小学生欺负与人格倾向关系的研究

【作者】 谷传华；

【导师】 张文新；

【作者基本信息】 山东师范大学 ， 教育心理学， 2001， 硕士

【摘要】 欺负是小学生中普遍存在的现象，是力量相对较强的一方对在未受激惹的情况下对力量相对弱小的另一方所进行的攻击。欺负现象的相对稳定性表明，欺负可能与某些特定的人格倾向相应。国外对中小学生进行的大量研究发现，欺负者、受欺负者（及欺负/受欺负者）分别具有不同的人格特点。在本研究中，运用嵌入式的匿名问卷法对中国小学生中欺负者、受欺负者与未卷入儿童的人格倾向的差异进行考察，并探查欺负、受欺负的频率或可能性与各种人格倾向的关系，以期为跨文化研究和欺负干预工作作出贡献。 以整群抽样法从济南市育新小学3、4、5、6年级抽取872名受测者。其中男童与女童基本相等，儿童年龄主要在7——11岁之间。以嵌入欺负、受欺负两个题目的艾森克个性问卷（幼年版）和自尊问卷进行施测，结果表明：1 小学男女生中，欺负者和受欺负者情绪稳定性得分、男生中的欺负者精神质得分均显著高于常模，受欺负者的精神质显著低于常模。2．多元方差分析表明，在艾森克个性问卷各维度的总体变异及精神质、情绪稳定性、掩饰性维度上，存在极其显著的类型主效应，而不存在显著的交互作用。欺负者的精神质得分极其显著地高于受欺负者、未卷入儿童，而受欺负者的精神质得分也显著高于未卷入儿童。卷入欺负者的情绪稳定性得分显著高于未卷入的一般儿童，其情绪更不稳定，而欺负者与受欺负者之间并未发现显著差异。同时，欺负者和受欺负者的掩饰性得分均显著低于未卷入儿童。另一方面，在内外向维度上，各类儿章之间未发现显著的差异。3．小学生的自尊得分具有极其显著的类型主效应，未卷入儿童的自尊极其显著地高于欺负者和受欺负者，但并未发现类型与年龄。性别之间存在显著的交互作用。斗．小学生欺负、受欺负的发生频率与精神质、情绪稳定性之问均存在极其显著的正相关，与自尊、掩饰性均存在显著或极其显著的负相关，而与内外向得分之间的相关未达到显著性水平。LOgistiC回归分析进一步发现，精神质、掩饰性／性别变量对预测欺负发生的概率，具有极为重要的作用，精神质水平越高，掩饰性越小，欺负他人的可能性越大；同时，男童欺负他人的概率显著高于女童。自尊、情绪稳定性和年龄对预测受欺负发生的概率作用极为显著，几童的自尊越低，情绪越不稳定，受欺负的可能性越大；同时，与门岁儿童相比，9岁几童更可能受欺负。 最后，本文对本研究的方法、结果及未来研究的取向进行了讨论，并对小学生的欺负干预问题提出建议。更多还原

【Abstract】 Bullying is prevalent among primary school students, and it is aspecial type of special aggression in which the more powerfuloppresses the less powerful without being provoked. The stability ofbullying implies that it is connected with certain characteristics ofpersonality. Significant diffeence in persona1ity between bu11ies,victims and those students not involved into bullying was found inlarge amount of research conducted among prirnary and secondaryschOol students outside China. In the preseni study, anonymousquestionnaires were administered to primary students, and thedifferences of personality’ characteristics of bullies, victims, andchildren not involved were examined. The re1ation between theprobabilities of bu1lying others and being bullied and personality wasalso investigated.872 students of grade 3, 4, 5, and 6, with approximately equalnumber of boys and girls, from Yu Xin primary schoo1 in Jinan Cityof Shandong Province participated in ti1e investigation. The age of thestudents was mainly between 7 and l 1 years old. Eysenck Persona1ityQuestionnaire 0unior) (EPQ) into which two questions about bullyingothers and being buIlied this term were inserted, and the Self EsteemQuestionnaire were administered to the panicipants. The results ofdata analysis were as fOllows.1. Conthested with the norm, the bu1lies and victims among primaryschool students scored significa-ntly higher on the NeuroticismSubscale, and the bullies among boys scored significantly higher onthe Psychoticism Subcale, and the victims scored significanily 1ower.IJI2. MAVOVA indicated that significant main effect of children’sstatus (bu1lies, victims and not involved) on the tota1 varia-nce ofpersonality dimensions of EPQ, a-nd on the Psychoticism Subscale,Neuroticism Subscale and Lie Subscale significant main effects ofchi1dren’s status were also fOund, however no interaction of age,gender and type was fOund. The scores of bullies on the PsychoticismSubscale were significantly higher than victims and those studentswho were not invo1ved, and victims’ scores on the subsca1e were alsosignificam1y higher tha-n those uninvolved. The scores of thoseuninvolved on the Neuroticism SubscaIe were signiflcant1y higherthan those u-ninvolved, and marked difference between bu11ies andvictims was not found. At the same time, Lie scores of buI1ies andvictims were significanly lower than those uninvo1ved. On the otherhand, there was no significan difference of Extraversion scoresbetween three types of chi1dren.3. The variable of children’ status has a significant effect onself esteem scores of primary schoo1 studeflts, and those uninvo1vedscored significant1y higher than those invo1ved, however significantinteraction was not found.4. Significam positive correlation was fOund between the frequencyof bullying others and being bullied and the scores on PsychoticismSubscale and Neuroticism Subsca1e, and significant negativecorrelation existed between the frequency of bu1lying others a-ndbeing bullied and SeIf esteem scores and Lie scores, however thecorrelation between the frequency a-nd Extraversion Scores was notfound to be significant. Logistic Regression Analysis trierindicated the very significant effect of the variables of psychoticism,1Vlie and gender on predicting the probabilities of bulling others, andthe effect of selfestCem, neuroticism and age on predicting theprobabilities of being bullied. The higher children scored on thePsychoticism Subscale, and the lower children scored on the LieSubscale, the more probably children would buIly others. At the sametime boys were more likely to bulIy others than girls. On the otherhand. the lower chi1dren scored on the Self eSteem QueStionnaire, andthe higher children scored on the Neuroticism Subscale, the moreprobable children would be bullied. At the saxne time, comPared tol1-year-old children, 9-year-old children were mo更多还原