---
source_url: "https://arxiv.org/abs/2605.18140"
ingested_date: 2026-05-19
sha256: placeholder
---

# Faculty Orientations Shape Adoption of AI in Research and Teaching

**Authors:** Timothy J. Atherton, Ian Descamps, Tova R. Holmes, Christina L. Vizcarra, Ning Sui, Max Webel, Jay J. Foley IV
**arXiv:** 2605.18140 [physics.ed-ph, cs.CY]
**Date:** 18 May 2026
**License:** CC BY 4.0

---

## Abstract

Despite the widespread availability of large language models (LLMs) in higher education, instructors vary substantially in their adoption and use of these tools, and the reasons for this variation remain poorly understood. A mixed-methods survey of 90 STEM faculty in the Research Corporation for Science Advancement (RCSA) Cottrell community examined relationships between AI use, attitudes, institutional context, and instructional practice. Exploratory factor analysis identified a coherent construct, **AI pedagogical orientation**, that strongly predicted self-reported AI use across research, teaching, and other professional activities. Qualitative analysis indicated that this construct reflected differing views about the role AI should play in disciplinary thinking, learning, and expertise development, rather than simply positive or negative attitudes toward AI. Institutional initiatives, demographic variables, and information sources showed comparatively weak associations with AI use. The results suggest that existing technology-adoption models may not fully explain adoption in contexts where technologies interact directly with disciplinary reasoning and knowledge production.

---

## 1. Introduction & Background

- **Context:** LLMs became widely available in 2022, offering dialogic interaction, summarization, coding assistance, and agentic task delegation.
- **Study focus:** How pedagogically motivated STEM faculty (RCSA Cottrell awardees) use LLMs in research, teaching, and professional work.
- **Technology-adoption frameworks:** Diffusion of Innovations (DoI) and UTAUT emphasize external conditions and user perceptions but may not account for how instructors interpret AI's epistemic role in disciplinary practice.
- **Computation analogy:** Integration of computation into physics education faced similar challenges; faculty perspectives and communities (e.g., PICUP) were crucial.

---

## 2. Methods

- **Population:** 572 RCSA Cottrell awardees — junior faculty selected for research quality and innovative education plans.
- **Survey:** Mixed-methods, Fall 2025. Likert-scale items, multiple-choice, free-response.
- **Response:** n=116 (20%); n=90 retained after filtering (16% of population).
- **Analysis:** Descriptive statistics, Exploratory Factor Analysis (EFA) on 36 binary/ordinal variables, qualitative coding (Practice, Epistemic, Affective, Structural), joint correlation of factor scores with codes and demographics.

---

## 3. Results

### 3.1 Descriptive
- Most common AI use: coding assistants (24 research, 10 teaching, 5 other). 22% used no AI at all.
- Research-teaching usage only partially correlated.
- 64% reported institutional strategic initiatives; only 30% had degree programs/certificates.
- Department-level initiatives rare (10% degree, 8% strategic).
- Top information sources: department colleagues (64%), discipline-specific news, popular press — NOT institutional support.

### 3.2 Exploratory Factor Analysis
- One dominant factor emerged: **AI pedagogical orientation** (9 items, consistent across rotation methods).
- This factor strongly predicted AI use across research, teaching, and other activities.
- Demographic variables (discipline, institution type, career stage) showed much weaker associations.

### 3.3 Qualitative Findings
- Faculty held fundamentally different views about AI's role in:
  - **Disciplinary thinking:** Should students learn to think with AI, or without it?
  - **Learning:** Is AI a prosthetic for learning, or a replacement?
  - **Expertise development:** What does it mean to be an expert when AI can perform tasks once central to disciplinary mastery?
- These were not simply "positive" vs. "negative" but reflected coherent pedagogical philosophies.

---

## 4. Implications
- Standard technology-adoption models (DoI, UTAUT) are insufficient for AI in education — they don't capture epistemic interpretations.
- Institutional initiatives alone won't drive adoption; faculty pedagogical orientation is the primary driver.
- Faculty development should focus on helping instructors articulate and refine their pedagogical orientation toward AI.