Evidano home
Master’s thesis published by East Tennessee State UniversityResearch SpotlightUniversity of Central Florida

Research Spotlight: How College Students Weigh AI’s Creative Promise and Ethical Risks

Evidano9 min read
Georgina Amankwa, Doctoral Student and Graduate Teaching Associate in the Nicholson School of Communication and Media at the University of Central Florida

A student editing a video late at night can now ask an AI tool to suggest a caption, smooth a transition, reorganize a script, or generate a visual concept before the next export finishes. The speed can feel transformative, yet the same workflow immediately raises harder questions about originality, ownership, trust, and what should remain distinctly human.

Those tensions sit at the center of Georgina Amankwa’s 2025 master’s thesis at East Tennessee State University, which examines how college students perceive AI in multimedia content creation. Rather than treating students as either enthusiastic adopters or wary skeptics, the study shows a more layered picture in which perceived usefulness can coexist with limited routine use and substantial ethical concern.

For this research spotlight, we invited Georgina Amankwa, who completed the study at East Tennessee State University, to share more about how students balance AI’s creative benefits with concerns about authenticity, intellectual property, and human creativity.

About the article

Research question

The study asks how college students perceive AI’s effectiveness in improving the quality and efficiency of multimedia content creation, what benefits they associate with AI, and how they understand its ethical implications. It focuses on the relationship between practical usefulness and concerns about creativity, authorship, fairness, privacy, and trust.

Key finding

Students generally recognized AI as useful for saving time, improving workflows, supporting content quality, and strengthening audience engagement, but consistent adoption remained uneven. The thesis also found persistent concerns about authenticity, intellectual property, plagiarism, job displacement, and the possibility that excessive reliance on AI could weaken human creativity.

Why this research matters

AI is increasingly woven into everyday creative work, from writing and editing to images, video, audio, and marketing content. What matters is not only whether these tools can perform a task, but whether people trust the output, understand the tradeoffs, and feel that their own creative agency remains intact.

College students are a useful group for examining that tension because many are active digital content creators and are encountering AI at the same time that academic, professional, and cultural norms are still developing. Amankwa’s study suggests that awareness alone does not translate into habitual use, which is important for educators, researchers, marketers, and technology developers who may otherwise assume that familiarity with AI means uncritical acceptance.

Amankwa’s work also invites a wider conversation about why student perspectives matter at this moment. Young adults are among the most active users of social platforms: Pew Research Center’s 2025 survey found that roughly half of U.S. adults ages 18 to 29 use TikTok at least once a day, and 93 percent of the students in Amankwa’s analyzed sample were between 18 and 24. Amankwa reflected on that significance: “College students are an especially important group because they sit at the center of today’s social media and content-creation culture. Young adults, from age 18 upward, are among the heaviest users of most platforms, and TikTok is a clear example. Because Gen Z has grown up with these platforms, they adapt quickly to fast-paced AI technologies and use them to support their content. Many college students also rely on AI to develop ideas and to support their writing and media creation, which makes their perspectives an early signal of where creative practice is heading.”

Key findings from the article

Across the quantitative and qualitative results, three contributions stand out. Together, they show that students can see clear practical value in AI while still placing conditions on how it should be used.

  • Use was uneven even though awareness was universal within the analyzed sample. Only 3 percent reported daily AI use for multimedia content, while 38 percent said they rarely used it. More frequent users tended to view AI more favorably, and the thesis reports a statistically significant relationship between usage frequency and perceived effectiveness.
  • Efficiency and creative support were central benefits. In the regression model, time saving was the strongest predictor of willingness to recommend AI, followed by perceived quality improvement and audience engagement. Open ended responses described AI as useful for idea generation, captions, script development, editing, visual assets, and routine organization, and one participant wrote, “AI helps me craft compelling captions when I have writer’s block.”
  • Ethical reservations remained substantial. Participants raised concerns about authenticity, originality, copyright, plagiarism, and possible job displacement, and the thematic analysis emphasizes that trust can limit fuller adoption. The thesis therefore frames AI as most useful when it supports rather than displaces human creative judgment.

The findings are notable precisely because they resist a simple story of enthusiasm or rejection. Amankwa reflected on the result that stood out most: “What surprised me most was how uneven adoption turned out to be. Every respondent was aware of AI, yet only 3 percent used it daily and 38 percent used it rarely. That gap between awareness and active adoption stood out because it shows that familiarity alone does not lead to routine use. Barriers such as accessibility, technical proficiency, and trust appear to matter as much as the tools’ capabilities.”

Methods and research approach

The study used a mixed methods survey design with students recruited through the East Tennessee State University SONA panel, primarily from psychology classes. A total of 190 responses were received, and 124 responses met the screening criteria for analysis. The questionnaire included multiple choice items, Likert scale measures, and open-ended questions about AI use, perceived effectiveness, benefits, ethical concerns, and future expectations.

The quantitative component used descriptive statistics and inferential analyses, including a chi square test, correlation analysis, and regression. The qualitative component used thematic analysis based on the Braun and Clarke framework to identify recurring patterns in open ended responses. The study also discusses credibility, transferability, dependability, confirmability, researcher reflexivity, supervisor feedback, and institutional ethics approval as part of its approach to rigor.

How Evidano supported the research

Evidano, then known as AILYZE, was used to organize and code open ended responses, and the data analysis section explains that AILYZE assisted in organizing themes and analyzing qualitative data. This work sat within a broader mixed methods process that also used an online survey platform for survey collection and SPSS for quantitative analysis.

The thesis also describes privacy and research ethics as factors in selecting AILYZE. It states that the system was chosen for a contained analysis environment, that submitted data were not used to further train the model, and that research data remained encrypted and inaccessible to external entities. Those statements are presented here as the researcher’s description of the workflow and safeguards used in the study.

AILYZE supported organization and analysis, but it did not replace scholarly judgment. Amankwa defined the research questions, selected the mixed methods design, interpreted the patterns, integrated quantitative and qualitative evidence, reflected on possible bias, and developed the final conclusions and recommendations.

This distinction between assistance and interpretation is especially important when AI enters a research workflow. Amankwa also reflected on responsible use: “I used AILYZE to organize and code the open-ended responses and to develop themes, and it also helped summarize emerging findings on how AI is approached in multimedia content creation. I chose it because it works in a closed system: participants’ data were not used to train the model and remained encrypted. Even so, the tool supported my judgment and did not replace it. I defined the research questions, interpreted the patterns, and reflected on my own possible bias. Researchers should choose tools with strong data protection, be transparent about how AI was used, and keep human interpretation at the center of the work.”

Broader implications

For researchers and educators, the study suggests that AI literacy should extend beyond learning how to operate a tool. Students also need opportunities to examine authorship, source attribution, privacy, bias, originality, and the limits of automated output. Training that combines practical skill with ethical reflection may help explain why people who understand the benefits of AI can still hesitate to use it consistently.

For content creators, marketers, researchers, and AI developers, the practical challenge is to turn that balance into everyday practice through transparency, clear ownership norms, careful oversight, and realistic expectations. Amankwa offered a takeaway for people navigating that challenge: “My hope is that researchers, educators, and content creators will aim for balance: using AI to support learning and creative work without letting it replace human originality, and pairing that use with transparency, AI literacy, and clear ethical guidance.”

The study’s limitations also point toward the next research questions. The analyzed sample was relatively small and came from one university, and the qualitative data came from open ended survey comments rather than interviews or focus groups. Cross institutional, cross cultural, and longitudinal research could test whether the same balance of enthusiasm and concern appears across different student populations and as AI tools continue to change.

Amankwa’s thesis ultimately argues for a balanced model of creative work in which AI can reduce routine effort and expand possibilities without becoming a substitute for human originality, context, and responsibility. That framing is useful well beyond multimedia production because it centers the question that many fields now face, not simply what AI can do, but how people can use it while preserving trust and meaningful human judgment.

About Georgina Amankwa

Georgina Amankwa, Doctoral Student and Graduate Teaching Associate in the Nicholson School of Communication and Media at the University of Central Florida

Georgina Amankwa completed an M.A. in Brand & Media Strategy at East Tennessee State University in 2025, where her thesis examined college students’ perceptions of AI in multimedia content creation. Her academic work connects media, marketing, technology adoption, and ethical questions about authenticity, intellectual property, and human creativity. During her time at ETSU, she served as a graduate assistant and presented this research at the 2025 Dr. Jay W. Boland Research Day. Before joining ETSU, she built a career in Ghana spanning marketing, administration, and human resources. She also holds an M.Sc. in Strategic Management and Leadership from Kwame Nkrumah University of Science and Technology and a B.Sc. in Marketing from the University of Professional Studies, Accra. Her thesis contributes an empirical student perspective to broader discussions about how AI can support creative work while preserving human judgment and responsible practice. Georgina is now a Ph.D. student at the University of Central Florida, where her research interests center on strategic communication and artificial intelligence.

About the authors

  • Georgina Amankwa (Department of Media and Communication, College of Business and Technology, East Tennessee State University, USA)

Read the article

Read the full research thesis, “Leveraging AI for Enhanced Multimedia Content Creation: College Students’ Perception on the Innovation,” published in East Tennessee State University Electronic Theses and Dissertations in 2025.

View the publication

We are grateful to Georgina Amankwa for her contribution to research on responsible AI in multimedia content creation and for referencing Evidano, then known as AILYZE, as part of the research process.

Other published studies and evaluations that used Evidano on comparable material.