# Interview Guide — Parent Mental Models of Progress

**Study**: Value Hub design probe — how parents want to understand their child's language progress, independent of any specific platform. The stimulus removes Novakid from the persona prompt entirely and probes parents' own mental models, desired signals, and imagined ideal experience.

**Method**: 8 personas (designed from the 6 archetypes + edge cases + cultural overlay), role-played by Claude Opus 4.6, each answering all 18 questions below. No Novakid-specific context is provided in the persona prompt. The persona does not know Novakid exists. The output is the *generative input* for design — not a source of empirical claims about the population.

**Critical disclosure (to appear on every derivative artefact)**:
> "These interviews are author-simulated to probe mental models. They are not real user research. No persona response counts as evidence about real parents. Every load-bearing design decision that rests on a persona insight carries a validation recommendation for real-user testing."

---

## Structure

The guide has five sections, each with a distinct purpose. Total: 18 questions. Runtime per persona: ~40 minutes of LLM transcript.

| Section | Purpose | Question count |
|---|---|---|
| A. Mental model of progress | How does this parent understand language learning in general? | 4 |
| B. Signals of evidence | What specific observations would convince them their child is progressing? | 4 |
| C. Imagined ideal hub | Given a blank canvas, what would they want to see, and when? | 4 |
| D. Tensions and trade-offs | How do they navigate conflicts (honesty vs reassurance, data vs story)? | 3 |
| E. Social and emotional frame | How does progress fit into their family and cultural context? | 3 |

No question mentions Novakid, dashboards-as-products, CEFR, teacher-platform interfaces, or any product-specific artefact. Every question is *about the child, the parent, and the abstraction of progress*.

---

## Preamble (read to the persona before Section A)

*"Thank you for joining us. I'm researching how parents think about their children's language learning — the way they notice progress, the evidence they rely on, the moments that feel meaningful. There are no right or wrong answers. I'm not testing you, and I'm not promoting a product. I want to understand how you think."*

*"Think about your own child — [persona details: name, age, how long they've been learning English, family context]. Answer from your own experience, and feel free to go into detail."*

---

## Section A — Mental model of progress

*Purpose*: surface how the parent understands language acquisition before any platform, tool, or metric is mentioned. Gets the parent's own conceptual map into view.

**A1.** When you think about your child's English learning journey, how do you picture it in your head? If you had to describe it as a shape, a metaphor, or a scene — what would it look like?

*[Probe: a ladder? a winding road? a garden growing? something else? Why that?]*

**A2.** Not all weeks feel the same. Some weeks you feel confident your child is progressing; some weeks you're not sure. What separates those two kinds of weeks for you? What changes in your own mind?

*[Probe: what evidence arrived, or didn't? what conversations happened, or didn't? what did you notice, or not notice?]*

**A3.** Imagine someone in your family — a parent, a partner, a sibling — asked you right now, "Is your child actually getting better at English?" How would you answer, and what would you point to as proof?

*[Probe: if they said "prove it," what would you show? a grade? a moment? a video? a feeling?]*

**A4.** Language teachers use a framework called CEFR — Pre-A1, A1, A2, B1, and so on. Without looking it up, how meaningful is this framework to you personally? Does knowing your child is "A2" help you understand them better, or does it feel abstract?

*[Probe: what do you wish you could replace it with, or add to it, so it spoke your language?]*

---

## Section B — Signals of evidence

*Purpose*: elicit the observable signals the parent already trusts. These are the proof-types the Value Hub must surface or connect to.

**B1.** Think of a specific moment in the last few months where you felt, in your gut, that your child had made real progress in English. Walk me through it. Where were you? What happened? What did your child do?

*[Probe: why that moment? why not a lesson score? what made this one count?]*

**B2.** Now think of the opposite — a moment when you worried your child wasn't progressing. What triggered that worry? Was it something that happened, or something that *didn't* happen?

*[Probe: what would have reassured you in that moment? what information would have helped?]*

**B3.** If you had to pick *three* concrete things — things you could see, hear, count, or observe — that would tell you your child is on a good trajectory, what would they be? Not abstract ("confidence") but concrete ("orders food in English at a restaurant").

*[Probe: are those three things the same across months? do they change as the child gets older?]*

**B4.** Parents sometimes build their own ways to track learning — a notebook, a spreadsheet, phone recordings, conversations at dinner. Do you have anything like that? If not, have you thought about it? What would make you start?

*[Probe: why is the parent DIY-tracking? what gap is it filling? what would make it unnecessary?]*

---

## Section C — Imagined ideal hub

*Purpose*: blank-canvas generative design probe. The persona isn't critiquing an existing product; they're inventing what they want.

**C1.** Imagine you had a magic notification that could tell you *one thing* every week about your child's English learning. One message, once a week. What would you want it to say? What would make you actually open it, instead of swiping it away?

*[Probe: what's in the message? who's the sender? what tone? what length? what's the first sentence?]*

**C2.** Now imagine you had a place — a screen, a page, a hub — where you could go to understand your child's learning in more depth. You'd visit it sometimes, not every day. When you opened it, what's the very first thing you'd want to see? Not a menu — the first piece of content.

*[Probe: is it a sentence, a number, a picture, a person? why that?]*

**C3.** On that same hub, if you scrolled down or clicked deeper, what would you want to find? What do you want to know that the first screen doesn't answer?

*[Probe: is it historical data? is it "what's coming next"? is it the teacher's voice? is it comparisons? what's the progression of depth?]*

**C4.** If this hub could *ask* you something back — a question it posed to you every so often to help it help your child better — what kind of question would you want to be asked? And how often?

*[Probe: what would feel helpful vs intrusive? what would you actually answer? what would you ignore?]*

---

## Section D — Tensions and trade-offs

*Purpose*: surface how this parent navigates the competing design tensions the Value Hub must balance.

**D1.** Some parents want to know the honest truth, even if it's uncomfortable. Others prefer a warmer framing that emphasises what's going well. Where do you fall on that spectrum — and does it change depending on what you're being told? For example, would you want honesty about a plateau? honesty about pronunciation struggles? honesty about whether the investment is paying off?

*[Probe: when does honesty help? when does it hurt? where's your line?]*

**D2.** Imagine two hypothetical dashboards for tracking your child's English. One shows you detailed data: words learned per week, charts, skill breakdowns. The other shows you a single sentence: "Mia is developing question-formation; expect more confident short answers this month." Which feels more useful to you — and why? And would your answer change for different kinds of questions you're trying to answer?

*[Probe: when do numbers help? when do stories help? can both coexist?]*

**D3.** Parents sometimes describe a tension between wanting to be involved in their child's learning and not wanting to become the homework parent. How does that tension feel for you? What's the right amount of involvement — for you specifically, not in some ideal abstract sense? And what would make that level of involvement sustainable?

*[Probe: what kinds of involvement feel meaningful? what kinds feel like work? what would you refuse to do?]*

---

## Section E — Social and emotional frame

*Purpose*: surface how this parent's progress-tracking behaviour fits into their family and culture — which determines which proof-types resonate and which social proofs matter.

**E1.** Describe a moment — real or imagined — where you felt genuinely proud of your child's English progress. What would you want to share with your family or friends? Who would you tell, and how? And would you tell them through words, a video, a screenshot, something else?

*[Probe: who is this performance of progress for? the parent themselves? the spouse? grandparents? peer network?]*

**E2.** As a parent investing in your child's English learning — your time, your attention, and probably money — what worries you most? What's the anxiety underneath the decisions you're making? And what would help reduce that worry, specifically and concretely?

*[Probe: is the worry about outcome, about cost, about parenting identity, about something else? what would relieve it?]*

**E3.** If your child's English learning went well for a year — really well — what would that look like in your family, your culture, your community? What would have changed? What would you tell yourself a year from now, looking back?

*[Probe: what's success, locally defined? how does this fit into the parent's life narrative about the child? what are the culturally-specific markers of "we made it"?]*

---

## Interview analysis framework

For each transcript, code along five dimensions. This is different from the old 45-persona analysis, which coded for pattern frequency ("40 of 45 said X"). Here, each persona is a *design probe* — what they say generates design hypotheses, but the evidence base for those hypotheses is the real-quote corpus + theory, not the persona responses themselves.

| Dimension | What to look for |
|---|---|
| 1. Active archetype | Which of the 6 jobs does this persona's responses most clearly activate? Do they shift jobs across the interview? |
| 2. Proof-type preference | Which proof-types do they cite most — external (school, olympiads, family), real-world (travel, home), structural (CEFR, exams), relational (teacher bond, peer comparison)? |
| 3. Mental-model metaphor | What shape/scene/metaphor did they pick for A1? Ladder? Garden? Journey? Building? This reveals how they think about progress. |
| 4. Tension resolution | For D1–D3, where did they land? Honesty or warmth? Numbers or stories? Involved or bounded? |
| 5. Cultural signal | For E1–E3, what cultural proof-types did they use? Social-comparative (best in class)? Individual-expressive (confidence)? Family-performative (make grandma proud)? |

The synthesis document (`synthesis.md`) will combine:
- The real quote corpus (primary empirical evidence)
- Theoretical frameworks (primary analytical lens)
- The persona transcripts (generative design input, *explicitly labelled as such*)

into a single narrative structured around the Information Bridge axis. Personas feed design hypotheses. The corpus + theory validate or discipline those hypotheses.

---

## Persona selection (6 primary + 2 edge, culturally distributed)

Selected to activate different archetypes as primary job, cover the edge cases identified in open coding, and span the seven market clusters from `cultural-analysis.md`.

| # | Name (working) | Primary job | Secondary job | Market | Edge case |
|---|---|---|---|---|---|
| 1 | Elena (Warsaw) | Navigator | Auditor | PL | — |
| 2 | Umut (Istanbul) | Reassurance-Seeker | Co-Pilot | TR | Father, not mother |
| 3 | Giulia (Milan) | Investor | Auditor | IT | Price-sensitive |
| 4 | Anastasia (Moscow) | Reassurance-Seeker | Outsourcer (shifts) | RU | — |
| 5 | Hans (Munich) | Co-Pilot | Navigator | DE | Non-native-English-speaking parent |
| 6 | Pilar (Madrid) | Co-Pilot | Reassurance-Seeker | ES | Child overcoming shyness |
| 7 | Ahmed (Riyadh) | Outsourcer | Investor | MENA | Long-term loyal, minimal engagement |
| 8 | Sophie (Lyon) | Navigator | Reassurance-Seeker | FR | Two children, different levels |

Each persona will be given a full profile (age, profession, child age, family structure, language-learning history of the parent themselves, relevant cultural context) before the interview, but **no Novakid-specific content**. The interview prompt explicitly instructs the LLM: *"You do not know any specific English-learning platform by name. You are thinking about your child's English learning in the abstract."*

---

## What this guide is NOT

- Not a poll. Responses are not counted.
- Not validation. Personas are not evidence about real parents.
- Not a product review. The stimulus excludes Novakid entirely.
- Not a replacement for real-user research. It is a scaffolding for design thinking that must be validated before production.

The correct posture toward the outputs of this guide: *"Here is what 8 thoughtful simulated parents with diverse contexts said they'd want if they could design this from scratch. The patterns in their answers generate testable hypotheses. The validation for those hypotheses lives in the quote corpus, the theory, and — before production — real parent interviews."*