Project – The Effect of Generative AI-Assisted Learning on Students’ Critical Thinking and Independent Problem-Solving Skills. A Study of Undergraduate students at selected universities in Ogun State.

CHAPTER ONE

INTRODUCTION

1.1 Background to the Study

Higher education has experienced significant changes following the integration of digital technologies into teaching, learning, research, and assessment. Universities increasingly employ technology-supported instructional approaches to provide students with access to information, encourage collaborative learning, and facilitate the development of competencies required in contemporary workplaces. Within this changing educational environment, artificial intelligence (AI) has emerged as an important technological development with the potential to influence how students acquire, interpret, and apply knowledge. Generative artificial intelligence represents a particularly significant advancement because it can produce human-like text, explanations, summaries, examples, and responses to complex questions. Unlike conventional digital resources that primarily provide access to existing information, generative AI systems can interact with users through conversational exchanges and generate responses tailored to specific prompts. Kasneci et al. (2023) explain that large language models offer opportunities to support personalised learning, writing, feedback, and educational dialogue, while also raising concerns about reliability, academic integrity, and students’ cognitive engagement. Similarly, UNESCO (2023) emphasises the need for a human-centred approach to generative AI in education, particularly regarding the protection of learners’ agency and the maintenance of meaningful human involvement in educational processes. These developments make it necessary to investigate not merely whether students use generative AI, but how such use influences the intellectual abilities that university education is expected to develop.

Generative AI-assisted learning refers to an instructional or self-directed learning process in which students use AI tools to support activities such as understanding difficult concepts, generating examples, explaining theories, brainstorming ideas, receiving feedback, and developing possible solutions to academic problems. Tools based on large language models can provide immediate responses to students’ questions and adjust explanations according to the prompts supplied by users. This capacity may help students overcome learning difficulties, explore alternative interpretations, and receive additional academic support outside conventional classroom hours. Lo (2023), in a rapid review of the emerging literature on ChatGPT in education, identifies potential benefits for personalised learning, teaching support, and educational productivity, while also highlighting concerns relating to inaccurate information, plagiarism, and overdependence. Tlili et al. (2023) similarly examine the opportunities and challenges associated with ChatGPT in education, including accessibility, ethical concerns, and the changing responsibilities of teachers and learners. However, the availability of AI-generated explanations does not automatically guarantee meaningful learning. Students must still evaluate the quality of the information, identify weaknesses in arguments, connect new knowledge with existing understanding, and determine whether suggested solutions are appropriate. Consequently, the educational value of generative AI-assisted learning depends partly on how students interact with these systems and whether their use promotes active intellectual engagement rather than passive acceptance of generated responses.

Critical thinking is a fundamental educational outcome because it enables students to analyse information, assess evidence, identify assumptions, evaluate arguments, make reasoned judgments, and reach defensible conclusions. At the university level, critical thinking extends beyond memorising lecture notes or reproducing information from textbooks; it requires students to question claims, compare explanations, recognise inconsistencies, and distinguish reliable evidence from unsupported assertions. Facione (1990), drawing on a Delphi study involving experts in critical thinking, identifies interpretation, analysis, evaluation, inference, explanation, and self-regulation as central dimensions of critical thinking. These abilities are particularly important in an educational environment where generative AI can produce fluent and persuasive answers that may nevertheless contain factual errors, misleading reasoning, or incomplete explanations. If students use AI-generated material as a starting point for questioning, verification, comparison, and independent judgment, the technology may create opportunities to practise critical thinking. Conversely, when students accept generated responses without examining their evidence or reasoning, their engagement with the underlying intellectual task may be reduced. Kasneci et al. (2023) highlight both the potential of large language models to facilitate critical engagement and the risks associated with inaccurate outputs and insufficient scrutiny. Therefore, understanding the relationship between AI-assisted learning and critical thinking is essential for determining whether these technologies support the development of independent and reflective university graduates.

Independent problem-solving is another important competency expected of university students. It involves recognising a problem, defining its requirements, identifying relevant information, developing possible approaches, selecting appropriate strategies, implementing solutions, and evaluating outcomes without relying excessively on external assistance. These abilities are relevant across disciplines, including education, business administration, engineering, computer science, social sciences, and the humanities. Jonassen (2000) argues that problem-solving is a central component of meaningful learning and that different types of problems require learners to develop suitable representations, reasoning strategies, and solution processes. Similarly, Zimmerman (2002) explains that self-regulated learners actively participate in planning, monitoring, and evaluating their learning activities. Generative AI can potentially support independent problem-solving by presenting alternative methods, offering hints, explaining errors, and helping students break complex tasks into manageable stages. Nevertheless, its usefulness depends on whether students retain responsibility for defining problems, choosing strategies, and judging the effectiveness of proposed solutions. When students routinely request complete answers before attempting tasks themselves, they may become less willing to tolerate uncertainty or persist through difficult intellectual challenges. The central educational concern is therefore whether AI-assisted learning functions as a scaffold that gradually strengthens students’ problem-solving competence or as a substitute for the effort required to develop that competence.

The possible effects of generative AI-assisted learning can be examined through constructivist and cognitive perspectives on learning. Constructivist approaches emphasise that learners develop understanding through active engagement, interpretation, reflection, and the integration of new information with existing knowledge. From this perspective, AI tools may be educationally valuable when students use them to explore concepts, test explanations, challenge assumptions, and construct their own understanding rather than simply reproduce generated content. Vygotsky’s (1978) account of learning highlights the importance of social interaction and guided assistance in the development of higher psychological functions. Although a generative AI system is not equivalent to a human teacher or knowledgeable peer, it may provide forms of temporary instructional support that help students engage with tasks they would otherwise find difficult. Such support becomes educationally meaningful when learners progressively develop the capacity to perform similar tasks independently. Sweller (1988), through cognitive load theory, also explains that learning can be affected by the demands placed on working memory during complex tasks. Generative AI may reduce unnecessary difficulty by clarifying instructions or organising information, but excessive assistance may also remove opportunities for learners to practise essential reasoning processes. These perspectives suggest that the influence of AI-assisted learning cannot be understood solely in terms of access to technology; it must also be examined in relation to the nature of assistance provided, the intellectual activities performed by students, and the extent to which learning transfers to independent tasks.

Research on AI-assisted education has expanded as universities consider how generative AI can be incorporated into teaching, learning, assessment, and academic support. Existing studies suggest that these systems may improve access to explanations and feedback, but their effectiveness depends on instructional design, learner behaviour, task characteristics, and the quality of generated information. Kasneci et al. (2023) discuss the opportunities offered by large language models while warning that their limitations require students and educators to maintain critical oversight. Lo (2023) similarly identifies promising educational applications alongside unresolved questions concerning accuracy, assessment, and responsible use. In a controlled study involving university students completing an essay-writing task, Noy and Zhang (2023) found that access to ChatGPT reduced completion time and improved the assessed quality of writing in that particular experimental setting. However, improvements in task performance or writing quality should not automatically be interpreted as evidence that students have developed stronger critical thinking or independent problem-solving abilities. Producing a better assignment with AI assistance is different from demonstrating the ability to analyse a new problem, construct an argument, or develop a solution without AI support. This distinction is important because universities must consider both immediate academic performance and the long-term development of intellectual competence. Consequently, research on generative AI-assisted learning should examine the processes through which students use AI and the extent to which the skills demonstrated during assisted activities remain evident when students work independently.

In Nigeria, the growing availability of digital learning resources presents universities with opportunities to reconsider conventional instructional practices and strengthen students’ engagement with academic content. Universities in Ogun State operate within a higher education environment in which students are expected to acquire disciplinary knowledge, analytical abilities, independent judgment, and practical problem-solving competence. Generative AI tools may provide additional academic support by helping undergraduates interpret difficult concepts, organise research ideas, explore alternative arguments, and receive explanations beyond scheduled lectures. However, the educational implications of these tools may vary according to students’ digital literacy, access to reliable internet services, lecturers’ guidance, institutional policies, assessment practices, and students’ willingness to verify AI-generated information. UNESCO (2023) cautions that educational institutions should address generative AI through appropriate pedagogical guidance, human agency, privacy protection, and institutional oversight. These considerations are relevant to Nigerian universities because the availability of an AI tool alone does not establish that its use improves educational outcomes. Furthermore, international findings cannot automatically be generalised to undergraduates in Ogun State, whose learning environments and patterns of technology use may differ from those represented in existing studies. A context-specific investigation is therefore necessary to examine how generative AI-assisted learning relates to students’ critical thinking and independent problem-solving skills in selected universities in Ogun State. Such an investigation can provide evidence to guide students, lecturers, and university administrators in developing practices that encourage responsible AI use while preserving the intellectual independence expected of university graduates.

1.2 Statement of the Problem

The increasing availability of generative AI tools has introduced new possibilities and challenges into university education. Undergraduate students can use these tools to obtain explanations, summarise academic materials, generate research ideas, develop written assignments, and receive suggestions for solving academic problems. Although such applications may facilitate access to learning support, their increasing use raises questions about the extent to which students remain actively involved in the intellectual processes underlying their academic work. Lo (2023) identifies opportunities for educational support alongside concerns about incorrect information, plagiarism, and inappropriate dependence on AI-generated responses. Similarly, UNESCO (2023) emphasises the importance of preserving human agency and ensuring that generative AI is used in ways that promote meaningful learning. The problem is that students may obtain apparently satisfactory academic answers without necessarily developing the analytical abilities and independent reasoning required to produce similar answers themselves. This creates a need to examine whether generative AI-assisted learning contributes to genuine intellectual development or primarily makes academic tasks easier to complete.

A major concern surrounding generative AI-assisted learning is its possible effect on students’ critical thinking skills. University education requires students to assess the credibility of information, examine competing arguments, identify assumptions, interpret evidence, and formulate logical conclusions. However, AI-generated responses can appear convincing even when they contain inaccurate information, incomplete reasoning, or unsupported claims. If students accept these responses without independent verification, their opportunities to practise analysis, evaluation, inference, and reflective judgment may be reduced. Facione (1990) identifies these abilities as essential dimensions of critical thinking, indicating that the development of critical thinking requires more than the acquisition of correct answers. Although generative AI may support critical thinking when students question its responses and compare them with credible academic sources, the extent to which such practices occur among undergraduates in selected universities in Ogun State requires empirical investigation. The problem is therefore not simply the availability of AI tools, but whether their use encourages students to examine information critically or makes them more inclined to accept generated content without sufficient evaluation.

Another concern is whether dependence on AI-generated explanations and solutions affects students’ capacity to solve academic problems independently. Independent problem-solving requires students to identify the nature of a problem, determine relevant information, formulate possible strategies, apply appropriate reasoning, and evaluate the results of their efforts. When students use AI to obtain hints, clarify difficult concepts, or compare alternative approaches, the technology may support the development of these competencies. However, when students consistently request complete solutions before attempting tasks themselves, they may have fewer opportunities to develop persistence, self-direction, and confidence in their own reasoning. Jonassen (2000) emphasises the importance of problem-solving processes in meaningful learning, while Zimmerman (2002) explains that effective self-regulated learning involves planning, monitoring, and evaluating one’s own performance. These perspectives suggest that the educational consequences of AI assistance depend partly on how much responsibility students retain for solving problems. Nevertheless, it cannot be assumed that frequent AI use necessarily weakens independent problem-solving; nor can it be assumed that access to AI automatically strengthens it. The extent and nature of this relationship among undergraduate students in selected universities in Ogun State remain important questions requiring systematic investigation.

Despite growing international discussion about generative AI in education, evidence of improved productivity, assignment quality, or access to academic assistance does not by itself establish that students develop stronger critical thinking and independent problem-solving skills. For example, Noy and Zhang (2023) demonstrated productivity and quality improvements in a controlled experiment involving professional writing tasks, but their findings do not directly establish the long-term effects of AI-assisted learning on undergraduate students’ independent reasoning. Similarly, Lo (2023) reviewed early research on ChatGPT in education and identified both potential benefits and significant concerns, demonstrating the need for further investigation into its educational implications. For selected universities in Ogun State, there is a need for context-specific evidence on students’ patterns of generative AI use and how those patterns relate to critical thinking and independent problem-solving. Without such evidence, lecturers and university administrators may find it difficult to formulate effective guidelines for incorporating AI into learning while maintaining academic independence and intellectual development. This study therefore seeks to investigate the effect of generative AI-assisted learning on students’ critical thinking and independent problem-solving skills among undergraduate students at selected universities in Ogun State.

1.3 Aim and Objectives of the Study

The aim of this study is to examine the effect of generative AI-assisted learning on students’ critical thinking and independent problem-solving skills among undergraduate students at selected universities in Ogun State.

The specific objectives are to:

1. Examine the extent to which undergraduate students in selected universities in Ogun State use generative AI tools for academic learning.

2. Determine the effect of generative AI-assisted learning on the critical thinking skills of undergraduate students in selected universities in Ogun State.

3. Assess the effect of generative AI-assisted learning on the independent problem-solving skills of undergraduate students in selected universities in Ogun State.

4. Examine the relationship between patterns of generative AI use and the development of critical thinking and independent problem-solving skills among undergraduate students in selected universities in Ogun State.

1.4 Research Questions

The following research questions will guide the study:

1. To what extent do undergraduate students in selected universities in Ogun State use generative AI tools for academic learning?

2. What effect does generative AI-assisted learning have on the critical thinking skills of undergraduate students in selected universities in Ogun State?

3. What effect does generative AI-assisted learning have on the independent problem-solving skills of undergraduate students in selected universities in Ogun State?

4. What relationship exists between patterns of generative AI use and the development of critical thinking and independent problem-solving skills among undergraduate students in selected universities in Ogun State?

1.5 Research Hypothesis

The following null hypothesis will be tested at a 0.05 level of significance:

H₀: Generative AI-assisted learning has no statistically significant relationship with the critical thinking and independent problem-solving skills of undergraduate students in selected universities in Ogun State.

The hypothesis will be evaluated using appropriate statistical techniques based on the study design, measurement instruments, and distribution of the collected data.

1.6 Significance of the Study

The findings of this study are expected to be beneficial to undergraduate students, university lecturers, university administrators, educational policymakers, curriculum developers, and future researchers.

1.6.1 Undergraduate Students

The study will help undergraduate students understand how their use of generative AI tools may influence their critical thinking and independent problem-solving skills. By identifying productive and potentially counterproductive patterns of AI use, the findings may encourage students to employ AI as a learning assistant rather than as a substitute for personal reasoning. Students may also become more conscious of the importance of verifying AI-generated information, comparing alternative explanations, and attempting academic problems independently before seeking technological assistance.

1.6.2 University Lecturers

The study will provide lecturers with insights into the implications of generative AI-assisted learning for students’ intellectual development. Its findings may help lecturers design teaching activities that encourage analytical reasoning, independent research, evidence evaluation, and authentic problem-solving. The findings may also support the development of classroom strategies in which students use AI to explore ideas and receive feedback while remaining responsible for the reasoning and conclusions presented in their academic work.

1.6.3 University Administrators

University administrators may benefit from evidence concerning the educational opportunities and challenges associated with generative AI. The findings may inform institutional discussions on responsible AI use, academic integrity, digital literacy, and appropriate assessment practices. They may also provide a basis for developing institutional guidelines that distinguish acceptable AI-assisted learning from practices that undermine independent academic work.

1.6.4 Educational Policymakers and Curriculum Developers

The findings may provide useful information for educational policymakers and curriculum developers concerned with integrating emerging technologies into higher education. By examining critical thinking and independent problem-solving alongside students’ AI-use patterns, the study may help inform policies that promote technological competence without neglecting essential cognitive abilities. Curriculum developers may also use the findings to incorporate AI literacy, source verification, analytical reasoning, and independent problem-solving activities into university programmes.

1.6.5 Future Researchers

The study will contribute to the growing academic discussion on generative AI and higher education by providing context-specific evidence from selected universities in Ogun State. It may serve as a reference point for future studies examining AI-assisted learning across different academic disciplines, institutions, and student populations. Researchers may also build on the study by investigating other outcomes, including academic performance, creativity, metacognitive awareness, self-regulated learning, and long-term retention of knowledge.

1.7 Scope of the Study

The study will focus on the effect of generative AI-assisted learning on students’ critical thinking and independent problem-solving skills among undergraduate students at selected universities in Ogun State, Nigeria.

Geographically, the research will be limited to selected universities within Ogun State. The specific institutions will be identified according to the researcher’s sampling plan, institutional access, and the availability of eligible undergraduate participants.

In terms of content, the independent variable will be generative AI-assisted learning, including the extent, frequency, and academic purposes of students’ use of generative AI tools. The dependent variables will be critical thinking skills and independent problem-solving skills. Critical thinking will be considered in relation to students’ abilities to analyse information, evaluate evidence, identify assumptions, make inferences, and reach reasoned conclusions. Independent problem-solving will cover students’ abilities to define problems, develop strategies, apply knowledge, evaluate solutions, and complete academic tasks without excessive dependence on external assistance.

The study will involve undergraduate students who have experience using generative AI for academic learning and, where the sampling design permits, students with lower or no regular use for comparison. Data may be collected through structured questionnaires and, if feasible, performance-based assessments of critical thinking and problem-solving. The final selection of instruments will depend on the research design and available resources.

The study will not assume that every form of AI use produces the same outcome. Rather, it will examine whether different patterns of use are associated with differences in the two cognitive skills. If the study adopts a cross-sectional survey design, its findings will primarily establish statistical relationships rather than conclusively demonstrate causation. Stronger causal conclusions would require an appropriate experimental or quasi-experimental design.

1.8 Operational Definition of Terms

Generative Artificial Intelligence (AI): In this study, generative AI refers to artificial intelligence systems capable of producing new content, including text, explanations, summaries, examples, and proposed solutions in response to users’ prompts.

Generative AI-Assisted Learning: This refers to students’ use of generative AI tools to support academic activities such as understanding concepts, generating ideas, receiving explanations, evaluating possible approaches, and working through academic problems.

Critical Thinking Skills: These refer to students’ abilities to analyse information, evaluate evidence, identify assumptions, assess arguments, make logical inferences, and formulate justified conclusions. These abilities may be assessed through appropriate questionnaire items or performance-based tasks.

Independent Problem-Solving Skills: These refer to students’ abilities to identify and define academic problems, select appropriate strategies, apply relevant knowledge, evaluate possible solutions, and complete tasks without excessive reliance on AI-generated answers or other external assistance.

Undergraduate Students: These are students enrolled in bachelor’s degree or other recognised undergraduate programmes at the selected universities in Ogun State.

Selected Universities: These refer to the universities in Ogun State that will be included in the study according to the researcher’s sampling procedures and institutional access.

AI Dependence: This refers to the extent to which students rely on generative AI tools to complete academic tasks, including situations in which they may accept generated responses without sufficient personal analysis or verification.

Academic Learning: This refers to the formal and informal learning activities through which undergraduate students acquire, interpret, apply, and evaluate knowledge related to their university programmes.

Effect: For this study, effect refers to the measurable influence or statistical association between generative AI-assisted learning and students’ critical thinking and independent problem-solving skills. The strength of any causal interpretation will depend on the research design employed.

Project – The Effect of Generative AI-Assisted Learning on Students’ Critical Thinking and Independent Problem-Solving Skills. A Study of Undergraduate students at selected universities in Ogun State.