The point that I am going to discuss is point number 16 from page 343. It cautions us that while testing the same hypothesis in different populations, even though the P values are on the opposite side of the cut-off value (0.05), we cannot always say that the results are conflicting (Greenland et al., 2016). In testing hypothesis, P value refers to the probability of the observed effects or effects that are more extreme under the condition that all our computed assumptions are correct (Greenland et al., 2016). The cut-off value is 0.05.
The difference in P value can be caused by other factors like study population and known standard error. Even though P values are on the different side of the cut-off point, their results would be in perfect agreement. Therefore, the point cautions us that we cannot always assume the results would be conflicting if the P values are different in testing the same hypothesis in different populations.
It is better to delve deeper to explore other factors like population characteristics and sample size differences. So, this point also highlights the importance of understanding the full context in interpreting the results from statistical tests in relation to answering the research question.
References
Greenland, S., Senn, S. J., Rothman, K. J., Carlin, J. B., Poole, C., Goodman, S. N., & Altman, D. G. (2016). Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. European journal of epidemiology, 31(4), 337–350. https://doi.org/10.1007/s10654-016-0149-3.