Popular searches

Request consultation

Do you need support in running a pricing or product study? We can help you with agile consumer research and conjoint analysis.

Looking for an online survey platform?

Conjointly offers a great survey tool with multiple question types, randomisation blocks, and multilingual support. The Basic tier is always free.

Back
List of articles Back

What is goodness of fit?

Goodness of fit assesses how well your survey report describes the respondents’ answers. Conjointly uses McFadden’s pseudo-R² to calculate goodness of fit; if the information used to calculate your survey report could be predicted to 100% accuracy, then the pseudo-R² would be 100%.

  • Strong goodness of fit (pseudo-R² value of over 65%) indicates that respondents have clear preferences for features
  • Weak goodness of fit (pseudo-R² value of under 45%) suggests that respondents’ choices are more arbitrary.
A survey report with a pseudo-R² value of 80%, indicating strong goodness of fit.

Goodness of fit can be calculated for the whole sample and for specific subgroups.

For another way to assess model quality, see how Conjointly tests predictive accuracy using holdout choice sets and holdout respondents.