Causal strength induction from time series data.

Area of Science:

• Cognitive Psychology
• Causal Inference
• Time Series Analysis

Background:

• Inferring cause-effect relations from time series data is challenging due to temporal trends.
• Unaccounted temporal trends can lead to incorrect causal judgments, including false positives or reversed causal directions.

Purpose of the Study:

• To investigate how people infer causal strength from time series data.
• To determine if individuals use a heuristic focusing on transitions (changes) to control for temporal trends.

Main Methods:

• Six experiments were conducted using time series data.
• Participants' judgments of causal strength were analyzed based on presented data formats (numerical vs. visual) and potential biases (primacy/recency).
• Focus was placed on whether participants utilized state information versus transition information.

Main Results:

• Participants utilize transitions (changes in variables) in addition to states for causal judgments, improving accuracy.
• The use of transitions is more pronounced when data is presented visually compared to numerically.
• Causal strength estimation relies more on the direction of change than the magnitude of change.

Conclusions:

• People employ a heuristic focusing on transitions to effectively infer causal strength from time series data.
• This transition-based heuristic aids in controlling for temporal trends, leading to more accurate causal inferences.
• Understanding this cognitive strategy offers insights into human causal reasoning with dynamic data.