Trends in education, 2015 (vol. 8), issue 1
TVV 2015, 8(1):298-302
NULL HYPOTHESIS VERSUS CONFIDENCE INTERVALS
- Katedra matematiky, Pedagogická fakulta UP, Žižkovo nám. 5, 771 40 Olomouc, ČR
The paper deals with the reason why researchers and scientists should shift emphasis in the statistical treatment of their research as much as possible from NHST (Null Hypothesis Significance Testing) to estimation based on ESaCI (Effect Sizes and Confidence Intervals). A basic problem of NHST is that it encourages dichotomous reasoning, which refers to the reject-or-do not-reject mindset that is widespread within the NHST theories. We would like to explain why a shift to estimation thinking is likely to be valuable, and provide evidence that confidence ESaCI can be more informative than NHST; ESaCI provide more complete information. On the other hand confidence intervals could be disappointingly long. They are long because it is often neither practical nor possible to use sufficiently large samples to get shorter confidence intervals. A way to avoid this disadvantage is to combine results from multiple studies to get better estimates.
Keywords: null hypothesis significance testing, effect sizes, confidence interval.
Published: July 1, 2015 Show citation
References
- CUMMING, G. Understanding the New Statistics: Effect Sizes, Confidence Intervals, and Meta-Analysis. New York: Routledge, 2012. ISBN 978-0-415-87967-5.
- RUPPERT, D., MATTESON, D. S. Statistics and Data Analysis for Financial Engineering. Berlin: Springer Texts in Statistics, 2015. ISBN 978-1-4939-2614-5.
Go to original source...
This is an open access article distributed under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0), which permits use, distribution, and reproduction in any medium, provided the original publication is properly cited. No use, distribution or reproduction is permitted which does not comply with these terms.