MoFaCTS Trial Types Reference

Documentation home View source on GitHub

Last Updated: 2026-10-09 | For: Teachers and content creators | Level: Intermediate

Trial types control presentation, response and feedback. The current runtime supports s, d, t, h, m and n. Assessment units use a fixed authored schedule; learning units choose study, drill or test trials through their learning model. A unit's title does not determine its trial type.

Types and feedback

Code	Type	Learner behavior	Answer feedback
`s`	Study	Read the question and displayed answer; advance after the study timer.	The answer is visible during study.
`d`	Drill	Type or select an answer.	Correct answers use `correctprompt`; incorrect answers/timeouts use `reviewstudy`.
`t`	Test	Submit an answer in the authored input mode.	No answer-feedback phase.
`h`	Button test	Select a button answer.	No answer-feedback phase.
`m`	Mandatory correction	After an incorrect answer, enter the displayed correct answer before continuing.	Uses the force-correction stage.
`n`	Timed correction	After an incorrect answer, enter the correct answer or wait for `forcecorrecttimeout`.	Uses the force-correction stage with a time limit.

Durations are milliseconds. Zero correctprompt or reviewstudy skips that outcome's ordinary feedback phase. Tests t and h show no answer feedback even if these durations are positive. Study uses purestudy. There are no separate correct/incorrect feedback visibility switches. History leaves feedback text and type empty when no answer-feedback phase occurs.

m requires a matching correction; n also accepts a matching correction but can leave that stage after its timeout. forceCorrectPrompt supplies the correction instruction, not a separate question answer or scoring rule. A wrong mandatory correction does not advance. Authors should verify the intended correction flow with their actual lesson settings.

Assessment schedule entries

An entry has four comma-separated fields:

stimulusIndex,inputMethod,trialType,location

The first field selects the zero-based stimulus within the cluster (or the supported random choice marker); location positions it relative to the template start. The input method is b for buttons or f for typed input. It is separate from the trial type. In particular, f is not a fast-study trial code. The old trial codes i and f are unsupported.

0,f,d,0
1,b,h,1
2,f,t,2

These entries illustrate drill, button test and typed test. Put them inside the assessment group's template structure; they are not a complete TDF. Prepared transitions and resume retain the authored input method, including a button-mode study introduction with one answer.

The lesson editor changes the TDF. Course assignment tools select uploaded lessons; they do not redefine trial logic. Download and edit the owning package, then review the update confirmation before applying it. See Assessment Sessions for offsets, group positions and schedule generation.

Correction settings

Setting	Purpose
`forceCorrectPrompt`	Text shown when correction is required.
`forcecorrecttimeout`	Maximum correction wait for type `n`, in milliseconds.
`forceCorrection`	Also requests correction after an incorrect response where the feedback flow permits it.
`correctprompt`	Ordinary correct-answer feedback duration.
`reviewstudy`	Ordinary incorrect-answer/timeout feedback duration.
`purestudy`	Study presentation duration.

There is no current trial-type contract that enables a Help button specifically for t or an instructional-test code. Use the lesson's actual authored instructional content and supported controls.

Sequence examples

These examples describe pedagogical order, not complete schedule templates:

Initial learning: s → s → d → d → t.
Spaced review: s → d → t → s → t.
Correction practice: s → n → d → m → t.
Mastery practice: s → d → m → m → t.

Timer Behavior

In the current source contract, drill, study, and feedback timers cover the learner-visible presentation envelope. The timer starts when the trial content begins fading in, and fade-out time is included before the next trial appears. Pre-recorded prompt audio counts against the response window, so content authors should set drill long enough for any required media. Generated TTS may hold an expired timer only until active playback completes; it does not restart the timer. Text entry and speech-recognition activity reset response timers.

Older deployed builds can start the response countdown after the question audio completes, even though latency measurement starts at reveal. A ten-second countdown can then appear as a 14-15-second no-input timeout when the audio lasts about four seconds. See Timer Behavior Depends on the Collection Build before interpreting historical records.

An internal trial type such as t is not determined by the unit's label: a unit named Posttest can contain drill-style trials. Exported 0/-1 timing sentinel values and study-only rows require type-specific interpretation. See Interpreting Trial Timing for timestamp origins, no-input substitutions, and duration formulas.