posted by George on .
The notion of statistical significance is frequently used in business to distinguish between true and spurious between-group differences (such as a difference between the response rates to two different promotions). Hence, when the difference between the baseline (no promotional stimulus) and promotion-stimulated brand purchase rate is statistically significant, such results are often interpreted as being indicative of the impact of the promotion of interest. Do you agree with this rationale? Why or why not?
I agree. A statistically significant difference shows that the promotion caused higher sales.
It would depend on level of significance used, which indicates that the probability that differences were due only to chance. Most social scientists use P = .05 or .01, but some use P = .10. Personally, I would not trust the last level.