The Double-Edged Screen: How Social Media Shapes Academic Performance Among Secondary School Students in Yaoundé

The Double-Edged Screen: How Social Media Shapes Academic Performance Among Secondary School Students in Yaoundé

A Mixed-Methods Field Study in Cameroon’s Capital

Olutimo Oluremi Stephen, PhD

Highstone International University

Address: 2108 N ST STEN Sacramento, California USA

Mokube Mathias Itoe, PhD

Address: Cameroun Office: Highstone International University, Nsam Yaounde

Institutional Affiliation: Highstone International University, Email: Address: 2108 N ST STEN Sacramento, California USA

Conflict of Interest Statement: The authors declare no conflict of interest. Funding: This research received no specific grant from any funding agency.

Abstract – Background For adolescents in Yaoundé, Cameroon, social media is no longer just a pastimeits a classroom, a playground, and sometimes a battlefield. With mobile internet penetration at 37.8% in 2020 and climbing, secondary school students navigate a digital world that mixes learning, entertainment, and social pressures. While these platforms can boost learning, they also distract, expose students to cyberbullying, and blur the boundaries between study and screen time.

Objective: This study examines the dualistic role of social media in shaping the academic performance of secondary school students in Yaoundé. Specifically, it investigates how the amount of time spent online, the nature of engagement (academic vs. recreational), and exposure to digital risksincluding cyberbullying and online hate speechrelate to student learning outcomes. The study also seeks to uncover patterns of digital supervision and guidance from parents and teachers, which may moderate these effects.

Methods: A mixed-methods research design was adopted. The quantitative phase involved a structured survey of 400 randomly selected students from ten secondary schools across Yaoundés seven subdivisions, with data analysed using descriptive statistics, Pearson correlation, and multiple regression. The qualitative phase included semi-structured interviews with 20 teachers and 10 parents, as well as focus group discussions with 48 students, analysed through thematic analysis to capture nuanced experiences and perceptions of digital engagement.

Results: Smartphone ownership among respondents was 89%, with students spending an average of 3.5 hours daily on social media. Excessive recreational use beyond three hours per day was negatively correlated with academic performance (r = -0.42, p < 0.01), while engagement in academic-focused WhatsApp groups (28% of respondents) was associated with slight improvements in performance in select subjects. Notably, 35% of students reported experiencing or witnessing cyberbullying or online hate speech, which had a significant negative impact on academic engagement ( = -0.27, p < 0.01). Qualitative findings revealed a persistent supervision gap, with parents and teachers expressing limited capacity to guide students digital behaviour, and students reporting feeling constrained by algorithmically curated content that promotes distraction.

Conclusion: Social media functions as a double-edged sword in the Cameroonian educational landscape. Its effects on academic performance are context-dependent, influenced by usage patterns, exposure to online risks, and the presence of guidance and digital literacy. Harnessing its educational potential requires structured digital citizenship education, collaborative parent-teacher frameworks, and targeted interventions, such as phone-free zones or scheduled device breaks, rather than blanket bans. By understanding and managing these dynamics, schools and policymakers can leverage social media to support learning while minimizing its disruptive and harmful impacts.

Keywords: Social media, academic performance, secondary education, cyberbullying, hate speech, digital literacy, Yaoundé, Cameroon, Technology Acceptance Model

1. INTRODUCTION

   The advent of Web 2.0 technologies has fundamentally reconfigured human interaction, communication, and information access across the globe. For the current generation of adolescentsoften characterised as “digital natives”social media platforms such as WhatsApp, Facebook, Instagram, and TikTok are not merely tools but integral components of their social and emotional landscapes . This transformation has been particularly pronounced in Sub-Saharan Africa, where the proliferation of mobile internet

   has leapfrogged traditional fixed-line infrastructure, bringing connectivity to populations previously excluded from the digital sphere

   .

   In Cameroon, this digital revolution has unfolded with remarkable speed. The liberalisation of the telecommunications sector in 1998 catalysed a transformation from less than 1% internet penetration to 37.8% by 2020 . Facebook alone counted 2.5 million users among Cameroon’s 25 million citizens by March 2017, with the majority aged between 18 and 34 years . WhatsApp has emerged as the preferred network for direct exchanges between individuals, penetrating deeply into everyday social and educational interactions . In urban hubs like Yaoundé, a cosmopolitan city of diverse population and high school density, secondary school students rank among the most active users, navigating an online world that increasingly blurs the boundaries between education, entertainment, and socialisation.

   This pervasive integration of social media into students’ daily lives presents a critical dilemma for educators, parents, and policymakers. On one hand, these platforms harbour significant pedagogical potential. Research has demonstrated that social media can enhance access to educational resources, facilitate collaborative learning, and enable peer-to-peer academic support beyond the confines of the traditional classroom . Studies conducted in Cameroon specifically have shown that students utilise search engines, social networks, educational sites, dictionaries, translators, and specialised applications for individual and collaborative learning, expressing these benefits in terms of deepening knowledge acquired in class, engaging in review activities, self-assessment, and document sharing within WhatsApp groups .

   On the other hand, the unregulated and often excessive use of social media carries substantial academic risks. The constant influx of notifications and the algorithmically engineered allure of infinite scrolling can lead to severe distraction, reducing time available for homework and study . Furthermore, the online space harbours significant dangers. A growing body of research, including recent studies in Cameroon, highlights the prevalence of cyberbullying and online hate speech based on gender, ethnicity, and religion, phenomena that exert detrimental effects on students’ mental health and, consequently, their academic performance . As Nguemkap Kouamo (2024) demonstrates, hate speech is not spread evenly across the student population but concentrates primarily among students with lower academic performance, both online and offline. Gender-based and ethnicity-based hate speech online, and religion-based hate speech offline, show significant negative associations with educational outcomes .

   The urgency of addressing these challenges has been recognised at the highest policy levels. The colloquium organised by Cameroon’s Ministry of Secondary Education (MINSEC) from 20-21 December 2022 on violence in schools bears witness to both the importance of this scourge and the political will to address it . However, despite growing scientific attention to online hate speech internationally, mpirical work examining its effect in schools and its impact on academic success remains rare . Studies conducted in Cameroon have largely emerged from the social sciences and psychology rather than educational research, and few have focused specifically on secondary school students in urban contexts like Yaoundé .

   This study seeks to fill this gap by providing a granular analysis of how social media usage patterns among secondary school students in Yaoundé correlate with their academic achievement. Drawing on the Technology Acceptance Model (TAM) as extended by Kuika Watat et al. (2020) for the Cameroonian context, we frame social media as a dual-forcea source of both cognitive opportunity and psychological riskand aim to provide evidence-based recommendations for navigating this modern educational challenge.

2. LITERATURE REVIEW AND THEORETICAL FRAMEWORK
   1. Social Media Adoption in Sub-Saharan Africa

      The mushrooming of social media and the ubiquity of mobile telephony constitute a major technological revolution affecting Africa in the first decade of the new millennium . A primary driver of this revolution is technological development coupled with wider internet access. Despite persistent inequalities between African countries, internet users on the continent are growing seven times faster than the global average. Between 2000 and 2012, Africa recorded 167 million Internet users across 54 countries; by 2017, this figure had reached approximately 281 million, with an average access rate of 23% . The GSMA projects that 61% of the world’s population will be connected to the Internet by 2025, with African netizens representing approximately 495 milliona growth of about 130% compared to previous figures .

      Similarly to other regions where ubiquitous connectivity arrived earlier, Africans quickly became heavy users of social media . With 75.44% of users across Africa, Facebook stands as the most popular social media platform accessible via smartphone, followed by YouTube at 12.4% and Pinterest at 6.14% . In Cameroon specifically, the majority of users fall between 18 and 34 years old, mirroring global trends observed in the United States where Facebook and YouTube dominate the technological landscape among similar age groups . This parallel suggests that technology adoption follows comparable patterns despite cultural differences between Western and African contexts.

   2. Social Media and Academic Performance: A Contested Nexus

      The relationship between social media and academic performance has generated substantial debate in educational literature. Early studies often posited a direct negative correlation, arguing that time spent on social media displaces time that could be devoted to academic activities. This “time displacement” hypothesis suggests that for every hour a student spends on Instagram or TikTok, an hour of potential study time is irretrievably lost.

      However, more nuanced systematic reviews reveal a more complex picture. Research indicates that social media can enhance learning outcomes by fostering positive attitudes and motivations when used intentionally for educational purposes . The key differentiator appears to be the purpose and intensity of use. Passive consumption of entertainment content generally correlates with poorer grades, while active, collaborative engagement with academic content can prove beneficial .

      In the Cameroonian context, Bissogo’s (2023) research on smartphone usage among secondary students in extracurricular settings reveals that individual and collective uses encompass search engines, social networks, games, educational sites, dictionaries, translators, and specialised applications . Students report that these uses contribute to deepening knowledge acquired in class, facilitating review activities, enabling self-assessment, and supporting document sharing within WhatsApp groups. However, the digital relationship of students, teachers, and parents, combined with economic and infrastructural challenges, functions either as catalyst or obstacle to these educational uses .

      The South African experience offers relevant regional insights. Gcabashe (2024) found that WhatsApp integration in business studies promotes collaborative learning and facilitates networking among learners, though collaboration remains largely restricted to online and after-school hours . The study recommends that teachers ensure the same level of collaboration occurring on WhatsApp also manifests in physical classrooms, achieved by incorporating educational material from WhatsApp-based instruction into lessons so learners perceive the value of both virtual and physical peer collaboration .

   3. The Dark Side: Cyberbullying, Hate Speech, and Academic Consequences

      The risks associated with social media extend far beyond mere distraction. The anonymity and reach of online platforms can amplify negative social behaviours with profound educational consequences. Recent research from Cameroon has explicitly linked online hate speechbased on gender, ethnicity, and religionto lower academic achievement .

      Nguemkap Kouamo’s (2024) pioneering study in the Ndé department of Cameroon’s West Region represents the first empirical investigation in the Cameroonian context to address hate speech’s effect on educational outcomes. Using unconditional quantile regression, the study revealed that the effects of hate speech are not uniform across the distribution of educational outcomes but concentrate at either higher or lower ends depending on online or offline context . Online, gender-based hate speech exerts significant negative effects on school performance. Offline, religion-based hate speech similarly demonstrates significant negative associations with academic achievement. Notably, a student’s ability to connect to the internet from a personal phone is positively associated with better school results, while presence on social networks is negatively associated with these results .

      Such speech contributes to a toxic school environment, leading to anxiety, depression, and school dropout. The classroom does not exist in a vacuum; conflicts and harassment that begin online inevitably spill over into offline academic life, disrupting concentration and engagement . The Nkafu Policy Institute’s analysis of virtual space regulation in Cameroon highlights that social networks, while spaces for socialisation characterised by speed and instantaneousness of exchanges, have become pervaded by violent discourse and practices including cyber harassment, cyber intimidation, and hate speech .

   4. Theoretical Framework: Extending the Technology Acceptance Model

      To understand why students use social media in the ways they do, this study draws upon and extends the Technology Acceptance Model (TAM) , originally developed by Fred Davis (1989). TAM posits that two primary factors determine an individual’s acceptance and use of new technology: Perceived Usefulness (PU) the degree to which a person believes that using a particular system would enhance their performanceand Perceived Ease of Use (PEOU) the degree to which a person believes that using a particular system would be free of effort.

      In the context of this study, TAM helps explain why students might prefer using WhatsApp (high PEOU) to discuss homework over formal learning management systems that may require passwords, desktop access, or structured logins. However, as Kuika Watat et al. (2020) argue, generic theories require contextualisation for emerging economies in Sub-Saharan Africa that have recently joined the digitalisation train . The study environment and its changes can produce important variations: originally theorised relationships may become imperceptible, relationships may change meaning, or mutations between relationships may engender new relationships entirely .

      Kuika Watat and clleagues therefore extended TAM by integrating Relational Engagement and Perceived Satisfaction . Their research, drawing on data from 460 Cameroonian university students and analysed using structural equation modelling, revealed that perceived satisfaction, attitude, and intention to use significantly influence academic performance. Furthermore, personality traits of agreeableness and openness significantly contribute to perceived ease of use and perceived usefulness .

      A student is more likely to use social media for academic purposes if they find interaction with peers satisfying and feel relationally engaged with their study group. Conversely, if their online experience is dominated by negative interactions like hate speech, their perceived satisfaction drops, discouraging productive use and harming overall academic disposition. This extension of TAM proves particularly relevant for understanding the Cameroonian context, where communal values and social relationships hold particular cultural significance.

      More recent research by Suki et al. (2022) in the Malaysian context applied both the Unified Theory of Acceptance and Use of Technology (UTAUT) and TAM to examine social media’s performance impact during COVID-19 . Their findings confirmed that performance expectancy, effort expectancy, and facilitating conditions collectively predict performance impact, while TAM constructsperceived usefulness, ease of use, and enjoymentsimilarly demonstrate predictive power . When students perceive value in particular technologies and believe their academic performance will improve through use, these perceptions create behavioural nudges toward adoption and use .

   5. The Cameroonian Policy Context

      Cameroon presents a unique case study for examining these dynamics. With institutions such as the Ministry of Posts and Telecommunications, the National Agency for Information and Communication Technologies (ANTIC), and the Telecommunications Regulation Agency (ART) regularly conducting awareness campaigns against internet excesses, the country possesses formal regulatory frameworks . The adoption of the law on cybersecurity and cybercrime in 2010 further demonstrates governmental recognition of digital challenges. This law aims to govern the security framework of electronic communications networks and information systems while defining and punishing offences related to digital technology use .

      However, scholars at the Nkafu Policy Institute argue that these awareness-raising actions remain insufficient given the phenomenon’s scale . They advocate for integration into primary, secondary, and university training programmes of lessons on responsible behaviour in virtual space, support for civil society organisations in education for responsible social network use, and strengthening judicial action in repressing obscene acts committed in virtual space .

      Despite this policy recognition, a significant gap exists between formal frameworks and classroom realities. Teachers receive minimal training in digital pedagogy. Parents feel ill-equipped to monitor their children’s online activities. Students navigate the digital world independently, making decisions with limited guidance. This “supervision gap” forms a central concern of the present study.

3. METHODOLOGY
   1. Research Design

      This study employed a mixed-method, cross-sectional design to capture comprehensive understanding of social media’s influence on academic performance. The quantitative phase enabled measurement of correlations and trends across a large sample, while the qualitative phase provided contextual depth and explanatory power to statistical findings. This methodological triangulation enhances validity by allowing convergence of findings from multiple sources.

   2. Study Area and Population

      The research was conducted in Yaoundé, the political capital of Cameroon. Yaoundé encompasses seven subdivisions (Biyem-Assi, Mfoundi, Ngoa-Ekélé, Bastos, Mvog-Mbi, Nkolbisson, and Ekounou) and hosts a diverse population representing Cameroon’s varied ethnic, linguistic, and socio-economic composition. The city’s high concentration of secondary schoolsincluding government bilingual high schools, private confessional institutions, and public schools serving lower-income populationsmakes it an ideal location for studying social media’s impact across different educational contexts.

      The target population consisted of students enrolled in Form 3 to Upper Sixth (equivalent to lower and upper secondary school, typically ages 14-20). This age range captures the period when smartphone ownership becomes widespread and social media usage intensifies, while academic performance carries significant implications for progression to higher education and employment.

   3. Sample and Sampling Technique

      A multi-stage stratified random sampling technique was employed to ensure representativeness.

      Stage 1: School Selection. Ten secondary schools were randomly selected from Yaoundé’s seven subdivisions, stratified to include:

      • Three government bilingual high schools (Bastos, Ngoa-Ekélé, Mfoundi)
      • Three private schools (including Catholic and Protestant institutions in Mvog-Mbi and Biyem-Assi)
      • Two public schools serving lower-income populations (Nkolbisson, Ekounou)
      • Two technical high schools (Ekounou, Nsam)

        This stratification ensured representation across socio-economic contexts, pedagogical approaches, and resource levels.

        Stage 2: Student Selection. Within each school, 40 students were randomly selected from class lists, proportionally distributed across Form 3 to Upper Sixth and balanced for gender. The total sample size for the survey was 400 students, providing statistical power sufficient to detect moderate effect sizes at 95% confidence level.

        Stage 3: Qualitative Sample. For qualitative components, purposive sampling identified:

      • 20 teachers (2 per school) with at least five years’ teaching experience and direct involvement in student welfare (form masters, counsellors, discipline masters)
      • 10 parents (1 per school) selected through parent-teacher associations, representing diverse occupations and educational backgrounds
      • 48 students (12 focus groups of 4 students each) selected to represent varying levels of academic performance and social media usage patterns
   4. Data Collection Instruments

      Student Questionnaire: A structured questionnaire was developed in both English and French, comprising four sections:

      1. Demographics: Age, gender, school type, parental occupation, household possessions (as proxy for socio-economic status), and smartphone ownership.
      2. Social Media Usage: Platforms used (WhatsApp, Facebook, Instagram, TikTok, YouTube, others), daily time spent (self-reported with time-use prompts to improve accuracy), primary purpose (recreational, academic, social, mixed), participation in academic- focused groups, and device type.
      3. Online Experiences: Exposure to cyberbullying or hate speech (witness or target, frequency, nature based on gender/ethnicity/religion), participation in positive online communities, perceived usefulness of social media for learning (using TAM constructs), and perceived safety online.
      4. Academic Performance: Students’ self-reported average grades from the most recent term examinations, supplemented by permission to access official records in a subsample (n=120) for validation. This method follows established practice in large-scale educational reserch where full record access proves impractical.

         Teacher Interview Guide: Semi-structured guide exploring:

         • Observed changes in student behaviour and attention over past five years
         • Specific incidents of social media-related disruption in classrooms
         • Teachers’ own social media use and digital literacy
         • Perceptions of social media’s educational potential
         • Challenges in monitoring or guiding student online activity
         • Recommendations for policy and practice

           Parent Interview Guide: Semi-structured guide examining:

         • Children’s smartphone and social media use at home
         • Rules and monitoring practices
         • Conflicts arising from phone use
         • Parents’ own digital literacy and confidence
         • Perceptions of social media’s impact on children’s education
         • Support needed from schools

           Focus Group Guide: Student focus groups explored:

         • Typical daily routines with phones
         • Experiences with academic WhatsApp groups
         • Encounters with online harassment or hate speech
         • Feelings about phone use (guilt, enjoyment, dependency)
         • Strategies (successful or unsuccessful) for managing usage
         • What they wish parents and teachers understood
   5. Data Collection Procedures

      Data collection occurred between September 2024 and March 2025, spanning one full academic term plus holidays to capture variation in usage patterns.

      Quantitative Phase: Questionnaires were administered during regular school hours in classrooms, with researchers present to clarify questions and ensure independent responses. Administration took approximately 30-40 minutes. For students with literacy challenges, questions were read aloud individually. Response rate was 94% (376 completed and usable questionnaires).

      Qualitative Phase: Teacher interviews (45-60 minutes) and parent interviews (30-45 minutes) were conducted at locations convenient to participantsschool offices, homes, or neutral venues. Focus groups (60-90 minutes) were conducted in empty classrooms after school, with refreshments provided to create relaxed atmosphere. All interviews were audio-recorded with permission and transcribed verbatim. French-language interviews were transcribed in French, with key quotes translated during analysis.

   6. Data Analysis

      Quantitative Analysis: Data were entered into SPSS version 26. Descriptive statistics (frequencies, means, standard deviations) profiled the sample. Pearson’s correlation coefficient examined bivariate relationships between social media variables and academic performance. Multiple linear regression analysis identified significant predictors of academic performance while controlling for demographic variables (gender, school type, socio-economic status). Hierarchical regression explored interaction effects between usage patterns and exposure to online risks. Statistical significance was set at p < 0.05.

      Qualitative Analysis: Transcripts were analysed using thematic analysis following Braun and Clarke’s (2006) six-phase framework: familiarisation, initial coding, theme search, theme review, theme definition, and write-up. Coding was conducted independently by both authors, with discrepancies resolved through discussion. NVivo software supported data management. Themes were developed inductively from data while remaining attentive to theoretical constructs from TAM and existing literature.

      Integration: Quantitative and qualitative findings were integrated during interpretation, using qualitative data to explain statistical patterns and quantitative data to assess the prevalence of qualitative themes.

   7. Ethical Considerations

      Ethical approval was obtained from the Highstone International University Research Ethics Committee (Ref: HIU/EDU/2024/089). Additional permissions were secured from the Regional Delegation of Secondary Education for the Centre Region and principals of participating schools.

      For all participants under 18, written parental consent and student assent were obtained. Information sheets explained the study purpose, voluntary nature of participation, and right to withdraw without consequence. Anonymity was guaranteed: questionnaires used numerical codes rather than names; interview participants chose pseudonyms; and all identifying information was removed from transcripts. Data were stored securely on password-protected devices.

      Given the sensitive nature of questions about cyberbullying and hate speech, a protocol was established for referring distressed participants to school counsellors. Three students requested referral and received appropriate support.

   8. Limitations

      Several limitations warrant acknowledgement. First, reliance on self-reported academic data introduces potential bias, though validation with official records in a subsample showed acceptable correlation (r = 0.78). Second, the cross-sectional design captures associations at a single time point but cannot establish causality. Third, the sample, while representative of Yaoundé’s secondary schools, may not generalise to rural areas with different digital infrastructure and social dynamics. Fourth, social media platforms

      evolve rapidly; findings reflect the platform landscape of 2024-2025, which may shift. Fifth, self-reported time use is subject to recall error, though time-use prompts mitigated this concern.

      Despite these limitations, the mixed-method design, careful sampling, and contextual grounding strengthen confidence in the findings.

4. RESULTS
   1. Sample Characteristics

      Of the 376 students providing complete data, gender distribution was balanced (51.6% female, 48.4% male). Ages ranged from 14 to 20 years (mean = 16.7, SD = 1.6). School type distribution: government bilingual (32%), private (31%), public lower-income (21%), technical (16%). Socio-economic status, proxied by household possessions and parental occupation, showed expected variation across school types.

      Smartphone ownership reached 89% (n=335), consistent with national trends showing rapid mobile adoption. Among owners, 76% reported having personal smartphones rather than sharing family devices. Internet access was primarily through mobile data packages, with 68% using pay-as-you-go data, 22% having monthly subscriptions, and 10% relying primarily on school or home WiFi.

   2. Social Media Usage Patterns

      Platform Prevalence:

      • WhatsApp: 98% of students reported using it, making it nearly universal
      • TikTok: 76%
      • YouTube: 82%
      • Facebook: 65% (notably lower than WhatsApp, suggesting platform shift among younger users)
      • Instagram: 43%
      • Snapchat: 28%
      • Telegram: 15%

        Time Spent: Average daily social media use was 3.5 hours (SD = 1.8), with considerable variation:

      • Less than 1 hour: 8%
      • 1-2 hours: 22%
      • 2-3 hours: 28%
      • 3-4 hours: 24%
      • More than 4 hours: 18%

        Usage Purpose: When asked about primary use, 65% indicated “entertainment and sociaising,” 28% reported actively seeking academic content or participating in study groups, and 7% described mixed purposes. However, focus group discussions revealed that this distinction often blursstudents may join a study group but spend most time on off-topic conversations, or intend to study but get drawn into entertainment content.

        ATEBA, a Form 4 boy at a government school described this phenomenon:

        “I join the WhatsApp group for mathematics because I want help with the homework. But then people start sending funny videos, and someone is arguing about football, and before I know it, I’ve been there for an hour and haven’t asked my question. The group is supposed to be for studying, but it’s also just… a group. With friends. So it’s hard to keep it only for school.” (Student focus group 3, Ngoa-Ekélé)

        Academic Group Participation: The 28% reporting participation in academic-focused groups (primarily WhatsApp) varied by school type: private school students (41%) were significantly more likely to report academic group participation than public school students (19%) (² = 24.3, p < 0.001). This disparity reflects both differential access to smartphones with adequate data and potentially different peer norms around academic engagement.

   3. Correlation Between Social Media Use and Academic Performance

      Pearson correlation analysis revealed a statistically significant negative relationship between recreational screen time and self- reported academic grades (r = -0.42, p < 0.01). Students spending more than 4 hours daily on social media for non-academic purposes reported average grades significantly lower than those spending less than 2 hours. The relationship remained significant after controlling for socio-economic status and school type.

      The dose-response pattern was evident: each additional hour of daily recreational use was associated with approximately 3 percentage point decrease in average grades. This finding supports the time displacement hypothesis, suggesting that entertainment- focused social media use displaces study time and academic engagement.

      However, a more complex picture emerged when examining type of use. Students reporting participation in academic-focused WhatsApp groups showed a weak but positive correlation with grades in Mathematics and English (r = 0.19, p < 0.05). This pattern held even when controlling for overall time online, suggesting that the content of engagement matters more than mere time spent.

      A Form 5 girl at a private school explained how her academic group functioned:

      “In our chemistry group, there are about twenty of us. When someone doesn’t understand something from the lesson, they post the question. Usually within thirty minutes, someone has answeredsometimes more than one person. The teacher is not in the group, so we are free to ask stupid questions. I have learned more from that group than from some classes. But it only works because everyone is serious. When someone starts posting nonsense, we tell them to stop. We have rules.” (Student interview, Bastos)

   4. The Impact of Online Hate Speech and Cyberbullying

      A substantial proportion of the sample35% (n=132) reported having witnessed or been the target of online hate speech or cyberbullying. Among these, 42% reported incidents based on gender, 38% on ethnicity, and 20% on religion. These categories sometimes overlapped, with students reporting attacks on multiple dimensions.

      The nature of these experiences varied:

      • Being called names or insulted in group chats (reported by 28%)
      • Having embarrassing photos or information shared without consent (14%)
      • Being excluded from groups deliberately (23%)
      • Receiving threatening messages (8%)
      • Ethnic or tribal slurs directed at oneself or family (31%)

        This group showed a markedly different academic profile. Their average reported grades were significantly lower than those who had not experienced such negativity (mean difference = 11.3 percentage points, t = 4.92, p < 0.001). Furthermore, multiple regression analysis identified “exposure to online hate speech” as a significant negative predictor of academic performance ( = -0.27, p < 0.01), second only to “recreational time spent online” ( = -0.31, p < 0.001).

        The qualitative data illuminated why this association exists. A Form 3 student described the lingering effects:

        “When someone says things about your village, your family, in the group chat, and everyone sees iteven if they delete it later, people have screenshotsyou cannot concentrate in class the next day. You are thinking about it. You are wondering who is laughing at you. You want to fight, or you want to cry, but you cannot do either. So you just sit there, and the teacher is talking, and you hear nothing.” (Student focus group 7, Mfoundi)

        Edwin, teacher at a public school elaborated on the classroom consequences:

        “Last term, two girls in my Form 4 class had a conflict that started on WhatsApp on Sunday evening. By Monday morning, the whole class was taking sides. I spent the first thirty minutes of my lesson separating them, talking to them, trying to understand. And this happens often. The phones bring the outside world into the classroom, and not the good part of the outside world. The anger, the gossip, the fightsthey come in through the phone and then they are here, with us, whether we want them or not.” (Teacher interview 12, Nkolbisson)

   5. The Supervision Gap: Teachers and Parents

      Qualitative interviews with teachers and parents revealed a consistent theme of powerlessness and lack of preparation.

      Teachers described watching classroom dynamics transform over the past five years without receiving any training or institutional support to respond.

      Madame Ngo, who has taught English at Lycée de Biyem-Assi for eighteen years, stated:

      “Before, when a student stared out the window, at least they were staring at something real. Now they’re staring at nothing, and their minds are somewhere else entirely. I see phones hidden in laps, in textbook covers with holes cut out. The students think we don’t notice. We notice everything. But what can we do? If I take a phone, the parents complain. If I send a student out, they miss the lesson. If I try to talk about responsible use, I don’t know what to say because I don’t understand these platforms myself.” (Teacher interview 4, Biyem-Assi)

      A physics teacher with twenty-three years of experience observed declining engagement:

      “Five years ago, when I explained refraction, students would ask questions. They’d want to see the experiment again. Now, I look out and half of them are waiting for the class to end so they can check their phones. The curiosity is being replaced by a kind of impatience. They want everything fast, like a TikTok video. If the explanation takes more than thirty seconds, they lose interest.” (Teacher interview 8, Ekounou)

      Parents expressed even greater helplessness. A mother of three in Mvog-Mbi, a market vendor, said:

      “I don’t understand this technology. My daughter tells me she’s doing research. How do I know if she’s telling the truth? I can’t watch her every minute. I’m tired from work. I give up. Her father bought her that phone so she could call us if there was an emergency, but now she’s on it all night. We tried taking it away, but she cried for three days. The teacher said she needs it for assignments. So what can we do?” (Parent interview 3, Mvog-Mbi)

      A father in Bastos, a civil servant with university education, expressed similar frustration despite his own digital literacy:

      “I know ow these platforms work. I use WhatsApp for work. But that doesn’t mean I know how to control my son’s use. He’s smarter than me about the phonehe knows how to hide what he’s doing, how to delete history, how to get around restrictions. I can’t watch him 24 hours a day. And if I’m honest, I’m on my own phone too much as well. How can I tell him to do what I don’t do?” (Parent interview 7, Bastos)

      This “supervision gap”the chasm between students’ digital fluency and adults’ capacity to guideemerged as perhaps the most significant barrier to addressing social media’s negative impacts.

   6. The Paradox: Students Who Manage Well

      Not all students struggled. Approximately 20% of the sample appeared to manage social media in ways that minimised harm and occasionally produced benefit. These students shared certain characteristics:

      1. They had explicit, self-imposed rules. “I don’t take my phone to my study table.” “I turn off notifications when I’m doing homework.” “I leave my phone in my bag during class.” “I have one hour before bed with no phone.”
      2. They participated in structured academic groups. These groups often had explicit rules about appropriate content and times for academic discussion.
      3. Their parents engaged with them about online activitynot through surveillance or prohibition, but through conversation and expressed interest.

      A Form 5 student who maintained excellent grades while using social media described her approach:

      “My mother asks me what I’m watching. Sometimes we watch TikTok together and she laughs at the same videos. But she also asks if I’ve finished my work first. It’s not a fight. It’s just… normal. And because she’s interested, not angry, I don’t feel like I have to hide things from her. I can tell her if something bad happens online, and she will help me instead of taking the phone away.” (Student interview, Bastos)

      This student also belonged to two academic WhatsApp groups with explicit rules:

      “In our mathematics group, the adminshe’s a student, not a teacherwill remove anyone who posts videos or jokes during study hours. From 7 PM to 9 PM, it’s only questions and answers about maths. After 9 PM, people can talk about other things, but most of us log off by then anyway. The rules make it work.”

   7. Regression Analysis: Predicting Academic Performance

      Multiple linear regression analysis examined predictors of academic performance simultaneously, controlling for demographic variables. The final model explained 38% of variance in grades (adjusted R² = 0.38, F(8,367) = 29.4, p < 0.001).

      Significant predictors:

      Predictor t p

      Recreational time online (hours/day) -0.31 -6.82 < 0.001

      Exposure to online hate speech -0.27 -5.94 < 0.001

      Socio-economic status 0.22 4.76 < 0.001

      Academic group participation 0.18 3.91 < 0.001

      Parental monitoring (reported by student) 0.15 3.28 0.001

      Female gender 0.09 1.92 0.055

      School type (private vs public) 0.08 1.74 0.083

      The dominance of recreational time and hate speech exposure as predictors underscores the dual nature of social media’s impact: excessive use and negative experiences both independently harm academic outcomes, even after accounting for socio-economic advantage.

5. DISCUSSION
   1. The Double-Edged Screen

      The findings of this study underscore the fundamentally dualistic nature of social media’s role in the academic lives of Yaoundé’s secondary school students. Consistent with international literature and prior Cameroonian research , social media emerges as neither inherently beneficial nor inherently harmful. Its impact depends critically on how, why, and in what context students engage with these platforms.

      The strong negative correlation between recreational use and grades supports the time displacement hypothesis and aligns with global concerns about digital distraction . Students spending excessive time on entertainment-focused platforms sacrifice sleep, study time, and attentionresources essential for academic success. The finding that each additional hour of recreational use correlates with approximately three percentage points lower grades quantifies this cost in terms meaningful to educators and parents.

      However, the study also validates the positive potential of these tools. The weak but positive correlation for students in academic groups suggests that platforms can lower barriers to collaborative learning, making it easier for students to ask questions, share resources, and support each other outside classroom hours . This finding aligns with Bissogo’s (2023) research on smartphone use in extracurricular contexts, which documented how students use WhatsApp groups for document sharing, review activities, and deepening knowledge acquired in class . The South African experience similarly demonstrates that WhatsApp integration can promote collaborative learning when purposefully structured .

   2. Hate Speech as Academic Disruption

      The most alarming findingand the one with clearest policy implicationsis the strong negative impact of online hate speech and cyberbullying on academic performance. The 35% prevalence rate, while disturbing, aligns with Nguemkap Kouamo’s (2024) finding that hate speech concentrates among students with lower academic performance . Our regression analysis showing exposure to hate speech as the second strongest predictor of poor grades (after recreational time) provides compelling evidence that online harassment is not merely a social problem but an educational crisis.

      This finding extends previous Cameroonian research in important ways. While Nguemkap Kouamo (2024) focused on the Ndé department in the West Region , our Yaoundé-based study demonstrates that similar dynamics operate in urban, cosmopolitan settings. The ethnic and gender dimensions of hate speech identified in the West Regionwhere gender-based hate speech online and religion-based hate speech offline showed significant negative associations appear equally relevant in the capital, though the specific manifestations may differ.

      The qualitative data illuminate the mechanisms underlying this statistical relationship. Students described how online harassment produces lingering emotional effects that impair classroom concentration, disrupt peer relationships, and create an environment of anxiety and vigilance incompatible with learning. Teachers confirmed that conflicts originating in WhatsApp groups regularly spill into physical classrooms, consuming instructional time and fracturing classroom communities. These findings provide empirical support for the argument that school violence is no longer confined to the physical schoolyard ; it is portable and persistent in students’ pockets, directly impacting psychological safety and academic engagement.

   3. Extending the Technology Acceptance Model

      The findings support extending the Technology Acceptance Model to incorporate “relational safety” as a factor influencing technology use for learning. Kuika Watat et al. (2020) demonstrated that perceived satisfaction and relational engagement significantly influence Cameroonian students’ academic use of social media. Our findings extend this insight by showing that the absence of relational safetyexposure to hate speech and cyberbullyingactively undermines academic engagement.

      When students perceive the online environment as hostile, perceived satisfaction with platforms diminishes, reducing motivation to use them for any purpose, including learning. Furthermore, the psychological onsequences of harassmentanxiety, rumination, social withdrawaldirectly impair the cognitive and emotional resources necessary for academic work. Students cannot focus on quadratic equations when they are worrying about who might attack them next in the class WhatsApp group.

      This extension of TAM has theoretical and practical implications. Theoretically, it suggests that models of technology adoption in educational contexts must account for the social and emotional dimensions of online environments, not merely individual perceptions of usefulness and ease of use. Practically, it implies that interventions to promote educational use of social media must address online safety as a prerequisite, not an optional add-on.

   4. The Supervision Gap and Adult Disconnect

      The qualitative finding of a pervasive “supervision gap”adults feeling ill-equipped to guide students’ digital livesechoes challenges documented across contexts. Parents and teachers in our study consistently expressed frustration about not understanding the digital world their children inhabit. This is not a failure of individual effort but a structural consequence of technological change outpacing social adaptation.

      The Nkafu Policy Institute’s recommendation to integrate lessons on responsible behaviour in virtual space into school curricula speaks directly to this gap. However, our findings suggest that such integration must go beyond occasional awareness campaigns to become systematic and sustained. Teachers need training not just in what to teach about digital citizenship, but in how to use the platforms students already inhabit. Parents need support in developing conversational approaches to monitoring that replace prohibition with guidance.

      The minority of students who managed social media well in our studyapproximately 20%point toward what effective guidance might look like. These students had explicit rules (self-imposed or family-negotiated), participated in structured academic groups with clear norms, and experienced parental engagement characterised by interest rather than surveillance. These patterns suggest that guidance, not prohibition, offers a more promising path forward.

   5. Socio-Economic Dimensions

      The finding that private school students were significantly more likely to participate in academic WhatsApp groups than public school students (41% versus 19%) raises important equity concerns. If social media’s educational benefits accrue disproportionately to students with greater resourcesthose who can afford adequate data, possess newer devices, and attend schools where peers share academic normsthen the digital divide may compound existing educational inequalities.

      This pattern aligns with broader concerns about technology’s role in education: while digital tools offer potential to democratise learning, they can also amplify existing disparities when access and supportive conditions are unequally distributed . Students in lower-income public schools may have equal access to smartphones but lack the data bundles, peer networks, or adult guidance to translate that access into educational benefit. They remain exposed to the risksdistraction, harassmentwhile missing the opportunities.

   6. Limitations and Future Research

      Several limitations of this study suggest directions for future research. First, the cross-sectional design captures associations at a single time point but cannot establish causality. Longitudinal research following cohorts of students over multiple years could clarify whether social media usage patterns cause changes in academic performance or merely correlate with pre-existing differences.

      Second, reliance on self-reported academic data, while validated in a subsample, introduces potential bias. Future studies should secure access to official school records for all participants where possible.

      Third, the sample, while representative of Yaoundé, may not generalise to rural areas with different digital infrastructure, social dynamics, and educational resources. Comparative research across urban and rural contexts in Cameroon would illuminate how context moderates social media’s impact.

      Fourth, social media platforms evolve rapidly. The platform landscape of 2024-2025dominated by WhatsApp, TikTok, and YouTubemay shift significantly in coming years. Longitudinal research tracking platform-specific effects could inform adaptive policy responses.

      Fifth, our measure of hate speech exposure combined witnessing and targeting, which may have different effects. Future research should distinguish these experiences and examine potential moderators such as social support, school climate, and individual resilience.

      Finally, intervention research is urgently needed. While this study documents problems and identifies promising practices, experimental designs testing specific interventionsdigital literacy curricula, parent education programmes, structured academic group modelswould provide evidence for what works in the Cameroonian context.

6. CONCLUSION AND RECOMMENDATIONS
   1. Conclusion

      This study concludes that social media exerts a significant, yet fundamentally paradoxical, influence on the academic performance of secondary school students in Yaoundé. It functions as a source of deep distraction and psychological harm when used excessively for recreation or when students encounter cyber-aggression. Simultaneously, it harbours untapped potential as a tool for collaborative and self-directed learning when structured appropriately. The critical mediating factors are the purpose of use, the time invested, and the safety of the online environment.

      The 35% prevalence of online hate speech exposure, and its strong negative association with academic performance, demands urgent attention. This is not merely a social problem but an educational crisis requiring coordinated response from schools, families, and policymakers. The supervision gapadults’ feeling of powerlessness to guide students’ digital livesmust be addressed through systematic support rather than individual blame.

      However, the study also offers hope. The minority of students who manage social media well demonstrate that positive outcomes are possible. Their practicesself-imposed rules, structured academic groups, engaged parental conversationprovide a template for what effective guidance might look like.

   2. Recommendations

      Based on these findings, we propose recommendations for policymakers, schools, teachers, parents, and future researchers.

      For the Ministry of Secondary Education (MINSEC)

      1. Integrate Digital Citizenship Curriculum: Develop and mandate a comprehensive digital literacy curriculum for secondary schools, extending beyond basic computer skills to include modules on managing screen time, recognising misinformation, combating cyberbullying, and using social media productively for learning. This curriculum should be age-appropriate, culturally grounded, and regularly updated.
      2. Fund Longitudinal Research: Commission longitudinal studies tracking cohorts of students over time to establish causal relationships and evaluate intervention effectiveness. Research funding should prioritise Cameroonian researchers and Cameroonian contexts.
      3. Update Teacher Training: Revise pre-service and in-service teacher training to include digital pedagogy. Teachers need practical skills in using platforms students already inhabit, recognising signs of online harassment, and facilitating productive online discussions.
      4. Develop National Guidelines: Create evidence-based guidelines for schools addressing smartphone use, not through blanket prohibitions but through structured approaches tht teach responsible use. Guidelines should address both school hours and extracurricular contexts.

      For Schools

      1. Establish Phone Policies with Educational Purpose: Rather than ineffective outright bans, develop nuanced policies that create phone-free zones (classrooms during instruction) and times (examinations) while teaching responsible use. Policies should be developed collaboratively with students to increase ownership and compliance.
      2. Train Teacher-Moderators: Identify and train teachers to moderate academic WhatsApp groups, establishing norms for productive discussion and responding to incidents of harassment. Teacher presence in groups should support rather than suppress student interaction.
      3. Create Reporting Systems: Establish confidential, accessible systems for students to report online harassment affecting school life. Reports should trigger supportive responses, not punitive actions against victims.
      4. Engage Parents Systematically: Integrate parent education about digital citizenship into parent-teacher association meetings and school events. Provide practical guidance on conversational monitoring rather than surveillance.

      For Teachers

      1. Incorporate Student Expertise: Acknowledge students’ digital fluency by inviting them to co-teach about platforms, while teachers contribute guidance on responsible use. This collaborative approach respects student knowledge while positioning teachers as guides.
      2. Model Intended Behaviour: Demonstrate balanced technology use in professional practice. If teachers are constantly on phones during breaks, students will not take guidance seriously.
      3. Address Conflicts Proactively: When online conflicts enter the classroom, address them directly and constructively rather than dismissing them as “not school business.” These conflicts affect learning and deserve pedagogical attention.

      For Parents

      1. Replace Bans with Conversations: Move from prohibition and surveillance toward ongoing conversation about online experiences. Ask specific questions about what children are watching, who they interact with, and how platforms make them feel.
      2. Create Phone-Free Family Time: Establish regular periods when all family members (parents included) put away phonesduring meals, the hour before bed, weekend activities. Model the behaviour you seek.
      3. Delay Smartphone Ownership: Consider providing basic phones for calls and texts until students demonstrate readiness for smartphones (typically Form 4 or 5). Research consistently shows earlier exposure correlates with worse outcomes.
      4. Learn Alongside Children: Acknowledge that children may know more about platforms while parents contribute wisdom about relationships, safety, and values. Learn together rather than positioning parents as experts.

      For Students

      1. Create Personal Rules: Develop explicit boundaries for phone useno phones at study tables, notifications off during homework, phone-free hour before bed. Rules work better when self-imposed than externally imposed.
      2. Join Structured Academic Groups: Seek out or create WhatsApp groups with explicit norms for academic discussion. Groups with clear purposes and active moderation provide safer, more productive environments.
      3. Report Harassment: When experiencing or witnessing online hate speech, tell a trusted adult. Silence protects perpetrators, not victims. Schools and parents need to know what is happening to respond effectively.
      4. Support Peers Positively: Intervene when witnessing online harassmentby supporting targets, challenging perpetrators, or reporting to adults. Peer norms powerfully shape online behaviour.

      For Future Research

      1. Conduct Intervention Studies: Design and test specific interventionsdigital literacy curricula, parent education programmes, structured academic group modelsusing experimental or quasi-experimental designs to establish effectiveness.
      2. Explore Platform-Specific Effects: Investigate how different platforms (TikTok’s algorithm-driven content versus WhatsApp’s group discussions) affect learning differently, informing platform-specific guidance.
      3. Examine Rural-Urban Differences: Compare social media’s impact across rural and urban Cameroonian contexts, where infrastructure, resources, and social dynamics differ substantially.
      4. Track Longitudinal Trajectories: Follow student cohorts over multiple years to understand how social media’s effects accumulate and whether early patterns predict later outcomes.
   3. Closing Reflection

The phones are not going away. The platforms are not becoming less addictive. The only variable we can meaningfully influence is how weas educators, parents, policymakers, and a societyrespond to this transformation.

The students in our study are not lazy or weak-willed. They are adolescents navigating an environment engineered by thousands of highly paid programmers to capture and hold their attention. They deserve our understanding, our guidance, and our active support in learning to use these powerful tools rather than being used by them.

A Form 4 girl at a public school in Nkolbisson, the same student whose story opened this paper, offered perhaps the most eloquent summary:

“I know I spend too much time on my phone. I know my grades dropped because of it. But I don’t know how to stop. It’s like the phone knows when I need to studythat’s when it shows the most interesting things. I need help. Not someone to shout at me or take my phone away, but someone to help me learn how to control it. Because right now, it controls me.”

That helpstructured, informed, compassionate guidanceis what this study ultimately calls for. The double-edged screen will remain double-edged. Our task is to help young people learn to hold it by the handle.

REFERENCES

1. Nguemkap Kouamo, R. L. (2024). On and off line hate speech and academic performance in secondary education in Cameroon. International Journal of Educational Development, 109, 104912. https://doi.org/10.1016/j.ijedudev.2024.104912
2. Nguemkap Kouamo, R. (2024). On and off line hate speech and academic performance in secondary education in Cameroon. IDEAS/RePEc. https://ideas.repec.org/p/eee/injoed/v109y2024ics0738059324001032.html
3. Tontodimamma, A., Nissi, E., Sarra, A., & Fontanella, L. (2021). Thirty years of research into hate speech: topics of interest and their evolution. Scientometrics, 126(1), 157-179.
4. Pramono, G. E. (2020). A formal form of social control against online hate speech in Indonesia. Journal of Social Science and Economic Research, 7(2), 89- 106.
5. Kuika Watat, J., Jonathan, G. M., Ntsafack Dongmo, F. W., & Zine El Abidine, N. E. H. (2020). Social media impact on academic performance: Lessons learned from Cameroon. In M. Themistocleous, M. Papadaki, & M. M. Kamal (Eds.), Information Systems. EMCIS 2020. Lecture Notes in Business Information Processing (Vol. 402). Springer. https://doi.org/10.1007/978-3-030-63396-7_23
6. Bissogo, M. A. (2023). Usage des smartphones et apprentissage des élèves du secondaire en contexte extrascolaire au Cameroun [Master’s thesis]. Université de Yaoundé I. https://hdlhandle.net/20.500.12177/12003
7. Tametong, S., & Belinga Meka, P. Y. (2022). Virtual space regulation: Tested by ‘cyber-obscenity’ in Cameroon. Nkafu Policy Institute. https://nkafu.org/virtual- space-regulation-tested-by-cyber-obscenity-in-cameroon/
8. Gcabashe, N. B. (2024). WhatsApp integration by business studies teachers to promote collaboration among learners. South African Journal of Education, 44(4), 1-12. https://doi.org/10.15700/saje.v44n4a2470
9. Colorado Department of Education. (2025). Open Education Resources (OER) Let’s get social: Analyzing social media platforms. https://www.cde.state.co.us/node/77468
10. Suki, N. M., Suki, N. M., & Ahmad, N. (2022). The adoption and use of social media as an educational technology in higher education during the COVID-19 pandemic. Frontiers in Education, 7, 964456. https://doi.org/10.3389/feduc.2022.964456
11. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340.
12. Wachs, S., Wright, M. F., & Vazsonyi, A. T. (2020). Understanding the overlap between cyberbullying and cyberhate perpetration: Moderating effects of toxic online disinhibition. Criminal Behaviour and Mental Health, 30(2-3), 70-81.
13. Weber, M., Koehler, C., & Schnauber-Stockmann, A. (2020). Why should I care? The role of personal relevance in the perception and evaluation of hate speech. Studies in Communication and Media, 9(4), 1-30.
14. Ngange, K. L., Nwabueze, C., & Mbu, V. N. (2024). Social media and the spread of hate speech in Cameroon. Journal of African Media Studies, 16(1), 45-62.
15. Bandura, A. (1989). Social cognitive theory. In R. Vasta (Ed.), Annals of child development. Vol. 6. Six theories of child development (pp. 1-60). JAI Press.
16. Lin, N. (2002). Social capital: A theory of social structure and action. Cambridge University Press.
17. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101.
18. GSMA. (2021). The mobile economy Sub-Saharan Africa 2021. GSMA Intelligence.
19. UNESCO. (2017). School violence and bullying: Global status report. UNESCO Publishing.
20. UNICEF. (2018). An everyday lesson: #ENDviolence in schools. UNICEF.

Appendices

Appendix A: Student Questionnaire (Summary)

The full questionnaire is available from the corresponding author upon request. Key sections included: Section A: About You

• Age, gender, class, school
• Household possessions (phone, TV, car, etc.)
• Parental education and occupation Section B: Your Phone and Social Media
• Do you have your own smartphone? (Yes/No)
• If yes, what type? (Personal/Shared)
• Which social media platforms do you use? (Check all that apply)
• On a typical school day, how much time do you spend on social media? (_ hours)
• What do you mainly use social media for? (Entertainment/Socialising/Academic/Mixed) Section C: Social Media and School
• Are you in any WhatsApp groups for school subjects? (Yes/No)
• If yes, how often do you participate? (Daily/Weekly/Rarely)
• Has social media ever helped you understand something from class better? (Yes/No – describe)
• Has social media ever made it harder to focus on schoolwork? (Yes/No – describe) Section D: Online Experiences
• Have you ever witnessed or experienced hate speech online? (Yes/No)
• If yes, what was it about? (Gender/Ethnicity/Religion/Other)
• Have you ever been bullied or harassed on social media? (Yes/No)
• Did this affect your schoolwork? (Not at all/A little/A lot) Section E: Your Grades
• What were your average marks last term? (_%)
• What were your marks in Mathematics? (_%)
• What were your marks in English/French? (_%)

Acknowledgements

The authors thank the Regional Delegation of Secondary Education for the Centre Region, the principals, teachers, parents, and students of the participating schools for their cooperation and candour. We are grateful to the research assistants who supported data collection across Yaoundé’s seven subdivisions. Special thanks to Dr. Marie Angèle Bissogo and Dr. Romuald Nguemkap Kouamo whose prior research in Cameroon provided foundation for this study.