Experimental Results on LLM Impacts on Work Performance and Responsibility

A 2026 study by Felix Kares examines the impact of large language model use on task performance and work-related experiences. The dataset contains PNG files summarizing results from two studies involving 40 software developers and 18 psychology graduate students. It was published on figshare under a CC-BY-4.0 license.

Use Cases

• Analyzing the relationship between prompt engineering advice and programming task performance based on the described experimental conditions.
• Studying longitudinal changes in perceived control and responsibility during AI-assisted work based on the 13-week study design.
• Comparing the effects of LLM assistance on writing versus data analysis performance based on the described task outcomes.
• Investigating the trade-off between task efficiency and cognitive engagement mentioned in the description.

Strengths

• Includes results from two distinct studies with specific participant counts (N=40 and N=18).
• Employs a longitudinal design over 13 weeks for one of the studies.
• Published under a permissive CC-BY-4.0 license for reuse.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small (318.8 KB), indicating limited scope, likely containing only summary figures.

Provenance

Source

figshare

Collection Method

Experimental studies involving task completion with and without ChatGPT.

Time Range

2026

Freshness

Last updated 2026-05-06 06:47:38; freshness should be verified.

Geography

null