(Equal Variances) From Table 17.3, the sample sizes are small, only a's are estimated, and the standard deviations are not significantly different

(see Exercise 19.2); therefore, we choose a two-sample t test with equal standard deviations, df = nl + n2 - 2 = 14. From Eq. (17.9),

From Table II, the p-value for t = 2.88 standard deviations from the mean for 14 dfis between 0.01 and 0.02. We have evidence to infer a difference in mean readings between ears. From statistical software, p = 0.012.

Sd= From Eq. (17.8), n1 + n2 1+1) [ [ (n-1)s+ (n-1)s n1+n2-2 x 0.882 +7 x 1.1227 8+8-2 = 0.5036. m1 m2 96.41 - 97.86 = -2.88. Sd 0.5036