AI agent skill · by @NeuroAIHub

Alternative Uses Task Designer

Domain-validated guidance for designing Alternative Uses Task (AUT) experiments measuring divergent thinking, with parameters for AI-augmented and traditional conditions

Archived 2026-10-23 · Source updated 2026-10-02 · License: AGPL-3.0. Check the original repository for the latest version.

SKILL.md instructions

# Alternative Uses Task Designer

## Purpose

This skill encodes expert methodological knowledge for designing Alternative Uses Task (AUT) experiments — the most widely used measure of divergent thinking in creativity research. It provides domain-specific parameter recommendations for stimulus selection, timing, condition design (including AI-augmented variants), online implementation, and quality control. A general-purpose programmer would not know the standard objects, timing constraints, scoring dimensions, or the critical design choices that determine whether an AUT experiment yields valid creativity data.

## When to Use This Skill

- Designing a study measuring divergent thinking or creative ideation
- Setting up an AUT experiment with AI-assisted conditions (e.g., ChatGPT, web search)
- Choosing appropriate objects, timing, and instructions for an AUT
- Adapting the AUT for online administration (MTurk, Prolific, Qualtrics)
- Planning attention checks and exclusion criteria for creativity studies

## Research Planning Protocol

Before executing the domain-specific steps below, you MUST:

1. **State the research question** — What specific question is this AUT study addressing?
2. **Justify the method choice** — Why AUT (not RAT, CAT, or other creativity tasks)? What alternatives were considered?
3. **Declare expected outcomes** — What pattern of results would support vs. refute the hypothesis?
4. **Note assumptions and limitations** — What does AUT assume about creativity? Where could it mislead?
5. **Present the plan to the user and WAIT for confirmation** before proceeding.

For detailed methodology guidance, see the `research-literacy` skill.


## ⚠️ Verification Notice

This skill was generated by AI from academic literature. All parameters, thresholds, and citations require independent verification before use in research. If you find errors, please [open an issue](https://github.com/HaoxuanLiTHUAI/awesome_cognitive_and_neuroscience_skills/issues).

## AUT Overview

The Alternative Uses Task (Guilford, 1967) asks participants to generate as many unusual uses as possible for a common everyday object within a fixed time limit. It is the standard measure of **divergent thinking** — the ability to generate multiple, varied, and novel ideas.

### Core Parameters

| Parameter | Default | Source |
|-----------|---------|--------|
| Time limit | **5 minutes** per object | Lee & Chung, 2024; Reiter-Palmon et al., 2019 |
| Number of objects | **1-3** per session | Silvia et al., 2008 |
| Response format | Open-ended text, one use per line | Reiter-Palmon et al., 2019 |
| Instructions emphasis | "unusual, creative, uncommon" uses | Guilford, 1967; Wallach & Kogan, 1965 |

### Standard Objects

Objects should be **concrete, familiar, and have many conventional uses** so that departing from typical uses requires genuine creative thinking.

| Object | Commonly Used In | Source |
|--------|-----------------|--------|
| Brick | Most widely validated | Guilford, 1967 |
| Paperclip | Classic Guilford item | Guilford, 1967 |
| Newspaper | Used in Lee & Chung, 2024 | Lee & Chung, 2024 |
| Cardboard box | Common alternative | Silvia et al., 2008 |
| Tin can | Common alternative | Wallach & Kogan, 1965 |
| Shoe | Frequently used | Reiter-Palmon et al., 2019 |

> **Avoid**: Objects that are already unusual (e.g., "kaleidoscope") or that have very few conventional uses (e.g., "toothpick"). The task requires a clear baseline of common uses to depart from.

## Condition Design

### Standard Conditions (Creativity Research)

```
Is the study examining AI's impact on creativity?
|
+-- YES --> Include at minimum:
| 1. AI-assisted condition (e.g., ChatGPT access)
| 2. No-assistance control
| 3. [Recommended] Web search control (Lee & Chung, 2024, Exp 2A/2B)
|
+-- NO --> Standard AUT with:
 1. Experimental manipulation (priming, mood, instructions)
 2. Control condition (neutral or baseline)
```

### AI-Augmented Design (Lee & Chung, 2024)

For studying AI's impact on creativity:

| Condition | Participant Instructions | Implementation |
|-----------|------------------------|----------------|
| **ChatGPT** | "You may use ChatGPT to assist you" | Embed ChatGPT in new browser tab; record interaction logs |
| **Web Search** | "You may use web search to assist you" | Allow Google/Bing access; record search queries |
| **No Assistance** | "Complete the task on your own" | Disable external tool access |

**Critical design decisions**:

- **Between-subjects** assignment to conditions (Lee & Chung, 2024) — avoids carryover effects
- **Random assignment** via survey platform (Qualtrics randomizer)
- **Cover story**: Frame as "idea generation study," not "creativity study" to reduce demand characteristics
- **Manipulation check**: Ask participants whether they used the assigned tool

## Online Implementation

### Platform Specifications

| Parameter | Recommendation | Source |
|-----------|---------------|--------|
| Platform | Qualtrics (survey) + MTurk/Prolific (recruitment) | Lee & Chung, 2024 |
| Sample size per condition | **100-200** for between-subjects AUT | Lee & Chung, 2024 (N=256 in Exp 2B) |
| Compensation | Prolific minimum + bonus for completion | Lee & Chung, 2024 |
| Estimated duration | **15-25 minutes** total session | Lee & Chung, 2024 |

### Attention and Quality Checks

1. **Attention check questions** — Embed 1-2 instructed-response items (e.g., "Please select 'Strongly Agree' for this item") (Oppenheimer et al., 2009)
2. **Seriousness check** — Post-task: "Did you take this study seriously?" (Lee & Chung, 2024)
3. **Gibberish detection** — Flag responses that are incoherent or clearly auto-generated
4. **Minimum response threshold** — Exclude participants with **<2 responses** (indicates disengagement)
5. **Duplicate detection** — Check for repeated responses within a participant
6. **Bot detection** — reCAPTCHA or honeypot fields; check completion time (exclude if <3 minutes)

### Exclusion Criteria (Lee & Chung, 2024)

- Failed attention check: **exclude**
- Self-reported not taking study seriously: **exclude**
- Completion time <3 minutes or >60 minutes: **flag for review**
- Fewer than 2 responses on AUT: **exclude**
- Non-native speakers (if language fluency is critical): **exclude or control for**

## Additional Measures

### Baseline Creativity

| Measure | Items | Duration | What It Captures | Source |
|---------|-------|----------|-----------------|--------|
| **RAT** (Remote Associates Test) | **15 items** | ~5 min | Convergent thinking | Mednick, 1962; Lee & Chung, 2024 |
| **Creative Achievement Questionnaire** | 10 domains | ~5 min | Real-world creative accomplishment | Carson et al., 2005 |
| **Creative Self-Efficacy Scale** | 3 items, 5-point Likert | <1 min | Belief in own creative ability | Tierney & Farmer, 2002 |

### Mediators / Moderators (Lee & Chung, 2024)

- **Creative self-efficacy** — 3-item scale (Tierney & Farmer, 2002): "I have confidence in my ability to solve problems creatively," "I feel that I am good at generating novel ideas," "I have a knack for further developing the ideas of others." 5-point Likert (1 = strongly disagree to 5 = strongly agree)
- **Task engagement** — Self-report items on effort and involvement
- **AI reliance** — Whether and how extensively participants used the AI tool

## Common Pitfalls

1. **Using "creative" in instructions without care**: Telling participants to "be creative" changes the scoring profile — it increases originality but may decrease fluency. Decide a priori and keep consistent across conditions (Nusbaum et al., 2014).

2. **Confounding fluency with originality**: Participants who generate more ideas statistically have a higher chance of producing rare ideas. Either control for fluency when analyzing originality, or use ratio-based measures (Silvia et al., 2008).

3. **Not controlling for AI-generated text**: In AI-augmented conditions, participants may copy-paste AI outputs. Record interaction logs and code whether responses are self-generated, AI-assisted, or directly copied (Lee & Chung, 2024).

4. **Ignoring the web search control**: Comparing ChatGPT only to no-assistance confounds AI-specific effects with general information access effects. Include a web search condition as active control (Lee & Chung, 2024, Exp 2A/2B).

5. **Insufficient sample size for between-subjects**: AUT effect sizes for condition differences are typically small-to-medium (d ≈ 0.3-0.5). Plan for N ≥ 100 per condition (Lee & Chung, 2024).

6. **Administering multiple objects sequentially without counterbalancing**: Practice effects and fatigue can confound results. Counterbalance object order across participants (Reiter-Palmon et al., 2019).

## Minimum Reporting Checklist

Based on Lee & Chung (2024) and Reiter-Palmon et al. (2019):

- [ ] Object(s) used and rationale for selection
- [ ] Time limit per object
- [ ] Exact wording of instructions (verbatim or cited)
- [ ] Condition descriptions and assignment method (random, counterbalanced)
- [ ] Sample size per condition with power justification
- [ ] Platform and recruitment source (MTurk, Prolific, lab)
- [ ] Attention check and exclusion criteria with exclusion counts
- [ ] For AI conditions: AI model and version, access method, interaction logging
- [ ] Scoring method used (fluency, flexibility, originality, semantic distance) — see `divergent-thinking-scoring` skill
- [ ] Inter-rater reliability for subjective scores (ICC or Cohen's kappa)
- [ ] Pre-registration status and link

## References

- Carson, S. H., Peterson, J. B., & Higgins, D. M. (2005). Reliability, validity, and factor structure of the Creative Achievement Questionnaire. *Creativity Research Journal*, 17(1), 37-50.
- Guilford, J. P. (1967). *The nature of human intelligence*. McGraw-Hill.
- Lee, B. C., & Chung, J. (2024). An empirical investigation of the impact of ChatGPT on creativity. *Nature Human Behaviour*. https://doi.org/10.1038/s41562-024-01953-1
- Mednick, S. A. (1962). The associative basis of the creative process. *Psychological Review*, 69(3), 220-232.
- Nusbaum, E. C., Silvia, P. J., & Beaty, R. E. (2014). Ready, set, create: What instructing people to "be creative" reveals about the meaning and mechanisms of divergent thinking. *Psychology of Aesthetics, Creativity, and the Arts*, 8(4), 423-432.
- Oppenheimer, D. M., Meyvis, T., & Davidenko, N. (2009). Instructional manipulation checks. *Journal of Experimental Social Psychology*, 45(4), 867-872.
- Reiter-Palmon, R., Forthmann, B., & Barbot, B. (2019). Scoring divergent thinking tests: A review and systematic framework. *Psychology of Aesthetics, Creativity, and the Arts*, 13(2), 144-152.
- Silvia, P. J., Winterstein, B. P., Willse, J. T., et al. (2008). Assessing creativity with divergent thinking tasks: Exploring the reliability and validity of new subjective scoring methods. *Psychology of Aesthetics, Creativity, and the Arts*, 2(2), 68-85.
- Tierney, P., & Farmer, S. M. (2002). Creative self-efficacy: Its potential antecedents and relationship to creative performance. *Academy of Management Journal*, 45(6), 1137-1148.
- Wallach, M. A., & Kogan, N. (1965). *Modes of thinking in young children*. Holt, Rinehart and Winston.

See `references/` for detailed instruction templates and object selection guide.