International Journal of Education, Pedagogy and Psychology
DOI: 10.64823/ijepp.2601005
Higher education in Cameroon faces simultaneous pressures: demand for modernization, adoption of pedagogical innovations (active learning, problem-based learning, technology-enhanced teaching), and the imperative to improve graduate outcomes despite resource constraints, large class sizes, and substantial variation in student preparedness. International development partners, regional policy frameworks, and internal university reforms have accelerated the diffusion of novel instructional methods. Yet, educators frequently report that student performance and mastery sometimes stagnate or decline in the short- to medium-term after implementing innovations, producing inefficient learning curves (slow initial gains, high variability, or regression). Understanding the cognitive mechanisms underlying these patterns is essential for designing effective, scalable reforms.
Innovative pedagogies and research-oriented curricula, although conceptually promising, often produce inefficient learning curves among underprepared or novice learners in Cameroonian higher education. This inefficiency manifests as protracted time-to-proficiency, superficial learning, and heightened dropout or failure rates in early cohorts after reform. The central problem addressed is: which cognitive mechanisms and contextual constraints primarily drive this inefficiency, and how can instructional design informed by CLT reduce extraneous load and accelerate schema acquisition? This paper is based on the hypothesis that when pedagogical innovations are implemented without explicit CLT-aligned design (pre-training, worked examples, scaffolded practice, modality choices, and segmentation), extraneous cognitive load and mismatches between task complexity and learner expertise produce inefficient learning curves in Cameroonian higher education. Conversely, interventions that reduce extraneous load and promote germane processing will measurably accelerate the learning curve.
Therefore , the paper aims to:
This paper addresses a pressing implementation gap: international and local enthusiasm for pedagogical innovation often lacks the instructional design specificity required to produce efficient learning. By framing the problem through CLT and offering implementable strategies suited to Cameroon's realities (bilingual instruction, large classes, limited technology), the study aims to improve learning outcomes, reduce wasted instructional investments, and guide policy priorities.
This study uses a mixed-methods approach combining (a) a systematic literature synthesis of CLT and multimedia learning research, (b) a diagnostic audit of instructional artifacts (syllabi, slide decks, lab manuals, e-learning modules) collected from representative Cameroonian faculties, and (c) qualitative interviews with instructors, administrators, and students across five universities (public and private) in Anglophone and Francophone regions. The conceptual analysis maps empirical evidence to observed practice to generate findings and design recommendations. Data sources and sampling include:
The study followed ethical protocols: informed consent, confidentiality, and anonymization of institutional identities. The research purpose was explained to participants, with no coercion or incentives affecting participation.
Widespread absence of pre-training increases intrinsic load for novices Many curricula introduced novel concepts and procedures without explicit pre-training of prerequisite terminology and component schemas. When learners lack foundational referents, the intrinsic cognitive load of complex tasks multiplies because elements cannot be chunked into higher-order schemas (Sweller et al., 1998). Examples include engineering modules that present system-level analyses before teaching key physics concepts and bioscience labs that require familiarity with laboratory equipment and measurement units not previously practiced. In interviews, students repeatedly reported being "thrown into" complex problems with insufficient orientation. The artifact audit found that only 9 of 42 modules included explicit pre-training sequences (glossaries, introductions to tools, or concept maps). The absence of pre-training forces working memory to handle both elemental concept acquisition and their integration, leading to overload and inefficient learning curves. CLT predicts that pre-training reduces element interactivity for novices by establishing schema fragments that can be combined during learning, thus accelerating acquisition and reducing early failure rates (Mayer & Moreno, 2003).
Lecture slides frequently presented dense blocks of text while instructors simultaneously read verbatim, creating redundancy and split-attention effects. Many slide decks included full paragraphs, complex diagrams placed remotely from explanatory labels, and decorative graphics that did not support core learning objectives. According to the redundancy principle and split-attention literature (Mayer, 2001; Chandler & Sweller, 1991), presenting identical verbal information in on-screen text plus narration competes for limited resources and reduces capacity for schema construction. In these contexts, initial student performance improvements were shallow and transient—the signature of an inefficient learning curve—because learners expended cognitive resources dealing with poor presentation rather than internalizing concepts. Audit metrics showed an average of 35% of slide content as extraneous across sampled materials.
Although problem-based learning (PBL) is widely adopted, many implementations omitted worked examples or scaffolded steps for novices. In classrooms where students first attempted open-ended problems with minimal guidance, many engaged in unproductive trial-and-error and cognitive search, leading to wasted working memory and little schema formation (Kirschner, Sweller, & Clark, 2006). Interview data indicated student frustration and frequent requests to "go back to showing us how to do it." The evidence suggests novices require explicit demonstrations (worked examples) to reduce problem-solving search and enable germane processing; absence of such elements predicts initial dips or stagnation in learning curves after PBL adoption.
Cameroon's bilingual context introduces additional language-related extraneous load when instructors switch languages or use technical terms in a language unfamiliar to some students. Where modules lacked bilingual glossaries, visual anchors, or pre-teaching of terminology, learners spent significant working memory capacity translating terms instead of processing conceptual relationships. The audit found poor translation practices: inconsistent use of technical vocabulary and reliance on spontaneous interpretation by peers. This language-based extraneous load particularly affected first-year cohorts with heterogeneous language backgrounds, slowing early gains and increasing variance in performance trajectories.
Classes of 100–300+ students are common; in these settings, instructors face constraints in providing individualized feedback. Without timely corrective feedback, learners continue to practice with misconceptions, which stabilizes incorrect schemas and prolongs the learning curve. The literature shows feedback as a key mechanism for guiding germane processing and preventing error consolidation (Hattie & Timperley, 2007). Interviewed instructors cited administrative burden and lack of teaching assistants; students reported delayed or absent feedback. The consequence is a slower progression from declarative to procedural knowledge and decreased transfer.
Many courses provided either full guidance or none, with little evidence of graduated support. Where scaffolds (worked examples, prompts, partial solutions) were used, they were not systematically faded as learners achieved competence. CLT and the expertise reversal effect indicate scaffolds must be tailored and gradually withdrawn; otherwise they either remain a crutch (if not faded) or fail novices (if removed too early) (Kalyuga, 2007). The audit found only sporadic use of completion problems and no consistent fading schedules across modules. This mismatch results in either prolonged dependence or sudden overload, both generating inefficient learning curves.
Assessments heavily weighted rote recall and single-step procedural tasks rather than integrated problem solving. Where summative assessments did not require schema construction, students prioritized study strategies that produced short-term performance but did not automate deeper schemas; consequently, skill retention and transfer were weak. The resulting learning curves show initial exam-oriented gains followed by rapid forgetting and poor application in later authentic tasks. CLT emphasizes aligning assessments with targeted schema complexity to promote germane processing and durable learning (Paas, Renkl, & Sweller, 2003).
Several e-learning modules and lecture animations placed explanatory text in separate documents or web pages, demanding that learners mentally integrate separated sources. The split-attention effect increases extraneous load and reduces learning efficiency (Chandler & Sweller, 1991; Mayer, 2001). In a key example, an engineering animation was hosted separately from its explanatory notes; students were required to toggle between screens, which interviewees described as "mentally tiring" and disruptive to comprehension. Such fragmentation of modalities predicts slow, effortful learning curves for novices.
Introduction of Learning Management Systems (LMS) and remote-conference tools improved availability of materials but often lacked simple onboarding, clear navigation paths, and low-bandwidth alternatives. Students spent time learning platform mechanics rather than content. The cognitive cost of navigating new systems is extraneous to learning and disproportionately affects low-resource students with limited device access. The artifact audit found many LMS pages with unclear signposting and inconsistent naming. The net effect: initial cohorts demonstrated low engagement metrics and slow content mastery.
Faculty who adopt new pedagogies often undergo a learning curve themselves; without explicit training in evidence-based design, instructors unintentionally increase extraneous load for students (e.g., poorly executed group activities, unclear problem statements, misaligned timings). Interviews revealed that many instructors report "more work" and initial poor student outcomes in the first 1–2 semesters after innovation adoption. This instructor-side learning curve contributes to the institutional learning curve, producing transient performance dips that are sometimes misinterpreted as failure of the pedagogy rather than implementation growing pains.
Laboratory sessions in STEM fields often required students to manage multiple instruments, procedures, and data collection tasks simultaneously, with insufficient stepwise guidance. This high element interactivity environment is intrinsically demanding; without thorough pre-training, worked examples, and instructor scaffolding, students experience overload, make procedural errors, and fail to abstract the underlying principles. Observed outcomes included repeated data collection errors and long delays achieving procedural fluency—manifestations of an inefficient learning curve for hands-on competencies.
Slide decks and e-learning modules frequently included decorative animations, background music, and extraneous images that did not support instructional goals. Such non-coherent elements consume perceptual and working memory resources and distract from meaningful processing (Mayer, 2001). Audit coding marked approximately 28% of multimedia elements as decorative. Students described modules as "noisy," and performance on conceptual questions lagged despite adequate exposure, consistent with coherence principle predictions.
Programs that intersperse topics without clear prerequisite sequencing force learners to revisit foundational concepts repeatedly. This fragmentation raises intrinsic load because learners encounter high element interactivity without requisite schemas. For example, a curriculum that alternated advanced modules with remedial content without scaffolding created repeated reorientation demands on students and contributed to prolonged time-to-proficiency. Careful task analysis and sequencing are required to smooth the learning trajectory.
Peer instruction can be powerful, but the study found it is less effective when structured groups contain uniformly novice students or lack guided prompts and corrective feedback. In such cases, peer discussion sometimes propagated misconceptions rather than clarified them, stabilizing erroneous schemas. Well-structured peer instruction with instructor scaffolds or worked-example complements, however, supported germane processing and accelerated learning; the difference underscores the need for scaffolding in peer modalities.
Graduate-level research training introduces complex, interacting skill sets (research design, statistical reasoning, academic writing) often taught simultaneously without staged scaffolding. Students reported feeling overwhelmed by concurrent deadlines and tasks, which hindered progressive mastery of research competencies. The complexity of research tasks implies high intrinsic load; targeted decomposition (sequential mastery of sub-skills) and worked examples (e.g., annotated research reports) help reduce cognitive burden and produce more efficient research-skill acquisition.
Teacher-education programs adopting minimally guided practicum experiences without worked demonstrations and reflective scaffolds produced trainee teachers who struggled to plan and execute lessons effectively. This indicates that novices require explicit modeling and scaffolded practice before independent classroom application. The result is a lag in classroom competence and a prolonged professional learning curve, with downstream impacts on secondary education quality.
Metacognitive strategies like self-explanation prompt learners to generate integrated schemas and identify knowledge gaps. The audit found sparse inclusion of self-explanation prompts in materials. Where present, students reported deeper understanding and better long-term retention. The absence of metacognitive scaffolds leads to less germane load devoted to schema-building and correspondingly inefficient learning gains.
Many programs increased contact hours or added activities after innovations were introduced, but these additional hours did not translate into improved learning because tasks were poorly aligned with schema-building goals. Time spent on administrative tasks, repetitive low-level drills, or unstructured projects consumed student effort without promoting automation of target schemas. The effect is inflated investment without commensurate learning-curve improvement.
Some faculty equated "active learning" with "minimally guided learning" and believed that reducing instructor direction uniformly increased engagement and learning. These misconceptions led to implementations that heightened extraneous load for novices. Where instructors received correct CLT-based training—differentiating guidance for novices vs. experts and employing worked examples—outcomes improved substantially. This highlights the importance of professional development grounded in cognitive science.
Pilot interventions that implemented a handful of CLT-aligned changes—for example, replacing verbose slides with integrated diagrams plus spoken explanations, adding pre-training modules, and providing worked examples—produced measurable acceleration in early learning curves in trial cohorts. Predictive simulations and pilot data suggest that relatively low-cost, high-leverage design changes can shift the learning trajectory substantially, reducing time-to-proficiency and variance across students.
The absence of pre-training for prerequisites forces working memory to process multiple novel elements simultaneously, increasing intrinsic load beyond manageable levels for novices (Sweller et al., 1998). Practically, this translates into slowed schema formation and prolonged errors. Institutions should mandate pre-training modules (glossary, concept maps, micro-lessons) for high-interactivity topics. Pre-training can be low-tech (printed concept maps) or brief peer-led primers and remains one of the most cost-effective strategies to accelerate initial learning slope.
Redundant textual information delivered both visually and verbally competes for phonological and visual working memory channels, reducing germane processing (Mayer, 2001). Adopting concise slides, using spoken narration for elaboration, and placing labels adjacent to diagram elements will reduce extraneous load. Faculty training workshops should demonstrate slide redesign templates and provide before/after exemplars.
PBL is valuable but must be introduced with worked examples and completion tasks for novices (Kirschner et al., 2006). Institutions should sequence PBL after students demonstrate competence on component tasks via worked-example practice. This staged approach prevents inefficient search behavior and stabilizes early performance gains.
In Cameroon, language operations are not neutral: switching languages without scaffolds imposes translation load. Modular glossaries, bilingual labels on diagrams, and consistent code-switching policies reduce unnecessary processing. Visual supports and pre-teaching of domain vocabulary in both languages are practical mitigations.
Large classes constrain individualized feedback, but scalable models (automated quizzes, peer-assessment with rubrics, targeted office-hour pools) can provide corrective information promptly. Embedding formative checks for common misconceptions prevents consolidation of errors and accelerates learning.
Scaffolds must be matched to learner expertise and faded systematically. Use diagnostic pre-tests to categorize students and deliver differential scaffolds: novices receive full worked examples and guided practice; intermediates receive completion problems; experts receive unguided problem-solving. Monitoring mastery criteria should trigger scaffold withdrawal.
Assessment design drives study behavior. Aligning summative tasks with integrative, multi-step problems incentivizes germane processing. Use rubrics that reward planning, reasoning, and transfer to ensure learners focus on schema construction rather than rote memorization.
Integrate explanatory text with animations (spatial contiguity) and present narrations synchronized with visualizations (temporal contiguity) to reduce extraneous integration demands (Mayer, 2001). Where bandwidth is limited, provide downloadable integrated PDFs or printed still frames with accompanying transcripts.
LMS pages, mobile-friendly navigation, and stepwise platform onboarding. Provide offline packets for students with intermittent access. Usability reduces extraneous navigational load and frees working memory for content.
Faculty transition requires structured professional development that includes CLT principles and concrete redesign templates. Mentoring, co-teaching, and staged rollouts reduce transient declines in instructional quality. Recognizing instructor learning curves prevents premature abandonment of effective pedagogies.
Decompose lab procedures into component skills, provide stepwise worked demonstrations, and scaffold instrument handling before full integrated tasks. Pair inexperienced students with an experienced peer or assistant initially to reduce supervisory load while preserving hands-on exposure.
Remove decorative elements that do not support learning objectives. Encourage designers to treat multimedia as cognitive workspaces—every added element carries a cognitive cost. Simpler, cleaner interfaces often yield better learning gains.
Conduct task analysis to order modules from low to higher element interactivity, ensuring prerequisites are scheduled and assessed before advancing. Sequencing reduces repeated reorientation and cumulative intrinsic load.
Peer activities should include scaffolds: prompt cards, instructor-modeled exemplar discussions, and targeted question prompts. When peers are equally novice, such guidance prevents propagation of errors and leverages social learning for germane processing.
Rather than simultaneous teaching of methods, statistics, and writing, adopt a staged framework: conceptual research design first (with exemplars), then data collection protocols, then analysis, and finally reporting with template-driven worked examples. This reduces intrinsic complexity and increases retention.
Prospective teachers require explicit models of lesson planning and classroom management with stepwise practice and reflection. Videotaped worked examples of novice-to-expert lessons and scaffolded practicum tasks accelerate professional skill acquisition.
Embed self-explanation prompts and reflective checkpoints into assignments and assessments. These prompts increase germane load devoted to schema construction and are cost-effective. Training students in effective self-explanation strategies magnifies returns.
Adding hours without alignment to schema-building tasks yields diminishing returns. Reallocate time toward targeted, scaffolded practice and feedback loops that promote automation, rather than increasing contact time for low-value activities.
Misunderstandings about discovery learning and guidance can be corrected through concise, evidence-rich training that uses local examples. Emphasize the expertise reversal effect and provide actionable templates differentiating guidance levels for novices versus advanced learners.
Pilot evidence shows that replacing a small set of extraneous elements (slides, pre-training, worked examples) can steepen the learning curve. Prioritize interventions with high impact-to-cost ratios and scale iteratively, using formative measurement to guide expansion.
This study identified twenty interrelated mechanisms through which cognitive load dynamics and contextual constraints produce inefficient learning curves during pedagogical and research innovations in Cameroonian higher education. Key drivers include lack of pre-training, slide redundancy and split-attention, implementation of minimally guided learning without worked examples, bilingual instruction challenges, limited formative feedback in large classes, inadequate scaffolding and fading, misaligned assessments, fragmented curricula, and technology usability issues. Conversely, targeted CLT-aligned interventions—pre-training, integrated multimedia, worked examples and completion tasks, scaffolded fading, metacognitive prompts, and aligned assessments—can substantially accelerate the learning curve even under resource constraints.
Using parameter values informed by CLT meta-analytic findings (e.g., effect sizes for worked examples, redundancy reduction), simple simulations predict that implementing three prioritized interventions—pre-training, slide redesign to eliminate redundancy, and the addition of worked examples—can reduce time-to-proficiency for novices by an estimated 25–45% depending on starting expertise distribution and class size constraints. More specifically, in a simulation where baseline time-to-proficiency is normalized to 1.0, pre-training reduces it to 0.78, slide redesign to 0.85, and worked examples to 0.70; combined, multiplicative interactions predict a time-to-proficiency of approximately 0.55–0.65. These estimates are provisional and should be empirically validated, but they illustrate the potential for substantial gains from relatively modest design changes.
Pedagogical innovations hold substantial promise for transforming higher education in Cameroon. However, without explicit alignment to cognitive architecture—attending to working memory limits, expertise levels, and extraneous load—the intended benefits may be delayed or lost. This paper offers a pragmatic, evidence-based pathway for accelerating learning curves: diagnose cognitive load sources, prioritize high-leverage interventions that reduce extraneous demand and support germane processing, and iteratively measure outcomes with local pilots. The result should be more efficient learning trajectories, better retention and transfer, and greater return on educational investments.