Students with learning disabilities (LD) have difficulty with their basic learning abilities, such as listening, speaking, reading, writing, and calculating, because of problems in their neuropsychological functions. LD also has negative effects on an individual’s learning process and social emotions. The self-efficacy and motivation of students with LD are lower than those of students without LD. Studies have reported that students with LD have poor mental health. The presence of an LD is a risk factor for mental health concerns, and the identification of mental health risks and protective factors for individuals with LD is necessary to design effective intervention models. Most previous relevant studies have recruited students without LD, and only a few have recruited those with LD.

A few studies have employed multiple methods that combine person-focused strategies with variable-focused strategies because they offer complementary methodological strengths. Person-focused models can be used to identify resilient people and understand how resilience develops by comparing the resilient with those who are not, namely those who are not faring well in adverse situations and have not experienced any threats to their development. A variable-focused approach focuses on the relationship between the indices of stress or adversity and the influencing factors of resilience and psychosocial functions. In summary, person-focused methods recognize the functions of a person as a living system, whereas variable-focused methods can be used to delve deeper into specific processes. The present study attempted to use these two approaches to attain a more comprehensive understanding of resilience among students with LD. In previous studies, students with LD were not considered the reference group; thus, we lack evidence on whether the resilience of students with LD differs from that of students with LD. Finally, this study explored the resilience mechanism of students with LD and can be used as a reference to determine how to provide support and assistance to students with poor mental health.

The study group of this research was composed of students with and without LD who were in grades 3 to 5 at elementary schools in Kaohsiung City. They were selected during the second semester of 2017 using the stratified purposeful sampling method. With parents’ permission, 802 students participated in this study, including 275 third graders, 269 fourth graders, and 258 fifth graders from 13 elementary schools. A total of 375 students had LD, including 113 third graders, 132 fourth graders, and 129 fifth graders from the resource classes in 55 elementary schools. Given the reading ability of students with LD, reducing the length of the questionnaire was necessary so that they could complete the survey. The study instruments included protective factors, mental health questionnaires, and academic achievement tests. The questionnaire on protective factors and mental health included five internal protective factors (i.e., physical self-concept, emotional self-concept, mathematics self-concept, Chinese self-concept, and self-efficacy) and three external protective factors (i.e., teacher support, peer support, and parental support), as well as mental health. The items in the questionnaire were scored from 4 (strongly agree) to 1 (strongly disagree), with higher scores indicating greater levels of performance. The academic achievement test included their achievements in Chinese and mathematics. These two instruments have demonstrated satisfactory reliability and validity. For the person-focused approach, students with LD were classified into four groups through cluster analysis on the basis of the indices of both their mental health and academic achievement. A two-step method was applied to cluster the students. The protective factors in the different clusters of students with LD were compared. Furthermore, this study compared the protective factors in different clusters to determine whether there were significant differences between students with and without LD. For the variable-focused approach, the direct and the indirect modes were used to examine which protective factors could be used to predict the mental health of students with and without LD. A path analysis was used to test the direct and indirect model between the protective factors and mental health. In the direct model, the researcher assumes that each protective factor is positively correlated with mental health whereas the risk factors (mathematics and Chinese achievement) are negatively correlated. The indirect model was used to examine the hypothesis that self-efficacy acts as a mediator in the relationship of social support and self-concept with mental health. The main hypothesis of the variable-focused approach was tested using an application for structural equation modeling and by employing AMOS 22.

Through the person-focused approach, the results obtained for the variables for students without LD were significantly higher than those for students with LD. In particular, the differences in their mental health and self-efficacy were large effect size, whereas the differences in their Chinese and mathematics achievements were more significant. After grouping the students through cluster analysis, four subgroups of students with LD were formed based on the following indicators: those high on mental health and high on academic achievement (positively adjusted, 29.26%); those high on mental health and low on academic achievement (resilient, 32.67%); those low on mental health and high on academic achievement (maladaptive, 10.80%); and those low on mental health and low on academic achievement (vulnerable, 27.27%). Students without LD were also divided into the same four groups (positively adjusted, 30.88%; resilient, 29.33%; maladaptive, 10.80%; and vulnerable, 11.24%). Regardless of whether the students had LD or not, no significant differences were observed in any protective factors for the two groups of students who were resilient and positively adjusted; however, a significant difference in academic achievement was observed. Between the two subgroups with similar low academic achievements, no significant differences in protective factors were noted. The present study discovered that students with LD and resilience had poor academic achievement but good mental health and the protective factors played a key role; among the two subgroups of students with LD with similar academic achievements, the protective factors for those at high risk were also higher.

The academic achievement of the subgroup of resilient students with LD was poor, but the students exhibited better psychological adaptation than those who were maladaptive, mainly because of the protective factors. Among the two subgroups of students with LD who had similar academic achievements, the protective factors of the highly vulnerable students were lower than those of the competent students. In this study, the proportion of resilient and competent students with LD was approximately 50%. The proportion of students without LD in the resilient and competent subgroups was similar to that of students with LD. The results suggest that the protective factors are important to the mental health of both students with and without LD. Academic and cognitive difficulties do not necessarily have negative effects on the mental health of all students with LD. Declining mental health can be prevented or slowed by using protective factors, which enable students with LD to achieve different adaptation results.

In the results of direct model in students without LD, emotional self-concept, Chinese self-concept, self-efficacy, and peer support predicted their mental health; by contrast, in students with LD, emotional self-concept, self-efficacy, and peer support, as well as parental support, were positively correlated with mental health. In the indirect model of students without LD, selfefficacy served as a mediator between emotional self-concept and parental support for mental health, whereas for students with LD, self-efficacy partially mediated the relationship between parental support and mental health. The results of the direct model indicated that mental health was significantly associated with protective factors. Therefore, regardless of whether the students had LD or not, protective factors enhanced the mental health of individuals, although the key factors differed between the two groups. In the indirect model, the study results revealed the mediating effect of self-efficacy among students without LD on their mental health parameters, including physical self-concept, emotion self-concept, Chinese self-concept, parental support, and peer support. These parameters directly influence the individuals’ mental health and are associated with mental health through self-efficacy. Among students with LD, emotional self-concept and peer support had a direct positive correlation with their mental health, whereas parental support enhanced their mental health by bolstering their self-efficacy and, in turn, improving their mental health. Additionally, the physical and Chinese self-concepts only had a direct positive correlation with self-efficacy. Thus, the self-efficacy of the students with LD is crucial for their mental health. The results indicated that the protective factors of students with LD, such as social support and self-concepts, can enhance mental health. Finally, based on the results, this study proposes suggestions for future research and practice.