Nástroje na zlepšenie štatistickej inferencie – analýza p-kriviek a ekvivalenčné testovanie
Tools for Better Statistical Inference. Analysis of P-Curves and Equivalence Testing
Author(s): Milan FicoSubject(s): Methodology and research technology, Evaluation research
Published by: Sociologický ústav - Slovenská akadémia vied
Keywords: Null hypothesis significance testing; publication bias; file-drawer effect; selective data reporting; hypothesizing after results are known; p-curves, equivalence testing;
Summary/Abstract: Tools for Better Statistical Inference. Analysis of P-Curves and Equivalence Testing. File-drawer effect, selective data reporting and additional data adjustment often lead to publication bias in social sciences. This text deals with two techniques diminishing consequences of these negative phenomena. P-curves can distinguish relevant publications from nonsignificant ones and help us to select information base for formulation of next research goals. Another presented technique is equivalence testing. Focused on the falsification of hypotheses, it helps better explore trivial results and makes them more informative. Both techniques support informative value of results so that the conclusions of statistical inference are more valuable. Sociológia 2020, Vol. 52 (No. 4: 323-353)
Journal: Sociológia - Slovak Sociological Review
- Issue Year: 52/2020
- Issue No: 4
- Page Range: 323-353
- Page Count: 31
- Language: Slovak
