Everyone gets a frog. Everyone has fun. Randomization happens at the feature layer — specifically, what social and narrative affordances each frog/steward gets access to in the app — not at the participation layer.
With 100 frogs, there's real statistical power here if conditions are cleanly assigned.
Assign conditions to frogs, not people. Each frog is a unit. When a steward scans and claims a frog, they inherit its condition. This is clean because:
Condition A: "Social Journey" frogs — stewards see a rich social feed: messages left by previous stewards, photos of the frog's journey, a "letter to the next steward" feature, ability to post updates visible to all past + future stewards of that frog.
Condition B: "Solo Journey" frogs — stewards see only the map, their own check-in log, and the frog's distance from Komoro. No social trace from other stewards. Functional but private.
Condition C (optional): "Spectator Social" frogs — stewards can see what previous stewards posted, but cannot contribute their own messages/photos. Read-only social feed. (Isolates consumption vs. contribution.)
What this tests: Does social accountability and narrative connection to other stewards increase stewardship quality (speed of handoff, distance moved toward Komoro, photo uploads, care behaviors)?
Reference: Gu et al. (2021) in Management Science ran exactly this kind of decomposition in a mobile app field experiment — they randomized users into full-crowdsourcing (submit + access content), submission-only, access-only, and control groups. Found that the full crowdsourcing condition increased user lifetime by 50% over control. Submission-only increased it 43%. This maps directly to your social feature tiers.
Reference: Chetty, Saez & Sandor (2014) tested social incentives vs. cash incentives vs. deadlines in a field experiment with journal referees. Social incentives (public posting of turnaround times) had persistent effects, especially on senior/tenured people — the ones who had the least extrinsic motivation. Analogous to your stewards who participate voluntarily.
Condition A: "Named + storied" frogs — each frog has a name, a personality blurb, a "dream" (e.g., "Kiko wants to see the sea before returning home"), and the app surfaces narrative prompts ("What did Kiko see today?").
Condition B: "Anonymous" frogs — frogs are identified only by number. Functional check-in, no narrative scaffolding.
What this tests: Does anthropomorphizing the object and adding narrative identity increase emotional attachment and stewardship effort? This connects to your "object totem" question without removing the frog from anyone.
Reference: The narrative transportation literature (Green & Brock, 2000) shows that immersion in a story changes attitudes and behavioral intentions. Giving the frog a story may create a sense of obligation via narrative empathy. Also connects to McAdams' work on narrative identity — stewards may integrate the frog's story into their own self-narrative.
Reference: The Tamagotchi effect / IKEA effect literature — people care more about things they've invested effort into and things that seem "alive." Nass & Moon (2000) showed people apply social rules to computers with minimal anthropomorphic cues. A named frog with a "dream" should trigger care responses.
Condition A: "Progress-rich" frogs — app shows a progress bar (distance to Komoro), milestone celebrations, how this frog ranks relative to others, estimated "hops" remaining.
Condition B: "Journey-focused" frogs — app emphasizes the journey itself: photos, places visited, cultural context about each location. No progress bar, no ranking. The frog is "exploring" rather than "racing home."
What this tests: Goal-gradient effect (Kivetz, Urminsky & Zheng, 2006) vs. intrinsic motivation. Does making the destination salient speed frogs up, or does it reduce the quality of the journey experience (rushed handoffs, less photo-taking)?
Reference: The goal-gradient hypothesis from behavioral economics — people accelerate effort as they approach a goal. But for Kaeru-chan, you may find the opposite: progress framing could make stewards feel urgency that undermines the contemplative/playful spirit. Testing this is genuinely novel.
Condition A: "Pen pal" frogs — when you hand off a frog, you and the next steward can exchange one message each. A brief, structured connection.
Condition B: "Thread" frogs — all stewards of a frog share a persistent group thread. Can post anytime, even after their turn.
Condition C: "Silent relay" — no direct steward-to-steward communication. Just the frog moves.
What this tests: Weak vs. strong social ties and their effect on relay speed, care, and community formation. Does persistent connection create a "frog family" that self-organizes?
Reference: Granovetter's strength of weak ties, but also Neustaedter, Tang & Judge (2013) on how geocaching communities form — they found the social experience (both online and physically colocated) was the primary motivator for continued participation, not the treasure-hunting itself.
Condition A: "Public steward" frogs — steward's name and photo appear on the public-facing map/timeline. Their journey is visible to anyone visiting the site.
Condition B: "Anonymous steward" frogs — steward contributes to the frog's journey data, but their identity is shown only as an icon or pseudonym on the public site. Full data captured backend.
What this tests: Does public recognition increase effort and pride, or does anonymity free people to be more playful and less performative? Privacy norms in Japan make this especially interesting.
Reference: Ariely, Bracha & Meier (2009) showed that extrinsic incentives in the form of public recognition can crowd out prosocial motivation — but the effect depends on the social context. In a Japanese cultural frame where modesty norms are strong, anonymous participation might actually increase effort.
This is the one that connects most directly to Akiya Collective's operational model and Michelle's open world quest work. The question: does giving someone an explicit identity as a "steward" — with a framed purpose, a named role, and a sense of mission — change their behavior compared to someone who simply... has a frog?
Condition A: "Called Steward" — Upon scanning the frog, the app delivers a full onboarding narrative. "You are now Kiko's Steward. Your mission: bring Kiko closer to home. Kiko is counting on you." The app uses the word "steward" throughout. The person gets a steward badge, a title ("Kiko's 7th Steward"), and framing that positions them as part of a lineage of caretakers. Responsibilities are made explicit: check in at least once, take a photo, move the frog forward, hand off to the next person.
Condition B: "Finder" — Upon scanning, the app says: "You found a frog! It's trying to get to Komoro. Tap here to log its location." No role language, no mission framing, no lineage. Purely transactional. The person can do everything the steward can do — the features are identical — but the narrative scaffolding is absent.
Condition C (stretch): "Chosen Steward" — Same as Condition A, but the app adds: "Kiko's previous steward [name] specifically chose you for this leg of the journey." This tests whether being selected by another person (social appointment) amplifies the stewardship effect beyond self-assignment.
What this tests: Whether role identity — the mere act of being named a "steward" — produces measurably different care behavior from having identical functional access without the label. This is the labeling effect applied to civic participation.
Reference: The classic identity labeling studies show that calling someone "a voter" (noun/identity) rather than asking them "to vote" (verb/action) increases turnout by ~15% (Bryan et al., 2011, PNAS). The mechanism is that identity labels activate self-consistency motivation — once you're "a steward," acting carelessly creates cognitive dissonance.
Reference: Thoits (2012) found that among volunteers supporting cardiac patients, the more prominent the volunteer role was to the person's self-identity, the better their mental and physical health outcomes — mediated by a greater sense of purpose and meaning. The stewardship label may function similarly: it doesn't just change behavior, it changes how people feel about the behavior.
Reference: Role theory (Biddle, 1986) argues that roles are the link between individual and society — having a named role integrates a person into a social structure and makes action meaningful. A person with no role is just passing through. A "steward" is participating in something.
Reference: The "responsibility for a group" experiment (ScienceDirect, 2020) found that making prosocial decisions on behalf of a group increased prosocial behavior — counter to the bystander effect. Being a steward "on behalf of Kaeru-chan" may activate this same mechanism.
Condition A: "Quest mode" — The app gives the steward a structured mission with clear objectives, sequenced steps, and completion criteria. "Step 1: Take a photo of Kiko at your current location. Step 2: Visit one landmark within 5km and check in. Step 3: Find a friend who can be Kiko's next steward. Step 4: Complete handoff." Progress bar shows completion.
Condition B: "Sandbox mode" — The app gives the steward full agency with no structure. "Kiko is with you now. Do whatever feels right." No steps, no progress bar, no suggested activities. Features are available (photo upload, check-in, handoff) but nothing is prompted.
Condition C: "Suggested but optional" — The app offers gentle nudges ("Some stewards like to show their frog a local landmark — want to try?") but frames everything as optional. Midpoint between quest and sandbox.
What this tests: The autonomy-structure tradeoff from Self-Determination Theory. SDT predicts that autonomy support increases intrinsic motivation, but research on serious games (Glasgow Caledonian, 2014) suggests SDT may need a 4th component — "purpose" — to predict engagement, and that structured missions can provide purpose without undermining autonomy. Rigby's "learner hero" framework (drawn from Campbell's Hero's Journey) argues that heroic narrative frames communicate that "the player's participation matters."
Why it matters for open world quests: If you're designing any place-based behavioral intervention, you need to know whether people do more (and feel better about it) when they're given a structured quest vs. left to self-direct. The answer probably isn't binary — it likely depends on personality type, familiarity with the context, and cultural norms. This experiment gives you real data on that spectrum in a Japanese cultural context.
Condition A: "You're part of a team" — The app shows the steward how many other stewards are active right now, a collective progress dashboard ("Together, our 100 frogs have traveled 3,247 km"), and frames each action as a contribution to a shared goal.
Condition B: "This is your journey" — The app focuses entirely on the individual steward-frog dyad. "You and Kiko." No mention of the larger network. The experience feels intimate and personal, not communal.
What this tests: Whether collective identity framing or individual ownership framing produces better outcomes. This has direct implications for how Akiya Collective frames membership — are you "part of a movement" or are you "the steward of this house"?
Reference: Cialdini's social proof research shows that knowing others are participating increases compliance. But the "personal responsibility" literature shows that diffusing responsibility across a group can reduce individual effort (bystander effect). Which wins in a voluntary stewardship context?
Condition A: "Time-pressured" — The app shows a gentle countdown: "Kiko has been with you for 3 days. Average stewardship is 5 days. Frogs that stay too long get homesick 🐸." Introduces soft social pressure without hard deadlines.
Condition B: "Take your time" — No temporal cues at all. The app emphasizes savoring the experience. "There's no rush. Kiko is happy wherever you are."
What this tests: Does time pressure speed up the relay at the expense of engagement quality? Or does open-ended time lead to stalled frogs? This maps directly to how you structure residency stays at Akiya properties and engagement windows for civic engagement interventions.
Condition A: "Legacy prompt" — After completing a handoff, the app asks: "Want to leave something for Kiko's next steward? A tip, a message, a photo of a good resting spot?" This creates a gift economy dynamic — each steward inherits and contributes to a growing bundle of care.
Condition B: "Clean handoff" — Handoff is purely logistical. Frog transfers, done.
What this tests: Whether prompting upstream generosity creates a cascade effect. If Steward #3 leaves a thoughtful message, does Steward #4 feel more obligated to do the same? This is the "pay it forward" hypothesis with measurable chain length.
Why it matters for Akiya Collective: This is the core mechanic of residency culture. If leaving something behind for the next person is built into the UX, does it become normative? Can you engineer a culture of care through interaction design?
The stewardship narrative conditions (#7) directly answer: how should we onboard new residents, members, and volunteers? If "Called Steward" frogs travel faster and are better cared for than "Finder" frogs, that's evidence that investing in narrative onboarding — giving people a title, a lineage, a mission — is worth the operational overhead. If there's no difference, you can simplify.
The mission structure conditions (#8) answer: should residencies be structured (here's your week, here are suggested activities) or open-ended (here's the space, do your thing)? The collective vs. individual framing (#9) answers: should marketing emphasize "join a community" or "this is your house"?
The temporal framing (#10) informs: how long should membership windows and residency stays be? And the reciprocity trigger (#11) tells you: does asking residents to leave something for the next person actually work?
Conditions #7, #8, and #9 together form a motivational design framework that you can apply to any place-based engagement system:
| Design Dimension | Spectrum | Kaeru-chan Tests |
|---|---|---|
| Role identity | Named role ↔ Anonymous participation | Steward vs. Finder (#7) |
| Mission structure | Guided quest ↔ Open sandbox | Quest vs. Sandbox (#8) |
| Social frame | Collective ↔ Individual | Team vs. Dyad (#9) |
| Temporal pressure | Urgent ↔ Timeless | Countdown vs. No rush (#10) |
| Legacy mechanics | Gift forward ↔ Clean exit | Pay-it-forward vs. Done (#11) |
This is a publishable framework. Each dimension can be tested independently or in factorial combination. The data from Kaeru-chan becomes the empirical backbone for a design theory of "spatial stewardship" — which is basically what Akiya Collective is building.
You mentioned not wanting to randomize "having a frog vs. not" — but you can still study the totem effect within the frog conditions. Compare behavioral intensity (photos taken, messages left, distance moved, handoff speed) between the "Named + Storied" frogs (#2) and the "Anonymous" frogs. If named frogs are cared for dramatically better, that's evidence that object attachment is driven by narrative identity, not just physical possession — which has implications for Meridian Objects and any ritual hardware you design.