Paradigms¶
The template ships three complete experiments under Assets/ResXR/Demo Experiments/. Each is a self-contained reference implementation of the flow model: it provides its own *_SessionManager, *_TaskManager, *_TrialManager, and a *_SceneReferencer, plus the data classes it logs. You configure a paradigm by editing fields in the Inspector; you extend one by copying its scripts. New to the Session → Task → Trial structure? Read Architecture first.
The SceneReferencer pattern
Each paradigm has a *_SceneReferencer component (a singleton you place once in the scene) holding public references to every scene object the flow needs — instruction panels, position marks, interactive objects, and timing values. The flow managers read these through XSceneReferencer.Instance.<field> instead of holding their own scene links, so all wiring lives in one place. To re-point an experiment at different objects, you edit the SceneReferencer, not the managers.
Binary Choice¶
A two-alternative forced-choice task. Two images appear in front of the participant; they reach out and touch one with their hand. The chosen side, the named stimulus, the hand used, and the reaction time are logged per trial.
How a choice is detected. Each option is a Choice component with a SpriteRenderer and a trigger collider. ChoicesManager shows both options and races their WaitForTouch() calls with UniTask.WhenAny; the first to be entered by a collider tagged Toucher wins. The hand is resolved from the touching collider's HandCollider (Left or Right).
Trial structure. Each task loads a folder of stimulus sprites and pairs them up; one trial is generated per pair, so the trial count is floor(spriteCount / 2). A fixation cross is shown between stimuli.
Configure it on these components:
| Component | Field | Meaning |
|---|---|---|
BinaryChoice_Task (one per task, in BinaryChoice_TasksConfiguration.tasks[]) |
taskName |
Task label, written into the logged events and tables. |
stimuliOrder |
RandomOrder (shuffled) or FixedOrder (alphabetical by sprite name). |
|
stimuliFolderPath |
A path inside a Resources/ folder (default BinaryChoice/StimuliPairs). Images must be imported as single Sprites; they are loaded with Resources.LoadAll<Sprite>. |
|
taskInstructions |
The InstructionsPanel shown before the task. |
|
BinaryChoice_SceneReferencer |
SecondsBetweenStimuli |
Fixation-cross duration between trials (default 0.5). |
instructionsDisplayTime |
How long task instructions stay up (default 3). |
|
fixationCross, choicesManager, generalInstructions, endInstructions |
Scene object references. | |
ChoicesManager |
choiceA, choiceB |
The two Choice slots. |
Logged data. Generic event markers (task_start:…, trial_start:…, fixation, session_start/session_end) plus two custom tables: ChoiceEvents (one row per trial — Task, Trial, OptionAName, OptionBName, Choice, ChosenOption, HandUsed, ReactionTime, displayTime, ChoiceTime) and StimulusBounds (one row per slot — the renderer and collider center/size of each option). See Data Output.
Scene: Binary Choice Experiment.unity.
Maze Navigation¶
The participant stands in a physical-scale maze and walks to reach a coin. Between trials the maze rotates 180° so the same physical space yields a new route.
How a trial ends. The trial awaits Coin.WaitForCoinPickup(), which resolves when the coin is touched. The coin debounces pickups (a one-second _acceptPickUps window) so that multiple fingers contacting it in the same instant count once, then plays its hide/press animation and a sound.
Trial structure. Maze_Task.numOfTrials (default 5) sets the trials per task; the trials carry no per-trial parameters. Before each trial a "Trial #N" panel is shown. Between trials (but not after the last), the view fades to black, Maze.Rotate180Degrees() flips the maze around its vertical axis, and the view fades back.
| Component | Field | Meaning |
|---|---|---|
Maze_Task (in Maze_SessionManager._tasks[]) |
numOfTrials |
Number of trials in the task (default 5). |
Maze_SceneReferencer |
startingPositionMark |
Where the participant begins each trial (PlayerPositionMark). |
coin, maze |
The target coin and the rotatable maze. | |
trialStartPanel, generalInstructions, endInstructions |
Instruction panels. | |
trialNumberPanelVisibleDuration |
"Trial #N" display time (default 3). |
|
Coin |
_facingForward |
Orientation of the coin mesh. |
Logged data. Event markers (trial_start:…, coin_pickup:…, maze_rotation:state=…, player_at_start_zone, …) plus the MazeTrialData custom table (one row per trial — Task, Trial, TrialName, MazeRotatedAtStart, the coin position CoinX/Y/Z, and the start-zone position StartZoneX/Y/Z). It records the same Task/Trial/TrialName columns as the shared TrialsData table, plus the Maze-specific columns.
Scene: Maze Experiment.unity.
Museum Viewing¶
A gallery experience with two task types, selected per task by Museum_Task.taskType:
- FreeExploration — the participant walks the gallery for
durationInSeconds. There is no interaction; head, position, and gaze are captured passively by the recorder. Implemented as a single trial that waits out the duration. - ImagesRating — a sequence of images is presented one at a time, each with a slider. The participant sets the slider and confirms; the raw and normalized ratings are logged. One trial is generated per image, in array order.
Before the rating task the player is repositioned to a marked spot and shown rating instructions. The Inspector drawer (MuseumTaskDrawer) hides durationInSeconds unless the task type is FreeExploration.
| Component | Field | Meaning |
|---|---|---|
Museum_Task (in Museum_SessionManager._tasks[]) |
taskType |
FreeExploration or ImagesRating. |
durationInSeconds |
Free-exploration length (shown only for that type). | |
Museum_SceneReferencer |
artworks (Renderer[]) |
The gallery pieces. Each GameObject's name becomes the ArtworkName in the logged bounds, so give them meaningful, unique names. |
artworkColliders (Collider[]) |
Colliders matching artworks in order. |
|
ratingTaskPlayerPositionMark |
Where the participant stands to rate. | |
imagesRating, plus the four instruction panels (welcomeInstructions, endInstructions, endOfExplorationInstructions, ratingInstructions) |
Scene references. | |
ImagesRating |
imagesToRate (SpriteRenderer[]) |
The images to present, in order. |
ratingPanel, ratingSlider |
The rating UI. The Slider exposes MinValue, MaxValue, NumOfIntervals, AllowContinuousValues. |
Logged data. Event markers (trial_start:…, free_exploration:…, …) plus three custom tables: ImageRatings (per rated image — Task, Trial, ImageName, RawRating, NormalizedRating; onset is presentation start, duration is deliberation time), SliderConfig (one row — the slider's range and interval settings), and ArtworkBounds (one row per artwork — name, renderer center/size, Euler rotation, and collider center/size).
Scene: Museum Experiment.unity.
Building your own¶
The fastest route to a new experiment is to copy the paradigm closest to yours and edit its flow scripts and SceneReferencer, or to start from the empty Flow Management scaffold described in Architecture. For logging your own per-trial fields, see Scripting & API → Custom tables and Extending the Template.