Premium
A DECISION MODEL FOR SELECTING ALTERNATIVE HYPOTHESES
Author(s) -
Solomon Susan L.
Publication year - 1975
Publication title -
decision sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.238
H-Index - 108
eISSN - 1540-5915
pISSN - 0011-7315
DOI - 10.1111/j.1540-5915.1975.tb01045.x
Subject(s) - selection (genetic algorithm) , statistical hypothesis testing , econometrics , affect (linguistics) , computer science , alternative hypothesis , statistical power , test (biology) , value (mathematics) , sample (material) , decision problem , statistics , mathematics , artificial intelligence , machine learning , psychology , null hypothesis , chemistry , chromatography , paleontology , communication , biology , programming language
One important aspect of experimental design is the decision whether to use a one‐ or two‐tailed hypothesis test. Assuming that the sample size and the level of significance have been specified, the selection of an alternative hypothesis can materially affect the conclusion drawn from the experiment. In the past, this selection was made on a subjective basis or for pragmatic reasons which can bias the experiment. This paper suggests an expected value approach, maximizing the expected power of the test, to place the selection of alternative hypotheses in an objective and reliable decision framework.