Abstract:

Artificial intelligence (AI) can in its best enhance learning. Challenges of AI are e.g. ethics, authorship, agency, transparency, privacy, and digital literacy of students and teachers. New AI applications are developed continuously. The University of Helsinki, Finland, has outlined that AI is an important work-life skill for students. With AI we mean here generative AI, large language models, such as ChatGPT and similar applications. The aim of this study is to examine gender differences in the AI readiness of agricultural students and their wishes for AI teaching.

First-year agriculture students (N=59) at the University of Helsinki, Finland, were examined in 2023. The share of females (f) was 47.5% (N=28) and that of males (m) 52.5 % (N=31). The AI readiness scale, modified from an earlier study, included 18 items under four themes: cognition (5 items), ability (7, in the final analysis 6 items), vision (3 items), and ethics (3 items). All items except the excluded one started with the words “I can…” indicating confidence on the statements. A five-point Likert scale was used in answering (1 = strongly disagree … 5 = strongly agree). In our earlier manuscript the 17-item scale was observed to be valid for our student group. After the Likert statements, an open question asked for wishes for the Degree Programme concerning AI teaching.

In the basic statistical parameters differences between genders were observed, although they were not statistically significant. Mean of the ethics theme was the greatest for both genders (3.51 f and 3.44 m), followed by vision (3.01 f and 3.30 m), ability (2.40 f and 2.75 m), and cognition (2.74 f and 2.97 m). The shares of strongly disagree and disagree responses together were greater for females than for males for all other items than those of ethics, in which the difference was the opposite. Furthermore, the shares of strongly agree and agree responses together were greater for males than for females, except again in the ethics theme in which the gender difference was the opposite. Majority of both female and male students had a positive attitude towards the use of AI in studying and research (the item left out from the ability theme: 3.50 f and 3.55 m, respectively). However, more female students (32.1%, N=9) than male students (16.1%, N=5) wished AI teaching.

As has been reported earlier about our student group as whole, the confidence of both female and male students was greater on the vision and ethics than on the cognition and ability statements. We cannot know if the approximate gender differences observed were more about real differences in e.g. skills or in self-efficacy in using AI. In earlier studies outside of AI, results concerning students’ gender in higher education have been varying, and gender has often been only a background variable without comparisons. As an example, it has been observed in other studies that female students sought more help than male students, which is a similar signal than in our study. In earlier studies it has been stated also that it is important to support and encourage each gender’s confidence in their own study strategies, and sometimes particularly female students. Concerning teaching of AI for the Finnish agricultural students, gender may not appear as a strongly significant practical factor, but supporting and encouragement of all students sounds relevant.

Keywords:

Artificial intelligence, large language models, higher education, university, agriculture.