The need for Bayesian hypothesis testing in psychological science
| Autor: | Wagenmakers, E.-J., Verhagen, J., Ly, A., Matzke, D., Steingroever, H., Rouder, J.N., Morey, R.D., Lilienfeld, S.O., Waldman, I.D. |
|---|---|
| Přispěvatelé: | Psychologische Methodenleer (Psychologie, FMG) |
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Value (ethics)
Frequentist probability Computer science Alternative hypothesis 05 social sciences Bayesian probability 050109 social psychology 050105 experimental psychology Statistical inference Test statistic Econometrics Probability distribution 0501 psychology and cognitive sciences Null hypothesis Social psychology |
| Zdroj: | Psychological Science under Scrutiny: Recent Challenges and Proposed Solutions, 123-138 STARTPAGE=123;ENDPAGE=138;TITLE=Psychological Science under Scrutiny Psychological Science Under Scrutiny |
| Popis: | This chapter explains why the logic behind p‐value significance tests is faulty, leading researchers to mistakenly believe that their results are diagnostic when they are not. It outlines a Bayesian alternative that overcomes the flaws of the p‐value procedure, and provides researchers with an honest assessment of the evidence against or in favor of the null hypothesis. The p‐value is the probability of encountering the value of a test statistic at least as extreme as the one that was observed, given that the null hypothesis is true. The logic that underlies the p‐value as a measure of evidence is based on what is known as Fisher's disjunction. The chapter focuses on an inherent weakness of p‐value: the fact that they depend only on what is expected under the null hypothesis H 0, what is expected under an alternative hypothesis H 1 is simply not taken into consideration. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|