Prospective Memory Task

Field	Value
Name	Prospective Memory Task
Version	v0.1.0
URL / Repository	https://github.com/TaskBeacon/T000077-prospective-memory-task
Short Description	Event-based focal versus nonfocal prospective-memory task embedded in lexical decision
Created By	TaskBeacon
Date Updated	2026-07-27
PsyFlow Version	local TaskBeacon checkout
PsychoPy Version	environment-managed
Modality	behavior
Language	Chinese instructions with English lexical stimuli
Voice Name	disabled

1. Task Overview

This task measures prospective memory: remembering to perform an intended action when a future event occurs while attention is occupied by an ongoing task. Participants make lexical decisions throughout. After a baseline block, they complete focal and nonfocal prospective-memory blocks in counterbalanced order.

The focal intention is triggered by the exact word PACKET. The nonfocal intention is triggered by the letter sequence TOR embedded in a word such as DOCTOR. The intended action is performed on the neutral WAITING screen after the lexical decision, which keeps ongoing-task accuracy and delayed-intention execution separable.

2. Task Flow

Block-Level Flow

Step	Participant-visible flow	Implementation detail
1	Lexical-decision instructions and practice	10 trials from assets/stimuli.csv; lexical feedback only
2	Baseline lexical-decision block	105 trials; no PM intention
3	Focal or nonfocal intention instruction	Order from participant parity; instruction stimulus matches block type
4	Embedded-figures distractor	distractor_instruction + embedded_figures; 120 s
5	Prospective-memory block	105 trials; targets at positions 25, 50, 75, and 100
6	Neutral break and second PM condition	Repeat steps 3-5 for the other cue focality
7	Completion and export	Summary screen; one reduced row per logical trial

Trial-Level Flow

Phase	Stimulus ID	Duration / response	Result
Lexical decision	letter_string	Up to 3 s; F nonword, J word	Ongoing-task accuracy and RT
Waiting	waiting	Up to 10 s; SPACE normally, / for PM cue	PM hit, miss, or false alarm
Continue	waiting	Conditional after /; SPACE, up to 10 s	Advances to next trial
Practice feedback	feedback_correct / feedback_incorrect	0.6 s; practice only	Lexical correctness

Controller Logic

Component	Rule
Adaptive controller	None
Trial planner	Deterministic build_plans() preserves target positions, lexicality ratio, and target identity
Counterbalancing	Odd participant IDs receive focal first; even IDs receive nonfocal first

Other Logic

Component	Rule
PM scoring	Hit, miss, and false alarm are scored independently of lexical accuracy
Data reduction	Each returned trial dictionary is one logical-trial row
Randomness	Concrete fillers are selected from a stable block seed before run_trial()

3. Configuration Summary

Settings below are from config/config.yaml.

a. Subject Info

Field	Value
Participant ID	Three-digit integer
Counterbalancing	Odd ID: focal first; even ID: nonfocal first

b. Window Settings

Setting	Value
Size	1280 × 800
Units	degrees
Background	#f4f5f7
Viewing distance	57 cm

c. Stimuli

Element	Participant-facing content	Response
Ongoing item	Uppercase English word or pronounceable nonword	J word; F nonword
Focal PM cue	PACKET	/ on following WAITING
Nonfocal PM cue	A word containing TOR	/ on following WAITING
Nontarget waiting	WAITING	SPACE
Distractor	Embedded-figures line drawing	Observe/search for 120 s

d. Timing

Phase	Human duration
Lexical response window	3.0 s
Waiting safety window	10.0 s
Post-PM continue window	10.0 s
Embedded-figures distractor	120.0 s
Practice feedback	0.6 s

e. Triggers

Range	Events
1 / 99	Experiment start / end
10 / 90	Block start / end
20-27	Lexical and PM-target onsets
30-32	Lexical responses / timeout
40-43	Waiting onset, PM response, advance, timeout
50 / 60	Practice feedback / distractor

f. Adaptive Controller

Parameter	Value
Adaptive controller	Not applicable
Fixed schedule helper	src/utils.py::build_plans

Run modes:

python main.py human
python main.py qa --config config/config_qa.yaml
python main.py sim --config config/config_scripted_sim.yaml
python main.py sim --config config/config_sampler_sim.yaml

4. Methods (for academic publication)

The task adapts the event-based prospective-memory protocol of Brewer et al. (2010). Participants first completed an ongoing lexical-decision baseline. Each trial presented one uppercase letter string, and participants classified it as a word or nonword using the J and F keys, respectively. After the response, a neutral waiting screen was shown and SPACE advanced the task.

Participants then encoded focal and nonfocal delayed intentions in counterbalanced blocks. For the focal intention, the exact word PACKET served as the PM cue. For the nonfocal intention, any word containing TOR served as the cue. Participants executed the intention by pressing / during the waiting screen after classifying the cue. A 120-second embedded-figures distractor separated intention encoding from task performance, and PM instructions were not repeated after the distractor. Each formal block contained 105 trials, with PM cues at positions 25, 50, 75, and 100.

Primary dependent variables are cue-detection accuracy by focality, PM false-alarm rate, ongoing lexical accuracy and latency, and PM-related ongoing-task cost relative to baseline. The unpublished filler lists and response deadlines were not available; the included filler pool and finite safety windows are documented as inferences in the evidence mappings.

Selected sources:

• Brewer, G. A., Knight, J. B., Marsh, R. L., & Unsworth, N. (2010). Memory & Cognition, 38, 304-311. https://doi.org/10.3758/MC.38.3.304
• McDaniel, M. A., Howard, D. C., & Butler, K. M. (2008). Memory & Cognition, 36, 716-724. https://doi.org/10.3758/MC.36.4.716
• McDaniel, M. A., Umanath, S., Einstein, G. O., & Waldum, E. R. (2015). Frontiers in Human Neuroscience, 9, 392. https://doi.org/10.3389/fnhum.2015.00392