Page 66 - Quantitative Data Analysis

Quantitative Data Analysis
Simply Explained Using SPSS

3. Populations are normally distributed

Robustness studies have shown that the independent t-test tends to
be fairly robust to violations of the assumption. The one exception
is with one-tailed tests with a skewed distribution - dominant tail
has more type I errors, opposite tail has fewer.

4. Populations have equal variances

If sample sizes are equal this assumption does not matter. If the
larger sample has the larger variance, the actual type I error rate is
less than .05, if the smaller sample has the larger variance, the type
I error rate is greater than .05.

Levene’s Test of Homogeneity of Variance

Equal Variance Assumption can be checked through Levene’s Test. A
non-significance Levene’s test indicates that variance is significantly
not difference thus it holds the assumption while a significant
Levene’s test indicates that there is a significant difference in the
variance thus assumption of homogeneity of variance violates.

The Theory and Applications of Statistical Inferences 50