1993
DOI: 10.1086/209329
|View full text |Cite
|
Sign up to set email alerts
|

Caution in the Use of Difference Scores in Consumer Research

Abstract: This article illustrates the common use of difference scores in consumer research and discusses a number of potential problems with using them. Difference scores often have problems in the areas of reliability, discriminant validity, spurious correlations, and variance restriction. The article concludes that difference scores should generally not be used in consumer research and offers alternatives that overcome their limitations. D ifference scores are often used as measures of constructs in consumer research… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

5
300
1
10

Year Published

1996
1996
2022
2022

Publication Types

Select...
9
1

Relationship

0
10

Authors

Journals

citations
Cited by 472 publications
(316 citation statements)
references
References 24 publications
(42 reference statements)
5
300
1
10
Order By: Relevance
“…Two criticisms are notable. One, charged by Peter et al (1993) and Brown et al (1993), relates to the indirect difference score approach. According to them, the difference score approach causes poor reliability and problems of variance restriction associated with the component scores.…”
Section: Service Qualitymentioning
confidence: 99%
“…Two criticisms are notable. One, charged by Peter et al (1993) and Brown et al (1993), relates to the indirect difference score approach. According to them, the difference score approach causes poor reliability and problems of variance restriction associated with the component scores.…”
Section: Service Qualitymentioning
confidence: 99%
“…The use of a covariate avoids the use of difference scores given such scores have attracted criticism in relation to their poor psychometric properties (e.g., Humphreys, 1993;Peter, Churchill, & Brown, 1993). For each analysis, all significant results are reported in Table 4 whilst the results of all follow-up tests of significant interactions are reported in Sections 3.3.1 to 3.4.2.…”
Section: Immediate Persuasion Outcomesmentioning
confidence: 99%
“…where s 2 1 and s 2 2 are the variances of the two measures, r 12 is the correlation between them, and R 1 and R 2 are their reliabilities (35) .…”
Section: Appendixmentioning
confidence: 99%