B Ma: Researcher — Prolific Citations Library

Explore 3 peer-reviewed studies by B Ma in Artificial Intelligence and AI Ethics (2024–2026). Discover research powered by Prolific's participant panel.

This page lists 3 peer-reviewed papers authored or co-authored by B Ma in the Prolific Citations Library, a curated collection of research powered by high-quality human data from Prolific.

Papers (3 of 3)

Moral Lenses, Political Coordinates: Towards Ideological Positioning of Morally Conditioned LLMs

Authors: C Yuan, B Ma, Z Zhang, B Prenkaj, F Kreuter, G Kasneci

Year: 2026

Published in: arXiv preprint arXiv:2601.08634, 2026•arxiv.org

Institution: Munich Center for Machine Learning, LMU Munich, Technical University of Munich

Research Area: Artificial Intelligence, AI Ethics, AI Alignment, Political Science, Computational Social Science

Discipline: Computer Science, Natural Language Processing (NLP)

This paper examines how large language models’ (LLMs) political outputs shift when you explicitly prime them with different moral values. Instead of just assigning fake personas (like “pretend to be liberal”), the authors condition models to endorse or reject specific moral values (e.g., utilitarianism, fairness, authority). They then measure how those moral primes move the models’ positions in...

DOI: https://doi.org/10.48550/arXiv.2601.08634
Eyewitness Misinformation Susceptibility Across Data Collection Contexts: Comparing Laboratory, Online, and Prolific Participant Responses

Authors: M Brassil, É Duncan, C Greene, B Mac Síthigh

Year: 2025

Published in: 2025 - osf.io

Institution: University College Dublin

Research Area: Eyewitness Memory, Misinformation Effect, Behavioral Research Methods, Online Data Collection Platforms

Discipline: Psychology

The study found that data collection contexts significantly influence susceptibility to eyewitness misinformation, with Prolific participants being less accurate and more susceptible compared to Laboratory or general online participants.

Methods: Two studies were conducted comparing eyewitness misinformation susceptibility across Laboratory, Prolific, and General Online participant groups under varying visual perceptual load conditions.

Key Findings: Eyewitness misinformation susceptibility and recall accuracy across Laboratory, Prolific, and General Online participant groups; the effect of visual perceptual load on recall accuracy.

Citations: 1
As Generative Models Improve, People Adapt Their Prompts

Authors: E Jahani, B Manning, J Zhang, H TuYe, M Alsobay, C Nicolaides, S Suri, D Holtz

Year: 2024

Published in: ArXiv

Institution: Massachusetts Institute of Technology, Microsoft Research, Stanford University, University of California Berkeley, University of Cyprus, University of Maryland

Research Area: Human-AI Interaction, Generative AI, Prompt Engineering

Discipline: Artificial Intelligence, focusing on Human-AI Interaction, Generative AI