Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Statistics›Robust Conjoint Analysis
Latent structureMultivariate analysis

Robust Conjoint Analysis

Also known as: robust CA, outlier-resistant conjoint analysis, robust stated preference analysis

Robust conjoint analysis decomposes respondent preferences for multi-attribute products or services into part-worth utilities while guarding against the distorting influence of outlying ratings or unusual respondents. It adapts classical conjoint estimation with robust regression or robust aggregation techniques so that conclusions about attribute importance remain trustworthy even when a minority of evaluations deviate markedly from the majority.

ScholarGate
  1. Latent structure
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Robust Conjoint Analysis
Conjoint AnalysisMixture ModelingRobust Canonical Correla…Robust Discriminant Anal…

When to use it

Use robust conjoint analysis when you expect that a non-trivial fraction of respondents may provide inconsistent or strategically biased ratings — for example in online panels with inattentive respondents, in markets with strong fringe segments, or when rating scales are long and prone to extreme use. It is also appropriate when pilot data or diagnostics from a classical conjoint run show that a small group of respondents disproportionately drives the aggregate utilities. Do NOT use it as a default replacement for classical conjoint without checking: if no outliers exist, robust methods carry no advantage and add interpretive complexity. Avoid it when the goal is to capture heterogeneity deliberately (use latent-class or hierarchical Bayes conjoint instead) because robust aggregation suppresses, rather than models, minority preference segments.

Strengths & limitations

Strengths
  • Protects aggregate part-worth estimates from distortion by inattentive, strategic, or idiosyncratic respondents.
  • Provides built-in respondent-level influence diagnostics that classical conjoint lacks.
  • Attribute importance scores and market simulations are more stable and replicable when outlying evaluations are present.
  • Compatible with standard conjoint designs (full-profile, adaptive) and evaluation formats (rating, ranking).
  • Produces results interpretable in the same way as classical conjoint, requiring no additional conceptual framework for stakeholders.
Limitations
  • Does not model preference heterogeneity — it estimates one aggregate utility set, so it cannot reveal distinct market segments.
  • The choice of robust loss function (Huber, bisquare) and tuning constant influences results; there is no universal default.
  • Software support is narrower than for classical conjoint; implementation often requires custom coding in R or Python.
  • When outliers are actually a meaningful minority segment, suppressing them leads to misleading importance scores for that group.

Frequently asked

How is robust conjoint different from latent-class conjoint?

Latent-class conjoint explicitly models preference heterogeneity by estimating separate utility sets for distinct respondent segments. Robust conjoint estimates one aggregate utility set but reduces the influence of extreme or inconsistent respondents. Use latent-class when you believe multiple genuine preference segments exist; use robust conjoint when you want a single representative utility estimate protected from response aberrations.

Which robust estimator should I use?

The Huber M-estimator is a common starting point because its loss function transitions smoothly from squared error (for small residuals) to absolute error (for large residuals), offering a balance between efficiency and robustness. The bisquare (Tukey) estimator provides stronger downweighting of extreme outliers but has slightly lower efficiency under clean data. The choice should be guided by inspecting the residual distribution from a classical conjoint run.

Does robust conjoint require a special design?

No. Standard conjoint designs — full-profile, fractional factorial, D-optimal or adaptive — work without modification. Robustness is introduced at the estimation stage, not the design stage.

What sample size is needed?

Robust estimation needs enough respondents to distinguish genuine outliers from random sampling variation. As a practical guideline, at least 80–100 respondents are recommended for an aggregate robust conjoint, more if you plan to inspect flagged respondents as a potential segment.

Can I run robust conjoint in R?

Yes. The MASS package provides rlm() for robust linear regression that can be applied profile by profile or to the stacked conjoint data matrix. The robustbase package offers additional M-estimators. Dedicated conjoint packages such as conjoint or support.BWS can provide the design and data-formatting infrastructure, with robust estimation substituted at the utility-estimation step.

Sources

  1. Croux, C., Filzmoser, P., & Oliveira, M. R. (2007). Algorithms for Projection-Pursuit Robust Principal Component Analysis. Chemometrics and Intelligent Laboratory Systems, 87(2), 218–225. DOI: 10.1016/j.chemolab.2007.01.004 ↗
  2. Green, P. E., & Srinivasan, V. (1978). Conjoint Analysis in Consumer Research: Issues and Outlook. Journal of Consumer Research, 5(2), 103–123. DOI: 10.1086/208721 ↗

How to cite this page

ScholarGate. (2026, June 3). Robust Conjoint Analysis. ScholarGate. https://scholargate.app/en/statistics/robust-conjoint-analysis

Related methods

Conjoint AnalysisMixture ModelingRobust Canonical Correlation AnalysisRobust Discriminant Analysis

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Conjoint AnalysisExperimental design↔ compare
  • Mixture ModelingStatistics↔ compare
  • Robust Canonical Correlation AnalysisStatistics↔ compare
  • Robust Discriminant AnalysisStatistics↔ compare
Compare side by side →

Similar methods

Bayesian Conjoint AnalysisConjoint AnalysisAdaptive Conjoint AnalysisTourism Product Conjoint AnalysisRobust Latent Class AnalysisChoice-Based ConjointRobust Correspondence AnalysisRobust Multiple Correspondence Analysis

Related reference concepts

Prior Elicitation and Sensitivity AnalysisConsumer Opinion & Attitude TestingMultivariate Multiple RegressionLatent Class AnalysisHyperpriors and ShrinkageNumerical Methods in Statistics

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Robust Conjoint Analysis (Robust Conjoint Analysis). Retrieved 2026-07-21 from https://scholargate.app/en/statistics/robust-conjoint-analysis · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Adaptations developed by robust statistics researchers building on Green and Srinivasan's conjoint framework
Year
1990s–2000s
Type
Preference decomposition / stated preference
DataType
Ranked or rated profiles of attribute combinations
Subfamily
Multivariate analysis
Related methods
Conjoint AnalysisMixture ModelingRobust Canonical Correlation AnalysisRobust Discriminant Analysis
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account