Making subjective judgments in quantitative studies: The importance of using effect sizes and confidence intervals
✍ Scribed by Jamie L. Callahan; Thomas G. Reio Jr
- Publisher
- John Wiley and Sons
- Year
- 2006
- Tongue
- English
- Weight
- 115 KB
- Volume
- 17
- Category
- Article
- ISSN
- 1044-8004
No coin nor oath required. For personal study only.
✦ Synopsis
At least twenty-three journals in the social sciences purportedly require authors to report effect sizes and, to a much lesser extent, confidence intervals; yet these requirements are rarely clear in the information for contributors. This article reviews some of the literature criticizing the exclusive use of null hypothesis significance testing (NHST) and briefly highlights the state of NHST reporting in social science journals, including Human Resource Development Quarterly. Included are an overview of effect sizes and confidence intervals-their definitions, a brief historical review, and an argument regarding their importance. The article concludes with recommendations for changing the culture of quantitative research within human resource development (HRD) to more systematically reporting effect sizes and confidence intervals as supplements to NHST findings.
Consumers of research often perceive quantitative studies as being objective presentations of research findings emerging from noisy and complex data. Commonly accepted practices associated with using null hypothesis significance testing (NHST; sole reliance on reporting p values) to convey research findings have contributed to this impression. However, this impression of objectivity obscures the fact that there are many levels of subjectivity in collected data, even through research using NHST, and this includes interpretations of the practical significance of research findings. As a result, there are some drawbacks to using NHST exclusively. Two approaches that can be used