In a metabolic study, four male swine were tested three times: when they weighed 30 kg, again
Question:
Suppose these data are analyzed by linear regression. With X = body weight and Y = nitrogen balance, preliminary calculations yield = 60 and = 18.7. The slope is b1 = 0.017, with standard error SEb1 = 0.032. The t statistic is ts = 0.53, which is not significant at any reasonable significance level. According to this analysis, there is insufficient evidence to conclude that nitrogen balance depends on body weight under the conditions of this study.
The above analysis is flawed in two ways. What are they? (Look for ways in which the conditions for inference are not met. There may be several minor departures from the conditions, but you are asked to find two major ones. No calculation is required.)
Step by Step Answer:
Statistics For The Life Sciences
ISBN: 9780321989581
5th Edition
Authors: Myra Samuels, Jeffrey Witmer, Andrew Schaffner