1. item fit: infit & outfit之MNSQ趨近1。
(ZSTD易受樣本大小影響,樣本大,容易超過|2|)
2. principal component analysis of Rasch residuals:
(1) 可被主要評量概念所解釋的變異佔總變異的 >60%(部份研究使用較寬鬆的標準為>50%)
(2) 殘差中第一個主成份可解釋的變異<5%的總變異
(3) 殘差中第一個主成份的特徵值 (eigenvalue) <3.0(較嚴格的標準為<2.0,因至少2個題目方能構成1個dimension,<2表示殘差中最主要的成分並不能形成第2個dimension)
References:
1. Linacre JM. A user's guide to winsteps. Chicago: Winsteps.com.; 2005.
2. Bond TG, Fox CM. Applying the rasch model: Fundamental measurement in the human sciences. NJ: Erlbaum: Mahwah; 2007.
3. Cervellione KL, Lee Y-S, Bonanno GA. Rasch modeling of the self-deception scale of the balanced inventory of desirable responding. Educational and Psychological Measurement. 2009;69:438-458
4. Shih MM, Rogers JC, Skidmore ER, Irrgang JJ, Holm MB. Measuring stroke survivors' functional status independence: Five perspectives. Am J Occup Ther. 2009;63:600-608
可列出參考文獻
回覆刪除