Introduction

Adaptation is a psychological process in which individuals actively develop and adjust themselves in response to changes in the objective environment and then harmonize and balance with the external environment (Piaget, 1976). Learning adaptation is an essential aspect of adaptation (Amirbagloie-Daryani et al., 2022). When college students face repeated and intense changes in the environment and learning patterns of university life, learning adaptation is critical to ensure their high-quality academic output as well as their healthy physical and mental development (Chenari et al., 2022; Conefrey, 2021; Lent et al., 2009). The problems following maladaptive learning are an important cause of limitations to college students’ mental health and academic performance and a serious barrier to educational progress (Chenari et al., 2022; Zhao, 2020). A recent qualitative study found that many college students perceive college learning and life as stressful and challenging and do not adapt well (Chenari et al., 2022). However, maladjustment in learning is a multifactorial phenomenon (Amirbagloie-Daryani et al., 2022). Hence, identifying the factors that influence college students’ learning adaptation is critical.

Past research has generally suggested that personal and environmental factors are two main factors that influence individual learning adaptation (Amirbagloie-Daryani et al., 2022; Chenari et al., 2022; Feng et al., 2006). Personal factors include personality traits such as optimism, gratitude, resilience, self-efficacy, and psychological capital (Chemers et al., 2001; Li et al., 2022; Sadoughi, 2018), while environmental factors involve social support, teacher support, and parent–child relationships (Lee, 2020; Li et al., 2022; Tomás et al., 2020). Hardiness is a positive personality trait that has been widely noted in recent studies as an important personal factor in coping with change and stress management (Maddi, 2002) and has been shown to have a positive predictive effect on learning adaptation (Ghanbari et al., 2013; Oktavia et al., 2019). In addition, studies have found that teacher support has the highest explanatory power for maladjustment in learning compared to other environmental factors (e.g., support from peers or parents) (Malecki & Demaray, 2003). Teacher support has already been proved to be more closely related to adolescent school adaptation (Longobardi et al., 2016; Tomás et al., 2020) and to have a significant positive impact on students' learning adaptation (Amai, 2021; Burns et al., 2018Claudio et al., 2016). It can thus be suggested that both hardiness and teacher support may have direct impacts on college students’ learning adaptation.

Meanwhile, research has shown that students with hardiness are more likely to receive teacher support (Duan & Wu, 2016), and teacher support has often played an important mediating role in recent empirical studies (Schweder & Raufelder, 2019; Wang et al., 2021). It can thus be inferred that teacher support may have a mediating role in the relationship between hardiness and learning adaptation among college students. An only child is defined as a child who does not have any siblings (Burke, 1956; Cai et al., 2012). The differences in health, personality traits, academic performance, and behavioral patterns brought by only-child (OC)/non-only-child (NOC) status have aroused great interest of Western and Chinese researchers (Burke, 1956; Falbo & Poston, 1993; Falbo et al., 1989; Jia et al., 2021; Kolm & Ytheir, 2006; Liu et al., 2022; Zheng et al., 2022). However, research findings on the difference between OC/NOC status in learning, adaptability, and interpersonal interactions remain divergent (Burke, 1956; Cameron et al., 2013; Falbo et al., 1989; Zheng et al., 2022). Since the One Child Policy (OPC) has been implemented in China for over 30 years (from 1982 to 2015), China has the largest number and proportion of only children worldwide (Falbo & Poston, 1993). The difference between OC/NOC status should be considered by researchers, both academically and practically. The current study would therefore examine the moderating role of OC/NOC status in the relationship between the indirect effects of hardiness on Chinese college students’ learning adaptation through teacher support.

Based on Social Cognitive Theory (SCT), in which the individual, environment, and behavior mutually interact (Bandura, 1986), the current study suggested that hardiness as a personal factor has a direct effect on college students' learning adaptation behavior, while it can indirectly impact learning adaptation through the environmental factor of teacher support. Empirical studies to date have paid less attention to the possible combined effects of hardiness and teacher support on college students' learning adaptation and their internal mechanisms of action, especially the lack of considering the difference in the OC/NOC status. Therefore, the current study constructed a moderated mediation model to explore these combined effects. It may not only enrich our understanding of how these variables are closely related to each other, but also provide a reference for universities to improve college students' learning adaptation.

Literature Review

Social Cognitive Theory

SCT, proposed by Bandura (1986), argues that personal, environmental, and behavioral factors mutually interact with each other. Currently, the SCT has been widely used in the studies of learning adaptation (Burns et al., 2018; Conefrey, 2021; Lent et al., 2009). Specifically, Conefrey (2021) considered high-impact practices as environmental factors, self-efficacy and belongingness as personal factors, and college students’ school adaptation as behavioral factors based on SCT. The results of the Conefrey (2021) study revealed that high-impact practices could positively influence college students’ school adaptation via enhancing self-efficacy and belongingness. Based on SCT, Burns et al. (2018) conducted a longitudinal study to explore the effects of social support from parents, peers, and teachers (environmental factors) on high school students’ self-efficacy, perceived control, adaptability, and personal best goals setting (personal factors); they also examined the effects of self-efficacy, perceived control, and adaptability on personal best goals setting and the effects of personal best goals setting on academic engagement and achievement (behavioral factors). Their results indicated that social support significantly predicted adaptability and self-efficacy; adaptability, self-efficacy, and teacher support significantly predicted the personal best goal setting; and personal best goal setting significantly predicted academic engagement and achievement (Burns et al., 2018). Based on SCT, Lent et al. (2009) found that academic self-efficacy (a personal factor) and environmental support (an environmental factor) positively predicted goal progress and academic adjustment (behavioral factors) among college students, and the learning adaptation positively predicted students’ life satisfaction.

Based on the above discussion, the current study considered hardiness as a personal factor, teacher support as an environmental factor, and the learning adaptation of college students as a behavioral factor. In the meantime, the OC/NOC status was included as a moderating variable, and a moderated mediation model was constructed to explore the influencing mechanism of hardiness on college students' learning adaptation via teacher support and the moderating effect of OC/NOC status.

Hardiness and Learning Adaptation

Hardiness as a positive personality trait in stress management was introduced in the psychology research by Kobasa (1979). On a basis of existentialist psychology, hardiness is defined as attitudes, beliefs, and behaviors that can help individuals manage stress and avoid physical and mental illness in the face of external stress and changes (Kobasa, 1979; Maddi, 2002). When environmental changes occur, an individual with hardiness typically has greater foresight and control and is adept in using positive factors and in adopting better response behavior, and regards the environmental changes as opportunities for growth, rather than negative acceptance and incompetence (Kobasa et al., 1981; Maddi, 2002). Similarly, prior studies have revealed that hardiness was significantly and positively correlated with student psychological health, learning engagement, learning adaptation, and academic achievement (Elhampour et al., 2019; Sadeghi & Einaky, 2020), demonstrating that hardiness is an indispensable personality trait for student adaptation and learning.

Learning adaptation refers to the psychological and behavioral process in which an individual makes positive and proactive self-adjustments to attain a balanced mind and behavior with the learning environment following changes in the external environment and in learning conditions (Feng et al., 2006). In particular, learning adaptation is a complex interaction between college students and the environment (Martinez-Lopez et al., 2019). Faced with the challenge of higher learning requirements in the higher education stage, college students need stronger self-management ability and greater autonomy to make proactive self-adjustment, so as to maintain a balanced and good learning state (Credé & Niehorster, 2012). For college students, the process of learning adaptation is also a socialization process; the ability of learning adaptation is critical for their development of physical and mental health as well as their academic performance (Chenari et al., 2022; Zhao, 2020). Thus, identifying the influencing factors for college students’ learning adaptation is essential. Positive psychology has explained the positive impact of positive personality traits on individual behavior (Csikszentmihalyi & Seligman, 2000; Sullivan, 2019). A previous empirical study has indicated that hardiness enhances the adjustment of university students (Ghanbari et al., 2013). Several studies have confirmed the positive predictive effect of hardiness on the career adaptability of both college students and employees (Coetzee & Harry, 2015; Ndlovu & Ferreira, 2019). Also, studies have shown that hardiness as an intrinsic resource enhances the overall development of individuals (Shukshina et al., 2019), and better relieves academic stress (Abdollahi et al., 2020). Hardy individuals exhibit strong self-adjustment abilities to proactively adapt to changes in the learning environment or conditions (Oktavia et al., 2019), which has a critical impact on students’ learning adaptation (Ghanbari et al., 2013; Soheili et al., 2021). Accordingly, Hypothesis 1 was established as follows:

H1: Hardiness has a significantly positive predictive effect on college students’ learning adaptation.

The Mediating Role of Teacher Support

Teacher support refers to a teacher providing care and guidance for student learning, understanding and considering students’ emotional changes, and treating students objectively and equally (Sakiz et al., 2012); it also denotes the assistance provided for student learning (Brophy & Good, 1970). While perceived supportive teacher behaviors influence students' psychological changes and learning activities, student performance affects teacher expectations and thus supportive teacher behaviors in turn, which is a bidirectional process (Braun, 1976). Teacher support can help students overcome problems of school maladjustment (Amirbagloie-Daryani et al., 2022; Malecki & Demaray, 2003). It is also a crucial predictor of college students' satisfaction and happiness (Suldo et al., 2009) as well as academic skills (Credé & Niehorster, 2012).

Previous studies have established a relationship model between hardiness and health and proved that hardiness prompts an individual to obtain greater social support and assistance, improving social adaptation (Maddi, 2002; Maddi & Khoshaba, 1994). Teacher support is a crucial element of social support (Malecki & Demaray, 2003; Moreira & Lee, 2020). Prior studies have also indicated that hardy individuals can obtain greater attention and higher expectations from their teachers, thus gaining greater teacher support academically and emotionally (Duan & Wu, 2016). SCT suggests that individuals, the environment, and behavior have mutual interactions (Bandura, 1986); prior studies have typically considered teacher support as a crucial environmental factor exerting a positive effect on an individual’s adaptation and learning behavior (Amirbagloie-Daryani et al., 2022; Burns et al., 2018). Besides, compared with other supportive environmental factors (e.g., support from peers or parents), teacher support is more important to students’ learning abilities and adaptation; thus, it is a crucial predictor of maladapted student learning (Malecki & Demaray, 2003; Tomás et al., 2020). An interview study from Amai (2021) found that teachers’ supportive behaviors with non-help-seeking students in secondary school were effective in identifying their mental health problems in a timely manner and in enhancing school adjustment. Teacher affirmations and positive support help students alleviate psychological crises and improve their learning adaptability (Amai, 2021; Reddy et al., 2003). Several studies have previously explored teacher support as a mediating factor (Schweder & Raufelder, 2019; Wang et al., 2021). Therefore, the current study suggested that hardiness might influence college students’ learning adaptation via teacher support. Accordingly, Hypothesis 2 was established as follows:

H2: Teacher support exerts a mediation effect on the relationship between hardiness and college students’ learning adaptation.

The Moderating Role of OC/NOC Status

The perspectives of parental differential treatment suggest that differences in parenting between OC/NOC families bring differences in personality traits, behaviors, and adjustment (Kolm & Ytheir, 2006). Previous studies on differences in OC/NOC status generally exhibit three different views. Firstly, an early study reported on teachers' appraisal of only children and non-only children, with teachers perceiving the main disadvantages of only children as being more conceited and aggressive (Fenton, 1928). Some studies noted that compared with non-only children, only children were more likely to lack the abilities of communication, peer cooperation, and adaptation (Minuchin, 2018); might be more pessimistic and less competitive (Cameron et al., 2013), and experience maladjustment (Falbo, 2018; Rosenkranz, 2004).

Secondly, in contrast, other studies reported different findings suggesting that only children could obtain more parental support and family resources (Lao & Dong, 2019; Marjoribanks, 1991), possess better intellectual skills (Falbo & Polit, 1986), and have stronger leadership skills (Smith, 1984) than non-only children. Studies have found that only children could exhibit better academic performance (Jia et al., 2021; Lao & Dong, 2019), perceive better school feelings (Jia et al., 2021) and more teacher support (Liu et al., 2022), and have higher achievement motivation, better interpersonal skills, and greater personal adaptability (Polit & Falbo, 1987).

Thirdly, Wang et al. (2020) have noted that interpersonal communication skills were slightly higher for only children when not controlling for socio-demographic factors (e.g., gender, family economic income, parental education); however, after controlling for socio-demographic factors, the differences in interpersonal communication between only children and non-only children were not significant. Another related study reported no significant differences in the adaptive and social skills among only and non-only children (Falbo & Polit, 1986). A comparative study on only children and non-only children in both Chinese and Western discovered that, except for a slight advantage for only children in the area of learning, there was essentially no difference between only children and non-only children in adaptive skills and personality traits, and the results were similar in China and the west (Poston & Falbo, 1990).

In conclusion, it can be seen from previous studies that opinions diverge greatly on the differences in perceptions of parental and teacher support, learning, and adaptabilities between only children and non-only children. However, a series of past empirical studies by Poston and Falbo have found no significant differences in personality traits between only children and non-only children in China and the West (Falbo & Poston, 1993; Falbo et al., 1989; Poston & Falbo, 1990). Their findings have been reconfirmed in recent empirical studies (Nair, 2020; Shanmugavinayagam et al., 2019). Thus, we speculate that only children and non-only children may not differ much in hardiness. Notably, the studies of Blake, (1981) and Minuchin, (2018) all indicated that only children may lack cooperation, interpersonal communication skills, and adaptability due to the lack of peer companionship and support from siblings. The only children were more likely to have maladaptive problems (Falbo, 2018) and needed more social and emotional support during growth (KOCATÜRK, 2021). Therefore, this study speculated that college students who were only children may be more likely to benefit from teacher support to improve their learning adaptation. Also, recent studies have focused on the essential moderating role of OC/NOC status (Jin et al., 2019; Liu et al., 2022). This study suggested that OC/NOC status may exert a moderate role in the second half of the mediating effect of hardiness on college students' learning adaptation through teacher support. Accordingly, hypothesis 3 was established as follows:

H3: The OC/NOC status moderates the influence of teacher support on college student learning adaptation.

In summary, this research established a hypothesis model (Fig. 1).

Fig. 1
figure 1

Hypothetical model

Method

Participants and Procedure

The current study adopted convenience sampling to investigate students from colleges A, B, and C in the Hunan province of China. These three colleges had been designated by the Ministry of Education of the People’s Republic of China as pilot normal colleges providing free-tuition education to train elementary school teachers, with a total of 13,000 teacher-training students who had been enrolled for free and adapt well. College A was an outstanding institution for teacher training and the first tuition-free pilot normal college set up in Hunan Province, with 5,200 teacher-training students. College B was a comprehensive normal university and the second batch of tuition-free normal institutions in Hunan, with 3,900 teacher-training students. College C was a comprehensive university located in a relatively remote area, with 3900 teacher-training students.

Following the Declaration of Helsinki (Goodyear et al., 2007), the current study recruited participants who were interested in the topic of this study and volunteered to participate from the three colleges mentioned above. The head teachers of classes responsible for distributing the questionnaire were trained about the questionnaire entries before recruitment. Moreover, before distributing the online questionnaires, the head teachers of classes were required to inform the participants of the following information:

  1. a)

    the introduction of the survey, the purpose of the study, and the confidentiality agreement;

  2. b)

    the questionnaires were submitted anonymously, and the data were processed confidentially;

  3. c)

    participants could refuse or withdraw from the study at any time if they had any concerns during the filling process.

After participants gave their informed consent, questionnaires were distributed through Questionnaire Star (www.wjx.cn), a widely used online questionnaire app in China. The filling online questionnaire processes were under the help of the head teachers of classes. Participants could scan a QR code (a two-dimensional barcode containing a link to the online questionnaire) or click on a link directly to fill in the online questionnaire and complete the survey.

In the pretest of this study, a total of 400 questionnaires were distributed. In the formal test, 1147 college students participated in the questionnaire survey after excluding those who had taken the pretest. Nine invalid questionnaires with short response times and missing values were excluded, and 1138 valid questionnaires were collected, with an effective rate of 99.2%. According to the formula proposed by Israel (1992) for calculating the sample size (sample size = z2 × p(1–p)/e2/1 + (z2 × p(1–p)/e2N), z = 2.58, p = 0.5, N = 13,000, and e2 = 0.0025), the required sample size of the formal test should be no less than 634. Therefore, the sample size of this study met the criteria. In the sample structure, 335 (29.4%) were male and 803 (70.6%) were female, 634 (55.7%) were urban and 504 (44.3%) were rural in family location, 778 (68.4%) were only children and 360 (31.6%) were non-only children. The imbalance between males and females in the sample might be because the sample schools were all normal colleges with a larger proportion of female students (Liu et al., 2022). The imbalance between only children and non-only children in the sample might be because China had implemented the OCP for more than 30 years with a larger proportion of only children (Falbo & Poston, 1993). Therefore, the sample structure of this study was consistent with the natural composition.

Data Analysis

First, SPSS (version 21.0, IBM, Armonk, NY, USA) was used for sample descriptive statistics (means and standard deviations), correlation analysis, scale reliability test, and common method variance (CMV) test. AMOS 21.0 (version 21.0, IBM, Armonk, NY, USA) was used for confirmatory factor analysis (CFA). Second, Hayes Process of Model 4 was used to test the mediation effect, and Hayes Process of Model 14 was used to test the moderated mediation effect. Finally, the bootstrap confidence interval (CI) was employed to confirm the significance of the mediation effect and the moderating effect (Hayes et al., 2017).

Meanwhile, studies pointed out that the mediation model should be divided into different types (Nitzl et al., 2016; Zhao et al., 2010):

  1. a)

    the full mediation: when the direct effect is insignificant and the indirect effect is significant, it indicates a full mediation;

  2. b)

    the partial mediation: when both direct and indirect effects are significant, it indicates a partial mediation. In the case of partial mediation, two sub-types of partial mediation can be further identified as the complementary partial mediation (the direct effect and indirect effect point in the same direction, the same positive or negative) and the competitive partial mediation (the direct effect and indirect effect point in different directions);

  3. c)

    the only direct effect: when a direct effect exists, but there are no indirect effects, it indicates the only direct effect;

  4. d)

    no effect: neither direct nor indirect effects exist.

The current study tested the type of mediation model accordingly. In addition, four possible models were constructed to determine the best model fitting the theoretical framework and reflecting the actual information through model competition (Hayes et al., 2017).

Measure

Hardiness Scale

The Hardiness Scale established by Lu and Liang (2008) was adopted. The scale included four dimensions—control, challenge, input and resilience—with 27 items. A 5-point Likert scale was used for ratings without reversed questions (see Appendix Table 7). An example item is “When I encounter difficulties, I always try to find solutions to them.” The item analysis on pretest samples demonstrated that (a) the critical ratio (C.R.) values for all items were greater than 3; (2) the correlation coefficients between each item and the total scale were greater than 0.4; and (3) the Cronbach’s α did not increase after the deletion of items. The item analysis results above met the criteria for retention of items (Nunnally & Bernstein, 1994), indicating that all the items in the scale could be preserved for the formal test. For the formal test samples, the Cronbach’s α of the total scale was 0.94, greater than 0.7 (Nunnally, 1978); the results of CFA indicated the standardized factor loadings ranging from 0.63 to 0.85, all of which were greater than 0.5 (Hair et al., 1992). Thus, the scale had favorable reliability and validity. The results for the model fit index are presented in Table 1 and indicated a favorable model fit for the models for this scale (McDonald & Ho, 2002).

Table 1 Model fit index of the scales

Learning Adaptation Scale

The Learning Adaptation Scale established by Feng et al. (2006) was used. The scale consisted of five dimensions—learning motivation, teaching mode, learning ability, learning attitude, and environmental factors—with 29 items. A 5-point Likert scale was used for ratings (see Appendix Table 8). An example item is “I feel I can adapt to the study in college.” Three items used reverse scoring among the items of the dimension of learning motivation. All items were reverse-scored for the dimensions of teaching mode and learning attitude. In the data analysis, reverse scoring was employed for these items. The item analysis on pretest samples demonstrated that (a) the C.R. values for all items were greater than 3; (2) the correlation coefficients between each item and the total scale were greater than 0.4; and (3) the Cronbach's α did not increase after the deletion of items. Since the item analysis results met the criteria for retention of items, all the items in the scale could be preserved for the formal test. For the formal test samples, the Cronbach's α of the total scale was 0.94, greater than 0.7; and the CFA results indicated that the standardized factor loadings ranged from 0.64 to 0.85, all greater than 0.5. Thus, the scale had favorable reliability and validity. The results for the model fit are presented in Table 1, indicating a favorable model fit of the models for this scale.

Student Perception of Teacher Supportive Behavior Questionnaire

The Student Perception of Teacher Supportive Behavior Questionnaire established by Ouyang (2005) was used. The scale consisted of three dimensions—learning support, emotional support, and ability support—with 19 items. A 5-point Likert scale was used for ratings. An example item is “My teacher often gives me encouragement in study and life.” The dimension of learning support had one reverse-scored item (1s5); thus, in the data analysis, reverse scoring was employed for this item. An item analysis following the pretest indicated that (a) there were two items (ls8 and ls9) in the dimension of learning support with C.R. values less than 3; (b) the correlation coefficients between the two items and the total scale were less than 0.4; (c) after deleting the two items, the Cronbach’s α of the pretest samples was increased. The item analysis results met the criteria for the deletion of items. Therefore, these two items were removed, and the remaining 17 items were adopted in the formal test (see Appendix Table 9). The Cronbach’s α of the total scale for the formal test was 0.93 and greater than 0.7; the results of CFA indicated that the standardized factor loadings ranged from 0.73 to 0.87, all of which were greater than 0.5. Thus, the scale had favorable reliability and validity. The results for the model fit are presented in Table 1, indicating a favorable model fit of the models for the scale.

Results

Common Method Variance (CMV) Test

Harman’s One-Factor Test was used to assess the CMV. The results of a verification based on unrotated factor analysis indicated that the KMO was 0.98 (> 0.8), and the Bartlett test of sphericity was significant (p < 0.001). The explanatory power of the first factor was 28.67%, which was below the critical value (40%) (Podsakoff et al., 2003), indicating that CMV was not significant in this study.

Descriptive Statistics and Correlation Analysis

Table 2 presents the descriptive statistics and the correlation analysis of three variables. The correlation analysis indicated a significantly positive correlation of hardiness with learning adaptation (r = 0.51, p < 0.001); hardiness was significantly and positively correlated with teacher support (r = 0.56, p < 0.001); teacher support was significantly and positively correlated with learning adaptation (r = 0.55, p < 0.001). The correlation coefficient between any two of the three variables was uniformly less than 0.8, indicating no collinearity problem and that the next step of linear regression analysis could be performed (Benesty et al., 2009).

Table 2 Descriptive statistics and correlation analysis

The descriptive statistics and correlation analysis of each dimension are shown in Table 3. The correlation analysis illustrated a significant low-to-medium correlation for each dimension and no high correlation dimension exists (Benesty et al., 2009).

Table 3 Descriptive statistics and correlation analysis of each dimension

Competition of Models

Four possible models were proposed based on suggestions from Hayes et al. (2017). Model 1 was a benchmark mediation model in which hardiness influences learning adaptation via teacher support (HA → TS → LA). Model 2 was a moderated mediation model in which the OC/NOC status moderated the first half of the mediating effect path (HA → TS). Model 3 was a moderated mediation model in which the OC/NOC status moderated the second half of the mediated effect path (TS → LA). Moreover, model 4 was a moderated mediation model in which the OC/NOC status moderated both the first and second half of the mediating effect path (HA → TS and TS → LA). As shown in Table 4, among the four possible models, the moderating effect of OC/NOC status only reached statistical significance in the second half of the mediation model (TS → LA), indicating that Model 3 was the best model that fits the theoretical framework of this study and reflected the actual information.

Table 4 Competition of models

The Mediating Role of Teacher Support

The Hayes Process of Model 4 was used to test the mediation effect of teacher support. Table 5 presents the results. In model 1, hardiness was a significant positive predictor for learning adaptation (B = 0.52, p < 0.001); in model 2, hardiness also significantly and positively predicted teacher support (B = 0.64, p < 0.001). Following the inclusion of teacher support in model 3 as a mediating variable, hardiness still significantly and positively predicted learning adaptation (B = 0.30, p < 0.001), but its predictive power was lower than that in model 1. Moreover, teacher support also significantly and positively predicted learning adaptation (B = 0.34, p < 0.001), demonstrating the partial mediation of teacher support for the influence of hardiness on learning adaptation. Bias-corrected nonparametric percentile bootstrapping was further used to verify the mediation effect of teacher support. The results demonstrated that (a) the indirect effect value was 0.22, with 95% CI ranging from 0.16 to 0.28, excluding 0; (b) the direct effect value was 0.30, with 95% CI ranging from 0.21 to 0.39, excluding 0; and (c) the mediation effect accounted for 41.97% of the total effect. Thus, the significance of teacher support’s partial mediation effect was again verified. Meanwhile, both the direct effect (B = 0.30, p < 0.001) and the indirect effect (0.64 × 0.34 = 0.22, p < 0.001) of hardiness on learning adaptation were positive. The direct effect and indirect effect pointed in the same direction, indicating a complementary partial mediation in the current study.

Table 5 Testing the mediation model of teacher support

The Moderated Mediation Model

In order to test whether the OC/NOC status had a moderating effect, the Hayes Process of Model 14 was used to test the moderated mediation model. The results, presented in Table 6, demonstrated that the interaction term between teacher support and the OC/NOC status significantly predicted learning adaptation (B = 0.11, p < 0.05), indicating that the OC/NOC status moderated the second half of the mediation model of hardiness affecting learning adaptation via teacher support. A test based on bias-corrected nonparametric percentile bootstrapping was conducted. The result indicated that (a) the OC/NOC status moderated the indirect effect of hardiness on learning adaptation via teacher support; and (b) the index of moderated mediation was 0.07 (LLCI = 0.14, ULCI = 0.01), with CI excluding 0, indicating the significance of the moderated mediation. This indirect effect of hardiness on learning adaptation via teacher support was stronger in the college students who were only children (B = 0.24, LLCI = 0.19, ULCI = 0.29) than in those who were non-only children (B = 0.16, LLCI = 0.10, ULCI = 0.23). A further simple slope analysis (Fig. 2) indicated that teacher support was a stronger predictor for the learning adaptation in the college students who were only children (simple slope = 0.37, t = 12.98, p < 0.001) than those who were non-only children (simple slope = 0.26, t = 6.17, p < 0.001).

Table 6 Testing the moderated mediation model
Fig. 2
figure 2

The moderating effect of OC/NOC status on the relationship between teacher support and learning adaptation

Discussion

Hardiness and Learning Adaptation

The results confirm H1 that hardiness has a significantly positive predictive effect on college students’ learning adaptation, consistent with prior studies (Ghanbari et al., 2013; Oktavia et al., 2019). This study verifies the relationship between hardiness and college students’ learning adaptation, demonstrating that hardiness could prompt college students to improve their abilities of learning adaptation. In addition, this study further extends the study of Zhao (2020) on strategies to improve learning adaptation among Chinese college students, indicating that excavating and cultivating students’ positive personality traits can help improve their learning adaptation. The findings of our study provide empirical research evidence and support the argument of positive psychology that hardiness is a positive personality trait that exerts a positive effect on individual behavior (Csikszentmihalyi & Seligman, 2000; Sullivan, 2019). This study maintains that college students with higher hardiness are more adept at responding to environmental changes and at using positive factors to conduct proactive self-adjustments instead of passive acceptance. They regard conditions of learning or environmental changes as opportunities and challenges for growth, and they thus work hard to enhance their psychological adaptation and accomplish learning adaptation quickly (Oktavia et al., 2019).

The Mediating Role of Teacher Support

The research results support H2 and reveal that teacher support exerts a complementary partial mediation effect on the relationship between hardiness and college student learning adaptation. It is indicated that college students with higher hardiness can get more expectations and support from teachers, thus improving their learning adaptability. The findings support the hardiness model (Maddi, 2002; Maddi & Khoshaba, 1994) and confirm that hardiness positively affects teacher support. The relationship between hardiness, teacher support, and learning adaptation has also been examined in a past study reporting that hardiness moderated the relationship between teacher support and learning adaptation (Chen & Tu, 2019). The previous studies exploring the effect of hardiness on adaptability may lack an examination of mediating variables (Payandeh et al., 2013). Therefore, the present study extends the past studies by further identifying a complementary partial mediation effect of teacher support in the relationship between hardiness and learning adaptation among college students. It suggests that teacher support is a critical facilitator for college students’ learning adaptation (Credé & Niehorster, 2012; Tomás et al., 2020). Teacher support can complement college students’ hardiness, effectively prevent their adaptability problems, and help them adapt to college studies and life as soon as possible. The findings also support the SCT that mutual interactions exist among the person, environment, and behavior (Bandura, 1986). Specifically, hardiness (a personal factor) influences college students’ learning adaptation (a behavioral factor) through teacher support (an environmental factor). Therefore, the results further enhance our understanding of the internal mechanisms between hardiness and college student learning adaptation.

The Moderating Role of OC/NOC Status

The research results endorse H3 and note that OC/NOC status moderates the relationship between teacher support and learning adaptation. More specifically, teacher support has a stronger effect on learning adaptation in college students who are only children. This study is the first to introduce the OC/NOC status as the moderator in the mediation model of the effect of hardiness on learning adaptation via teacher support. The findings support the perspectives of parental differential treatment. That is, parenting differences between the OC/NOC families result in differences in an individual's ultimate adaptability and behavior (Kolm & Ytheir, 2006). It is consistent with the results of a prior study (Liu et al., 2022). This finding may result from the fact that college students who are only children grow up in comparatively monotonous environments. They usually lack support from and collaboration spirits with siblings and companions while growing up, leading to inferior interpersonal communication skills and adaptability (Blake, 1981; Minuchin, 2018) and requiring more support and assistance from their teachers (KOCATÜRK, 2021) for learning adaptation. College students who are non-only children get more support from siblings as well as exchanges and cooperation with companions; thus, they are more adaptable.

Conclusion, Suggestions and Limitations

Overall, this study explored the effects of hardiness on college students’ learning adaptation, the mediating role of teacher support, and the moderating role of OC/NOC status. It is verified that hardiness not only directly influenced college students’ learning adaptation, but also indirectly affected learning adaptation via the mediating role of teacher support. The OC/NOC status moderated the effect of teacher support on learning adaptation. The effect of teacher support on learning adaptation was stronger for only children college students than for non-only children college students. These findings support the SCT, the hardiness model, and the perspectives of parental differential treatment, while enriching our understanding of the combined effects of hardiness, teacher support, and the OC/NOC status on college students’ learning adaptation.

Based on the results, the following suggestions are proposed: First, colleges and universities can incorporate the hardiness training course into the introductory mental health curriculum and enhance college students’ hardiness level through class learning and activities. Second, college teachers should give full play to their supportive role in discovering their students’ difficulties and discomfort in study and life and then provide encouragement and help in time. Third, while cultivating and enhancing students’ hardiness, colleges and universities should also organize training, workshops, and seminars to strengthen teachers’ abilities and provide better support. Fourth, college teachers should consider the differences in OC/NOC status regarding students’ learning adaptation abilities and give different attention to them.

There are two major limitations to this study. First, this study only conducted a questionnaire survey among students in three colleges in Hunan Province of China, and future studies may consider further expanding the geographical scope of the sample. Second, this study was a cross-sectional study. Although it revealed the predictive relationships between variables, it cannot determine their causal relationships. Future research may consider combining longitudinal and experimental studies.