Generative AI use by blind and low vision software professionals in the workplace
Data files
Nov 20, 2025 version files 258.87 KB
-
Participant_Information.csv
2.98 KB
-
Participant_Information.pdf
59.72 KB
-
README.md
5.59 KB
-
Study_1_Interview_Protocol_Workarounds_and_GenAI.pdf
94.01 KB
-
Study_2_Interview_Protocol_GenAI_Utilization_in_the_Workplace.pdf
96.56 KB
Abstract
The software development workplace poses numerous technical and collaborative accessibility challenges for blind and low vision software professionals (BLVSPs). Though Generative AI (GenAI) is increasingly adopted within the software development industry and has been a rapidly growing topic of interest in research, to date, the unique perspectives of BLVSPs have yet to be consulted. We report on a qualitative study involving 39 semi-structured interviews with BLVSPs about what the introduction of GenAI has meant for their work. We found that BLVSPs used GenAI for many software development tasks, resulting in benefits such as increased productivity and accessibility. However, significant costs were also accompanied by GenAI use, as they were more vulnerable to hallucinations than their sighted colleagues. Sometimes, organizational policies prevented use. Based on our findings, we discuss the higher-risks and higher-returns that BLVSPs had to carefully weigh when deciding whether and when to use GenAI tools for work.
Dataset DOI: 10.5061/dryad.gmsbcc313
Overview
This dataset consists of four files relevant to the research conducted for the "Game Changer" or "Overenthusiastic Drunk Acquaintance"? Generative AI Use by Blind and Low Vision Software Professionals in the Workplace paper, accepted into the 48th International Conference on Software Engineering (ICSE).
The interview-based study consisted of semi-structured interviews with 39 Blind and Low Vision Software Professionals (BLVSPs) about their use of GenAI in the workplace.
The study's goal was to understand how BLVSPs used GenAI in the workplace and identify the benefits and challenges of AI use. Specifically, the study answered two research questions:
- RQ1: How do BLVSPs perceive and utilize generative AI in the workplace?
- RQ2: How do generative AI tools impact BLVSPs’ work and career prospects?
The dataset is provided to supplement the paper’s study design section and to aid other researchers interested in repeating or running a similar study.
The participant interview transcripts are not included as they contain sensitive and identifiable data.
The accompanying data consists of:
- Anonymized demographic data from the 39 participants involved in the study. Due to space constraints, these are ommitted from the related paper.
- Two interview protocols containing the questions used to interview participants. These allow other researchers to repeat the study or to conduct a similar study.
Study design
The Study Design section of the associated paper provides the context for the data files. It describes the recruitment and interview processes, data collection, data analysis, and ethical concerns. The data analysis section describes the generation and use of the provided codebook to support the overarching thematic analysis.
All participants in the study met the recruitment criteria. Namely, they (i) self-identified as being blind or having low vision, (ii) worked or had worked as a software professional (e.g., engineer, product manager, accessibility specialist), and (iii) had at least one year of professional work experience as a software professional.
This study extended a prior study that explored the use of custom tools by BLVSPs at work. Thirty participants took part in that first study. The initial interview protocol included questions about the use of GenAI, and it is this part of the dataset that informs this study. Subsequently, five participants from the prior study were re-interviewed for this study, along with a further nine participants recruited for this study. These two rounds of interviews led to thirty-nine participants involved in this study on the use of GenAI in the workplace by BLVSPs.
It is recommended to read the related paper to better understand the study's full context before using the provided data files.
Description of the data and file structure
The data provided is:
- The interview protocol that was used in the first set of interviews about custom tools, including some questions on the use of GenAI.
- The interview protocol used in additional GenAI-focused interviews.
- Anonymized participant demographic data (Participant_Information.pdf and as a machine readable file Participant_Information.csv)
The first interview protocol file (Study_1_Interview_Protocol_Workarounds_and_GenAI.pdf ) contains the questions used in the first interview. The GenAI related questions in this protocol are highlighted.
The second interview protocol (Study_2_Interview_Protocol_GenAI_Utilization_in_the_Workplace.pdf) has questions focusing on GenAI use at work.
The interview protocols are free to reuse and amend.
The data files included in this dataset are:
| Data File | Description |
|---|---|
| Participant_Information.pdf | Anonymized participant information |
| Participant_Information.csv | Same anonymized participant information, but in a machine-readable form |
| Study_1_Interview_Protocol_Workarounds_and_GenAI.pdf | Interview questions used in the first set of interviews |
| Study_2_Interview_Protocol_GenAI_Utilization_in_the_Workplace.pdf | Interview questions used in the second set of interviews |
Sharing/Access information
The dataset is licensed under a CC0 1.0 Universal (CC0 1.0) Public Domain Dedication license.
Human subjects data
The data contains a table of de-identified demographic information about the interview participants. Each participant was provided with a unique participation identifier to be referred to in research papers and in the included data file. No personal identifiable information is shared. The study followed the University of California, Irvine's (UCI) Human Research Protection protocols and was approved by UCI's Institutional Review Board (IRB). All participants provided informed consent to participate in the study.
Data Preparation
The researchers developed the interview protocols before the semi-structured interviews. The Study_1 protocol was used to interview 30 participants in an earlier study that subsequently formed part of this one. The Study_2 protocol was used to interview nine additional participants and to re-interview some of the previous 30 participants.
Data Collection
The demographic data in this dataset were collected during the study's recruitment process.
The accompanying research paper provides further details on the research method, including participant recruitment.