Current location - Trademark Inquiry Complete Network - Futures platform - Good as water, hello! The factor analysis I did showed that the distribution of problems in each dimension was too different. What should I do?
Good as water, hello! The factor analysis I did showed that the distribution of problems in each dimension was too different. What should I do?
1. The last two dimensions have only one topic or two topics, so it is difficult to become one dimension. Generally speaking, it takes more than three questions to become a dimension, so we should consider the modification, addition and deletion of the topic;

2. If there are any changes, generally speaking, it is necessary to conduct another test, then do factor analysis, and then determine the dimension;

3. Another thing to consider is the theoretical framework. If these two topics are missing, we should consider whether they can be deleted.

4. I don't know whether you revised the questionnaire yourself or used someone else's ready-made questionnaire. If it is ready-made and has been used and verified by many people, you can contact the previous questionnaire reviser to see if there are any other reasons;

5. In addition, there is the answer of the subject. If you have obvious answers at will, you need to exclude the data of these cases for further processing, so you should pay attention to the extreme value, and the attitude of answering questions can be reflected in the test paper;

6. How reliable is it? You can find the internal consistency coefficient and split-half reliability. Oh, look at the situation now.