Hypothesis Testing Discussion
1. In 50 to 100 of your own words, summarize this article.
The article discusses the undue consequences of so much emphasis on hypothesis testing and the use of P values to dichotomise significant or non-significant results as against how other alternative statistical approaches of interpreting study results have been neglected.
The writer suggested that confidence intervals should be reported instead of standard errors to enable a move away from the current emphasis on statistical significance which has led to a suspected shift in emphasis away from the basic results of research findings
In the introduction, the authors describe a scenario involving mean concentrations of a blood constituent. They point out that readers should want to know "how much the illness modified the mean blood concentrations... rather than only the level of statistical significance." Do you agree with this statement? Why or why not?
I agree with the writer’s statement.
We should be more concerned with the effects of the studies done rather than just being happy with the level of statistic significance. Statistic significance does not go down to tell us the importance of the findings rather it simply tells us to what extent the statistics is reliable and that there is a good chance that we are right in finding that a relationship exists between two variables. But statistical significance is not the same as practical significance and significant finding implications may have no practical application. The researcher must always examine both the statistical and the practical significance of any research finding.
3. In the section titled Presentation of study results: limitation of P values, the authors state that simply presenting P values "encourages lazy thinking." What do you think is meant by this statement? Do you agree with it? Why or why not?
This could mean that P values are essential when defining two alternative outcomes ie “significant and not significant” and is usually seen as not really helpful when estimating the probabilities in a research.
I agree to the fact that P values encourage lazy thinking. Using P value obviously does not give detailed information of the research finding but just gives a clear self defined significant or not significant answer which actually in most cases does not call for critical thinking. For this reason, the confidence interval which is more informative is advocated to be used more often.
After reading this article, do you have any hesitations about doing hypothesis testing? Explain why or why not.
I don’t have any hesitation on the use of hypothesis testing.
I agree that one of the problems in hypothesis testing is the difficulty of reasoning with uncertain evidence and the natural human tendency to take recourse to behavioral rules.
Obviously, in experimental and observational data on human behavior there is bound to be variable that the evidence produced by these data is uncertain and this requires us to reason provisionally if our conclusions are certain rather than tentative; to be concise rather than accurate in reporting our arguments and findings.
Therefore rather than hesitating on the use of hypothesis testing, better interpretations of hypothesis testing should be taught deeply while the tests should be used with better understanding and interpretation.
Ada