諮詢主題 | 1. scaling success計算方式。 2. 確認SF-36之相關係數算法 | |
諮詢時間 | 100/2/23 2pm-3pm | 預計諮詢時數：1小時 |
諮詢內容 | 已知概念 | 希望釐清的概念 |
1. 於已讀的文獻，計算scaling success的方式為，項目與所屬的次量表相關係數，顯著地高於項目與其他(不所屬)次量表的相關係數2個標準差，一個標準差等於1/√樣本數。。 2. 相關係數算法 (1) 項目與所屬的次量表相關係數計算方式為，看Cronbach’s alpha之corrected item- total correlation。 (2) 項目和其他次量表相關係數計算方式為，看項目與其他次量表總分之pearson correlation。 (3) 依據量表設計假設，項目與所屬次量表的相關要高，與其他的次量表的相關要低，但Bodily pain、General health、Social functioning，此三個次量表的數值不符合假設。 | 1. 確認計算scaling success的方式是否正確。 2. 確認SF-36之相關係數算法是否正確(項目與所屬次量表相關係數，及項目和其他次量表相關係數)。 | |
2011年2月21日 星期一
統計諮詢: SF-36之資料分析
訂閱：
張貼留言 (Atom)
It's good that you mentioned some weird findings like 2.(3). However, you're not confident about the rest of the analyses. May I suggest that please specify what are the problems (concerns) on the analyses/results. I think it's not good to ask Yen Lee just to check whether you're correct. You should find some problems (or questions),
回覆刪除Maybe, this is my first time to do real analysis by SPSS. I am kind of afraid that I am doing the wrong analysis, when I see the weird results.
回覆刪除I think I will may ask Yen Lee that what I should do when I see weird results. For this time, what should I do for checking the findings of correlation from her perspectives?
Yes, it's ok!
回覆刪除May I ask you to try to do something to check your data and results before the meeting with Yen Lee. Not just asking her opinions or to check them for you.