For t-test, different sample sizes cause a problem or not. Suppose I have 500 female and 117 Males. Should I use equal sample size (117 Male and 117 Female) or can use all data (500 Females and 117 Males). What are implications on power and p-value due to equal or unequal sample size. Any reference or justification for my example, should I use equal or unequal sample size.