Timing statistics: how long respondents take

Timing statistics: how long respondents take

LimeSurvey can record how much time each respondent spends on every question, every question group, and the survey as a whole — and, separately, you can put a countdown timer on a question. This article covers both.

Step 1: turn on Save timings — before you activate

Timing is not recorded by default. In the survey editor, open the Settings panel (gear icon), go to Notifications & data, and switch Save timings on.

Do this before activating the survey. Like the other data-recording options, Save timings locks once the survey is active — you cannot switch it on later and recover timings for responses already collected.

Step 2: view the timing statistics

With timings enabled, Timing statistics appears in the classic Responses & statistics toolbar. It shows one row per response, with sortable columns:

  • ID — the response ID.
  • Total time — how long the whole survey took, in seconds.
  • One column per group and per question — the time spent on each, in seconds.

Each row's Action column lets you view the response details, edit the response, or delete it — but note that the timing values themselves cannot be edited.

Two things worth knowing:

  • Group columns only show values when the survey has at least two groups. With a single group, the group time simply equals the total time, so no separate figure is displayed.
  • The timings are stored in the response table. To actually analyse them — averages, outliers, speeders — export your responses and work in a spreadsheet or statistics tool.

Detecting speeders

The most common use of timing data is finding speeders — respondents who race through so fast their answers cannot be trusted. A practical recipe:

  1. Sort the Timing statistics table by Total time, ascending — the suspects are at the top.
  2. For a defensible rule, export the responses and compute the median completion time (median, not average — a few very slow respondents distort the average). Flag everyone under one third of the median.
  3. Check per-question times too: a respondent with a normal total may still have blasted through your key questions in a second or two.
  4. Before removing anyone, cross-check for straightlining — identical answers down every row of an array. Fast and straightlined is a confident removal; fast but varied may just be a decisive respondent.

Prevention works better than cleanup: include an attention-check question ("Please select 'Agree' for this row") in a long array, and for panel samples route failed checks to a screen-out quota so speeders do not consume your sample (see Panel integration). Note that the question timer below enforces a maximum time only — LimeSurvey has no built-in minimum-time gate.

Putting a countdown timer on a question

Separately from recording times, you can limit them. The time_limit question attribute starts a countdown (in seconds) as soon as the question's page loads — set it to 240 for a four-minute limit. When the timer runs out, the question either moves on to the next page automatically or becomes read-only, depending on how you configure the accompanying timer options.

Timers are available on a subset of question types: Short, Long and Huge free text, List (Radio), List (Dropdown), and Text display. Like other advanced attributes, they are set in the classic question editor.

A practical pattern: put the timer on a Text display "question" to time an instruction or stimulus page — show a text or image for exactly 30 seconds, then move on.

Related articles