Session: Integrating Chatgpt into Social Work Research: A Workshop on Prompt Engineering, API Integration, and Recommendations for Ethical Use (Society for Social Work and Research 28th Annual Conference - Recentering & Democratizing Knowledge: The Next 30 Years of Social Work Science)

All in-person and virtual presentations are in Eastern Standard Time Zone (EST).

SSWR 2024 Poster Gallery: as a registered in-person and virtual attendee, you have access to the virtual Poster Gallery which includes only the posters that elected to present virtually. The rest of the posters are presented in-person in the Poster/Exhibit Hall located in Marquis BR Salon 6, ML 2. The access to the Poster Gallery will be available via the virtual conference platform the week of January 11. You will receive an email with instructions how to access the virtual conference platform.

64 Integrating Chatgpt into Social Work Research: A Workshop on Prompt Engineering, API Integration, and Recommendations for Ethical Use

Schedule:
Friday, January 12, 2024: 8:00 AM-9:30 AM
Congress, ML 4 (Marriott Marquis Washington DC)
Cluster:
Organizer:
Bryan Victor, PhD, Wayne State University
Speakers/Presenters:
Bryan Victor, PhD, Wayne State University, Brian Perron, PhD, University of Michigan-Ann Arbor and Rebeccah Sokol, PhD, University of Michigan - Ann Arbor
Social work researchers are increasingly exploring the use of natural language processing (NLP) tools to enhance their research projects. One promising tool is ChatGPT, a state-of-the-art NLP model trained by OpenAI. This workshop will provide an overview of ChatGPT, its use in scientific inquiry broadly, and its relevance to social work research. Participants will learn about the benefits and limitations of ChatGPT and its potential applications to their own initiatives.

A primary focus of the workshop will be on prompt engineering, the process of designing effective prompts to elicit desired responses from ChatGPT. Participants will learn how to create prompts that are tailored to their research and evaluation needs. In particular, we will demonstrate how to use prompt engineering for cleaning data, writing analytic code, supporting qualitative research, and improving clarity of research and dissemination materials. With respect to data cleaning, participants will learn how to create prompts that can correct data entry errors or inconsistencies. In the analytic code context, participants will learn how to quickly generate code across statistical software packages, thereby increasing efficiency in workflow. For qualitative research support, participants will learn how to use prompts to support the analysis of textual data, prompting ChatGPT to classify (i.e., code) data based on specific instructions and conceptual descriptions. Participants will also learn how to create prompts to improve the clarity of research and dissemination materials, including increasing the readability of consent forms and questionnaires, generating plain-language summaries of research results for the general public, and using ChatGPT as a copy-editor for academic writing.

Secondly, we will show participants how to use ChatGPT’s application programming interface (API) to enhance the utility of the model. By leveraging the API, ChatGPT can access and process a wealth of data from various sources such as databases, online platforms, or social media, providing researchers with tailored insights and analyses. This integration not only streamlines research processes but also enables social work researchers to conduct large-scale studies, identify patterns and trends, and generate evidence-based recommendations for policy and practice.

In addition to exploring the technical dimensions of ChatGPT, the workshop will also emphasize the ethical considerations and best practices for using ChatGPT in social work research. We will offer guidelines around transparency in reporting, ensuring that the use of AI-driven methods and/or content generation is clearly disclosed and detailed in scientific communications. We will delve into the potential biases of ChatGPT, as well as strategies for mitigating their impact on research outcomes. We will also examine privacy and data protection concerns.

By fostering a comprehensive understanding of these ethical issues alongside its technical capabilities, this workshop aims to equip social work researchers with the knowledge and skills necessary to responsibly and effectively incorporate ChatGPT into their research practices.

Note: Much of the content for this abstract was generated by ChatGPT with editing from the presenters.

See more of: Workshops