Table 1.
Descriptive statistics for depression symptoms.
Females | Males | Gender Difference | ||||||
---|---|---|---|---|---|---|---|---|
|
||||||||
Symptoms | M | SD | M | SD | t | p-value | d | 95% CI for d |
Age 11, CDI | 1.11 | 1.85 | 0.97 | 1.52 | 0.742 | 0.458 | 0.08 | [−0.14, 0.31] |
Age 13, CDI | 1.64 | 2.29 | 0.93 | 1.62 | 3.468 | 0.001 | 0.36 | [0.15, 0.56] |
Age 15, CDI | 2.03 | 2.57 | 1.04 | 1.76 | 4.177 | 0.000 | 0.45 | [0.23, 0.66] |
Age 18, CDI | 1.93 | 2.26 | 1.62 | 2.41 | 1.232 | 0.219 | 0.14 | [−0.08, 0.36] |
Data presented for all participants at each time point (n = 311, 376, 337, and 325 at ages 11, 13, 15, and 18, respectively). P-values are associated with independent samples t-tests for gender differences at each age. d = (Mf−Mm) /sw. Mf = mean for females, Mm = mean for males, and sw = pooled within gender standard deviation. Positive d values indicate that females reported more depression symptoms than males. At ages 13 and 15 the data violated the equal variance assumption (females had significantly more variance), resulting in use of Welch’s t-test and the associated p-value. All of the models examining gender differences in depression symptoms violated the General Linear Model (GLM) assumption of normality; however, the coefficients are still the best, unbiased, efficient estimators among linear solutions (Cohen et al., 2003).