2019
DOI: 10.1136/gpsych-2019-100069
|View full text |Cite
|
Sign up to set email alerts
|

Post hoc power analysis: is it an informative and meaningful analysis?

Abstract: Power analysis is a key component for planning prospective studies such as clinical trials. However, some journals in biomedical and psychosocial sciences ask for power analysis for data already collected and analysed before accepting manuscripts for publication. In this report, post hoc power analysis for retrospective studies is examined and the informativeness of understanding the power for detecting significant effects of the results analysed, using the same data on which the power analysis is based, is sc… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
162
0
2

Year Published

2019
2019
2024
2024

Publication Types

Select...
9
1

Relationship

0
10

Authors

Journals

citations
Cited by 244 publications
(164 citation statements)
references
References 8 publications
0
162
0
2
Order By: Relevance
“…However, a higher rate of withdrawal than anticipated meant that 445 subjects were available for PP analysis. We did not perform a post hoc power estimation, as these are not considered useful [26].…”
Section: Discussionmentioning
confidence: 99%
“…However, a higher rate of withdrawal than anticipated meant that 445 subjects were available for PP analysis. We did not perform a post hoc power estimation, as these are not considered useful [26].…”
Section: Discussionmentioning
confidence: 99%
“…A drop-out of 30% was added and final sample size required was 302. Post hoc power was calculated and was reported >85% (Cohen, 1988;Hertzog, 2008;Zhang et al, 2019).…”
Section: Sampling Procedures and Sample Sizementioning
confidence: 99%
“…To understand the extent of these limitations, we carried out retrospective power analysis (using G * power) and observed that the correlation values obtained in Study 1 were within the critical range, but the beta errors exceeded the acceptable limit (Banerjee et al, 2009). Some researchers observe that the retrospective power analysis violates the key assumptions of a random sample (Zhang et al, 2019); however, the results of Study 1 should be interpreted within the limitation of small sample size and its impact on the power to detect significant correlations. Similarly, the non-random assignment of participants to the two groups also limits generalization; however, the yoga group represents a population that self-selects and seeks yoga training for cognitive benefits.…”
Section: Limitations and Future Directionsmentioning
confidence: 99%