Snacks After Class: choosing a useful activity with honest expectations

Research date: September 19, 2026. EAG research report, first edition.

Prepared for Education Angel Group with AI-assisted research and writing. EAG has an interest in developing the entity discussed here. This is a focused evidence review and proposed product rationale, not an independent evaluation. No child interviews, destination testing or Snacks After Class outcome study was conducted.

1. Proposition and audience

Snacks After Class is EAG's proposed student-facing umbrella for practical-skill games and creative worlds. Its immediate job is to help a student understand an activity, decide whether it fits, and reach the correct experience with an accurate expectation of what happens first. The name does not describe a food program. The defensible proposition is informed orientation among distinct activities; learning benefits must be established for the activities themselves. [A0]

The umbrella has several audiences with different information needs. An exploring student may know the kind of thing they want to do without knowing the family names. A returning student may simply want to reach an activity again. A supporting adult needs enough information to judge requirements and suitability. The person maintaining the catalog needs evidence that its descriptions and destinations remain correct. A design that makes exploration pleasant but leaves returning visitors stranded has only partly addressed the proposition.

The EAG product definition gives the umbrella ownership of orientation, availability information, expectations and handoff. Each child experience retains its own behavior, data, permissions and identity. Snack's Tube, Create After Class, Sim, Trade, Launch, Game and individual experiences remain distinct. Orchard is a separate portfolio entity. The EAG product definition does not establish a final taxonomy or certify that a first destination is available today. This report preserves those unresolved inputs instead of manufacturing a launch catalog. [A0]

EAG's proposed contribution should therefore be judged by the quality of a decision, not the number of cards viewed. A student who accurately recognizes that an activity needs equipment they do not have, then leaves without disclosing personal information, has made a useful decision. A student who clicks an attractive illustration expecting a playable world and encounters a description of a future idea has encountered a failure even if analytics record a successful outbound click.

The scope is deliberately small enough to test: one understandable choice, one verified destination, one first action and a reliable way back. There is no research basis here for a universal account, a shared game engine or a single mascot identity across the family. Those would introduce separate questions without first proving that the orientation service is useful.

2. Questions the report can answer

The central question is how children can choose a suitable creative or playful activity without unnecessary choice burden or misleading educational promises. Five narrower questions make that question researchable.

First, what must a student know before entering an activity? A topic label alone may not distinguish playing, making, watching or experimenting. A useful description must explain what the student actually does. Second, does reducing the number of visible options reliably improve children's decisions? The available choice experiment challenges any universal answer. Third, which educational claims are justified when games show different results for practiced tasks and transfer? Fourth, what alternatives already serve topic discovery or creative learning? Finally, what would demonstrate that an EAG umbrella adds enough value to justify its additional step?

These questions require different evidence. Experiments can inform the limits of claims about choice or learning. Accessibility guidance can inform how destination purpose is expressed. First-party catalogs can show that a discovery approach exists. EAG product definition defines ownership and intended behavior. None of those sources is a substitute for observing the proposed umbrella with its intended users.

The report uses these differences to constrain recommendations. It does not estimate the prevalence of unmet demand, determine an optimal age range, forecast adoption or assign a financial value to successful routing. The precise audience ages and first child destination remain unresolved. Findings from toddlers, preschoolers and middle-school students cannot be collapsed into a single claim about what all children need.

3. Need and context: the cost of an unclear choice

The need considered here is a plausible product problem, not a measured population deficit. A family of activities can require a visitor to translate unfamiliar names into intended actions, compare requirements and remember where to return. The umbrella could reduce that work if it provides information the child sites do not make easy to compare. It could also increase work by adding another page, introducing new categories or requiring a student to answer questions before seeing useful options.

Choice burden has several possible components. There is the number of alternatives, the effort needed to understand each one, uncertainty about the desired outcome, and the cost of correcting an unsuitable selection. These are analytically different. A short list of obscure labels could be harder to use than a longer list of familiar activities. A student who knows a destination might prefer direct access over a guided chooser. EAG should treat these as competing design hypotheses, not assume a large catalog is inherently harmful.

Castelo, Kim and Carlson's two sticker-choice studies included 106 and 52 toddlers in a Midwestern U.S. metropolitan area. They did not detect choice overload under their operational definition, with arrays extending to 26 and 53 options. The samples skewed toward affluent, highly educated families. Choosing between a sticker array and a singleton differs materially from choosing a game with reading, time and device requirements. The result does not measure every form of confusion, regret or decision cost. S1: Study 1, Study 2 and Discussion

The EAG inference is limited: a maximum of three or six destinations would be a design choice to test, not a scientific rule supplied by this study. The EAG product definition's rejection of a wall of every idea remains valid product authority, but its rationale should be understandable orientation and truthful availability, rather than an invented developmental threshold.

Another possible burden is expectation repair. A student may be willing to learn a new control scheme when that is part of making something, but disappointed by the same work when expecting a quick game. The content may be sound while the invitation is misleading. For this entity, a description of the first action is more operationally useful than a general promise to develop creativity or problem-solving. Whether students understand that description remains an empirical question.

4. Existing alternatives and the proposed gap

These equivalents accompany the reader figures. Source IDs refer to the report references. A0 identifies EAG's product definition reviewed for the original report, not an external study or proof of availability. Other source IDs resolve to this report's reference and access record.

Figure 1. Existing routes serve different jobs

Existing alternatives and a proposed direct-link comparator; no ranking of educational quality.

PBS KIDS | S6: Topic-based game discovery. Math catalog with named activities and an adult route. HTML structure only; no game runtime or efficacy audit.

Scratch Foundation | S7: Creative project support. Family guidance for meaningful projects and getting unstuck. Provider approach, not independent effect evidence.

Direct destination | EAG comparison: Known-activity access. A direct link may suffice for a returning student. The proposed umbrella must justify its added step.

Sources: S6, S7, A0.

Figure 1. Source-described alternatives and an EAG comparison hypothesis. These are different jobs, not a ranking of learning effectiveness. Text and data alternative.

PBS KIDS' math-game page groups named activities under a mathematics heading and offers Games, Videos and a Grown-Ups route. The retrieved HTML establishes this catalog structure. It does not establish that every game ran successfully, that a particular device is supported, or that catalog use improves mathematics. S6: Math Games page

Scratch Foundation's family guide encourages meaningful creative projects, tinkering and support when a learner becomes stuck. It also points to activity resources. This is a first-party account of its approach, not independent evidence of broad cognitive gains. The editor, accounts and community were not tested for this report. S7: Creative Learning Guide for Families

These alternatives remove any justification based on the supposed absence of topic discovery or creative support. They also reveal a useful distinction: choosing a game to play and choosing a project to make are not interchangeable invitations. The umbrella could help a student notice that distinction across its own family. It should not suggest that a familiar provider's research, reputation or characters endorse an EAG destination.

A direct link is another serious comparator. If a returning student already knows the activity, a bookmark, teacher-provided link or child site's own entry page may do the job with fewer steps. This is a proposed comparison, not an audit of a particular classroom practice. An umbrella that mainly intercepts known destinations may add little value. Its strongest potential use is a visitor who has a goal but cannot yet identify the right family member, or whose previously known destination has changed.

Offline creative activity and adult guidance also belong in the decision space. EAG need not turn every interest into a digital session. A student may want to sketch a scene, build with available materials or ask someone how a task works. This report does not estimate the relative effectiveness of those alternatives. Their relevance is that declining a digital path can be a sensible choice, and a useful service should not interpret every departure as lost engagement.

The proposed gap is therefore specific: a maintained, comprehensible route among the approved EAG family that explains differences and current requirements. If the family contains only one usable destination, a simple explanation and direct link may be sufficient. More categories would need to earn their place through actual ambiguity that they resolve.

5. Evidence and counterevidence: activity is not transfer

Figure 2. Learning findings depend on the task and package

Qualitative source findings across different interventions and ages; no pooled effect or product prediction.

Touchscreen study | S2: Learning and transfer varied. Small preschool comparison; watching sometimes supported transfer. Age and content matter.

Fraction trial | S3: Practice did not establish wider transfer. Abstract-only evidence distinguishes the trained task from untrained fraction outcomes.

Science-media package | S4: Specific positive outcomes; broader uncertainty. Supported bundle improved structure and friction measures. Broader outcomes were non-significant; no isolated portal effect.

Sources: S2, S3, S4.

Figure 2. Qualitative evidence comparison across different ages, content and interventions. No pooled effect or predicted EAG gain is calculated. Text and data alternative.

Schroeder and Kirkorian studied 44 preschoolers, ages 3 to 5.5, who each played one touchscreen game and watched another. Outcomes separated direct learning, near transfer and transfer to real objects. Results varied with age and content; for older children's growth task, far-transfer improvement appeared in the watching condition. Character familiarity was not associated with measured learning. Small samples, brief exposure and floor or ceiling effects limit generalization. This is not a verdict against all interactive learning or characters. S2: Participants, Results and Discussion

Nuraydin, Stricker and Schneider report a randomized trial with 188 fifth- through eighth-graders in two fraction-game conditions and an active control. The abstract reports that younger grades improved on the trained number-line task without gains on untrained fraction measures or arithmetic; older grades showed no intervention effects. Full methods and effect tables were unavailable, so no numerical effect is charted. The intervention used feedback without direct instruction. S3: original abstract

Positive counterevidence also matters. Grindal and colleagues randomly assigned 454 children aged four and five from low-income households to a Cat in the Hat package or an alternative treatment for eight weeks. The package included videos, games and offline activities. Specific structure/stability and friction/movement outcomes improved; broader science and material-property estimates were not statistically significant at conventional levels. Volunteer recruitment, supplied tablets, preloaded content, parent reminders and adapted measures limit generalization. The bundle cannot isolate a game or discovery page; usage associations were exploratory. S4: executive summary, pp. 26-27 and 37

Taken together, these studies justify neither blanket enthusiasm nor blanket dismissal. The following interpretation is EAG's synthesis: the educational unit of analysis should be the specific activity, learner group, support conditions and outcome. A family label such as practical skills cannot inherit every positive result from digital learning research. Equally, a null result for one fraction game does not show that a creative world cannot be valuable.

For catalog language, the distinction has immediate consequences. Describing what a child will practice is different from claiming what they will learn. Describing a trained task is different from promising general mathematical understanding. A successful transfer test is different from evidence of retention months later. Enjoyment can be a legitimate outcome in its own right, but it should be named as enjoyment rather than presented as proof of academic improvement.

There is also an attribution problem between umbrella and child. If students learn after using a child experience, the result could reflect that experience, teacher support, prior knowledge or selection into use. The umbrella's contribution would require a comparison that changes the orientation route while holding the destination opportunity sufficiently comparable. Until such evidence exists, the appropriate claim is that the umbrella aims to help students choose and enter an activity accurately.

6. Proposed mechanism: informed entry with a usable return

Figure 3. Orientation and learning have separate evidence

EAG conceptual mechanism; no inevitable causal progression or assumed private data transfer.

Proposed umbrella contribution. Identify a desired activity → Understand first action and requirements → Choose a verified destination or decline → Enter and return with orientation intact. Hypothesis: fewer mismatches and less recovery effort. Not measured.

Separate child-level question. Use the actual activity → Assess the practiced skill → Test untrained tasks and later retention. Entry and engagement alone do not establish learning or transfer.

Sources: A0.

Figure 3. EAG conceptual mechanism, not an observed causal sequence. The student's decision can end in declining an activity. Text and data alternative.

The proposed mechanism begins with an intention expressed in ordinary activity language. The student compares a small amount of relevant information, predicts the first action, checks requirements and decides whether to continue. After entry, the child experience owns the activity. The umbrella should remain understandable when the student returns, changes their mind or discovers a mismatch. This is a product hypothesis drawn from EAG product definition, not a validated sequence. [A0]

Each step can fail independently. A label can be understandable while its availability is stale. A destination can be live while its first action differs from the description. A game can be well designed while the return route loses the student's context. Separating these failures prevents an aggregate completion number from concealing which part of the promise was broken.

The predicted benefit is less avoidable mismatch and recovery effort. It is not necessarily a faster decision for every student. Someone may reasonably spend longer comparing activities because the information exposes meaningful differences. Someone else may use a direct destination link immediately. A useful evaluation should ask whether the decision was informed and whether the route respected the student's intention, rather than rewarding speed without regard to understanding.

The mechanism also depends on maintenance. An operator must know which child owns the destination and what evidence supports the current description. A functioning link alone cannot verify that an activity remains suitable for the described audience or that it still opens to the same first action. When that evidence is missing, an honest unknown or unavailable state is part of the service. Hiding uncertainty would undermine the very expectation-setting job the umbrella is meant to perform.

Return continuity should preserve orientation without silently accumulating a profile. Remembering the page context needed to go back is different from collecting interests or progress across children. The EAG product definition permits anonymous orientation and retains child-specific permissions when sign-in is required. No cross-child transfer of private information is assumed. [A0]

7. Design concepts that follow from the evidence

Figure 4. A destination invitation must answer four questions

EAG proposed invitation questions follow its product definition and link-purpose guidance; live facts remain unverified.

Activity: What will I actually do first? Use a readable job and first action. An icon reinforces recognition; it is not proof of gameplay.

Availability: Can I do that there now? A working link is insufficient. Retain child ownership and an honest unknown or unavailable state.

Requirements: What will I need? Verify account, input, device and adult-support requirements where relevant before making claims.

Recovery: How do I return or choose again? Preserve orientation after a mismatch or interruption. A no-fit exit is a valid outcome.

Sources: A0, S5.

Figure 4. Proposed information needed for a truthful destination invitation. These are research-informed questions, not verified live product facts. Text and data alternative.

A useful activity description could pair an ordinary verb with a concrete first action. For example, a future description might distinguish arranging elements to create something from controlling a character to complete a challenge. These examples illustrate information structure; they do not name approved destinations or imply either activity currently exists. The exact taxonomy should emerge from the real child inventory and user comprehension, not from an attractive set of category names.

W3C's explanation of Link Purpose in Context supports making link purpose available through its text or programmatically associated context. It also describes the usefulness of consistent destination naming and accessible image-link treatment. This is guidance for one accessibility criterion, not full conformance certification or a learning intervention. S5: Intent, Benefits and Examples

The EAG design inference is that an icon should reinforce a readable destination and activity description. It should not carry the entire decision or look like a working game when it is only a link. The eventual visual treatment can be playful while the handoff remains literal about what will happen. This report authorizes no character assignment, final icon or shared visual world.

Requirements should appear before they become a surprise. Relevant facts might include whether the destination needs an account, sound, a particular input method or adult support. Each fact requires child-owned verification. An unknown requirement must not be converted into a reassuring default. The invitation should also distinguish being technically reachable from being available for the intended action.

A chooser could allow an undecided student to compare activities while retaining direct access for someone who already knows the destination. This is a candidate pattern, not a prescribed interface. A guided question flow may help some users, but it can also impose a reading burden or suggest that EAG is assessing personal ability. No sensitive profile is necessary merely to describe options. The first study should compare simple alternatives before adding recommendation machinery.

Recovery deserves the same clarity as entry. A withdrawn destination should explain the situation in student-readable terms and offer a return or alternative without repeated relaunching. If there is no suitable activity, the service should allow an uncomplicated exit. Neither case needs urgency, scarcity cues or a demand to create an account to continue browsing.

8. Equity, accessibility and operating burden

The source set does not establish that one navigation pattern serves every age, language or access need. A student can recognize a picture while misunderstanding its activity. Another may understand the words but be unable to operate a destination with the available device. A third may need adult help that is not available at that moment. These are plausible constraints to test, not prevalence estimates.

Equity therefore begins with accurate requirements and genuinely usable alternatives. The umbrella should not make students disclose school, location or private interests to learn whether an activity fits. It should avoid treating a need for help, a slower decision or an opt-out as a deficit. Supporting adults need concise information, but their route should not make the student's own choice invisible.

Technical access has two layers. A lightweight orientation page can work while a child game fails to load, consumes too much data or requires an unavailable input. Conversely, an accessible child can remain difficult to discover through an inaccessible chooser. Later verification should distinguish these layers with real devices, keyboard and assistive-technology use, reduced motion and interrupted sessions. No such testing was conducted here. A text link is a useful starting point, not proof that the complete journey is accessible.

The positive science-media study should not be used to erase access constraints. Its supported exposure differs from a student independently finding an activity on an existing device. The EAG implication is to test ordinary entry conditions rather than assume that evidence from a supplied-device study predicts everyday reach. The same restraint applies to age: evidence from preschool learning does not settle whether the proposed family labels make sense to older students.

Maintenance is a continuing cost rather than a one-time catalog task. Someone must reconcile changed URLs, unavailable destinations, account requirements and descriptions that no longer match the first action. If that work cannot be sustained, a smaller verified directory may be more trustworthy than a broad stale one. The report does not estimate staffing or prescribe a refresh interval; those decisions require the actual rate of child-site change and the consequences of an incorrect claim.

A final risk is educational overstatement through aggregation. A collection may feel more authoritative than each individual entry. Broad banners about future-ready skills could make weakly supported child claims seem established by the family. The umbrella should retain the specificity and uncertainty of each activity's evidence. Shared navigation must not become shared proof of effectiveness.

9. Evaluation and criteria for changing direction

The first evaluation should ask whether students can predict and control the journey. With appropriate consent and a defined audience, a bounded study could present a realistic intention, let participants choose or decline, and ask them to explain the first action and requirements before entry. Then it could observe the actual destination and return. No study is authorized or represented as completed by this report.

The primary outcomes would be accurate expectation, suitable entry or informed rejection, and recovery from a mismatch. A wrong prediction is informative even when the student completes the click. An unavailable destination recognized before entry differs from one discovered after several steps. Assistance needed from an adult should be recorded as part of the interaction cost, not silently supplied and ignored.

Compare the umbrella against an ordinary direct-link list and the relevant child entry pages. For topic exploration, a first-party catalog can inform comparison tasks without implying identical content or audiences. For creative work, the comparison should preserve the difference between selecting a ready-made game and beginning an open-ended project. Recruiting only enthusiastic existing EAG users would limit conclusions about unfamiliar visitors.

Before measurement, specify which findings would change the design. If students understand activity descriptions but consistently confuse family names, revise naming or grouping. If known-destination users take unnecessary steps, preserve a shorter route. If requirements cause repeated failed entries, repair the information and child availability process. If the umbrella adds no useful understanding over direct links, simplify it. These are proposed decision rules; no arbitrary percentage threshold is disguised as an evidence-based standard.

Learning requires a separate study at the child-experience level. Define the practiced skill, measure relevant prior knowledge, select an appropriate comparison and distinguish performance within the activity from untrained tasks and later retention. If the umbrella itself is claimed to improve learning, its added contribution also needs a suitable comparison. A dashboard combining visits and game scores cannot establish that causal claim.

Evaluation should also include the people least well served by the initial design. Report differences in comprehension, assistance and successful recovery across the intended ages and access conditions where sample sizes permit responsible interpretation. Do not infer subgroup success from a favorable overall average. Keep data collection proportionate: these questions can often be investigated without creating permanent cross-child profiles.

10. Methods, limitations, references and corrections

This report uses a focused, question-led source review conducted on September 19, 2026. Search avenues covered child choice, game-learning transfer, topic catalogs, creative-learning alternatives and accessible link purpose. Original research and institutional pages were preferred. This is not a systematic review, comprehensive market inventory or current destination audit. No Arkansas-specific discovery study was established, and no national study is presented as Arkansas prevalence.

The full choice and touchscreen articles, original EDC evaluator report, W3C explanation and two provider pages were reopened. The EDC PDF initially failed but became accessible through its official landing-page download. The fraction article remained abstract-only: direct PubMed extraction failed, while a targeted search returned the complete original abstract and metadata. The DOI did not provide full text. Claims from that source are limited accordingly.

Publication and retrieval dates are separate. The EDC report is dated November 2019 despite its 2024 upload path. The fraction article appeared online January 22, 2022 and in the May 2022 journal volume. The touchscreen study collected data from October 2014 through April 2015. Exact collection dates were not established for the toddler-choice, fraction and science-package studies. Undated product pages are described as retrieved in September 2026, not newly published research.

A correction boundary concerns statistical interpretation: the EDC report's discussion on printed page 27 describes a p-value as a probability that a finding is an error. That wording is not adopted here. Non-significant estimates remain inconclusive; they are not proof of no possible benefit. No comprehensive correction or retraction registry review was performed. Future corrections should identify the affected claim and figure and preserve the difference between source correction and a changed EAG proposal.

The remaining gaps are substantial but specific: intended age range, final taxonomy, first verified child, ordinary-device access, destination maintenance, comparative usefulness and any educational effect. The evidence supports testing a truthful orientation service. It does not establish that Snacks After Class improves learning, that its children are publicly available, or that the proposed family is the best option for a particular student.

References:

• A0. Education Angel Group, Snacks After Class product definition, reviewed September 19, 2026 for this report. Defines intended scope, not implementation or effectiveness.

• S1. Castelo, Kim and Carlson (December 14, 2023), More is more: toddlers do not show choice overload. Full original article.

• S2. Schroeder and Kirkorian (September 13, 2016), When Seeing Is Better than Doing. Full original article.

• S3. Nuraydin, Stricker and Schneider (2022), No transfer effect of a fraction number line game. Original abstract and metadata only.

• S4. Grindal and colleagues (November 2019), Early Science and Engineering. Full evaluator report, EDC and SRI International.

• S5. W3C WAI, Understanding Link Purpose in Context. Guidance retrieved September 19, 2026.

• S6. PBS KIDS, Math Games. First-party catalog HTML retrieved September 19, 2026.

• S7. Scratch Foundation, Creative Learning Guide for Families. First-party resource page retrieved September 19, 2026.

Figure text and data alternatives

Complete semantic alternatives accompany each figure in this reader and its text download. Source IDs refer to the references in this report. A0 identifies EAG's product definition reviewed for the original report, not an external study or proof of availability. Other source IDs resolve to this report's reference and access record.

Corrections and entity context

Readers identifying an error can contact EAG with the section, disputed statement and supporting source. Snacks After Class provides the public entity context, not an external citation for EAG's product claims.