Research ICT Africa

Abstract Identity-based discrimination – unfair treatment by systems and individuals based on social status – is a recognised determinant of health and wellbeing, yet evidence centred on young people living in Global South contexts of exclusion, inequality and oppression remain limited. This article draws on a rapid review of 53 studies examining discrimination related to gender, sexual orientation, race/ethnicity, (dis)ability, geography, and religion. Outcomes were assessed using the Gender and Adolescence Global Evidence (GAGE) framework spanning six domains: psychosocial wellbeing, bodily integrity, education, physical health, economic empowerment, and voice and agency. Most studies reported adverse psychosocial outcomes, with fewer documenting effects on bodily integrity, education, or physical health. Very few addressed economic disempowerment or barriers to voice and agency. The review highlights important gaps and identifies five priorities: expanding evidence on economic impacts; examining protective factors and interventions through intersectional approaches; comparative analyses between Global North and South youth; extending geographical, age and setting coverage; and advancing longitudinal and intersectional research that capture the cumulative effect of discrimination on Global South youth.

Digital technologies are transforming electoral processes across West Africa. Social media, digital ID systems, and online results interfaces offer new avenues for transparency and civic participation, but they also threaten democratic integrity. This article examines the region’s digital-political ecosystem, focusing on how digital platforms are shaping electoral communication in contexts marked by democratic fragility, ethnic divisions, and political instability. Biometric voter registration and electronic results transmission have improved electoral administration, yet at the same time social-media disinformation, internet shutdowns, and surveillance undermine fundamental democratic freedoms. This article also draws attention to emerging digital-governance frameworks that are applicable in West Africa, including UNESCO’s 2025 Model Policy Framework for Information Integrity in West Africa and the Sahel and the AAEA’s 2023 Principles and Guidelines for the Use of Digital and Social Media in Elections in Africa. The effectiveness of digital governance, which requires a balancing of state security concerns against protections for human rights and free expression, will play a central role in determining whether West Africa’s digital transformation strengthens or undermines electoral legitimacy and political stability.

ABSTRACT Our study explores the degree of congruence between citizen and police officer experiences of police sexual violence (PSV). We use the content analysis of 327 newspaper stories of police sexual violence (PSV) in South Africa, comparing the experiences of 92 citizens and 27 police officers who were PSV victims. Our analysis implies that police officers who commit PSV tend to “specialize” and focus on one type of victim, be they citizens or fellow police officers. Citizens tend to be more likely to experience rape, while police officers are more likely to experience sexual harassment. Citizens are typically victimised by non-commissioned police officers, while police officers are more likely to be victimised by their supervisors who are either commissioned officers or senior management. The limited information about the criminal process indicates that police officers victimising citizens are more likely to be charged with multiple offences than police officers victimising other police officers.

This study examines the determinants of new product development (NPD) performance and retailer engagement among South African small and medium-sized enterprises (SMEs). It investigates how strategic orientation, NPD research, team collaboration, and retailer–supplier relationship quality shape innovation outcomes in a resource-constrained retail environment. A positivist quantitative research design was adopted, and data were collected from 74 SME owner–managers using a structured online questionnaire. Descriptive, correlation, and regression analyses were conducted in SPSS (v28.0). Findings show that NPD research is the only statistically significant predictor of retailer engagement (β = 0.61, p < 0.05). Structured customer validation, market testing, and evidence-based learning therefore play a decisive role in improving SMEs’ credibility with retail buyers. Although strategic alignment and cross-functional collaboration show positive associations with NPD outcomes, neither exerts a significant effect on retailer accessibility. The results highlight a persistent “intent–execution gap” in SMEs: while strategic awareness and innovation culture are strong, weak institutionalisation of research and collaboration limits NPD effectiveness. Retailer–supplier relationships remain characterised by information asymmetry, limited feedback, and procedural opacity, reflecting structural power imbalances in South Africa’s concentrated grocery sector. Strengthening research-driven NPD routines, shared market intelligence, and transparent supplier-development mechanisms can improve SME access to dominant retail channels. The study contributes to the Resource-Based View (RBV) and Dynamic Capabilities Theory (DCT) by demonstrating how research intensity and learning routines enhance SMEs’ market legitimacy within asymmetric value chains.

Despite the increasing popularity of Valorant in esports, there is limited research done on how team compositions affect match results when considering various maps and changing game metas. To address this gap, the study aims to identify the optimal team compositions by investigating the match data from the Valorant Champions Tour between 2021 and 2024. It was observed that the teams lack data driven strategies which adapt to agent role changes, map dynamics and meta changes. The study uses quantitative strategies to analyse role distributions and real game performance of Duelists, Controllers, Initiators and Sentinels combinations. The results show that teams that used Duelists with Controllers, Sentinels and Initiators won more often than teams that used only Duelists in 2024. This demonstrates that competitive teams are changing their priorities. The study shows that effective agent roles depend on the map. Duelists are better on Bind and Controllers perform better on Ascent. The findings indicate that Controllers and Initiators were the most popular choices in 2024. The research adds to the knowledge about rolebased success and changing team dynamics in competitive Valorant and it also fills a gap in data driven esports research on strategy and team composition.

This study examines the transformative potential and ethical challenges of artificial intelligence integration in educational assessment systems. Through comprehensive analysis, the research identifies significant value propositions of AI-enabled evaluation, including enhanced data-driven assessment capabilities, sophisticated process evaluation mechanisms, and improved value-added assessment models that transcend traditional limitations. However, the investigation reveals profound ethical concerns surrounding algorithmic evaluation hegemony, privacy erosion, human agency marginalisation, and systemic fairness challenges stemming from algorithmic bias and opacity. The study proposes a multi-dimensional optimisation framework encompassing institutional development, value-oriented integration of instrumental and humanistic rationality, and robust ethical guardianship mechanisms. The findings emphasise the critical need for harmonious human-machine collaboration in educational contexts, ensuring that technological advancement serves rather than supplants fundamental educational values whilst addressing equity concerns and preserving human dignity in assessment practices.

The National Basketball Association provides strong potential for predictive analytics due to its global fan base, structured season and rich historical data. This research evaluates the ability of machine learning models such as logistic regression, support vector machines and random forests to predict the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{2 0 2 3 - 2 0 2 4}$</tex> NBA Championship winner. The research used 20 years of team performance statistics to select predictive features including field goals percentage, two-point percentage, three-point percentage, assists, rebounds and blocks through correlation-based feature selection. The models achieved evaluation through accuracy, precision, recall and F1 score measurements. Logistic regression produced the most balanced and interpretable results despite the class imbalance in the dataset because it correctly predicted Boston Celtics as the championship winner at 94.99 % probability. Random forest produced the highest accuracy rate of 93 %, yet it did not correctly predict the winner because it performed poorly on minority class detection. Support vector machine yielded intermediate performance. The success of logistic regression demonstrates its effectiveness for sports datasets with imbalanced outcomes and proves that conventional team statistics remain valuable for predictions. The research findings expand sports analytics knowledge while showing how machine learning systems can assist professional basketball teams with strategic planning. The developed framework can be adapted to support performance management and talent development in emerging basketball leagues. Future research could use advanced metrics together with neural networks and other alternate modelling techniques to boost predictive power and solve binary classification challenges.