I thought if the test statistic is greater, we reject the null hypothesis and this mean it is not heteroskedastic. Did I misread someplace?
Your procedure was correct, but your understanding of the null and alternative hypotheses is incorrect. The null for this test is homoskedasticity (no heteroskedasticity). If you reject the null, it’s because you have evidence of non-constant variance (heteroskedasticity).
Null = No Heteroskedasticity, while,
Alt = There is Hetero,
so be careful, I also experienced such when I was doing practice problem.
Cheers.