Two drugs A B are to be compared with respect to their effec

Two drugs A, B are to be compared with respect to their effectiveness in preventing premature birth. Suppose 15 pairs of pregnant women are available for the study. In each pair the women are similar with respect to maternal age, race, weight and prior pregnancy history. One women within the pair receives drug A the other receives drug B. The table gives the birth weight (in lbs) and the difference (D) in birth weight within pair.

Sample means and standard deviation are also displayed.

How would you test the hypothesis that the effect on birth weight is the same for the two drugs?

What are the null and alternative hypotheses?

State the test procedure you would use.

Carry out your test procedure and assess the magnitude of the p-value from the information below.

If the test is carried out at the 5% level of significance what is your conclusion?

The following percentiles of the t- distribution with 14 DF: t.998 = 3.44, t.999 = 3.79.

Calculate a 95% confidence interval of the mean difference in birth weight (drug A minus drug B). You will need t.975(14) = 2.15.

If the data were viewed as two samples of birth weights, one for drug A and the other for drug B, would a 2-sample procedure be appropriate? If not, what assumptions would be violated?

Solution