- 最后登录
- 2016-11-27
- 注册时间
- 2003-1-21
- 威望
- 250
- 金钱
- 16832
- 贡献
- 11934
- 阅读权限
- 255
- 积分
- 29016
- 日志
- 4
- 记录
- 0
- 帖子
- 1438
- 主题
- 69
- 精华
- 0
- 好友
- 380
签到天数: 3 天 [LV.2]偶尔看看I - 注册时间
- 2003-1-21
- 最后登录
- 2016-11-27
- 积分
- 29016
- 精华
- 0
- 主题
- 69
- 帖子
- 1438
|
明,你放心,我记得见过好几个研究,都说4,5,6,7个选择其实影响不大。我在Proquest帮你找了一下,找到下面这个博士论文,好让你安心: Impact of the Number of Response Categories on Frequency Scales: An Examination of Information Obtained, Reliability, and Factor Structure A Dissertation Presented for the Doctor of Philosophy Degree The University of Memphis
The study examined the impacts of altering the number of response categories to frequency items. Specific issues addressed were inclusion or exclusion of the neutral category, scale reliability, factor structure, and information obtained. Five versions of a questionnaire were developed with scales from the College Student Experiences Questionnaire (Pace, 1994) used with the permission of Indiana University. Each version contained identical questions, however, individual versions contained either 3, 4, 5, 6, or 7 response categories. Undergraduates at The University of Memphis participated in the study and completed a total of 1,162 questionnaires The data revealed several benefits from increasing the number of response categories. The advantages included increased scale reliability, greater total variance explained in factor analysis, and decreased selection of the neutral category. In several instances, however, the magnitude of gains was small and the majority of the gain occurred when the number of categories increased from 3. The data suggested that the choice of either 4, 5, 6, or 7 categories may not impact the analysis to the same extent, however, the pattern of gains continued. Because of the artificial shift it scale means caused by inclusion of a neutral category when the scale mean occurs away from the middle of the scale, researchers may wish to exclude the middle category. In which case, the results suggest a slight advantage in offering 6 response categories over only 4 categories. The most noticeable was a gain of 4.2% in total variance explained in the factor analysis. Additionally, form 6 showed an average increase of 2% in scale reliabilities over form 4; and only a 0.44% average increase in scale standard deviations. Therefore, researchers selecting or preparing questionnaires for college students should consider offering 6 response categories.
|
|