Interactions with generative AI chatbots: unveiling dialogic dynamics, students’ perceptions, and practical competencies in creative problem-solving

The assigned CPS task was the creation of an innovative research proposal. We found that there were significant differences in the dialogic exchanges observed between the two types of interaction. Student-GAI chatbot interactions featured more knowledge-based dialogue and elaborate discussions, with less subjective expression compared to student-peer interactions. Notably, students contributed significantly less dialogue when interacting with a GAI chatbot than they did during peer interactions. The dialogic exchanges arising from student-GAI chatbot interactions tended to follow distinct patterns, while those from student-peer interactions were less predictable. The students perceived interacting with a GAI chatbot as more useful and easier than interacting with peers. Furthermore, they exhibited higher intention levels when utilising a GAI chatbot to tackle the CPS task compared to engaging in discussions with their peers. Ultimately, practical performance was significantly enhanced through interactions with a GAI chatbot. This study implies that the prudent use of GAI-based techniques can facilitate university students’ learning achievement.

Researchers hold different perspectives regarding the key components of CPS. For example, Parnes (1981) proposed a five-stage creative problem-solving model. This includes: discovering facts, discovering problems, seeking ideas, seeking solutions, and seeking acceptance. Stanish and Fberle (1997) developed a six-stage problem-solving model comprising: identifying confusion, searching for information, identifying the problem, collecting ideas, searching for countermeasures, and accepting the ideas. Furthermore, Isaksen et al. (2000) proposed an eight-stage creative problem-solving model: searching for opportunities, searching for information, identifying problems, stimulating ideas, developing solutions, seeking acceptance, evaluating the process, and designing the process. Lin and Cho (2011) explored a dynamic system model for evaluating CPS in mathematics, involving six properties that are used as predictors. Xia (2021) identified three important steps in CPS: problem identification, problem definition, and problem-solving

Productive dialogue, including with GAI chatbots, consistently exhibits one or several of the following characteristics:

• Capturing key ideas, formulas, and techniques, and integrating them with existing knowledge (e.g., Nystrand et al., 2003; Song, 2024);

• Free expression of ideas, leading to the development of subjective perspectives (e.g., Hennessy et al., 2021; Howe et al., 2019);

• Elaboration, explanation, and clarification that make a concept or an idea more explicit, thorough, and comprehensive (e.g., Alexander, 2017; Tai & Chen, 2023);

• Seeking coordination that pursues integrated lines of inquiry and connection between contributions (e.g., Howe et al., 2019; Wilkinson et al., 2023);

• Recognising and evaluating varied viewpoints, building on and clarification of one’s own or another’s contribution (e.g., Alexander, 2017; Littleton & Mercer, 2013);

• Exploration beyond what is being learned, and making links to the broader context (e.g., Song et al., 2021; van der Veen et al., 2017).

The study addresses three research questions:

RQ1: What are the dialogic dynamics in student-GAI chatbot interactions in creative problem-solving? How do they differ from those observed in student-peer interactions?

RQ2: What are students’ perceptions of interacting with a GAI chatbot in creative problem-solving?

RQ3: How effective is student-GAI chatbot interaction in enhancing practical creative problem-solving competencies?

The CI-PCD consists of six core coding dimensions: prior knowledge, subjective expression, elaboration, coordination, speculation and construction. As can be seen from the literature review, dialogue enables students to grasp essential concepts, formulas, and methods, while integrating them with their prior knowledge. It should foster an environment where they can freely articulate their thoughts, producing subjective expression. A productive dialogic approach prompts students to defend their responses with reasoning, thus resulting in elaboration. Teachers should guide students in identifying connections and synthesising knowledge, which promotes coordination. Furthermore, engaging in comprehensive discussions that venture beyond the scope of learned material encourages speculation. Dialogue should be a collaborative and cumulative process where participants build upon each other’s ideas in a constructive manner, leading to construction. Detailed descriptions and examples of the coding dimensions are shown in Table 1. These elements of dialogue exchanges represent various levels of cognitive engagement. Nystrand et al. (2003) differentiates between dialogic contributions that demand ‘high’ versus ‘low’ levels of cognition.

the introduction of a GAI chatbot could enhance university students’ learning experiences. They perceived the chatbots as more useful and easier to use than peers, and they showed a higher intention to use the chatbot in the future in CPS.

All interviewees from the experimental group reported that chatbots were easy to use. For instance, one interviewee noted, ‘Using the chatbot was straightforward, without unnecessary complexity or pressure. I simply entered commands as I would in a regular conversation.’ Another interviewee echoed, ‘The chatbot’s ease of use was striking. The key is that instructions should be provided gradually, clearly, and specifically. The quality of the questions asked must be high to achieve satisfying results.’ Additionally, interacting with chatbots was reported to be beneficial for refining research questions, summarizing existing studies, suggesting research ideas, and proposing methodological strategies. For example, one interviewee shared, ‘The chatbot can suggest research questions based on current literature. When I inquired about the differences between concepts, it was able to discuss the nuances of each and compare them comprehensively.’ Furthermore, an interviewee mentioned, ‘The chatbot could even outline a research design exploring interpersonal relationships in graduate student dormitories from a spatial perspective. It quickly developed a proposal with research questions and methodology based on the limited information I provided.’ Moreover, the interviewees expressed their strong intention to continue using chatbots in the future. As one interviewee put it, ‘The chatbot was more flexible than other tools and capable of engaging in follow-up conversations. It could also provide personalized assistance, which makes me want to use it again.’ Another interviewee agreed, ‘The chatbot’s strength lies in its ability to provide a large amount of information in a short time, making it very efficient.