Correlation Analysis with Confidence Intervals

Description

Measures how strongly two UX measures move together, with an interval showing how sure you can be.

Correlation quantifies association: does SUS track completion, does NPS rise with ease ratings, do two tasks share difficulty. Pearson r suits continuous measures, Spearman ranks suit skewed or ordinal data, and phi suits two binary variables. Sauro and Lewis emphasise reporting confidence intervals (via the Fisher z transformation) because small-sample correlations are very unstable: r = 0.5 from 20 people could plausibly be anywhere from about 0.07 to 0.77. They also give sample size formulas for detecting a correlation of a given size and benchmarks for magnitude in UX data.

Attributes

Loss type
Active
Stage
Evaluation
Effort
Hours

Steps

  1. Plot the two variables in a scatterplot first to check for non-linearity and outliers.
  2. Choose Pearson, Spearman or phi to fit the data types.
  3. Compute r and its p-value.
  4. Transform to Fisher z, form the interval and transform back.
  5. Interpret magnitude in context and avoid causal language.

Sources

  • Quantifying the User Experience (Sauro & Lewis)
  • Companion Book to Quantifying the User Experience (Lewis & Sauro)
  • Practical Statistics for Data Scientists (Bruce & Bruce)
  • Myatt, G., Making Sense of Data: A Practical Guide to Exploratory Data Analysis and Data Mining (2007)
  • Gratton, C. & Jones, I., Research Methods for Sports Studies (2010)
  • Veal, A. J. & Darcy, S., Research Methods in Sport Studies and Sport Management (2014)

Also known as

  • Pearson correlation
  • Phi coefficient
  • Spearman correlation
  • Partial correlation

Categories

  • Analytics & Behaviour
  • Statistics & Modelling