International Journal of Contemporary Africa Research Network

ROLE OF XAI IN BRIDGING THE INTERPRETABILITY GAP ACROSS HIGH-STAKES DOMAINS

2026-07-31T03:09:02-07:00

Explainable Artificial Intelligence (XAI) has emerged as a critical field addressing the "black-box" nature of complex machine learning models, ensuring transparency and trust in AI-driven decision-making. This paper explores the role of XAI in bridging the interpretability gap across high-stakes domains such as healthcare, finance, and education. This paper review key techniques—including SHAP (SHapley Additive exPlanations), LIME (Local Interpretable Model-agnostic Explanations), and attention mechanisms—the study highlights how XAI enhances human understanding of AI predictions while maintaining model accuracy. The paper also, emphasizes a human-centric approach, arguing that effective XAI must align with users’ cognitive and emotional needs to foster trust and actionable insights. Future directions was submitted in the paper to include integrating causal reasoning, developing domain-specific explanation methods, and establishing robust governance frameworks to ensure ethical AI deployment. The study concludes that XAI’s success hinges on balancing technical rigor with usability, ensuring AI systems are not only accurate but also transparent, equitable, and meaningful for end-users. This synthesis underscores XAI’s transformative potential in creating collaborative human-AI partnerships across critical sectors.

Academic Integrity in the Age of AI: Educator and Student Opinions on Plagiarism Detection Tools

2025-12-22T14:01:05-08:00

This study examines educators' and students' perceptions of AI-powered and traditional plagiarism detection tools in Nigerian universities, focusing on their effectiveness in identifying AI-generated content. Using a descriptive survey design, data was collected using a questionnaire from 120 lecturers and 450 students across six federal universities. The instrument was validated and a reliability test was conducted yielding a reliability coefficient of 0.84. the data collected were analysed using descriptive statistics of mean and standard deviation and inferential statistics of independent t-test. The results revealed significant differences in confidence levels, with lecturers demonstrating stronger belief in both AI-powered and traditional detection tools compared to students. The findings highlight a critical perception gap, where educators' institutional experience fosters trust in these systems, while students exhibit skepticism due to concerns about accuracy and fairness. The study concludes that overreliance on detection technologies without addressing stakeholder perception gaps may undermine academic integrity efforts. Recommendations include among others include adopting multi-layered assessment strategies combining AI detection with human evaluation, fostering dialogue and writing skill development is advised, while developers should improve algorithmic transparency.

PERCEIVED INFLUENCE OF ARTIFICIAL INTELLIGENCE ON THE TEACHING AND LEARNING OF VOCATIONAL EDUCATION AMONG LECTURERS AND POSTGRADUATE STUDENTS

2025-12-25T15:01:39-08:00

This study investigated the Perceived Influence of Artificial Intelligence on the Teaching and Learning of Vocational Education among Lecturers and Postgraduate Students at the University of Uyo. A descriptive survey research design was adopted for the study. Two research questions were posed and two null hypotheses were formulated and tested at a 0.05 level of significance. The population comprised all lecturers and postgraduate students in the Faculty of Education. A sample of 90 lecturers and 300 postgraduate students was selected using a stratified random sampling technique. The instrument for data collection was a researcher-developed questionnaire titled "Perceived Influence of Artificial Intelligence on Teaching and Learning of Vocational Education (PIAITLVE)". The instrument was validated by three experts in Vocational Education and Measurement and Evaluation, and its reliability was established with a Cronbach's alpha coefficient of 0.81. Simple percentages were used to analyze the research questions, while Pearson Product Moment Correlation (PPMC) was used to test the hypotheses. The findings revealed a strong, statistically significant positive relationship between the perceived influence of AI and the teaching of Vocational Education among lecturers (r = .849, p < 0.05). A significant positive relationship was also found between the perceived influence of AI and the learning of Vocational Education among postgraduate students (r = .699, p < 0.05). It was concluded that AI has a substantial perceived influence on enhancing both instructional delivery and knowledge acquisition in vocational education. It was recommended that the university administration should prioritize investment in AI infrastructure and facilitate comprehensive training programs to maximize the beneficial integration of AI in vocational training.

EXPOSURE AND INTENTION OF NURSING STUDENTS TO INTEGRATE ARTIFICIAL INTELLIGENCE (AI) INTO HEALTHCARE

2025-12-25T14:11:52-08:00

The integration of Artificial Intelligence (AI) into healthcare is transforming clinical practice globally. However, a significant gap exists in understanding the preparedness of nursing students in low- and middle-income countries, like Nigeria, for this digital revolution. This study assessed the exposure to AI technologies and the intention to adopt AI in clinical practice among nursing students in Akwa Ibom State, Nigeria. Two research questions were raised to guide the study and two research hypotheses were formulated and tested. A cross-sectional survey design was employed, recruiting 400 nursing students from two accredited colleges via a multi-stage sampling technique. Data were collected using a structured, self-administered questionnaire that assessed sociodemographic characteristics, exposure to 13 AI applications, and intention to adopt AI using an 8-item scale. Data were analyzed using descriptive statistics and one-way Analysis of Variance (ANOVA) with SPSS version 26.0. Findings revealed a generally low level of exposure to AI technologies (Mean score = 8.2/13), with exposure primarily limited to basic digital tools like electronic record keeping (62.5%). Exposure to advanced AI applications, such as robot-assisted procedures (6.3%) and predictive analytics (13.8%), was minimal. However, students demonstrated a high level of intention to adopt AI, with the majority willing to undertake AI training (87.5%) and confident in their ability to adapt (83.0%). ANOVA results showed a statistically significant difference in exposure levels by year of study (p < 0.001), but no significant difference in intention to adopt AI across academic levels (p = 0.458). Despite limited practical exposure, Nigerian nursing students exhibit strong positive intentions toward adopting AI. There is an urgent need to bridge the gap between intention and competency by integrating AI education into nursing curricula and investing in the digital infrastructure of clinical training environments to prepare a future-ready nursing workforce.

AWARENESS STATUS AND PERCEIVE SOCIETAL EFFECTS OF ARTIFICIAL INTELLIGENCE IN AGRICULTURE BY FARMERS, EXTENSION AGENTS AND LECTURERS IN AKWA IBOM STATE, NIGERIA

2025-12-23T01:18:32-08:00

This study examined the awareness status and perceived societal effects of Artificial Intelligence (AI) in agriculture among farmers, extension agents, and lecturers in Akwa Ibom State, Nigeria. Two research questions guided the study. One hypothesis was formulated and tested at .05 level of significance. The design of the study was a descriptive survey. Using a multi-stage sampling technique, the study selected 319 respondents (123 farmers, 56 extension agents, and 140 lecturers) from three agricultural zones (Uyo, Eket, and Ikot Ekpene). Data were collected through structured questionnaires assessing socio-demographic characteristics, awareness of AI technologies (e.g., drones, precision farming tools, decision-support systems), and perceptions of AI’s societal effects using a 4-point Likert scale (1 = Strongly Disagree, 4 = Strongly Agree). Descriptive statistics (frequencies, percentages, mean scores) and inferential statistics (Analysis of Variance, ANOVA) were employed for data analysis, with post-hoc tests conducted to identify group differences. Results revealed significant disparities in AI awareness, with lecturers (92.1%) and extension agents (89.3%) demonstrating higher familiarity than farmers (58.5%). Perceptions of AI’s societal effects varied markedly: farmers expressed concerns about job displacement (mean = 2.08) and cultural misalignment (mean = 1.92), while lecturers emphasized productivity benefits (mean = 3.85) and overall positive impact (mean = 3.71). ANOVA results confirmed significant differences (p < 0.05) across all societal effect dimensions, rejecting the hypothesis of no differences among stakeholders. The study concludes that AI adoption in Akwa Ibom’s agricultural sector requires stakeholder-specific strategies addressing farmers’ socio-cultural concerns while leveraging extension agents and lecturers as change agents. Recommendations include targeted farmer education, strengthened extension services, and inclusive innovation platforms to ensure equitable and culturally sensitive AI integration.

ANALYSIS OF PERFORMANCE OF ChatGPT AND Meta AI ON PROJECT WRITING

2026-07-31T03:11:47-07:00

This analysis compares the project writing capabilities of two leading AI language models: OpenAI’s ChatGPT (GPT-4o) and Meta’s LLaMA 3.1 405B. ChatGPT demonstrates superior performance in structured writing tasks, offering excellent coherence, formatting, and adaptability across various tones and academic styles. It excels in research support, outline generation, and multimodal integration, making it ideal for general users seeking efficient and polished output. In contrast, Meta’s LLaMA 3.1 405B provides greater customization and domain-specific adaptability through full fine-tuning, open-source flexibility, and self-hosting potential. While its native writing polish and formatting support are less refined than ChatGPT’s, it is better suited for enterprise or institutional settings that require private infrastructure and tailored model behavior. Overall, ChatGPT is optimal for ease of use and quality output, whereas LLaMA 3.1 offers deeper control for technical and specialized applications.

AWARENESS, UTILIZATION, PERCEPTION AND CHALLENGES: INTEGRATION OF ARTIFICIAL INTELLIGENCE-BASED CHATGPT INTO SCIENCE EDUCATION

2025-12-23T00:48:51-08:00

This research investigates the awareness, utilisation, perceptions, and challenges faced by secondary school science teachers in the Akwa Ibom North East Senatorial District, Nigeria, regarding the integration of artificial intelligence-based ChatGPT into science instruction. The study employed a descriptive survey research design. The Science Teachers’ Awareness and Perceptions of AI-based ChatGPT Questionnaire (STAPACQ) was used to collect data from a stratified random sample of 189 science teachers. Quantitative data were analysed using descriptive statistics (mean and standard deviation). The findings revealed a low level of awareness (Mean = 2.03) and a very low level of utilisation (Mean = 1.41) of ChatGPT among the teachers. However, teachers' perceptions of its potential impact were generally positive (Mean = 3.31), indicating optimism about its benefits for teaching efficiency and student engagement, alongside concerns about its potential to hinder critical thinking. The challenges to integration were perceived as very high (Mean = 3.66), with unreliable internet connectivity, lack of technical knowledge, concerns about information accuracy, and absence of institutional support identified as the most significant barriers. The study concludes that while science teachers are open to the potential of AI tools, systemic and infrastructural constraints severely limit their adoption. It emphasises the critical need for targeted professional development, robust policy frameworks, and substantial investment in school infrastructure to facilitate the effective integration of AI technologies in science education.

Artificial Intelligence integration in Student Counseling among Guidance and Counseling Lecturers in Nigeria Universities

2026-01-23T13:32:53-08:00

In the rapidly evolving landscape of higher education, Artificial Intelligence (AI) is emerging as a transformative force, presenting unprecedented opportunities to revolutionize student support services. This study investigated the integration of AI-driven tools specifically chatbots and virtual counseling platforms in student counseling services across Nigerian universities. The research focused on assessing the extent of adoption, effectiveness, and challenges associated with AI applications in guidance and counseling. The study employed a descriptive survey design, with a population comprising guidance and counseling lecturers from 60 universities across Nigeria's six geopolitical zones. A stratified random sampling technique was used to select 300 respondents. Data were collected using a structured 20-item questionnaire titled "Artificial Intelligence Integration in Student Counseling Questionnaire (AIISCQ)," validated by experts in Educational Technology and Counseling Psychology. The instrument, based on a 4-point Likert scale (Very High=4 to Very Low=1), demonstrated high reliability with Cronbach's Alpha coefficients of 0.83 and 0.85. Descriptive statistics (mean and standard deviation) were used for data analysis, with a benchmark mean of 2.50 distinguishing between high and low adoption levels. Findings revealed significant AI adoption in specific counseling functions, including at-risk student identification (Mean=2.92), multilingual support (Mean=2.89), and automated urgent case alerts (Mean=3.02). However, limitations persisted in areas requiring human nuance, such as mental health screening (Mean=2.45) and video counseling (Mean=2.38). The study concluded that while AI enhances counseling accessibility and efficiency, its optimal use requires balanced integration with human expertise. Key recommendations include institutional investments in AI infrastructure, counselor training programs in digital competencies, and the development of national ethical frameworks for AI in counseling. These findings provide critical insights for universities, policymakers, and counseling professionals navigating the digital transformation of student support services in Nigeria's higher education sector.

AI UTILIZATION IN EDUCATIONAL ASSESSMENT: THE LECTURERS VIEW ON INNOVATIVE ASSESSMENT AND ACADEMIC INTEGRITY

2025-12-21T14:30:11-08:00

This study investigated Nigerian university lecturers' perceptions of Artificial Intelligence (AI) in educational assessment, focusing on academic integrity concerns and innovative assessment practices. The research employed a correlational design with a population of 3,083 university teachers, from which a sample of 322 professors, associate professors, and senior lecturers were randomly selected from federal, state, and private universities across Nigeria's six geopolitical zones. Data were collected using the validated University Teachers' Perception and Utilization of AI Questionnaire (UTPUAIQ), a 20-item, 4-point Likert scale instrument with high reliability (split-half coefficient = 0.88). The Google Survey Approach was utilized for data collection, with rigorous follow-up procedures ensuring a strong response rate. The method of data analysis was Pearson Product Moment Correlation. The study revealed a significant negative relationship between positive perceptions of AI use and academic integrity concerns, indicating that lecturers who view AI favorably report fewer integrity worries; and a significant positive relationship between AI tool perception and innovative assessment concerns, suggesting that technologically-engaged educators critically evaluate implementation challenges. Both null hypotheses were rejected at the 0.05 significance level. The study recommended among others that University Lecturers should actively participate in training programs that enhance their understanding of AI tools in assessment, focusing on both technical skills and ethical considerations.

PERCEIVED INFLUENCE OF ARTIFICIAL INTELLIGENCE ON ADMINISTRATIVE TASKS IN THE TERTIARY INSTITUTION IN AKWA IBOM STATE

2025-12-25T14:54:01-08:00

The paper under analysis herein is the perceived influence of artificial intelligence on administrative tasks of University of Uyo, Uyo. There were three hypotheses and three research questions leading to the study. The research design based on this study was cross-sectional survey. The study sample was a population of 2780 non-academics staff in the University of Uyo, Uyo. Data collection will be based on a self-develop structured questionnaire called: Perceived Impact of Artificial Intelligence on Administration of University Makurdi Questionnaire (PIAIYMQ). The research questions were answered in terms of data analysis with the use of descriptive statistics through mean values of scores and the standard deviation. The hypotheses were tested at 0.05 level of significance by using the Chi-Square Statistics. The results indicated that the perceived impact of AI on admission process in University of Uyo, Uyo is huge. The second discovery indicated that the perceived impact of AI on decision making process in University of Uyo, Uyo is significant. The third finding indicated that there is a considerable perceived effect of AI on resource administration in University of Uyo, Uyo it was advised among others that the university needs to develop a complex implementation plan of embodying AI in university procedures, making sure stakeholder buy-in, foreseeable difficulties including the safety of information and ethical concerns should be solved. The university must make investments on training programs with the aim of ensuring the establishment of AI literacy among the staff and faculty to empower them to make appropriate use of AI tools in decision-making, resource administration, and even admissions.

DEMOGRAPHIC DIFFERENTIALS ON PERCEPTIONS OF ARTIFICIAL INTELLIGENCE UTILIZATION FOR IMPROVED AGRICULTURAL PRODUCE AMONG FARMERS IN SOUTH-SOUTH NIGERIA

2025-12-23T01:24:48-08:00

This study examined demographic differentials in farmers' perceptions of Artificial Intelligence (AI) utilization for improved agricultural production in South-South Nigeria. Four research questions guided the study. Three null hypotheses were formulated and tested at .05 level of significance. The study adopted a descriptive survey research design. A multi-stage sampling procedure was employed to select 384 crop farmers (316 valid responses) across urban and rural communities in Delta, Rivers, and Akwa Ibom states. Data were collected using a structured questionnaire assessing five perception dimensions: awareness, perceived benefits, ease of adoption, trust, and future adoption intent, measured on a 4-point Likert scale. The instrument was validated by three experts. Descriptive statistics, independent t-tests, and one-way ANOVA were used for analysis. Results revealed moderately positive perceptions overall, with urban, male, and younger (18–35 years) farmers showing significantly higher acceptance than rural, female, and older counterparts. Key barriers included low awareness, perceived complexity, and rural-urban disparities in infrastructure access. It was recommended among others that Agricultural Extension unit should Implement age- and gender-sensitive training programs with differentiated training modules.

EFFECT OF JIGSAW TEACHING METHOD ON QUADRATIC EQUATIONS ACADEMIC ACHIEVEMENT AND RETENTION IN SECONDARY SCHOOLS IN ORUK ANAM LOCAL GOVERNMENT AREA

2026-07-31T03:19:21-07:00

The study examined the effect of jig-saw methods on Mathematics students’ Academic Achievement and Retention in Quadratic Equation in Oruk Anam Local Government Area Akwa-Ibom state. Two purposes of the study and two corresponding research questions as well as two null hypotheses were formulated to direct the study. Quasi experimental research design was employed in carrying out the study. pretest-posttest control group design was employed. Two (2) intact classes were drawn from a population of two thousand three hundred and eleven (2,311) students across three senior secondary schools, yielding a sample size of one hundred and fifty (150) SS II Mathematics students. Mathematics Achievement and Retention Test (MART) was the major instrument for data collection, face and content validity was carried out. The data generated from the study instrument was analyzed using means and standard deviation and Analysis of co-variance (ANCOVA). The hypothesis was tested at .05 alpha level of significant. The result shows that there is a significant difference between the mean academic achievement score of students taught quadratic equation using jig-saw and those taught using lecture method of teaching. The result also shows that there is no significant difference between the mean academic achievement score of male and female students taught quadratic equation using jigsaw method. It was recommended Mathematics and other science teachers should change from lecture method to jig-saw and problem-based to teach science in secondary school.

AWARENESS, PERCEPTION AND ADOPTION OF ARTIFICIAL INTELLIGENCE (AI) TECHNOLOGIES AMONG AGRICULTURAL ENTREPRENEURS IN NIGERIA

2025-12-23T01:11:03-08:00

This study investigated the awareness, perception, and adoption of artificial intelligence (AI) technologies among agricultural entrepreneurs in Akwa Ibom State, Nigeria. A systematic random sampling method was employed to select six local government areas (Uyo, Ikot Ekpene, Eket, Abak, Oron, and Etinan) known for their high agribusiness activity. Data were collected from 150 respondents (25 from each LGA), comprising crop, poultry, and fish farmers, using a structured questionnaire titled "Awareness, Perception, and Adoption of Artificial Intelligence Technologies Among Agricultural Entrepreneurs in Akwa Ibom State." The instrument was validated by three experts in agricultural extension, agribusiness, and information technology, and its reliability was confirmed with a Cronbach’s alpha coefficient of 0.85. Descriptive statistics (mean, percentages) and inferential statistics (chi-square test) were used to analyze the data. Results revealed moderate awareness of basic AI tools like weather forecasting (mean=2.90) but low awareness of advanced applications such as soil monitoring (mean=2.15). Farmers perceived AI as beneficial for productivity (mean=3.05) but expressed skepticism about labor reduction (mean=2.75). Adoption rates were low, with only 15% regularly using weather apps and below 5% adopting precision tools like automated irrigation. A significant association (χ²=18.65, p<0.001) was found between awareness and adoption, indicating that knowledge gaps are a critical barrier. The study concludes that while AI holds transformative potential, its adoption in Nigeria is hindered by limited awareness, cost, and infrastructure. Recommendations include targeted training for extension agents, subsidized AI tools, and improved rural digital infrastructure. These findings provide actionable insights for policymakers to accelerate AI integration in Nigerian agriculture.