Statistics for Business: Decision Making and Analysis, 3rd Edition

1. ... directed to the old page spent ${\overline{x}}_{\text{old}}=\text{\$}253$ on average (s_old = $130); those directed to the new page spent ${\overline{x}}_{\text{new}}=\text{\$}328$ on average (s_new = $161). Software computed the standard error of the difference to be $17.30 with 295 degrees of freedom.

   1. Should we be concerned with the possibility of confounding?¹⁰

      ¹⁰ It is always good to be concerned about confounding, but the randomization removes the concern.

   2. Does the new page generate statistically significantly higher sales than the old page? State the null hypothesis and whether it’s rejected. (Assume that these samples are large enough to satisfy the sample size condition.)¹¹

      ¹¹ Yes, at the usual 5% tolerance for an error, t = (328 − 253)/(17.3) ≈ 4.3. The p-value is much less than 0.05.

   3. A manager claims that ...