Timing Parameters

Experiments

Testable provides fine-grained control over every temporal aspect of a trial. All timing values are in **milliseconds** unless noted otherwise.

ITI: inter-trial interval

The blank interval between the end of one trial and the start of the next. The screen is empty (or shows the background colour) during this time.

ITI = 500

ISI: inter-stimulus interval

The blank interval between stimuli within a single trial during sequential presentation. Relevant when you have multiple stim columns and want a gap between them.

For trials with three stimuli:

ISI = 300; 300

This places a 300 ms gap after the first stimulus and another 300 ms gap after the second stimulus.

Tip

Use the shorthand ms*number when ISIs are identical. E.g. 1000*3 is the same as 1000;1000;1000.

Grouping stimuli into screens: stimScreens

When some stimuli should share a screen, state the grouping in the stimScreens column and keep the timing columns as plain value lists. Write one count per screen, joined with +:

stimScreens = 2+1
presTime = 500
ISI = 0

With three stimuli: the first two appear together for 500 ms, then the third alone until the response (presTime takes one value per screen, or one fewer to leave the last screen up until the participant responds; ISI takes one value per gap). Two shorthands cover even splits: stimScreens = 2 makes screens of two until the stimuli run out, and 2*4 is four screens of two. With stimScreens set and no timing written at all, screens run for 1000 ms with 500 ms gaps and the last screen waits for the response.

To randomise the stimulus order while keeping the screen structure, see the stimOrder column: random(screens) shuffles whole screens and random(within screens) shuffles inside each screen.

The older notation: empty ISI slots

Before stimScreens, the grouping was written into ISI itself by omitting a value:

ISI = 500;;1000

With four stimuli: 500 ms after stim1, then stim2 and stim3 appear on the same screen (no ISI), then 1000 ms before stim4. This notation still works, but stimScreens says the same thing more clearly and keeps grouping and timing in separate columns; prefer it for new files.

ISI = 0

Using ISI = 0 shows the next stimulus instantly after the previous one ends:

ISI = 500;0;1000

With four stimuli: 500 ms after stim1, stim2 appears, stim3 appears instantly, then 1000 ms before stim4.

presTime: stimulus presentation time

How long each stimulus is shown before it disappears. Set it per stimulus using semicolons:

presTime = 500; 1000

The first stimulus shows for 500 ms, the second for 1000 ms.

Tip

Presentation times are rounded to the nearest screen refresh frame (typically 16.67 ms at 60 Hz). If exact timing matters, check the presTime_ms column in your results file to see the actual achieved presentation times.

responseWindow: when participants can respond

The responseWindow column controls when responses open within a trial and, optionally, for how long.

The default

With no responseWindow column, or the cell left empty, responses open when the last stimulus appears and stay open indefinitely, even after the stimulus disappears, until a response is given.

Setting a deadline

A plain number gives the participant a deadline:

responseWindow = 2000

Responses open at the last stimulus as usual and close 2000 ms later. If nothing was pressed, the response is recorded as timeout and the trial auto-advances.

Choosing when responses open

Write a start position instead of, or in front of, the deadline:

Example	Behaviour
responseWindow = 2000	Opens at last stimulus, closes 2000 ms later
responseWindow = afterStim	Opens only after the last stimulus is hidden (or audio/video playback ends), stays open until response
responseWindow = afterStim_2000	Same, plus a 2000 ms deadline starting when the window opens
responseWindow = lastStim	Opens at last stimulus onset and closes when it is hidden; a default response is recorded if none was given
responseWindow = withStim	Open only while stimulus screens are displayed
responseWindow = allStim	Opens at trial start, closes when the last stimulus is hidden
responseWindow = allTrial	Opens at trial start, no automatic close unless a deadline is added
responseWindow = 500_2000	Both parts numeric: opens 500 ms after trial start, closes 2000 ms later
responseWindow = 0	Opens at trial start, no deadline

Caution

Start positions are case-sensitive, and an unrecognised value (including typos like afterstim) means the response window never opens at all, so the trial collects nothing.

Minimum response time: min and max

Use min: to block responses for an initial period, max: as a deadline, or both:

• responseWindow = min:2000: responses do not count for the first 2000 ms. Key presses and clicks made during the minimum are buffered and applied the moment it ends; text boxes and the advance button are disabled and then re-enabled.
• responseWindow = min:2000;max:10000: the same, plus a timeout at 10 s.

These work with any response modality, not just text boxes. When a text-box trial times out with a partial answer, the recorded response has ;timeout appended.

Feedback timing

How long feedback stays up (feedbackTime) and where it appears (feedbackOptions) are covered on the Feedback page, together with everything else about feedback.

Holding the response window open

To hold the response window open even after a response is given, until the full responseWindow duration has elapsed, add the waitResponseWindow flag to the trial’s responseOptions cell. Useful when you want a consistent total trial duration regardless of when the participant responds.

For audio and video trials where the full clip should play before the trial can advance, open the response window only after the stimuli finish, using the afterStim start position described above.

Achieved timing in results

Testable reports the actually achieved timings back into the results file (presTime_ms, ITI_ms, ISI_ms, frame counts, and more), so you can verify and filter trials where the display rounded your requested times to the screen’s refresh rate. These columns are documented in Results Reference: Timing Diagnostics.