Related Experiment Videos
Post hoc tests following analysis of variance
Research in Nursing & Health
|June 1, 1985
Summary
Researchers reviewed nursing literature and found inconsistent reporting of post hoc comparisons for mean differences. This paper provides guidelines for selecting appropriate post hoc tests, focusing on Type 1 and Type 2 error control.
Area of Science:
- Nursing Research
- Biostatistics
- Statistical Analysis
Background:
- Post hoc comparisons of mean differences are crucial in nursing research for detailed analysis.
- Inconsistent reporting of these comparisons in nursing literature suggests a need for clearer guidelines.
Purpose of the Study:
- To critically examine the features of various post hoc procedures.
- To offer evidence-based guidelines for selecting appropriate post hoc methods.
- To elucidate the control of Type 1 and Type 2 errors by different procedures.
Main Methods:
- Literature review of nursing research focusing on statistical reporting.
- Analysis of critical features of common post hoc comparison methods.
- Comparative evaluation of Type 1 and Type 2 error control.
Main Results:
- Post hoc comparisons are often warranted but inconsistently reported in nursing studies.
- Different post hoc procedures offer varying levels of Type 1 and Type 2 error control.
- Guidelines for selection based on error control needs are presented.
Conclusions:
- Standardized reporting of post hoc comparisons is needed in nursing research.
- Understanding error control is vital for appropriate selection of statistical tests.
- This paper provides a framework for researchers to make informed decisions on post hoc analyses.