Full Program »

Acceptance of ChatGPT Tool for Learning Programming in Higher Education (20 min)

Finding faster solutions to programming tasks, writing, and correcting program code, among others, is a challenge for many current and future programmers, but also for educators and other stakeholders tasked with assisting them in the process of learning programming. For this purpose, artificial intelligence tools are increasingly used, for instance, large language models (LLM) like ChatGPT, which can be viewed as an innovation in the educational process within the context of education. It's essential to underline that a quality-generated prompt is needed to produce code using large language models to obtain an appropriate programming solution, which is also related to the user's programming competency level. Otherwise, anyone could program. Understanding the factors of (non)acceptance of artificial intelligence tools for learning programming among students is especially significant for those aiming to implement this innovative technology in programming education and a genuine business programming environment.

This presentation will present the results of a survey on students' opinions (N=109) regarding using ChatGPT for learning programming at a higher education institution. The analysis of the collected data in this study showed that most students agree that the advantage of using ChatGPT for programming learning is its ability to support the user at any time (83.49%), quick access to information (80.73%), better productivity (70.64%), facilitating code understanding through simple examples (69.72%), and that it increases their interest in learning programming (54.13%). Furthermore, the majority of students in this study believe that ChatGPT can only provide an accurate answer if given the full context (65.14%) and that it encourages dependence on its use (57.80%). Interestingly, one-third of the respondents (34.86%) feel that using the ChatGPT tool induces negative feelings, while 16.51% were undecided. Most respondents (80.73%) expressed their intention to use ChatGPT during future programming learning, and 71.56% stated they intend to use it more frequently. The results of the path analysis using the PLS-SEM method show that the intention to use the ChatGPT tool for learning programming significantly affects the increase of negative emotions when thinking about using artificial intelligence tools, while the intention of higher usage of ChatGPT negatively affects the increase of negative emotions. In other words, one can conclude that the experience of working with this innovative tool is related to emotions towards it. In this context, the impact of using the tool on negative emotions changes depending on the time horizon of the intention to use. Based on all those mentioned above, it is necessary to consider ways of developing programming competencies in students using ChatGPT from both a technical and pedagogical perspective and to conduct further research in different environments. Traženje načina bržeg rješavanja programerskih zadataka, pisanja i ispravaka programskog koda i drugo, izazov je za mnoge sadašnje i buduće programere, ali i za obrazovne djelatnike i ostale dionike koji imaju zadaću pomoći istima u procesu učenja programiranja. U tu svrhu sve više se koriste alati umjetne inteligencije, primjerice veliki jezični modeli (eng. large language model - LLM) poput ChatGPT-a, koji se u kontekstu obrazovanja mogu promatrati kao inovacija u nastavom procesu. Važno je naglasiti da je za stvaranje koda putem velikih jezičnih modela potreban kvalitetno generairan upit (eng. prompt) kako bi se za rezultat dobilo adekvatno programersko rješenje, što je vezano i uz razinu programerskih kompetencija korisnika. U suprotnom, svatko bi mogao programirati. Razumijevanje činitelja (ne)prihvaćanja alata umjetne inteligencije za učenje programiranja kod studenata posebno je značajno za sve one koji imaju za cilj implementaciju ove inovativne tehnologije u nastavi programiranja i stvarnom poslovnom programerskom okruženju. U ovom izlaganju predstavit će se rezultati istraživanja mišljenja studenata (N=109) o upotrebi ChatGPT-a za učenje programiranja na jednoj visokoškolskoj instituciji. Rezultati analize prikupljenih podataka u ovom istraživanju pokazali su da najviše studenta se slaže da je prednost primjene ChatGPT-a za učenje programiranja njegova mogućnost pružanja potpore korisniku u bilo koje vrijeme (83,49%), brz pristup informacijama (80,73%), bolja produktivnost (70,64%), olakšavanje razumijevanja koda putem jednostavnih primjera (69,72%) te da povećava njihovu zainteresiranost za učenje programiranja (54,13%). Nadalje, većina studenata u ovom istraživanju smatra da ChatGPT nije u mogućnosti odgovoriti točno na postavljeno pitanje ukoliko mu se ne postavi cjelokupan kontekst (65,14%) te da potiče ovisnost o potrebi njegova korištenja (57,80%). Posebno je zanimljivo što jedna trećina ispitanika (34,86%) smatra da pomisao na korištenje alata Chat GPT u njima potiče negativne osjećaje, dok njih 16,51% je bilo neodlučno. Namjeru za korištenjem ChatGPT-ja tijekom učenja programiranja u budućnosti iskazala je velika većina ispitanika (80,73%), dok je njih 71,56% iskazalo da imaju namjeru isti više koristiti. Rezultati path analize PLS-SEM metodom pokazuju da namjera korištenja alata ChatGPT za učenje programiranja u signifikantno utječe na povećanje negativnih emocija pri pomisli na korištenje alata umjetne inteligencije, dok namjera višeg korištenja Chat GPT-a negativno utječe na povećanje negativnih emocija. Drugim riječima, može se zaključiti da se iskustvo u radu s ovim inovativnim alatom dovodi u vezu s emocijama prema istome. Pri tome, utjecaj korištenja alata na negativne emocije mijenja se s obzirom na vremenski horizont namjere korištenja.

Darko Etinger
Fakultet informatike Sveučilišta Jurja Dobrile u Puli

Darko Etinger is an Associate Professor and Dean of the Faculty of Informatics at Juraj Dobrila University of Pula. His research interests include Business Process Management (BPM), process mining, IT management, modeling, and simulation. He is leading the project of launching the first Croatian online university degree in Informatics at the Faculty of Informatics in Pula. His teaching activities include holding and conducting lectures in Business Process Management, Business Information Systems, Modeling and Simulation, IT Management, and Artificial Intelligence. He is a member of professional organizations such as IEEE, ACM, and HDOI (Croatian Operations Research Society).

Snježana Babić
Fakultet informatike Sveučilišta Jurja Dobrile u Puli

Snježana Babić is an Associate Professor at the Faculty of Informatics at Juraj Dobrila University of Pula. In 2006, she completed the E-learning Academy, focusing on E-learning Tutoring at CARNET. She has worked as a member and/or leader on several professional and scientific projects in e-learning. As a (co)author, she has written over 70 professional and scientific papers in journals and proceedings from international conferences in the field of innovation acceptance and new educational technologies, and human-machine interaction. She has mentored over 90 professional and graduate theses of informatics students at the polytechnic and university level. She is a member of IEEE, the Croatian Association for Information, Communication, and Electronic Technology (MIPRO), and one of the leaders of the Computers in Education section, the president of the Cyberneticians Society Rijeka. She received an award for her selfless effort and contribution to the technical culture of the City of Rijeka in 2014 in the field of promoting algorithmic thinking and early learning of programming with the support of the Mema method by the original author Prof. Dr. Marina Čičin-Šain. She is also a member of the ERASMUS+ project team "CybeMEMA".

Saša Mamula
Fakultet informatike Sveučilišta Jurja Dobrile u Puli

Saša Mamula, B.Sc. in Informatics, is currently a student in the final year of the graduate program at the Faculty of Informatics at Juraj Dobrila University of Pula. Concurrently with his education, he works as a programmer at the company IN2. The programming languages he works with are C#, C++, and in his free time, Java and Javascript when creating mobile or web applications.

Participants' prior knowledge / Potrebno predznanje: Osnovna razina znanja o učenju i poučavanju programiranja i bez predznanja.

Learning outcomes / Ishodi učenja: Očekuje se da će sudionici moći kritički prosuditi potrebu za polaznikovim učenjem programiranja podrškom ChatGPT-ija u nastavi informatike na svim obrazovnim razinama.