[Paper Review] Gender, Age, and Technology Education Influence the Adoption and Appropriation of LLMs

Fiona Draxler, Daniel Buschek+5 more|arXiv (Cornell University)|Oct 10, 2023

Artificial Intelligence in Healthcare and Education13 citations

TL;DR

The study analyzes how gender, age, and technology education shape LLM adoption in a representative US sample of 1,495 adults, showing a gender gap, an age effect, and the leveling role of technology education, with insights into usage scenarios and adoption barriers.

ABSTRACT

Large Language Models (LLMs) such as ChatGPT have become increasingly integrated into critical activities of daily life, raising concerns about equitable access and utilization across diverse demographics. This study investigates the usage of LLMs among 1,500 representative US citizens. Remarkably, 42% of participants reported utilizing an LLM. Our findings reveal a gender gap in LLM technology adoption (more male users than female users) with complex interaction patterns regarding age. Technology-related education eliminates the gender gap in our sample. Moreover, expert users are more likely than novices to list professional tasks as typical application scenarios, suggesting discrepancies in effective usage at the workplace. These results underscore the importance of providing education in artificial intelligence in our technology-driven society to promote equitable access to and benefits from LLMs. We urge for both international replication beyond the US and longitudinal observation of adoption.

Motivation & Objective

• Assess whether gender and age predict LLM usage in a representative US sample.
• Examine how interaction effects between gender and age influence adoption.
• Explore whether technology education moderates gender gaps in LLM adoption.
• Identify common usage scenarios for novices versus expert users and barriers to adoption.

Proposed method

• Online survey with socio-demographic and LLM usage questions.
• Bayesian regression (logistic models) to predict LLM usage and frequency.
• Inductive coding and topic modeling for open-ended usage scenarios and non-use reasons.

Experimental results

Research questions

• RQ1 Does gender predict LLM usage, and is there an interaction with age?
• RQ2 Does younger age predict higher LLM adoption, and how does age interact with gender in adoption patterns?
• RQ3 Does having education in technology (especially college-level tech degrees) level the gender gap in LLM usage?
• RQ4 What are the typical usage scenarios for novices versus expert users?
• RQ5 What reasons do non-users give for not using LLMs?

Key findings

• 41.5% of participants reported having used an LLM.
• There is a gender gap with female users less likely to have used LLMs than male users (34.3% vs 49.1%).
• Usage likelihood decreases with age; younger groups show higher adoption, with some midlife groups using more than the youngest group in frequency.
• Among college graduates, the gender gap is mitigated for technology-related fields (almost equal usage by gender when studying technology).
• Experts are more likely to use LLMs for professional tasks (formal texts and coding) than novices, who are more exploratory.
• Non-users cite lack of knowledge, perceived need, ethical concerns, and preference for human writing as barriers.