# Theoretical Frameworks — April 2026 **Purpose**: give the Value Hub research a theoretical backbone beyond review amplification. Each framework is sourced, summarised, and explicitly mapped to the Novakid design problem. These frameworks **anchor** the analysis — themes from review coding are interpreted *through* them, not instead of them. --- ## 1. Self-Determination Theory (SDT) — Deci & Ryan **Source**: Ryan & Deci 2000 ([SDT documents](https://selfdeterminationtheory.org/SDT/documents/2000_RyanDeci_SDT.pdf)); Niemiec & Ryan 2009 ([Autonomy, competence, and relatedness in the classroom](https://selfdeterminationtheory.org/SDT/documents/2009_NiemiecRyanTRE.pdf)); [Frédéric Guay 2022, Canadian Psychology](https://journals.sagepub.com/doi/10.1177/08295735211055355). ### Core claim Human motivation and well-being rest on three innate psychological needs: - **Autonomy** — the need to experience behaviour as volitional, self-endorsed. - **Competence** — the need to feel effective in one's interactions. - **Relatedness** — the need to feel connected to others who matter. Satisfying these needs produces intrinsic motivation and well-being. Thwarting them produces controlled motivation, anxiety, withdrawal. ### Application to parents (not just the child learner) Parents have the same three needs *about their role as parents* of a language-learning child: - **Parent autonomy**: I can make informed decisions about my child's learning — pace, teacher, approach — without feeling at the mercy of an opaque system. - **Parent competence**: I understand my child's progress well enough to feel effective as a parent — I can answer "is it working?" confidently. - **Parent relatedness**: I feel connected to the teacher, the platform, and a community of similar parents. ### Novakid-specific implications - Auto-renewal that can't be easily cancelled **violates parent autonomy**. - Dashboard that shows data without explaining what it means **thwarts parent competence**. - One-directional teacher feedback (parent cannot reply) **blocks parent relatedness**. - Teacher change without consultation **violates all three at once** — no agency (autonomy), can't evaluate new teacher (competence), connection broken (relatedness). ### Design lens derived The Value Hub should systematically support parent autonomy (clear controls), competence (interpretable progress), and relatedness (bi-directional communication, community). --- ## 2. Expectancy-Value Theory (EVT) — Eccles model **Source**: [Eccles' expectancy-value model, ASCN Higher Ed summary](https://ascnhighered.org/ASCN/change_theories/collection/evt.html); [Expectancy-value theory in education, ERIC](https://files.eric.ed.gov/fulltext/ED630343.pdf); [Wikipedia: Expectancy-value theory](https://en.wikipedia.org/wiki/Expectancy-value_theory). ### Core claim Achievement-related choices (e.g., to renew a subscription, to continue effort) are jointly driven by: - **Expectancy** — "How likely is it that this will work?" - **Subjective task value**, decomposed into: - *Attainment value*: how important is the outcome to who I am as a parent? - *Utility value*: how useful is it for other goals (school, emigration, career)? - *Intrinsic value*: how enjoyable is the process for my child? - *Cost*: what do I pay (time, money, stress) to pursue it? ### Parent-level cascade (Eccles) Parents' beliefs → parents' behaviours → children's motivational beliefs → children's behaviours. Parents' expectancy and value beliefs **transmit** to the child. ### Application to renewal decision A parent's decision to renew Novakid at month 6 is not "is my child improving?" — it's: **`Renew = Expectancy(improvement continues) × [Attainment(being a good parent) + Utility(future payoff) + Intrinsic(child enjoys it) − Cost(money, stress, logistics)]`** If any term collapses, the product breaks, even if the others are fine. ### Why the current dashboard fails EVT - It can't prove **expectancy** (no evidence of trajectory, only of activity). - It can't defend **attainment** (I'm being a good parent by investing here). - It ignores **utility** (no link to school English, exam alignment, future mobility). - It rarely captures **intrinsic** signal (the child's own enthusiasm is invisible to the dashboard). - It amplifies **cost** through billing opacity. ### Design lens derived The Value Hub must provide evidence for **each EVT term separately**, not just an overall "progress" number. A parent renewing needs to be able to answer all four questions in ≤30 seconds. --- ## 3. Jobs-To-Be-Done (JTBD) — Christensen school **Source**: [Jobs-to-be-Done Radio: Education Industry](https://jobstobedone.org/radio/jobs-to-be-done-in-the-education-industry/); Christensen, *Competing Against Luck*, 2016. ### Core claim (Bob Moesta on education) > "Kids want to make progress. Kids want to get through school to get to college or to get into a job. Progress — not money — is the currency students spend." Consumption blocks — not motivation deficits — explain dropout. Students who drop out aren't lazy; they've decided progress is easier *elsewhere*. ### Applied to Novakid parents — candidate jobs Functional jobs the Value Hub might be "hired" for: - **J1**: *Help me believe the money is working* — investment justification. - **J2**: *Help me see my child's effort without nagging them* — passive observation. - **J3**: *Help me decide whether to renew, pause, or cancel* — judgment support. - **J4**: *Help me contribute without becoming the homework manager* — bounded participation. - **J5**: *Help me explain my child's progress to family* — storytelling tool. - **J6**: *Help me catch problems early enough to fix them* — anomaly detection. - **J7**: *Help me align the platform with school / exam / life goals* — external integration. Emotional jobs: - **E1**: *Reduce my anxiety about whether I'm doing the right thing as a parent.* - **E2**: *Protect my pride when asked by relatives if the money is worth it.* - **E3**: *Feel connected to my child's learning journey, not excluded from it.* Social jobs: - **S1**: *Signal to family/peers that I'm investing in my child's future.* - **S2**: *Give my child pride in their progress (shareable milestones).* ### Design lens derived Each Value Hub surface should be testable against: *which job is this hired for?* If a screen does not clearly serve one of these jobs, it's decoration. If a job has no screen, the product has a gap. --- ## 4. Parental Educational Anxiety — Guo et al. (2024, 2025); Wu et al. (2022) **Source**: [Frontiers in Psychology, Guo et al. 2024](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2024.1380363/full); [Sage, Guo et al. 2025, PAES scale](https://journals.sagepub.com/doi/10.1177/07342829251335362); [Wu et al. 2022, Frontiers](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2021.764824/full). ### Core claim Parental educational anxiety is a measurable, multi-dimensional construct with three sub-dimensions: - **Learning & development anxiety** (grades, admission, future career) - **Physical & mental health anxiety** (bullying, stress, wellbeing) - **Educational resource anxiety** (school quality, teachers, tutoring) Measured by the Parental Educational Anxiety Scale (12 items; Cronbach α = 0.874). ### Transmission mechanism Parent anxiety → parenting style → child anxiety. Specifically: - **Overprotective parenting** transmits 49% of the effect. - **Rejection** transmits 15%. - **Emotional warmth** *reduces* transmission by 36%. Correlation between parent and child anxiety: *r* = 0.301 (significant). ### Tutoring as intervention - **Academic tutoring** has a small "placebo effect" on parent anxiety (β = −0.033) — it relieves anxiety even if modest effect on child. - **Arts/sports tutoring** shows a larger direct reduction in child learning anxiety (β = −0.058). ### Implications for Novakid Novakid partially functions as a **parental-anxiety-relief product**, not only a language product. This is why: - Parents pay even when progress is unclear ("at least I'm doing something"). - Guilt about unused lessons (Lingokids campaign) is intense — unused lessons mean *failed anxiety relief*, not just wasted money. - Dashboards that amplify uncertainty (unclear metrics, unexplained plateaus) are *actively harmful* to the product's anxiety-relief function. ### Design lens derived Every dashboard decision should be tested against: *does this increase or decrease parent anxiety?* A "more information" dashboard can decrease competence anxiety but increase judgement anxiety. Design must balance these — typically by pairing data with interpretation. --- ## 5. Dashboard usability for non-experts — applied research **Sources**: [UXPin dashboard design principles 2025](https://www.uxpin.com/studio/blog/dashboard-design-principles/); [Pencil & Paper UX pattern analysis](https://www.pencilandpaper.io/articles/ux-pattern-analysis-data-dashboards); systematic review of 75 studies summarised in industry literature (information overload = #1 issue, 46.7% of users); F-pattern eye-tracking research (Nielsen Norman). ### Core empirical findings - **Information overload** is the most prevalent dashboard failure mode (46.7% of users cite it). - Users scan dashboards in **F-pattern** (top-left → top-right → down-left): the primary KPI must live in the top-left, largest font, highest contrast. - Non-expert users require **interpretation layers**, not just data: raw numbers without context create anxiety rather than clarity. - Dashboards "viewed weekly or less" (per parent-psychology research) are **read, not interacted with** — static labelled views outperform heavy interactivity. ### The paradox for parents Parents check dashboards rarely (≤ weekly per existing research), and bring low domain expertise (they aren't language teachers). This is the *worst-case* combination for traditional dashboards: - Rare check = low familiarity with layout → every visit is a relearning event - Low expertise = unable to interpret raw metrics → need interpretation, not data ### Design lens derived - **Top-left headline must be an interpretation, not a metric**. "Your child is on track toward A2" beats "Lessons: 14, Skills: 23, Vocab: 156". - **Every metric needs a benchmark** (vs. typical pace, vs. last month, vs. plateau-is-normal). - **The dashboard's first job is answering "how are we doing?" in one glance** — before any drill-down. --- ## 6. Parent mental model of CEFR — Dinolingo framing **Source**: [Dinolingo: CEFR for Parents](https://dinolingo.com/cefr-for-parents-how-to-understand-your-childs-english-progress/); British Council plateau research (existing). ### Core claim CEFR is a teacher-professional framework. When parents encounter it, they lack: - **Structured benchmarks** for what each level means observationally. - **Developmental sequencing** (comprehension before production). - **Vocabulary** to discuss plateau phases. Parents typically hold **binary mental models** ("improving vs not improving") and rely on **home observation** ("is my child speaking more?"), which are poor matches for CEFR's ordinal, plateau-heavy nature. ### Observable-behaviour mapping (Dinolingo's translation) | CEFR level | Observable parent signal | |---|---| | Pre-A1 / A1 | Understands short songs/videos in English | | A1–A2 | Responds to simple questions in English | | A2–B1 | Uses complete sentences about past events | | B1+ | Spontaneous self-expression, idiomatic phrasing | ### Design lens derived The dashboard must **translate CEFR (teacher-framework) into observable parent-visible behaviours**. Don't display "B1.2" — display "Your child can now tell a story about last weekend; expect this to get richer over the next month." --- ## 7. Baumrind's parenting style typology (for persona design, not archetype axis) **Source**: Baumrind 1966, 1971; meta-analyses including Larzelere et al. 2013. ### Four styles (modern reading) - **Authoritative** (high warmth + high demand): the design-aligned parent — wants both structured tracking and emotional connection with child's progress. - **Authoritarian** (low warmth + high demand): wants control, clear metrics, willing to override child's preferences. - **Permissive** (high warmth + low demand): wants child happiness; suspicious of pressure-based dashboards. - **Uninvolved** (low warmth + low demand): outsources entirely; dashboard used rarely, only in crisis. ### Use in this research **Not** a classification axis for archetypes. Used as a **design dimension** — the dashboard should serve authoritative and authoritarian parents differently (more depth vs more structure) without alienating permissive parents (not shaming non-engagement). --- ## 8. Behavioural economics — relevant biases ### Endowed progress effect (Nunes & Drèze 2006) Progress visibly underway motivates continuation; progress invisible motivates quitting. Loyalty card studies show: a 10-stamp card where 2 are pre-stamped ("you've already started!") outperforms an 8-stamp card with 0 stamps, despite identical remaining effort. **Novakid implication**: *visible* progress indicators — even on the same underlying trajectory — will materially affect churn. ### Sunk-cost bias, flipped Traditional sunk-cost bias predicts people over-commit to failing investments. In subscription contexts, the opposite often holds: parents *cancel* when they can't see value *because* they're afraid of throwing more money at something invisible. "Sunk cost" only works when value is visible. **Novakid implication**: invisible progress doesn't create retention — it creates pre-emptive exit. ### Loss aversion (Kahneman & Tversky) Framing the cancellation decision as "lose progress" vs "save money" has measurable different outcomes. Parents respond more to loss framing. **Novakid implication**: renewal UX should emphasise what the child loses by stopping, not what the parent saves. --- ## 9. British Council intermediate plateau research **Source**: [British Council — Getting past the B2 plateau](https://www.teachingenglish.org.uk/professional-development/teachers/understanding-learners/getting-past-b2-plateau); widely replicated in SLA literature. ### Core claim All language learners hit a plateau where measurable production stops growing even as comprehension, nuance, and automaticity continue. This is a known, documented phase of language acquisition — **not** a product failure. ### Parent implications Parents report considering cancellation during plateau phases (existing 40-of-45 signal + real quotes). Zero parents report being **told** plateau is normal. ### Design lens derived The dashboard has a specific retention obligation: **name the plateau, teach its existence before it's encountered**. The moment a parent first sees flat data is the last moment to intervene — a pre-registered "plateau will likely start around month 4, and here's why it's normal" message turns a churn trigger into a credibility signal. --- ## How these frameworks combine (preview) The Value Hub problem is **not** "design a better dashboard". It is: *design a system that simultaneously supports parent autonomy/competence/relatedness (SDT), provides evidence for all four value terms (EVT), serves 7 distinct jobs (JTBD), reduces rather than amplifies parental anxiety (PEAS), respects dashboard-usability constraints of rare-use low-expertise users, translates teacher-frameworks into parent-observable behaviours, and proactively names known stuck points (plateau).* That is the actual scope the design must navigate. The expanded review corpus tells us *where* parents struggle; these frameworks tell us *why* and *what structural interventions matter*.