标题:
应用项目反应理论对友谊质量量表修定—以结构方程模型、层面理论多种技术支持修定量表的质量Modifying Friendship Quality Questionnaire with Item Response Theory Approach——Validating with Facet Theory and Structure Equation Techniques
作者:
赵守盈, 臧运洪, 杨建原
关键字:
项目反应理论, 友谊质量量表, 层面理论, 最小空间法Item Response Theory; Friendship Quality Questionnaire; Facet Theory; The Smallest Space Method.
期刊名称:
《Advances in Psychology》, Vol.1 No.3, 2011-12-19
摘要:
本文应用项目反应理论的区分度、难度和信息函数峰值三个参数对Parker & Asher (1993)的友谊质量量表进行修订，目的是使修订后的量表能更精确地调查中国青少年友谊质量的现状。运用SPSS15.0软件对数据进行管理，运用MULTILOG 7.03软件对项目进行参数分析，运用结构方程、层面理论及最小空间法对修订的量表进行验证性检验。结果：1) 友谊质量量表符合单维性检验，可以根据项目反应理论进行修订。2) 新量表的区分度a值和难度b值具有合理的取值范围。3) 新量表的测验信息函数峰值变小。4) 新旧量表的各因子之间显著高正相关。5) 经结构方程、层面理论及最小空间法分析验证：具有26个项目的新量表四因子结构清晰。四因子分别是：肯定与关心、帮助与指导、亲密坦露与交流和冲突解决与陪伴。与原量表相比没有冲突和背叛因子。修订后的量表具有更高的信、效度，内部一致性信度为0.875，方差解释率为44.380%。经正式施测，修订后的量表可以有效地调查中国青少年友谊质量的现状。
The item discrimination, difficulty and information function peak of the item response theory are used to revise the friendship quality questionnaire produced by Parker & Asher (1993), the purpose is that this questionnaire revised is more accurate to survey the status of friendship quality of chinese youth. SPSS15.0 software is used to manage data, using MULTILOG 7.03 software to analysis parameter. Structural equation, facet theory and the minimum space method are used to confirm the questionnaire revision. Results are as follows: 1) The friendship quality questionnaire is one-dimensional which can be revised by item response theory. 2) The item discrimination a, difficulty b of new questionnaire are with reasonable scopes. 3) The test information function peak of the new questionnaire is smaller. 4) The factors of old and the new questionnaire have significantly high positive correlation. 5) The results confirmed are that: The structure of four factors with 26 items is clear. Four factors are: affirmation and care, help and guidance, revealed intimately and exchange and conflict solution and accompany. Compared with the original questionnaire, the new one has not conflict and betrayal factor. The questionnaire revised has higher reliability and validity, consistency reliability is 0.875 and variance explained is 44.380%. After measuring officially, the questionnaire revised can survey the status of friendship quality of Chinese youth effectively.