Research Intro
Every day, we make choices under uncertainty — for example, deciding whether to use a navigation app or a paper map. Beyond the decision itself, we often ask: “How confident am I in this choice?”
This project explores the cognitive and metacognitive processes behind such decisions, focusing on task switching — an unexplored area. Specifically, this research examined how switching between tasks shapes confidence judgement in problem-solving and whether this effect varies across cues (visual signals) embedded in each task. The study highlights how task design and switching work together with cues to shape self-assessment and metacognitive performance.
Research Focus
Differences in Metacognitive Judgments → Decision-making and reasoning under varying cognitive load during task switching.
Task-Switching Effects → Task-switching conditions affect performance and confidence ratings. Addresses a largely unexplored area of metacognition.
Core Domains
-
Metacognition: Monitoring performance and adjusting strategies based on confidence.
-
Human Factors & Cognitive Load: Mental effort required for task switching.
-
Information Architecture: Visual cues (shape, color, congruency) guiding reasoning.
Goal
Impact of task switching on performance (success rates) and confidence, and how visual cues shape self-evaluation under changing cognitive demands.
Theoretical Recap
Task Switching & Cognitive Load
Metacognition Framework
-
Task switching = shifting between different tasks → reconfiguration of cognitive resources.
-
Switch cost = slower responses and increased errors.
⚙️ Task switching was manipulated using:
-
Separated tasks: no switching = AAA, BBB
-
Predictable = AA, BB, AA, BB
-
Unpredictable: Random sequences
Metacognition involves monitoring one’s own performance, effort and strategy > I measured metacognitive monitoring is measured via confidence judgments in decision making.
Main Hypohesis:
Task switching – which is associated with grater cognitive load and mental effort – will affect metacognitive monitoring, as well as cognitive performance.
Experimental Design Overview
Pilot Study
Experiment 1
N=192
Experiment 2
N=228
Two experoments | 3 task-switching conditions | 2 reasoning tasks
Analysis
-
Data were analyzed using R
-
Analysis method: Hierarchical multi level regression models
Predictors Variables: Heuristic Cues
⌛ Response Time
🔲 Task Based Cues:
-
Perimeter–Area Congruency: Perimeter–Area Size Comparison
-
Difference in Edges: Edge Count vs. Size
Switching Conditions
-
Separated tasks
-
Predictable switching
-
Unpredictable switching - associated with the highest cognitive load
Experiment 2
-
Color manipulation: tasks in same/different color (when different = more challenging)
-
Similar tasks (both shapes) = switching more challenging
Key Findings
📉 Experiment #1
Perimeter-Area Congruency 🔲
(Task's Global Cue)
-
Success rates higher for congruent shapes (solid green);
-
Confidence unaffected (Purple bars - same in all conditions).
-
Unpredictable switching reduced success for incongruent shapes → switch cost: higher cognitive load challenges global processing.
Response Time & Effort ⏳
-
Success (green dashed line) remained stable across all task-switching conditions, regardless of response time.
-
Confidence (solid purple line) declined as response time increased, but only under unpredictable switching.
💡 Confidence for items requiring more effort (longer time) = lower in the unpredictable switching condition
💡Upredictable switching + more difficult cues (incongruent shapes) + global processing = more challenge.
Cognitive Load & Heuristic Cues
→ Inform strategies to enhance self-regulation & decision-making; heuristics under high cognitive load shape performance assessment.
Task Specification & Design
→ Managing task complexity and minimizing unnecessary cue overlap could reduce cognitive load and improve task performance.
Metacognitive Strategies
→ Awareness of task switching costs can help individuals allocate mental resources more effectively, especially for complex tasks / cues.
Main Conclusion & Implications
Confidence is blind across cognitive load conditions
-
Participants’ confidence remained similar across task-switching conditions, even though performance varied.
-
Harder tasks led to lower accuracy, but users could not reliably judge task difficulty.
-
Gap between perceived and actual performance
-
→ Highlights limits in metacognitive awareness.
-
→ Implications for decision-making under cognitive load.
-
These findings highlight how the interplay between cognitive load and heuristic cues can inform strategies for enhancing self-regulation, improving performance, and identifying weak points in complex reasoning tasks.