The User Experience Questionnaire (UEQ): A Practical Semantic Differential

TL;DR: UEQ is a 26-item questionnaire that measures reliably the Pragmatic Quality and Hedonic Quality of a product. Five other sub-dimensions have been proposed but literature is torn in their validity. UEQ provides a benchmark for comparing a system scores with normative data for a collection of other systems.


Defining UX has been, and to some extent still is, tricky. In their 2006 publication, Laugwitz, Schrepp, and Held introduced the User Experience Questionnaire (UEQ), originally developed in German (and later adapted into English). As a working definition of UX, the UEQ builds on Hassenzahl’s theory that users perceive a product through Pragmatic and Hedonic qualities. Pragmatic Quality emphasises how well a product supports users in achieving task-related goals with effectiveness and efficiency, whereas Hedonic Quality emphasises the product’s role in psychological well-being, stimulation, and identity expression. UEQ has shown promising psychometric properties, but more research is needed to evaluate the dimensions it is measuring.

Setting up the UEQ

The questions of the UEQ

The UEQ consists of 26 items, presented as pairs of opposite adjectives on a 7-point semantic differential scale ranging from -3 to +3, with 0 representing neutrality. Each item asks participants to rate the product somewhere between two extremes. Some items are positively worded (where higher scores indicate better UX), while others are negatively worded (where higher scores indicate worse UX).

The labels of the 26 semantic differential items are:

  1. Annoying – Enjoyable
  2. Not understandable – Understandable
  3. Creative – Dull
  4. Easy to learn – Difficult to learn
  5. Valuable – Inferior
  6. Boring – Exciting
  7. Not interesting – Interesting
  8. Unpredictable – Predictable
  9. Fast – Slow
  10. Inventive – Conventional
  11. Obstructive – Supportive
  12. Good – Bad
  13. Complicated – Easy
  14. Unlikable – Pleasing
  15. Usual – Leading edge
  16. Unpleasant – Pleasant
  17. Secure – Not secure
  18. Motivating – Demotivating
  19. Meets expectations – Does not meet expectations
  20. Inefficient – Efficient
  21. Clear – Confusing
  22. Impractical – Practical
  23. Organized – Cluttered
  24. Attractive – Unattractive
  25. Friendly – Unfriendly
  26. Conservative – Innovative

The translations and some adaptations can be found on the UEQ website.

A screenshot of the google forms (link in the caption) containing the UEQ.
The UEQ presented in Google Forms.

Calculating the UEQ score

The rules

Positively worded items are: 

1, 2, 6, 7, 8, 11, 13, 14, 15, 16, 20, 22, and 26.

Negatively worded items are: 

3, 4, 5, 9, 10, 12, 17, 18, 19, 21, 23, 24, and 25.

The overall score of the UEQ (the product’s score)

There is no single “overall UEQ score”. Instead, the questionnaire is interpreted through three overarching qualities and six sub-dimensions. However, if we are really after an overall score we can use the score of the Attractiveness scale, which also should be approximately (Pragmatic + Hedonic) / 2.

We will assume in the calculation formulas below that all replies are in the range of 1 to 7.

The product’s Attractiveness score

To get the Attractiveness score for each participant:

  1. Add items 1, 14, and 16 
  2. From the number 24, subtract items 12, 24, and 25 
  3. Add the two numbers from the previous steps, divide by 6, and subtract 4

Or in one formula:

\text{participant\_attractiveness\_score} = \frac{item_{1} + item_{14} + item_{16} + 24 - (item_{12} + item_{24} + item_{25})}{6} - 4

To get the overall Attractiveness score, calculate the mean of all participants’ scores.

The product’s Pragmatic Quality score

To get the Pragmatic Quality score for each participant:

Or in one formula:

  1. Add items 2, 8, 11, 13, 20, and 22
  2. From the number 48, subtract items 4, 9, 17, 19, 21, 23
  3. Add the two numbers from the previous steps, divide by 12, and subtract 4
\text{participant\_pq\_score} = \frac{item_{2} + item_{8} + item_{11} + item_{13} + item_{20} + item_{22} + 48 - (item_{4} + item_{9} + item_{17} + item_{19} + item_{21} + item_{23})}{12} - 4

To get the overall Pragmatic Quality score, calculate the mean of all participants’ scores.

The product’s Hedonic Quality score

To get the Hedonic Quality score for each participant:

  1. Add items 6, 7, 15, and 26
  2. From the number 32, subtract items 3, 5, 10, and 18
  3. Add the two numbers from the previous steps, divide by 8, and subtract 4

Or in one formula:

\text{participant\_hq\_score} = \frac{item_{6} + item_{7} + item_{15} + item_{26} + 32 - (item_{3} + item_{5} + item_{10} + item_{18})}{8} - 4

To get the overall Hedonic Quality score, calculate the mean of all participants’ scores.

The five sub-dimensions of the UEQ

The Pragmatic and Hedonic dimensions are further broken down into 5 sub-dimensions. Even though these dimensions were reported in the initial publications, recent research has shown mixed results in replicating them (e.g. research by Schankin et. al). The five sub-dimensions were introduced to help pin-point which aspects of the product need to be improved, for the time being, we suggest interpreting these dimensions with caution.

Pragmatic Quality dimensions:

Perspicuity (ease of learning and clarity):

To get the Perspicuity score for each participant:

  1. Add items 2, and 13
  2. From the number 16, subtract items 4, and 21
  3. Add the two numbers from the previous steps, divide by 4, and subtract 4

Or in one formula:

\text{participant\_perspicuity\_score} = \frac{item_{2} + item_{13} + 16 - (item_{4} + item_{12})}{4} - 4

To get the overall Perspicuity score, calculate the mean of all participants’ scores.

Efficiency (task completion and productivity): 

To get the Efficiency score for each participant:

  1. Add items 20, and 22
  2. From the number 16, subtract items 9, and 23
  3. Add the two numbers from the previous steps, divide by 4, and subtract 4

Or in one formula:

\text{participant\_efficiency\_score} = \frac{item_{20} + item_{22} + 16 - (item_{9} + item_{23})}{4} - 4

To get the overall Efficiency score, calculate the mean of all participants’ scores.

Dependability (predictability, control, and trust):

To get the Dependability score for each participant:

  1. Add items 8, and 11
  2. From the number 16, subtract items 17, and 19
  3. Add the two numbers from the previous steps, divide by 4, and subtract 4

Or in one formula:

\text{participant\_dependability\_score} = \frac{item_{8} + item_{22} + 16 - (item_{17} + item_{19})}{4} - 4

To get the overall Dependability score, calculate the mean of all participants’ scores.

Hedonic Quality dimensions:

Stimulation (excitement, motivation, and engagement):

To get the Stimulation score for each participant:

  1. Add items 6, and 7
  2. From the number 16, subtract items 5, and 18
  3. Add the two numbers from the previous steps, divide by 4, and subtract 4

Or in one formula:

\text{participant\_stimulation\_score} = \frac{item_{6} + item_{7} + 16 - (item_{5} + item_{18})}{4} - 4

To get the overall Stimulation score, calculate the mean of all participants’ scores.

Novelty (innovation and creativity):

To get the Novelty score for each participant:

  1. Add items 15, and 26
  2. From the number 16, subtract items 3, and 10
  3. Add the two numbers from the previous steps, divide by 4, and subtract 4

Or in one formula:

\text{participant\_novelty\_score} = \frac{item_{15} + item_{26} + 16 - (item_{3} + item_{10})}{4} - 4

To get the overall Novelty score, calculate the mean of all participants’ scores.

Understanding the UEQ’s score

Benchmarks

The UEQ provides a benchmark dataset based on 21175 evaluations of 468 products across industries. These benchmarks can be used to compare the results from the product we are evaluating with the products in the dataset and help put a “label” to our numbers. The authors of the UEQ have provided a percentile rank (similar to SUPR-Q) that can be interpreted per subscale as shown on the tables below:

Benchmark for Attractiveness

PercentileMeanLabel
< 25%Bad
25%0.69Bellow Average
50%1.18Above Average
75%1.58Good
90%1.84Excellent

Benchmark for Perspicuity

PercentileMeanLabel
< 25%Bad
25%0.72Bellow Average
50%1.2Above Average
75%1.73Good
90%2Excellent

Benchmark for Efficiency

PercentileMeanLabel
< 25%Bad
25%0.6Bellow Average
50%1.05Above Average
75%1.5Good
90%1.88Excellent

Benchmark for Dependability

PercentileMeanLabel
< 25%Bad
25%0.78Bellow Average
50%1.14Above Average
75%1.48Good
90%1.7Excellent

Benchmark for Stimulation

PercentileMeanLabel
< 25%Bad
25%0.5Bellow Average
50%1Above Average
75%1.35Good
90%1.7Excellent

Benchmark for Novelty

PercentileMeanLabel
< 25%Bad
25%0.16Bellow Average
50%0.7Above Average
75%1.12Good
90%1.6Excellent

For example a product that has an attractiveness score of 1.21, is perceived as having better attractiveness than 50% of the products in the dataset, and worse than 25%. We can label it then as “Above Average” regarding its attractiveness.

The Excel file provided by UEQ can be used both as a shortcut to the calculation and to the benchmarks.

Factors that affect UEQ

FactorInfluenceStudies
GenderPossibly no significant influenceInitial data from Schrepp et al. and initial data for UEQ-S Kollmorgen et al.
Usage frequencyPossibly some influence on Attractiveness, PQ and HQInitial data for UEQ-S from Kollmorgen et al.
Self-reported knowledge of the productPossibly some influence on Attractiveness, PQ and HQInitial data for UEQ-S from Kollmorgen et al.
Duration of usePossibly some influence on Attractiveness, PQ and HQInitial data for UEQ-S from Kollmorgen et al.

Leave a Comment

Your email address will not be published. Required fields are marked *

To top
WordPress Cookie Plugin by Real Cookie Banner