Sample Size Estimator for the UEQ

This Excel file tries to provide some assistance in the planning of studies based on the UEQ. A typical and
critical question in the planning phase is how many participants are required to reach a certain accuracy of
the measured scale values.
This sheet provides two methods to get some insights concerning the number of participants you need to
recruit for your study. The first (Worksheet "Direct_STD_Entry") requires that you have done a similar
study in the past (similar or same product, same way to recruit participants, same user group, sufficient
sample size). In this case the standard deviations from this previous study is a good basis to estimate the
required sample size for a required accuracy (this is measured by the width of the confidence interval).
Since the UEQ has six scales use the highest standard deviation of a scale from this previous study to be on
the save side for the estimation.
If you do not have access to results from a previous study, the second method (Worksheet
"Level_of_Agreement") allows you to get a rough idea based on the level of agreement you assume (do
you expect that your participants differ strongly in their opinion concerning the product or do you expect
that they more or less share the same opinion). Of course, this method is less reliable and has a higher risk
to result in an over- or underestimation of the required sample size.
More background information concerning these methods can be found in the UEQ Handbook.
Be careful, the methods provided in this Excel are just heuristics that should help to make an educated
guess about the required sample size. The estimation is based on several assumptions that can be wrong
in some circumstances.
Use standard deviation measured in a previous study
Use this worksheet if you have data from a previous study concerning the product
with a sufficient sample size and a similar target group. In this case simply pick the
highest standard deviation observed for the six UEQ scales and use this as a best
guess to estimat the required number of participants.
Enter this value in the cell highlighted in yellow below. The table will then show the
widths of the 95% and 90% confidence interval for typical values of the sample size.
Choose the sample size that matches your target width for the confidence interval.

Standard deviation: 0.92

Standard Width of 95% Width of 90%

N
Error Confidence Interval Confidence Interval
10 0.29 1.14 0.96
20 0.21 0.81 0.68 Standard error and widt
30 0.17 0.66 0.55
1.20
40 0.15 0.57 0.48
50 0.13 0.51 0.43 1.10
60 0.12 0.47 0.39
70 0.11 0.43 0.36 1.00
80 0.10 0.40 0.34
0.90
90 0.10 0.38 0.32
100 0.09 0.36 0.30 0.80
110 0.09 0.34 0.29
0.70
120 0.08 0.33 0.28
130 0.08 0.32 0.27 0.60
140 0.08 0.30 0.26
150 0.08 0.29 0.25 0.50
160 0.07 0.29 0.24 0.40
170 0.07 0.28 0.23
180 0.07 0.27 0.23 0.30
190 0.07 0.26 0.22
0.20
200 0.07 0.26 0.21
210 0.06 0.25 0.21 0.10
220 0.06 0.24 0.20
230 0.06 0.24 0.20 0.00
10 20 30 40 50 60 70 80
240 0.06 0.23 0.20
250 0.06 0.23 0.19 Standa
ndard error and width of confidence interval for confidence levels 95% and 90%

40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

Standard Error Confidence level 95% Confidence level 90%

Base the estimation on the expected level of agreement
Use this worksheet if you do not have data from a previous study. Below you can see estimations for
the width of the confidence interval that are based on a larger set of 123 analysed UEQ studies. For
each of these studies the standard deviation of all the UEQ scales was calculated. This data set was then
split into three equally sized subsets based on the size of the standard deviation. The means over these
subsets are used for the estimation below. The size of a standard deviation of a scale mean can be
interpreted as level of disagreement. The higher the standard deviation is, the more participants
disagree concerning their UX impression.
If you have some idea about the level of disagreement you can expect, then choose the corresponding
level. If not, then it is best to orient your estimation on the medium level of agreement.

Low Agreement Medium Agreement High Agreement

Std.Dev. = 1.17 Std.Dev. = 0.93 Std.Dev. = 0.66
N
Width 95% Width 90% Width 95% Width 90% Width 95% Width 90%
Conf. Interval Conf. Interval Conf. Interval Conf. Interval Conf. Interval Conf. Interval
10 1.45 1.22 1.15 0.97 0.82 0.69
20 1.03 0.86 0.82 0.68 0.58 0.49
30 0.84 0.70 0.67 0.56 0.47 0.40
40 0.73 0.61 0.58 0.48 0.41 0.34
50 0.65 0.54 0.52 0.43 0.37 0.31
60 0.59 0.50 0.47 0.39 0.33 0.28
70 0.55 0.46 0.44 0.37 0.31 0.26
80 0.51 0.43 0.41 0.34 0.29 0.24
90 0.48 0.41 0.38 0.32 0.27 0.23
100 0.46 0.38 0.36 0.31 0.26 0.22
110 0.44 0.37 0.35 0.29 0.25 0.21
120 0.42 0.35 0.33 0.28 0.24 0.20
130 0.40 0.34 0.32 0.27 0.23 0.19
140 0.39 0.33 0.31 0.26 0.22 0.18
150 0.37 0.31 0.30 0.25 0.21 0.18
160 0.36 0.30 0.29 0.24 0.20 0.17
170 0.35 0.30 0.28 0.23 0.20 0.17
180 0.34 0.29 0.27 0.23 0.19 0.16
190 0.33 0.28 0.26 0.22 0.19 0.16
200 0.32 0.27 0.26 0.22 0.18 0.15
210 0.32 0.27 0.25 0.21 0.18 0.15
220 0.31 0.26 0.25 0.21 0.17 0.15
230 0.30 0.25 0.24 0.20 0.17 0.14
240 0.30 0.25 0.24 0.20 0.17 0.14
250 0.29 0.24 0.23 0.19 0.16 0.14
 Width of the 95% and 90% confidence intervals for different levels of agr
1.60

1.50

1.40

1.30

1.20

1.10

1.00

0.90

0.80

0.70

0.60

0.50

0.40

0.30

0.20

0.10

0.00
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 20

95% CI Low Agreement 90% CI Low Agreement 95% CI Medium Agreemen

90% CI Medium Agreement 95% CI High Agreement 90% CI High Agreement
0.00
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 20

95% CI Low Agreement 90% CI Low Agreement 95% CI Medium Agreemen

90% CI Medium Agreement 95% CI High Agreement 90% CI High Agreement
different levels of agreement

160 170 180 190 200 210 220 230 240 250

95% CI Medium Agreement

90% CI High Agreement
160 170 180 190 200 210 220 230 240 250

95% CI Medium Agreement

90% CI High Agreement