臺灣青少年社會關係認知特點：多情境、多維度探討

Taiwanese Adolescents' Cognition of Social Relationships: Multibackground and Multidimensional Exploration

本研究有二個主要目的，一是瞭解臺灣青少年對不同情境社會關係的認知結構與維度，以及臺灣青少年對維度認知的方向性；二是瞭解不同年級、不同性別臺灣青少年的社會關係認知特點。研究對象為四、六、八年級（9－15歲）學生。研究首先通過訪談獲取臺灣青少年社會關係清單，運用分類法考察臺灣青少年對家族、學校、同儕三種情境社會關係的認知空間與潛在維度；之後針對較重要的幾種關係，以9點利克特量表方式，進一步具體考察臺灣青少年對關係維度量尺（義務程度、回報程度、服從程度、情感程度）的評定情況。研究結果顯示：（一）臺灣青少年社會關係認知結構在三種情境中皆表現出情感維度；家族情境還有成人兒童、輩份、性別、姻親維度，學校情境還有成人兒童、職責／身份維度，同儕情境還有生活背景、職責／身份維度。社會關係不同維度間有所交織。（二）臺灣青少年對社會關係維度的評定表現出方向性。（三）六、八年級社會關係認知結構比四年級多出情感維度和姻親維度，八年級家族情境中還存在性別維度；在具體維度評定時，六年級則在許多方面表現出不同於四、八年級的特點。（四）無論是宏觀的社會關係概念結構還是具體的維度評定，女生均比男生表現出更明顯的情感取向。

Research on Taiwanese adolescents' social relationships has mostly focused on the types, qualities, and states of the relationships as well as their correlation with factors such as adolescents' self-esteem, deviant behavior, and mental health. Relatively few studies have examined social relationships as cognitive objects to investigate their cognitive structure or dimensions. A few studies on cognition have been limited to family background or parent-child relationships and lacked attention to the whole social relationship network. How adolescents perceive their social relationships plays a crucial role in their development. Fourth, sixth, and eighth grade students are in transition periods of adolescence, during which their cognitive competence develops. Moreover, their peer relationships and friendships develop rapidly, and their relationship with their family changes. Additionally, the three grade levels span from primary school to junior high school, and the students' social relationships may differ depending on the environment. According to the interpretation-based view and embodied cognition view, memes, living environment, and growth experience play roles in the development of adolescents and affect their cognition of social relationships. Furthermore, adolescents' social roles change dynamically according to the situation. Social relationships are based on various dimensions, with any one social relationship having more than one dimension. Moreover, social relationships involve at least two parties, and their dimensions may be directional (e.g., self-first or other-first). Therefore, multibackground, multidimensional, and comprehensive research on the social relationships of individuals of various categories (e.g., different ages and sexes) and backgrounds is necessary. The aim of the present study was twofold: (1) to examine the cognitive structures and dimensions of social relationships among different backgrounds and the directionality of dimensions among Taiwanese adolescents and (2) to examine the cognitive characteristics of Taiwanese adolescents of different grade levels and sexes. Two studies were conducted. In the first, 35 students in fourth, sixth, and eighth grades were surveyed. They were asked to write as many relationship terms of address from family, school, and peer backgrounds as possible based on their own experience. This formed the source of the social relationship list of these Taiwanese adolescents. The lists contained 46 different terms of address under the family background, 55 under the school background, and 44 under the peer background. Subsequently, 321 students in grade 4 (9.78 ± 0.66 years, 54 boys and 43 girls), grade 6 (11.43 ± 0.55 years, 51 boys and 40 girls), and grade 8 (13.59 ± 0.61 years, 72 boys and 61 girls) participated in the formal study to categorize the relationship terms of address from each background. After completing the categorization , participants wrote the categorization criteria they used for each background. The number of categories and the criteria of each participant were recorded in SPSS 23.0 to obtain descriptive statistics. Next, the specific categorization criteria of each participant for each background were organized into a dissimilarity matrix. Then, the matrices were superimposed by grade and sex. Nine matrices were obtained for the grade levels, with each grade having three matrices (family, school, and peer backgrounds), and six matrices were obtained for the sexes, with both boys and girls having three matrices each (family, school, and peer backgrounds). These dissimilarity matrices were input into SPSS 23.0, and multidimensional scaling was employed for statistical analysis. Three evaluators labeled the dimensions in the conceptual structures according to the multidimensional scaling analysis results, categorization criteria of the participants, and labeling in previous studies. The kappa coefficients exceeded 0.971. The second study focused on father-child, mother-child, and grandparent-grandchild relationships for family background; teacher-student and classmate relationships for school background; and classmate relationships and friendships for peer background. Dimension scales such as obligation, reciprocation, obedience, and affection were selected based on relevant studies. On a 9-point Likert-type scale, 287 students in grade 4 (9.82 ± 0.68 years, 54 boys and 38 girls), grade 6 (11.42 ± 0.54 years, 49 boys and 35 girls), and grade 8 (13.53 ± 0.60 years, 61 boys and 50 girls) rated the dimension scales of the 6 relationships for both directions (self-first and other-first). Participants' assessments were input into SPSS 23.0. The following analyses of variance (ANOVAs) were conducted with repeated measures on the first two factors for each of the dependent measures (obligation, reciprocation, obedience, and affection) in family and peer relationships: 3 (types of family relationship) × 2 (directions of dimension scale) × 3 (grade levels of participants) × 2 (sexes of participants) and 2 (types of peer relationships) × 2 (directions of dimension scale) × 3 (grade levels of participants) × 2 (sexes of participant). Furthermore, 2 (directions of dimension scale) × 3 (grades of participants) × 2 (sexes of participants) ANOVAs were used with repeated measures on the first factor for each of the dependent measures in teacher-student relationships. The results were as follows: (1) Taiwanese adolescents' conceptual structure of social relationships exhibited both similarities and differences among family, school, and peer backgrounds. Affection was an obvious dimension in all three backgrounds. In addition, adult-child, generation (bei), gender, and in-law dimensions in the family background; adult-child and rank (identity) dimensions in the school background; and life context and rank (identity) dimensions in the peer background were revealed. (2) Taiwanese adolescents' assessments of the dimension scales (obligation, reciprocation, obedience, and affection) had the following characteristics: first, the assessment degrees of mother-child relationships were all the highest among family relationships, and those of friendships were all higher than those of classmate relationships. Second, the assessments of the dimension scales exhibited directionality. The obligation of elders (teachers) to younger generations (students) was higher than that of younger generations (students) to elders (teachers), and the obedience of younger generations (students) to elders (teachers) was higher than that of elders (teachers) to younger generations (students). The affection of teachers for students was higher than that of students for teachers, but no such directionality existed in family relationships. The reciprocation of participants toward peers was higher than that of peers toward participants. (3) Grade-level characteristics were different in the categorization task and the specific dimension scale assessment. For the conceptual structures of social relationships, the sixth and eighth graders exhibited larger affection dimensions in all three backgrounds as well as a larger in-law dimension in the family background compared with the fourth graders. The eighth graders also exhibited gender dimension in the family background. The eighth graders attached greater importance to affection in the categorization task, exhibiting a higher level of cognitive development. However, in the specific dimension assessment, their assessments of affection in family relationships and of obligation in teacher-student relationships were all lower than those of the fourth and sixth graders. In terms of the directionality of dimensions assessment, the sixth graders exhibited some differences with the fourth and eighth graders. The sixth graders' assessments of reciprocation in family relationships and assessment of obedience in teacher- student relationships exhibited more obvious directionality. For the fourth and eighth graders, assessments of reciprocation of friendships were higher than those of classmate relationships, whereas the difference was nonsignificant in the sixth grade. (4) Girls exhibited more obvious affective tropism than did boys in both macro conceptual structure and the specific dimension assessment. In terms of family background conceptual structure, girls exhibited adult-child, in-law, and affection dimensions, whereas boys exhibited adult-child, in-law, and gender dimensions. Girls' assessments of obligation, reciprocation, obedience, and affection in teacher-student relationships were all higher than those of boys. Moreover, girls' assessments of reciprocation and obedience in both family and teacher-student relationships exhibited more obvious directionality than those of boys. Furthermore, girls' assessments of obligation, reciprocation, and obedience in peer relationships exhibited more obvious directionality than those of boys. (5) Taiwanese adolescents did not cognize their social relationships and the dimensions in isolation-social relationship dimensions overlapped and were interwoven. The macro cognitive structure and the specific dimension assessment exhibited both commonalities and differences, as did the cognitive characteristics of social relationships from different backgrounds as well as those of Taiwanese adolescents of different grade levels and sexes. The results suggest that multibackground, multidimensional, and multilevel research on the cognition of social relationships is necessary. Furthermore, the results verify and enrich theoretical perspectives such as dynamic-contextualist metatheory, the convoy model, relational model theory, the embodied cognition view, and the interpretation-based view. This research also has crucial supplementary significance in the research field of Taiwanese adolescents' social relationships.