Logit regression PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
 Joanna Harma posted on Monday, December 31, 2007 - 6:01 am
I ran logit regression to test the effect of 'fathers' occupation' and 'asset index' on the 'school choice'. Initially the effect of occupation was insignificant but when I added 'parents perception of government schools' the effect of occupation became signigicant and effect of asset index also increased. Occupation and perception variables are not correlated with each other but are correlated with dependent variable. Do you know why this would happen? How reliable is occupation as an explanatory variable of school choice in this case?
 Linda K. Muthen posted on Monday, December 31, 2007 - 7:53 am
I would need more information to say. Do you have the same observations in both analyses? It seems odd that the covariates are correlated zero.
 Joanna Harma posted on Monday, December 31, 2007 - 8:59 am
Same observations has been used in both the steps of regression. Correlation between two covariates is 0.015 which is highly insignificant. I was thinking of possible suppression effect but for that two covariates need to be correlated.
 Linda K. Muthen posted on Monday, December 31, 2007 - 2:21 pm
You need to send the inputs, data, outputs, and your license number to support@statmodel,com.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message