Restricted principal components analysis for marketing research

Wayne Desarbo, Robert E. Hausman, Jeffrey M. Kukitz

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

PurposePrincipal components analysis (PCA) is one of the foremost multivariate methods utilized in marketing and business research for data reduction, latent variable modeling, multicollinearity resolution, etc. However, while its optimal properties make PCA solutions unique, interpreting the results of such analyses can be problematic. A plethora of rotation methods are available for such interpretive uses, but there is no theory as to which rotation method should be applied in any given social science problem. In addition, different rotational procedures typically render different interpretive results. The paper aims to introduce restricted PCA (RPCA), which attempts to optimally derive latent components whose coefficients are integer-constrained (e.g.: {-1,0,1}, {0,1}, etc.). Design/methodology/approachThe paper presents two algorithms for deriving efficient solutions for RPCA: an augmented branch and bound algorithm for sequential extraction, and a combinatorial optimization procedure for simultaneous extraction of these constrained components. The paper then contrasts the traditional PCA-derived solution with those obtained from both proposed RPCA procedures with respect to a published data set of psychographic variables collected from potential buyers of the Dodge Viper sports car. FindingsThis constraint results in solutions which are easily interpretable with no need for rotation. In addition, the proposed procedure can enhance data reduction efforts since fewer raw variables define each derived component. Originality/valueThe paper provides two algorithms for estimating RPCA solutions from empirical data.

Original languageEnglish (US)
Pages (from-to)305-328
Number of pages24
JournalJournal of Modelling in Management
Volume2
Issue number3
DOIs
StatePublished - Jan 1 2007

Fingerprint

Marketing research
Principal component analysis
Interpretive
Integer
Empirical data
Modeling
Car
Business research
Psychographics
Design methodology
Latent variables
Multicollinearity
Buyers
Social sciences
Branch and bound algorithm
Combinatorial optimization
Coefficients

All Science Journal Classification (ASJC) codes

  • Decision Sciences(all)
  • Strategy and Management
  • Management Science and Operations Research

Cite this

Desarbo, Wayne ; Hausman, Robert E. ; Kukitz, Jeffrey M. / Restricted principal components analysis for marketing research. In: Journal of Modelling in Management. 2007 ; Vol. 2, No. 3. pp. 305-328.
@article{34587decb61c48fc9769934ad4be7fd5,
title = "Restricted principal components analysis for marketing research",
abstract = "PurposePrincipal components analysis (PCA) is one of the foremost multivariate methods utilized in marketing and business research for data reduction, latent variable modeling, multicollinearity resolution, etc. However, while its optimal properties make PCA solutions unique, interpreting the results of such analyses can be problematic. A plethora of rotation methods are available for such interpretive uses, but there is no theory as to which rotation method should be applied in any given social science problem. In addition, different rotational procedures typically render different interpretive results. The paper aims to introduce restricted PCA (RPCA), which attempts to optimally derive latent components whose coefficients are integer-constrained (e.g.: {-1,0,1}, {0,1}, etc.). Design/methodology/approachThe paper presents two algorithms for deriving efficient solutions for RPCA: an augmented branch and bound algorithm for sequential extraction, and a combinatorial optimization procedure for simultaneous extraction of these constrained components. The paper then contrasts the traditional PCA-derived solution with those obtained from both proposed RPCA procedures with respect to a published data set of psychographic variables collected from potential buyers of the Dodge Viper sports car. FindingsThis constraint results in solutions which are easily interpretable with no need for rotation. In addition, the proposed procedure can enhance data reduction efforts since fewer raw variables define each derived component. Originality/valueThe paper provides two algorithms for estimating RPCA solutions from empirical data.",
author = "Wayne Desarbo and Hausman, {Robert E.} and Kukitz, {Jeffrey M.}",
year = "2007",
month = "1",
day = "1",
doi = "10.1108/17465660710834471",
language = "English (US)",
volume = "2",
pages = "305--328",
journal = "Journal of Modelling in Management",
issn = "1746-5664",
publisher = "Emerald Group Publishing Ltd.",
number = "3",

}

Restricted principal components analysis for marketing research. / Desarbo, Wayne; Hausman, Robert E.; Kukitz, Jeffrey M.

In: Journal of Modelling in Management, Vol. 2, No. 3, 01.01.2007, p. 305-328.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Restricted principal components analysis for marketing research

AU - Desarbo, Wayne

AU - Hausman, Robert E.

AU - Kukitz, Jeffrey M.

PY - 2007/1/1

Y1 - 2007/1/1

N2 - PurposePrincipal components analysis (PCA) is one of the foremost multivariate methods utilized in marketing and business research for data reduction, latent variable modeling, multicollinearity resolution, etc. However, while its optimal properties make PCA solutions unique, interpreting the results of such analyses can be problematic. A plethora of rotation methods are available for such interpretive uses, but there is no theory as to which rotation method should be applied in any given social science problem. In addition, different rotational procedures typically render different interpretive results. The paper aims to introduce restricted PCA (RPCA), which attempts to optimally derive latent components whose coefficients are integer-constrained (e.g.: {-1,0,1}, {0,1}, etc.). Design/methodology/approachThe paper presents two algorithms for deriving efficient solutions for RPCA: an augmented branch and bound algorithm for sequential extraction, and a combinatorial optimization procedure for simultaneous extraction of these constrained components. The paper then contrasts the traditional PCA-derived solution with those obtained from both proposed RPCA procedures with respect to a published data set of psychographic variables collected from potential buyers of the Dodge Viper sports car. FindingsThis constraint results in solutions which are easily interpretable with no need for rotation. In addition, the proposed procedure can enhance data reduction efforts since fewer raw variables define each derived component. Originality/valueThe paper provides two algorithms for estimating RPCA solutions from empirical data.

AB - PurposePrincipal components analysis (PCA) is one of the foremost multivariate methods utilized in marketing and business research for data reduction, latent variable modeling, multicollinearity resolution, etc. However, while its optimal properties make PCA solutions unique, interpreting the results of such analyses can be problematic. A plethora of rotation methods are available for such interpretive uses, but there is no theory as to which rotation method should be applied in any given social science problem. In addition, different rotational procedures typically render different interpretive results. The paper aims to introduce restricted PCA (RPCA), which attempts to optimally derive latent components whose coefficients are integer-constrained (e.g.: {-1,0,1}, {0,1}, etc.). Design/methodology/approachThe paper presents two algorithms for deriving efficient solutions for RPCA: an augmented branch and bound algorithm for sequential extraction, and a combinatorial optimization procedure for simultaneous extraction of these constrained components. The paper then contrasts the traditional PCA-derived solution with those obtained from both proposed RPCA procedures with respect to a published data set of psychographic variables collected from potential buyers of the Dodge Viper sports car. FindingsThis constraint results in solutions which are easily interpretable with no need for rotation. In addition, the proposed procedure can enhance data reduction efforts since fewer raw variables define each derived component. Originality/valueThe paper provides two algorithms for estimating RPCA solutions from empirical data.

UR - http://www.scopus.com/inward/record.url?scp=70349457769&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349457769&partnerID=8YFLogxK

U2 - 10.1108/17465660710834471

DO - 10.1108/17465660710834471

M3 - Article

VL - 2

SP - 305

EP - 328

JO - Journal of Modelling in Management

JF - Journal of Modelling in Management

SN - 1746-5664

IS - 3

ER -