| Zusammenfassung |
Statistical significance testing is a central technique for everyday empirical-quantitative work in media psychology. Its most common form, the null hypothesis significance test represents a well-established method of legitimating knowledge claims drawn from statistical analyses based on random samples. Nevertheless, intended epistemological meaning of the procedure is commonly misunderstood and its results are often misinterpreted. Furthermore, a prevalent view of regarding the “significance” of statistical estimations as providing a measure for the meaningfulness of study results has also led to various forms of questionable research and publication practices. Nonetheless, significance testing is still a well-accepted technique in our field. This entry explains the different versions, their interpretation, and tries to abolish misunderstandings. It also gives an overview on extensions and alternatives that have been historically developed, as well as practical recommendations for meaningful reporting of significance test results. |
