||This study aims at revising the Chinese-version Multidimensional Sense of Humor Scale (Ho, Chik, & Thorson, 2008) using Rasch analysis, and to compare differences between two kinds of data collection—web-based questionnaire and paper-and-pencil questionnaire. Participants were 634 (sample 1) and 587 (sample 2) undergraduate students respectively for item revision and validation. Sample 2 were grouped into 4 groups to test differences between the two kinds of data collection. Rasch Rating Scale Model analysis was used to assess model-data fit, and MANCOVA was employed to compare the differences between the web-based and paper-and-pencil questionnaires.|
The results indicated that three dimensions of the revised 18-item scale—humor production, attitudes towards humor, and coping humor—showed good model-data fit. All items had no statistically significant differential item functioning in genders (boys and girls) and in ways of testing (the web-based and the paper-and-pencil). Reliabilities for humor production, attitudes towards humor, and coping humor were .89, .81, and .81 respectively. The revised scale was further examined through construct-related validity through general self-efficacy and emotional adaptive ability. Furthermore, only the dimension of attitudes towards humor had a statistically significant difference in ways of testing. Suggestions for future research on developing instrument of sense of humor were proposed.