This is the result of my goodness of fit test in AMOS:
P = 0.046
CMIN/DF = 1,473
RMSEA = 0,069
CFI = 0,969
GFI = 0,921
TLI = 0,953
Most of the fit indices show that the model is fit except for P-value. Do I have to revise my model or is this good enough?
There is no consens among reviewers. I would send it because of good RMSEA.
Ignore P value here if others values of Model fit are good such as CMIN/Df to be 1-3, GFI, CFI,TLI,NFI >.95 and RMSEA
Hi Dewi,
The p value is significant in most cases. Therefore, it is suggested to consider the value of chi square / df instead of p value. The use of GFI value in the field is not recommended due to its sensitivity to sample size. CMIN / df, CFI values indicate perfect fit. TLI and RMSEA indicate good fit. The results show that confirmatory factor analysis is successful.
Hello,
I am facing almost the same problem:
P Value =,001
CMIN/DF=1619 TLI=,957
GFI=,917 CFI=,968
AGFI=,870 RMSE=,060
where does it fit?
with thanks,
Muhammad Anwar Salam, as you mentioned about GOF and P-Value (.000) may consider good model fit. Is there any article to support this statement. In my case, my CFI, GFI, TLI, RMSEA, DF meet the threshold but p is .ooo but my SV is asking me to meet .05 for P then processed. If any ref you have please share.
Thanks
Suliman Hossin you can read my articles and cite for support.
Anwar, M., Rehman, A. U., & Shah, S. Z. A. (2018). Networking and new venture’s performance: Mediating role of competitive advantage. International Journal of Emerging Markets, 13(5), 998-1025.
Anwar, M., Zaman Khan, S., & Ali Shah, S. Z. (2018). A Study of the Relationship Between Innovation and Performance Among NPOs in Pakistan. Journal of Social Service Research, 1-15.
P-value is very sensitive to non-normality and outlier. Check Mardia's coefficient which shows the multivariate normality in AMOS.
There are some remedies such as using modification analysis in AMOS or removing the outliers.
Usually, the CMIN/DF
- Badges
- Science topic
- Similar topics
- Philosophy Of Science
- Models