The power of analysis: Statistical perspectives. Part 1

Ann E Pulver, John J. Bartko, John A. McGrath

Research output: Contribution to journalArticle

Abstract

Failure to consider statistical power when achieving apparently "negative" results prevents accurate interpretation of the results. A nonsignificant result can be obtained when one includes an insufficient number of subjects to permit observation of a true effect (low power to detect an effect), or when one has an adequate number of subjects, but a meaningful effect does not exist (high power, no effect); one can also have a situation of lower power and no real effect. Without considering power, one is unable to distinguish a "negative" experiment from an inadequate one. This article examines 154 published nonsignificant t-test results. When power is calculated with an effect size equal to a standardized difference of unity, over 50% of the tests have inadequate power.

Original languageEnglish (US)
Pages (from-to)295-299
Number of pages5
JournalPsychiatry Research
Volume23
Issue number3
DOIs
StatePublished - 1988
Externally publishedYes

Fingerprint

Observation

Keywords

  • power
  • Statistical methods
  • Type II error

ASJC Scopus subject areas

  • Psychiatry and Mental health
  • Biological Psychiatry
  • Psychology(all)

Cite this

The power of analysis : Statistical perspectives. Part 1. / Pulver, Ann E; Bartko, John J.; McGrath, John A.

In: Psychiatry Research, Vol. 23, No. 3, 1988, p. 295-299.

Research output: Contribution to journalArticle

Pulver, Ann E ; Bartko, John J. ; McGrath, John A. / The power of analysis : Statistical perspectives. Part 1. In: Psychiatry Research. 1988 ; Vol. 23, No. 3. pp. 295-299.
@article{f92332da7f314fc49bd19b11290f0287,
title = "The power of analysis: Statistical perspectives. Part 1",
abstract = "Failure to consider statistical power when achieving apparently {"}negative{"} results prevents accurate interpretation of the results. A nonsignificant result can be obtained when one includes an insufficient number of subjects to permit observation of a true effect (low power to detect an effect), or when one has an adequate number of subjects, but a meaningful effect does not exist (high power, no effect); one can also have a situation of lower power and no real effect. Without considering power, one is unable to distinguish a {"}negative{"} experiment from an inadequate one. This article examines 154 published nonsignificant t-test results. When power is calculated with an effect size equal to a standardized difference of unity, over 50{\%} of the tests have inadequate power.",
keywords = "power, Statistical methods, Type II error",
author = "Pulver, {Ann E} and Bartko, {John J.} and McGrath, {John A.}",
year = "1988",
doi = "10.1016/0165-1781(88)90020-0",
language = "English (US)",
volume = "23",
pages = "295--299",
journal = "Psychiatry Research",
issn = "0165-1781",
publisher = "Elsevier Ireland Ltd",
number = "3",

}

TY - JOUR

T1 - The power of analysis

T2 - Statistical perspectives. Part 1

AU - Pulver, Ann E

AU - Bartko, John J.

AU - McGrath, John A.

PY - 1988

Y1 - 1988

N2 - Failure to consider statistical power when achieving apparently "negative" results prevents accurate interpretation of the results. A nonsignificant result can be obtained when one includes an insufficient number of subjects to permit observation of a true effect (low power to detect an effect), or when one has an adequate number of subjects, but a meaningful effect does not exist (high power, no effect); one can also have a situation of lower power and no real effect. Without considering power, one is unable to distinguish a "negative" experiment from an inadequate one. This article examines 154 published nonsignificant t-test results. When power is calculated with an effect size equal to a standardized difference of unity, over 50% of the tests have inadequate power.

AB - Failure to consider statistical power when achieving apparently "negative" results prevents accurate interpretation of the results. A nonsignificant result can be obtained when one includes an insufficient number of subjects to permit observation of a true effect (low power to detect an effect), or when one has an adequate number of subjects, but a meaningful effect does not exist (high power, no effect); one can also have a situation of lower power and no real effect. Without considering power, one is unable to distinguish a "negative" experiment from an inadequate one. This article examines 154 published nonsignificant t-test results. When power is calculated with an effect size equal to a standardized difference of unity, over 50% of the tests have inadequate power.

KW - power

KW - Statistical methods

KW - Type II error

UR - http://www.scopus.com/inward/record.url?scp=0023945897&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0023945897&partnerID=8YFLogxK

U2 - 10.1016/0165-1781(88)90020-0

DO - 10.1016/0165-1781(88)90020-0

M3 - Article

C2 - 3290931

AN - SCOPUS:0023945897

VL - 23

SP - 295

EP - 299

JO - Psychiatry Research

JF - Psychiatry Research

SN - 0165-1781

IS - 3

ER -