國中生人格特質、師生互動關係與偏差行為之關係

Abstract

本研究旨在瞭解師生互動關係的概況、探討師生背景變項及學生人格特質與師生互動關係間的關係、以及師生互動關係及滿意度與學生偏差行為表現間的關係，乃以國中學生800人為研究對象，再採用人格特質量表、師生互動關係及滿意度量表及偏差行為量表等研究工具，獲取所需資料，再以描述統計、卡方考驗、多變量變異數分析、區別分析、皮爾森積差相關、單因子變異數分析、二因子變異數分析、scheffe’之事後考驗、多元逐步迴歸分析等統計方法進行分析。\n 本研究之主要發現如下：\n1.當前國中師生互動關係偏向「影響力」較多，「接近性」中等的情形，若進一步分為四種類型來看，又以「低影響低接近型」及「高影響高接近型」的互動關係最多；「低影響高接近型」最少。\n2.性別不同的學生與教師間的師生互動關係（影響力、接近性及類型），並無顯著差異存在。\n3.教師對來自不同家庭社經地位學生的「影響力」及師生互動關係類型，並無顯著差異存在。然而，教師與來自高社經地位家庭的學生的「接近性」，較多於來自中社經地位家庭者。\n4.性別不同的教師與學生間的師生互動關係（影響力、接近性及類型），並無顯著差異存在。\n5.不同性別組合的師生，其間互動關係的「影響力」及師生互動關係類型，並無顯著差異存在。然而，男學生與男教師在互動時，其間的「接近性」，顯著高於男學生與女教師，且女學生與男教師在互動時，其間的「接近性」，顯著高於男學生與女教師者。\n6.學生的五大人格特質與師生互動關係之「影響力」間存有顯著正相關的關係；學生的和善性、嚴謹自律性、外傾支配性及聰穎開放性，與師生互動關係之「接近性」間存有顯著正相關的關係；且以學生人格特質區別師生互動關係類型的正確率為42.5%。\n7.當教師對學生之影響力或接近性越少時，學生會表現越多的偏差行為。且「低影響低接近型」的師生互動關係，學生的偏差行為表現顯著多於「高影響高接近型」的師生互動關係。\n8.當教師對學生之影響力或接近性越多時，學生對師生互動關係越滿意。且「高影響高接近型」的師生互動關係，學生滿意度高於其他三種類型的互動關係；而「低影響低接近型」者，學生滿意度低於其他三種類型的互動關係。\n9.師生互動關係與滿意度會交互影響學生偏差行為表現。\n10.對整體學生而言，能有效預測其偏差行為表現的變項各有不同，預測力在17.8%至18.0%之間。\n11.對男學生而言，能有效預測其偏差行為表現的變項各有不同，預測力在18.8%至19.6%之間。\n12.對女學生而言，能有效預測其偏差行為表現的變項各有不同，預測力在16.3%至17%之間。\n 最後，本研究根據研究結果進行討論，並對教師、學校及未來研究，提出具體建議，供後續實務工作及研究參考。

The main purposes of this study were to investigate the styles of teacher-student interaction in junior high school and to explore the relationships among personalities of junior-high-school students, teacher-student interaction, students’ satisfaction and misbehaviors.\n The participants included 800 junior high school students in Taiwan. The employed instruments were Personality traits Inventory, Teacher-Student Interaction Inventory, and Activity Experiences Inventory. The applied analysis methods were Descriptive statistics, Chi-square test, MONOVA, Discriminant Analysis, Pearson Correlation, One-way ANOVA, Two-way ANOVA, Scheff’s Method and Multiple Regression.\n The main findings in this study were as following:\n1.Teacher-student interaction in junior high school was high “Influence” and middle “Proximity”, and further, the most frequently styles of teacher-student interaction were “Low-influence Low-proximity”, and “High-influence High- proximity”; the least is “Low-influence High- proximity”.\n2.For students, there were no significant gender differences on teacher-student interaction.\n3.There were no significant SES differences on influence and styles of teacher-student interaction, however, SES had negative influences on proximity of teacher-student interaction.\n4.For teachers, there were no significant gender differences on teacher-student interaction.\n5.There were no significant gender differences on influence and styles of teacher-student interaction, however there were significant differences on proximity of teacher-student interaction, more specially, interaction of male students and male teachers had higher proximity than male students and female teachers, and interaction of female students and male teachers had higher proximity than male students and female teachers.\n6.Personalities of students and influence of teacher-student interaction had significant positive correlations. Agreeableness, conscientiousness, extraversion and openness to experience of personalities of students and proximity of teacher-student interaction had significant positive correlations, and personalities of students can correctly discriminate 42.5% styles of teacher-student interaction.\n7.Influence and proximity of teacher-student interaction had negative influences on students’ misbehaviors, in addition, the style of “Low-influence Low- proximity” leads to high students’ misbehaviors.\n8.Influence and proximity of teacher-student interaction had positive influences on students’ satisfaction, and students mostly satisfy the style of “High-influence High- proximity”, and dissatisfy the style of “Low-influence Low- proximity.\n9.Teacher-student interaction and students’ satisfaction effect reciprocally student’s misbehaviors.\n10.For whole students, many different variables could effectively predict student’s misbehaviors about 17.8% to 18%.\n11.For male students, many different variables could effectively predict student’s misbehaviors about 18.8% to 19.6%.\n12.For female students, many different variables could effectively predict student’s misbehaviors about 16.3% to 17%.\n Finally, after discussion, the researcher proposed some suggestions for educational instruction and future studies.