Description
Converting multi-item questionnaire responses into reliable composite scores by applying documented reverse-coding, aggregation and missing-item rules.
Scale scoring turns item responses into the construct score used in analysis. Correct scoring requires reverse-coding negatively worded items, applying the published algorithm (for example SUS: odd items minus 1, 5 minus even items, sum multiplied by 2.5 to give 0-100), deciding how many missing items are tolerable (a common rule is the person-mean of available items when at least 80% are answered), and choosing between sum, mean or factor scores. Rigour comes from scoring in code, checking reliability before using the composite, verifying scores fall within the theoretical range, and only summing items that a factor analysis supports as unidimensional. UX teams frequently mis-score standard instruments or average items across distinct factors, which is why the practice is vestigial.
Attributes
- Discipline
- Survey & Psychometrics
- Loss type
- Vestigial
- Stage
- Evaluation
- Data
- Quantitative
- Effort
- Hours
Steps
- Document each scale's items, response range, reversed items and published scoring algorithm.
- Reverse-code negatively worded items in code (e.g. new = (max + min) minus old).
- Apply a missing-item rule, such as person-mean imputation when at least 80% of a scale's items are present, otherwise set the score to missing.
- Compute the composite (sum, mean or published transformation such as SUS x 2.5) and check that all scores fall within the theoretical range.
- Check internal consistency (e.g. omega or alpha of 0.70 or above) and unidimensionality before using the composite.
- Where norms exist, convert to percentiles or grades (e.g. SUS 68 is roughly the 50th percentile).
Sources
- factor_analyzer
- semopy
- pyreadstat (SPSS import)
- scikit-learn (MICE)
- pingouin (reliability)
- pandas
- Brooke, J. (1996) SUS: A quick and dirty usability scale
- Sauro, J. & Lewis, J. R. (2016) Quantifying the User Experience, 2nd ed.
- DeVellis, R. F. & Thorpe, C. T. (2022) Scale Development: Theory and Applications, 5th ed.
- The Inquiry Series, C3 — Survey Pipeline in Python — From Raw Responses to Structural Equation Model
Also known as
- Composite scoring
- Summated rating scales
- Scale score computation
Categories
- Analytics & Behaviour
- Survey & Psychometrics