At what sample size do latent variable correlations stabilize?

André Kretzschmar, Gilles E. Gignac

Research output: Contribution to journalArticle

Abstract

We conducted a Monte-Carlo simulation within a latent variable framework by varying the following characteristics: population correlation (ρ = 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, and 1.00) and composite score reliability (coefficient omega: ω = 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90). The sample sizes required to estimate stable measurement-error-free correlations were found to approach N = 490 for typical research scenarios (population correlation ρ = 0.20; composite score reliability ω = 0.70) and as high as N = 1000+ for data associated with lower, but still sometimes observed, reliabilities (ω = 0.40–0.50). We encourage researchers to take into consideration reliability, when evaluating the sample sizes required to produce stable measurement-error-free correlations.

Original languageEnglish
Pages (from-to)17-22
Number of pages6
JournalJournal of Research in Personality
Volume80
DOIs
Publication statusPublished - 1 Jun 2019

Fingerprint

Sample Size
Population Characteristics
Research Personnel
Research
Population

Cite this

@article{f7768b6bdc4b4919b8add6f0e9998fbe,
title = "At what sample size do latent variable correlations stabilize?",
abstract = "We conducted a Monte-Carlo simulation within a latent variable framework by varying the following characteristics: population correlation (ρ = 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, and 1.00) and composite score reliability (coefficient omega: ω = 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90). The sample sizes required to estimate stable measurement-error-free correlations were found to approach N = 490 for typical research scenarios (population correlation ρ = 0.20; composite score reliability ω = 0.70) and as high as N = 1000+ for data associated with lower, but still sometimes observed, reliabilities (ω = 0.40–0.50). We encourage researchers to take into consideration reliability, when evaluating the sample sizes required to produce stable measurement-error-free correlations.",
keywords = "Accuracy, Correlation, Reliability, Sample size, Simulation",
author = "Andr{\'e} Kretzschmar and Gignac, {Gilles E.}",
year = "2019",
month = "6",
day = "1",
doi = "10.1016/j.jrp.2019.03.007",
language = "English",
volume = "80",
pages = "17--22",
journal = "Journal of Research in Personality",
issn = "0092-6566",
publisher = "Academic Press",

}

At what sample size do latent variable correlations stabilize? / Kretzschmar, André; Gignac, Gilles E.

In: Journal of Research in Personality, Vol. 80, 01.06.2019, p. 17-22.

Research output: Contribution to journalArticle

TY - JOUR

T1 - At what sample size do latent variable correlations stabilize?

AU - Kretzschmar, André

AU - Gignac, Gilles E.

PY - 2019/6/1

Y1 - 2019/6/1

N2 - We conducted a Monte-Carlo simulation within a latent variable framework by varying the following characteristics: population correlation (ρ = 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, and 1.00) and composite score reliability (coefficient omega: ω = 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90). The sample sizes required to estimate stable measurement-error-free correlations were found to approach N = 490 for typical research scenarios (population correlation ρ = 0.20; composite score reliability ω = 0.70) and as high as N = 1000+ for data associated with lower, but still sometimes observed, reliabilities (ω = 0.40–0.50). We encourage researchers to take into consideration reliability, when evaluating the sample sizes required to produce stable measurement-error-free correlations.

AB - We conducted a Monte-Carlo simulation within a latent variable framework by varying the following characteristics: population correlation (ρ = 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, and 1.00) and composite score reliability (coefficient omega: ω = 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90). The sample sizes required to estimate stable measurement-error-free correlations were found to approach N = 490 for typical research scenarios (population correlation ρ = 0.20; composite score reliability ω = 0.70) and as high as N = 1000+ for data associated with lower, but still sometimes observed, reliabilities (ω = 0.40–0.50). We encourage researchers to take into consideration reliability, when evaluating the sample sizes required to produce stable measurement-error-free correlations.

KW - Accuracy

KW - Correlation

KW - Reliability

KW - Sample size

KW - Simulation

UR - http://www.scopus.com/inward/record.url?scp=85063723259&partnerID=8YFLogxK

U2 - 10.1016/j.jrp.2019.03.007

DO - 10.1016/j.jrp.2019.03.007

M3 - Article

VL - 80

SP - 17

EP - 22

JO - Journal of Research in Personality

JF - Journal of Research in Personality

SN - 0092-6566

ER -