Abstract
This study investigates how educators can utilise generative AI tools like DALL-E to create representative, inclusive imagery for educational settings; offering an
alternative to image search methods like Google Image and clipart libraries. Through image elicitation interviews with eight educators on a postgraduate course, the research examines the impact of prompt specificity, questioning the potential balance between detailed and minimal prompts in achieving authentic representation. The findings reveal that while generative AI can produce representative imagery, biases persist, often requiring an iterative approach to refine prompts for accurate depiction. Interestingly, prompt specificity was not universally seen as a critical requirement, and some perceived successful results emerged from brief open-ended prompts. An iterative decision-making framework is recommended to guide educators in developing inclusive images effectively, enhancing engagement and diversity in instructional resources. The study underscores generative AI's potential, limitations, and needs for refinement, offering strategies for educators to leverage these tools to create more inclusive and representative visuals for their educational presentations
and materials.
alternative to image search methods like Google Image and clipart libraries. Through image elicitation interviews with eight educators on a postgraduate course, the research examines the impact of prompt specificity, questioning the potential balance between detailed and minimal prompts in achieving authentic representation. The findings reveal that while generative AI can produce representative imagery, biases persist, often requiring an iterative approach to refine prompts for accurate depiction. Interestingly, prompt specificity was not universally seen as a critical requirement, and some perceived successful results emerged from brief open-ended prompts. An iterative decision-making framework is recommended to guide educators in developing inclusive images effectively, enhancing engagement and diversity in instructional resources. The study underscores generative AI's potential, limitations, and needs for refinement, offering strategies for educators to leverage these tools to create more inclusive and representative visuals for their educational presentations
and materials.
| Original language | English |
|---|---|
| Pages (from-to) | 27-40 |
| Journal | Education in Practice |
| Volume | 6 |
| Issue number | 1 |
| Publication status | Published - 25 Feb 2025 |