ISSN 0439-755X
CN 11-1911/B
主办:中国心理学会
   中国科学院心理研究所
出版:科学出版社

心理学报 ›› 1996, Vol. 28 ›› Issue (2): 120-125.

• • 上一篇    下一篇

霍式中国职业兴趣量表构想效度的验证性因素分析(Ⅱ):多特质多方法矩阵的LISREL模型比较检验

白利刚;凌文辁;方俐洛   

  1. 中科院心理所
  • 发布日期:1996-06-25 出版日期:1996-06-25
  • 通讯作者: 白利刚

CONFIRMITORY FACTOR ANALYSIS TO THE CONSTRUCT VALIDITY OF THE CHINESE VOCATIONAL INTEREST INVENTORY OF HOLLAND TYPE(Ⅱ):THE CONTRAST OF THE MTMM-LISREL,MODELS'

Bai Ligang;Lin Wenquan;Fang Liluo(Institute of Psychology Academia Sinica,Beijing,100012)   

  • Online:1996-06-25 Published:1996-06-25

摘要: 霍式中国职业兴趣量表通过验证性因素分析的交叉证实检验,显示出具备良好的会聚效度和区分效度,但同时也显示出一定的来自方法偏误的影响。在本研究中,研究者采用多特质多方法的Lisrel模型各成分间的比较检验法,对模型中特质因素、方法因素及方法偏误因素等各成分的作用做了进一步的区分,以考察二者各自影响作用的大小。结果表明,霍式中国职业兴趣量表构想效度良好,来自方法偏误的影响作用不大。

关键词: LISREL模型, 构想效度, 方法偏误

Abstract: The Chinese Vocational interest Inventory of Holland Type was cross-validated by applying confirmitory factor analysis.The results showed that the convergent and discriminant validity was good,but also some effects of method-bias.In this study we tested the effects of traits,methods and method-bias respectively by applying the contrast of the MTMM-LISREL Models.The results suggested substantial support that the inventory's RIASEC construct validity is good,the effects of method-bias is limited.

Key words: Lisrel model, construct validity, method-bias