testing for the presence of heteroscedasticity

[Azlin Ahmad]

Hi.

Can I know command STATA to test heteroskedasticity with instrument variable?

Currently, I'm using command

ivhettest

but this command not function well.

Thanks in advance

[Carlo Lazzaro]

Azlin:
https://www.statalist.org/forums/help #9 explains why phrasings like "it doesn't work" are not that helpful.

[Joro Kolev]

There are two way here:

1. You can explain why -ivhettest- "does not function well" and we can try to fix this.

2. You can perform the test manually. Generate the residuals, and regress them on some functions of the all exogenous variables. You can use just levels of the exogenous variables, or levels, squares, and cross terms, like in the standard White test.