al., 2005). This conclusion is important when researchers wish to distinguish the two components of hope. For example, Franke, Huebner, and Hills (2017) evaluated the role of Pathways thinking separately from Agency in a test of a theory of positive emotions with adolescents.

Second, consistent with previous studies (Ling et al., 2015; Ling et al., 2016; Merkaš & Brajša-Žganec, 2011), significant differences emerged between the high and low hope groups on the BMSLSS, REES, MSPSS, and LOT; that is, adolescents with high (vs. low) hope reported the expected higher levels of life satisfaction, self-esteem, social support, and optimism. These findings also provide supportive evidence for the validity of the CHS.

Third, the reliability coefficients for the CHS total score were acceptable (α ≥ .800) for the total, male, and female sample. However, the coefficients for the Agency and Pathways subscales were somewhat lower than the widely accepted .70 cutoff for research (Nunnally & Bernstein, 1994). These findings are consistent with previous studies, suggesting that the internal consistency of the Agency and Pathways subscales could be considered adequate, particularly given the small number of items (i.e., three) in each subscale (Jovanović, 2013). The mean inter-item coefficients in our study were well between the lowest (.10) and the highest cutoff values (.50), also supporting the internal consistency of the Chinese version of the CHS (Bentler, 1990). Finally, considering that the 1 month test–retest reliabilities exceeded .70, the reliability of the CHS overall appears acceptable.

Fourth, the results of our study illustrate that the CHS can be utilized to assess hope among both female and male Chinese adolescents. The MGCFA results revealed that both the one-factor and two-factor structures were invariant across gender, suggesting that Chinese boys and girls interpret the CHS similarly and that the CHS measures the same constructs across gender. Therefore, the CHS may be used in mean-level comparisons of Chinese male and female adolescents.

Finally, tests of latent mean differences demonstrated no statistically significant differences across gender on the total score for the one-factor structure or on the Agency subscale for the two-factor structure among Chinese adolescents. These equivalent mean scores indicated that both Chinese boys and girls reported equivalent levels of hope toward achieving their desired goals. This comparability may be due to the possibility that goals representing similar standards are more motivating to students and especially enhance their sense of agency (Gherasim, Butnaru, & Mairean, 2013; Merkaš & Brajša-Žganec, 2011). However, men scored higher on the Pathways subscale than women, which might reflect differences in Chinese parents’ expectancies and parenting styles for male and female adolescents (Wang et al., 2013). In China, men are more likely to be raised to be strong, capable, and independent individuals as well as to be rewarded for outperforming others (Su, Mcbride, & Xiang, 2015), whereas women are raised to be kind-hearted, honest, and considerate (Zeng et al., 2017). These differences, in turn, may lead Chinese men to be more likely than women to develop higher expectations for pursuing and persisting at individual goals (White, 2010). Thus, Chinese male adolescents may become more active than women in seeking ways to overcome obstacles and achieve their goals.

Limitations

Limitations of this study should be noted. First, participants’ ages ranged from 11 to 16 years. Given that hope is a malleable construct (Callina et al., 2014); Merkaš & Brajša-Žganec, 2011), a wider age range might be considered in future research to investigate the generalizability of gender differences across different age groups. Second, although a large sample was utilized, its homogeneity was relatively high because participants were all from one province. Studies with more nationally representative samples should be conducted to confirm the generalizability of our findings. Third, all data were collected through self-report measures. Future research should employ multi-method approaches, such as incorporating ratings of knowledgeable others. Finally, cross-sectional research designs, such as the design of this study, do not capture the dynamic nature of adolescents’ possible changes in hope levels. More longitudinal research is needed to investigate the stability of hope among youth and particularly to investigate its antecedents, correlates, and consequences across differing age, gender, and cultural groups.

Conclusion

Our results provided promising preliminary evidence for the reliability and validity of the Chinese version of the CHS. Both the one-factor and two-factor models were supported, providing preliminary support for the use of both subscale scores as well as its total score with Chinese adolescents. Moreover, the results of measurement invariance tests suggested that the CHS may be used in cross-gender comparisons among Chinese adolescents. No gender differences for the CHS Total or Agency factor, whereas men scored higher than women on the Pathways factor. Pending replication, such differences suggest implications related to differential risk levels and intervention efforts for promoting hope in male and female early adolescents in China.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Key Program for Research on Women’s Theory and Practice of Hunan Province (Grant No. 18ZDB04).

ORCID iD

Mingtian Zhong

References

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological bulletin, 107(2), 238. doi:10.1037/0033-2909.107.2.238

Bernardo

A. B. I.

(2015). Hope in early adolescence: Measuring internal and external locus-of-hope. Child Indicators Research, 8(3), 699-715. doi:10.1007/s12187-014-9254-6

Brown

T. A.

(2006). Confirmatory factor analysis for applied research. New York: Guilford Press.

Browne

M. W.

Cudeck

(1992). Alternative ways of assessing model fit - Institute for Social and Economic Research (ISER). Clinical & Diagnostic Laboratory Immunology, 7(2), 161-167. doi:10.1128/CDLI.7.2.161-167.2000

Callina

K. S.

Johnson

S. K.

Buckingham

M. H.

Lerner

R. M.

(2014). Hope in context: Developmental profiles of trust, hopeful future expectations, and civic engagement across adolescence. Journal of Youth and Adolescence, 43(6), 869-883. doi:10.1007/s10964-014-0096-9

Chan

S. M.

(2012). Depressive mood in Chinese early adolescents: Relations with shyness, self-esteem and perceived social support. Asia-Pacific Psychiatry, 4(4), 233-240. doi:10.1111/j.1758-5872.2012.00179.x

Cheung

G. W.

Rensvold

R. B.

(2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 233-255. doi:10.1207/S15328007SEM0902_5

Ciarrochi

Heaven

P. C. L.

Davies

(2007). The impact of hope, self-esteem, and attributional style on adolescents' school grades and emotional well-being: A longitudinal study. Journal of Research in Personality, 41(6), 1161-1178. doi:10.1016/j.jrp.2007.02.001

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

10.

Daphna

Coon

H. M.

Markus

(2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological bulletin, 128(1), 3-72. doi:10.1037/0033-2909.128.1.3

11.

Dixson

D. D.

(2017). Hope across achievement: Examining psychometric properties of the children’s hope scale across the range of achievement. Sage Open, 7(3), 60-70. doi:10.1177/2158244017717304

12.

Edwards

L. M.

Ong

A. D.

Lopez

S. J.

(2007). Hope measurement in mexican american youth. Hispanic Journal of Behavioral Sciences, 29(2), 225-241. doi:10.1177/0739986307299692

13.

Franke

K. B.

Huebner

E. S.

Hills

K. J.

(2017). Cross-sectional and prospective associations between positive emotions and general life satisfaction in adolescents. Journal of Happiness Studies, 18, 1075-1093. doi:10.1007/s10902-016-9763-8

14.

French

B. F.

Finch

W. H.

(2008). Multigroup confirmatory factor analysis: Locating the invariant referent sets. Structural Equation Modeling: A Multidisciplinary Journal, 15(1), 96-113. doi:10.1080/10705510701758349

15.

Gherasim

L. R.

Butnaru

Mairean

(2013). Classroom environment, achievement goals and maths performance: Gender differences. Educational Studies, 39(1), 1-12. doi:10.1080/03055698.2012.663480

16.

Gray

K.M.

Taffe

J.R.

Keating

C. M. S. L.

(2012). The trajectory of behavior and emotional problems in autism. American Journal on Intellectual and Developmental Disabilities, 117(2), 121–133. doi:10.1352/1944-7588-117-2.121

17.

Guse

Bruin

G. P.

Kok

(2016). Validation of the children's hope scale among South African adolescents. Child Indicators Research, 9(3), 757-770. doi:10.1007/s12187-015-9345-z

18.

Hendricks-Ferguson

(2006). Relationships of age and gender to hope and spiritual well-being among adolescents with cancer. Journal of Pediatric Oncology Nursing Official Journal of the Association of Pediatric Oncology Nurses, 23(4), 189. doi: 10.1177/1043454206289757

19.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural equation modeling: a multidisciplinary journal, 6(1), 1-55. doi:10.1080/10705519909540118

20.

Jovanović

(2013). Evaluation of the children’s hope scale in serbian adolescents: Dimensionality, measurement invariance across gender, convergent and incremental validity. Child Indicators Research, 6(4), 797-811. doi:10.1007/s12187-013-9195-5

21.

Ling

Huebner

E. S.

Liu

W.-L.

Zhang

Xiao

(2015). The origins of hope in adolescence: A test of a social-cognitive model. Personality and Individual Differences, 87, 307-311. doi:10.1016/j.paid.2015.08.016

22.

Ling

Huebner

E. S.

Zeng

(2016). A person-oriented analysis of hope in Chinese adolescents. Personality and Individual Differences, 101, 446-450. doi:10.1016/j.paid.2016.06.048

23.

Liu

(2008). Assessing measurement invariance of the teachers’ perceptions of grading practices scale across cultures. Paper presented at NERA Conference Proceedings 2008, Rocky Hill, CT.

24.

Merkaš

Brajša-Žganec

(2011). Children with different levels of hope: Are there differences in their self-esteem, life satisfaction, social support, and family cohesion? Child Indicators Research, 4(3), 499-514. doi:10.1007/s12187-011-9105-7

25.

Nunnally

J. C.

Bernstein

I. H.

(1994). Psychometric theory (3rd ed.). New York: McGraw Hill.

26.

Pulido-Martos

Jiménez-Moral

J. A.

Lopez-Zafra

Ruiz

J. R.

(2014). An adaptation of the children’s hope scale in a sample of spanish adolescents. Child Indicators Research, 7(2), 267-278. doi:10.1007/s12187-013-9223-5

27.

Rosenberg

(1989). Society and the adolescent self-image. 1989, rev. Middletown, CT: Wesleyan University Press Google Scholar.

28.

Savahl

Casas

Adams

(2016). Validation of the children’s hope scale amongst a sample of adolescents in the western cape region of South Africa. Child Indicators Research, 9(3), 701-713. doi:10.1007/s12187-015-9334-2

29.

Scheier

M. F.

Carver

C. S.

Bridges

M. W.

(1994). Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-esteem): A reevaluation of the life orientation test. Journal of personality and social psychology, 67(6), 1063. doi:10.1037//0022-3514.67.6.1063

30.

Seligson

J. L.

Huebner

E. S.

Valois

R. F.

(2003). Preliminary validation of the brief multidimensional students’ life satisfaction scale (BMSLSS). Social Indicators Research, 61(2), 121-145. doi:10.1023/A:1021326822957

31.

Shadlow

J. O.

Boles

R. E.

Roberts

M. C.

Winston

(2015). Native American children and their reports of hope: Construct validation of the children’s hope scale. Journal of Child and Family Studies, 24(6), 1707. doi:10.1007/s10826-014-9974-8

32.

Silverberg

S. B.

(1990). Psychological well-being of parents with early adolescent children. Developmental Psychology. 26(4), 658-666. doi:10.1037/0012-1649.26.4.658

33.

Snyder

C. R.

Cheavens

Michael

S. T.

(2005). Hope theory: History and elaborated model. In Eliott

J. A.

(ed.), Interdisciplinary perspectives on hope. Nova Science Publishers.

34.

Snyder

C. R.

Harris

Anderson

J. R.

Holleran

S. A.

Irving

L. M.

Sigmon

S. T.

Harney

(1991) The will and the ways: Development and validation of an individual-differences measure of hope. Journal of Personality and Social Psychology, 60(4), 570-585. doi:10.1037/0022-3514.60.4.570

35.

Snyder

C. R.

Hoza

Pelham

W. E.

Rapoff

Ware

Danovsky

Stahl

K. J.

(1997). The development and validation of the Children’s Hope Scale. Journal of Pediatric Psychology, 22(3), 399-421. doi:10.1093/jpepsy/22.3.399

36.

Steenkamp

J. B. E. M.

Baumgartner

(1998). Assessing measurement invariance in cross‐national consumer research. Journal of Consumer Research, 25(1), 78-107. doi:10.1086/209528

37.

Mcbride

R. E.

Xiang

(2015). College students’ achievement goal orientation and motivational regulations in physical activity classes: A test of gender invariance. Journal of Teaching in Physical Education, 34(1), 2-17. doi:10.1123/jtpe.2013-0151

38.

Tian

Zhang

Huebner

E. S.

(2015). Preliminary validation of the brief multidimensional students’ life satisfaction scale (BMSLSS) among Chinese elementary school students. Child Indicators Research, 8(4), 907-923. doi:10.1007/s12187-014-9295-x

39.

Tsaousis

Kazi

(2013). Factorial invariance and latent mean differences of scores on trait emotional intelligence across gender and age. Personality and Individual Differences, 54(2), 169–173. doi:10.1016/j.paid.2012.08.016

40.

Vandenberg

R. J.

Lance

C. E.

(2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3(1), 4-70. doi:10.1177/109442810031002

41.

Vassar

Bradley

(2010). A reliability generalization study of coefficient alpha for the life orientation test. Journal of personality assessment, 92(4), 362-370. doi:10.1080/00223891.2010.482016

42.

Wang

M. F.

Yuan

C. C.

Yang

(2013). Parents’ child-rearing sex-role attitudes and its relationship with their gender schematicity. Chinese Journal of Clinical Psychology, 21(4), 646-649. doi:10.16128/j.cnki.1005-3611.2013.04.018

43.

White

S. A.

(2010). Goal orientation and perceptions of the motivational climate initiated by parents. Pediatric Exercise Science, 8(2), 122-129. doi:10.1123/pes.8.2.122

44.

Wong

S. S.

Lim

(2009). Hope versus optimism in Singaporean adolescents: Contributions to depression and life satisfaction. Personality and Individual Differences, 46(5), 648-652. doi:10.1016/j.paid.2009.01.009

45.

Zeng

Ling

Huebner

E. S.

Lei

(2017). The psychometric properties of the 5-item gratitude questionnaire in Chinese adolescents. Journal of Psychiatric and Mental Health Nursing, 24(4), 203-210. doi:10.1111/jpm.12372

46.

Zhao

Sun

(2011). Reliability and validity of the Chinese version of children’s hope scale. Chinese Mental Health Journal, 25, 454-459. doi:10.3969/j.issn.1000-6729.2011.06.010

47.

Zhao

Wang

Kong

(2014). Exploring the mediation effect of social support and self-esteem on the relationship between humor style and life satisfaction in Chinese college students. Personality and Individual Differences, 64(8), 126-130. doi:10.1016/j.paid.2014.02.026

48.

Zhao

Zhu

(2009). Comparison study on basic rights between Chinese, Japanese, Korean, and American high school students. China Youth Study, 6, 1-7. doi:10.19633/j.cnki.11-2579/d.2009.06.017

49.

Zimet

G. D.

Dahlem

N. W.

Zimet

S. G.

Farley

G. K.

(1988). The multidimensional scale of perceived social support. Journal of personality assessment, 52(1), 30-41. doi:10.1207/s15327752jpa5201_2