Two participants can finish the same task and describe very different experiences. One finds the route immediately; the other succeeds after backtracking and wondering whether a payment has gone through. The Single Ease Question, usually shortened to SEQ, captures a participant’s judgement of how easy or difficult a task felt immediately after attempting it.
It uses a seven-point response scale, conventionally running from very difficult at 1 to very easy at 7. This makes it a task-level measure of perceived ease, distinct from the System Usability Scale, which gathers a broader assessment of a system.
Ask while the task is still fresh
Present the question after each selected task attempt, including unsuccessful attempts, before a long discussion changes how someone remembers the experience. MeasuringU’s account of the SEQ’s development explains the research behind the measure and the evolution of its wording.
Use an established version consistently and record the wording, endpoint labels and direction of the scale. A useful current formulation asks how easy or difficult it was to complete the task. Keep any follow-up question separate from the rating so that participants can explain an answer without being encouraged to change it.
For an illustrative appointment-booking study, a participant might give a low rating despite securing the desired slot. Their explanation could reveal uncertainty about whether the booking was confirmed. The score tells you that the experience felt difficult; the accompanying evidence helps locate the cause.
Read the score with the observed behaviour
The SEQ measures perception, not correctness. Someone who unknowingly books the wrong appointment can find the process easy, while someone who succeeds may find it demanding. Read the responses alongside task success rate, errors and observations rather than using one measure to stand in for all the others.
Patterns of disagreement can be particularly informative. If participants give high ease ratings while repeatedly choosing an unsuitable option, the design may be making an incorrect interpretation feel plausible. If they succeed but rate the task poorly, the route may impose work that a completion figure conceals.
A rating cannot identify which screen or interaction caused the difficulty. That explanation still requires evidence from the session or focused follow-up research.
Interpret averages without inventing a pass mark
Report the mean with the number of responses and an appropriate expression of uncertainty. Retain the distribution when it helps: a middle average could describe generally middling experiences or a split between people who found the task very easy and very difficult.
MeasuringU’s interpretation guidance connects scores with reference data and descriptive labels. Such benchmarks provide context, but the suitability of a comparison depends on the task and study. There is no single score that certifies every product as usable, and a seven-point mean should not be relabelled as a percentage of satisfied customers.
Use repeated measurement to examine a change
When testing a revised design, keep the task, scale and recruitment approach sufficiently comparable to interpret the result. If the second group attempts an easier task or already knows the product, a higher score may not be attributable to the revision.
Within the same study, remember that several ratings from one participant are related observations. Analysis across tasks or versions needs to account for the study design. A useful report combines the numerical result with what changed in people’s attempts, giving the team a reason to trust an improvement beyond the movement of the average.