When the data are normally distributed and variances are equal.

When the sample size is very large ($n > 1000$).

When the data violate the assumption of normality or contain significant outliers.

When the groups have identical means.