Which of the following best explains why the t-test for a population mean $\mu$ is considered robust, while the chi-square test for a population standard deviation $\sigma$ is not?

The t-test works only for perfectly normal distributions, while the chi-square test works for all shapes.

Both tests are equally robust to violations of normality assumptions.

The chi-square test is preferred because it uses degrees of freedom, while the t-test does not.

The t-test gives accurate results even when the population is slightly non-normal, especially with large samples, but the chi-square test is very sensitive to non-normality.