This preview shows page 1. Sign up to view the full content.
Unformatted text preview: lt of the
statistical test (i.e., linking the conclusion to the result of the hypothesis test).
Reject the null hypothesis because P-value is less than stated α (or because Pvalue is very small, or because test statistic falls in the rejection region). There is
convincing evidence that the mean mental score of babies who used walkers is
different from the mean score for babies who did not use walkers.
If both an α and a P-value are given, the linkage is implied. If no α is given, the
solution must be explicit about the linkage by giving a correct interpretation of the
P-value or explaining how the conclusion follows from the P-value.
If the P-value in part 3 is incorrect but the conclusion is consistent with the
computed P-value, part 4 can be considered as correct.
NOTE: A confidence interval approach will earn full credit for
• correct hypotheses at outset or, implicitly, in conclusion,
• correct procedure (by name or formula) and assumptions checked,
• correct mechanics, including specification of a (reasonable) confidence
level, degrees of freedom specified (if appropriate)
• 2-sample t interval, unpooled, 95%, df=102 or 53: (−15.2, −4.8)
• 2-sample t interval, pooled, 95%, df=107: (−15.2, −4.8)
• 2-sample z interval, 95%: (−15.1, −4.9)
• correct conclusion in context: "Since 0 is not in the 95% confidence interval,
there is a significant difference between the mean mental skill scores of
babies with walkers and babies without at the α=.05 level of significance."
part (b): No. This was an observational study, and a causal relationship can not be inferred
from an observational study.
• It is sufficient to say any of:
• "no; observational study" (or “no; not controlled experiment”).
• "no; no randomization in grouping" or "no; parents choose which babies
• “no” and then cite a plausible confounding variable and indicate how it is
confounded with the formation of the groups.
• It is not sufficie...
View Full Document