If other factors are held constant, how does increasing the number of scores in each sample affect the value of the independent measures t statistic and the likelihood of rejecting the null hypothesis?