Simple steps to do Principal Component Analysis (with supporting literatures) Varimax Rotation SPSS
Вставка
- Опубліковано 9 вер 2022
- Just providing an example for your reference. Principal Component is one of the methods in Exploratory Factor Analysis to explore factor(s) in the a latent variable.
Uji konstruk validity!
#efa
#ExploratoryFactorAnalysis
#spss - Наука та технологія
I am trying to perform factor analysis using SPSS, varimax. But the Rotation failed and come up this message:"Rotation failed to converge in 25 iterations. (Convergence = .000)." Why this happened?
Many possible reasons: 1) Method: The choice of rotation method can impact convergence. You can try changing the rotation method, such as "PROMAX," and "OBLIQUE." 2) Increasing the Number of Iterations: In some cases, you might need to increase the maximum number of iterations allowed in the EFA procedure. If 25 iterations are insufficient, you can set a higher limit, like 50 or 100 or more (until give you warning). 3) Number of Factors: The number of factors specified for extraction can affect convergence. If you have specified a large number of factors, reducing the number of factors to extract may help with convergence. 4) Correlation Matrix: Ensure that your correlation matrix is well-conditioned. Highly correlated variables can cause convergence issues. Consider excluding highly correlated variables.
Please do you have any academic references about the value of factor loading after rotation ( for example why you specifically chose 0.4)
Factor loading is classified based on their magnitude: (a) greater than > .30 (minimum consideration level, (b)>.40 (more import- ant), and (c) > .50 (practically significant; Hair, Black, Babin, Anderson, & Tatham, 1998).
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (1998). Multivariate data analysis: A global perspective (Vol. 5). Upper Saddle River, NJ: Prentice Hall.
Use that reference on my own work: doi.org/10.1177/2377960819831468
@@JokoGunawanPhD I think it's Hair 2009 so we can based on this reference ??
@@JokoGunawanPhD thank you very much. I saw some authors use the cutoff of 0.30 or 0.32 that's why I asked you. I'm very thankful for your answer.
Can the Principal Component Analysis method be applied to test items in the EDUQUAL customer service model?
That’s possible if you translate items, reduce or add items that may change the eduqual constructs/dimesions
@@JokoGunawanPhD Thank you for your prompt response. I propose changing the Eduqual model to the local language and adding another domain with five items.
Yes, that needs PCA
Thnak you very much@@JokoGunawanPhD