Consider a correlation coefficient of .095 between variables A and B. The correlation coefficient was calculated from a sample of 1100 respondents and is statistically significant, p < .0001. Consider a correlation coefficient of .891 between variables C and D. The correlation coefficient was calculated from a sample of 15 observations and is statistically nonsignificant, p > .10.
So, when should marketing analysts consider a correlation strong enough to even consider given the issue raised above?
For testing correlation the null hypothesis will be
against the alternative
., where
is the correlation coefficient of the population. A strong
positive correlation means
is close to 1. Then the correlation will be significant. Here the
correlation between C and D is close to 1 than that of A and B
which is somewhat close to 0. So the correlation between C and D is
strong enough to even consider when comparing to A and B.
Get Answers For Free
Most questions answered within 1 hours.