in a two sample test of means with unknown and unequal population variances, the degrees of freedom is calculated as 9.76. what is the number you should use for degrees of freedom when using the t-distribution table?