Statisticians issue warning over misuse of P values


TOK TOPICS: Reason, Ethics, Mathematics, Natural Sciences, Human Sciences.


Case Description: The American Statistical Association (ASA) warns against the rampant misinterpretation and misuse of P value — a common test for judging the strength of scientific evidence — leading to unreliable research findings in Medicine and Social sciences . In a detailed explanation ASA further asserts how P value by itself cannot determine whether a hypothesis is true or whether results are important. Thus the researchers should avoid drawing scientific conclusions or making policy decisions based on P values alone.

TOK LINK : The statistical analysis has played a pivotal role in predicting and validating models of scientific theories. Still, in recent years there have been reproducibility issues at various stages of experiment , analysis and publication leading to the statisticians community including ASA to intervene and enforce Ethical standards for Statistics. Besides increasing the transparency and openness in sharing of data and analysis , the research community wants to enforce stricter requirements to assess the underlying assumptions of the models in terms of context and relevance.

Sample KQs:

What role does Ethics play in pursuing the ‘scientific method’?

How far should our policies and decisions depend on statistical analysis?