Runstr: the case for participation, supportive relationships and school resources

Prepared by Education Angel Group (EAG), with AI-assisted research and drafting. Research date: September 19, 2026. EAG is developing Runstr and has an interest in its success. This report examines the rationale for the proposed product; it is not an independent evaluation of Runstr.

1. What Runstr is, who it serves and what remains proposed

Runstr is intended to connect participation, profiles, paid supporter relationships, points, student rewards and school purchases. Its four profile types are Student, Teacher, Team and School. Participants would share Check-ins, and permitted supporters could follow and respond. Text remains a complete participation option alongside optional voice. These are EAG's product definitions, not findings that the service is operating successfully or improving outcomes.

The school-facing purpose is to help schools obtain supplies, equipment, programs and student experiences through monthly profile memberships. Memberships are not donations. They do not constitute unrestricted personal cash, cash-like points or a guaranteed allocation to a school. A school-directed resource and a student's personal prize have different beneficiaries and approval requirements. Staff personal benefit and automatic institutional acceptance are outside the intended model.

The economic rules and some reward-entitlement decisions remain unresolved. This report therefore cannot state a membership price, school allocation, prize value, conversion rate or return on investment. Its nearest-term product example is narrower: a participant creates a recoverable Check-in in an authorized profile and a permitted supporter can understand it. That interaction would demonstrate neither a functioning membership economy nor successful resource delivery.

Overview. Evidence supports taking school resource needs, family communication and unequal access seriously. It does not establish that combining them in Runstr will work. Some communication interventions have improved selected school outcomes; other outcomes were unchanged. Incentives can affect behavior, but reward type, context and the outcome measured matter. Runstr's case rests on whether a paid relationship can provide sufficient value to supporters, remain voluntary and useful to participants, and produce accepted school resources after costs. Each part needs its own evidence.

2. The questions this report answers

Seven questions organize the review:

1. What resource need is documented nationally and in Arkansas, and what remains unknown locally?

2. What can communication improve, and who bears the work of producing and responding to it?

3. Could optional voice make participation easier, without excluding people who need or prefer text?

4. What do existing alternatives already provide, including recurring support?

5. Who can participate, and could differences in access or supporter networks widen inequalities?

6. When might rewards support participation, and when might they reduce interest or create pressure?

7. Who may authorize access, accept resources and determine personal reward eligibility?

Confidence differs across these questions. Official surveys provide useful descriptions of spending and access for their measured years. Randomized studies provide stronger evidence about the particular interventions they tested. Frameworks help identify desirable relationships without estimating effects. Vendor documentation establishes described capabilities, not comparative effectiveness.

The weakest link is the combined Runstr proposition. None of the retained sources tests a student-authored Check-in connected to this paid supporter, points, prize and school-resource model. That gap matters more than whether any one ingredient has favorable research. A successful communication channel could still have poor economics. A successful membership could still impose unreasonable work or distribute benefits unevenly.

3. Need and context: resources are local, while the available numbers are broad

Teacher spending is one sign that classroom provision can depend on personal resources. The revised NCES report found that, among public-school teachers who spent unreimbursed personal money on classroom supplies in 2014-15, the mean was $478 and the median $297. The report was revised in March 2021 to correct sample weights. These historical, self-reported amounts are not a current national shortfall or an estimate of what teachers would request through Runstr. Maura Spiegelman, S1

More recent public expenditure totals answer a different question. Census reports current spending of $17,619 per pupil in fiscal year 2024, compared with $16,526 in 2023. These are nominal public-school expenditure figures, not inflation-adjusted measures of adequacy; current spending also differs from capital outlay. Rising aggregate spending cannot tell a reader whether a particular classroom has the materials it needs. U.S. Census Bureau, S2

Arkansas's March 2026 resource-allocation study provides a closer institutional context. Its instructional-material expenditure series reports $165.8 million from all funds in fiscal year 2023 and $156.9 million in 2025, including $70.7 million from foundation funds in 2025. Superintendent responses describe additional needs across resource categories. The analysis also identifies limitations in mapping accounting codes to the funding matrix and separating commingled funds. These figures describe spending, not a verified list of unfilled requests. Arkansas Bureau of Legislative Research, S3

These text equivalents accompany the report figures. Source IDs refer to the report references.

Figure 1. Three measures, three different questions

Resource context uses separate units and denominators. Historical teacher spending, national per-pupil expenditure and Arkansas materials spending do not estimate a Runstr-addressable shortfall.

Teacher spending | 2014-15: $478 mean; $297 median. Unreimbursed supply spending among U.S. public-school teachers who spent personal money. Historical self-report, revised weights in March 2021. Source: S1.

National current spending | FY2024: $17,619 per pupil. U.S. public elementary-secondary systems. Nominal dollars; excludes capital outlay. This is expenditure, not school-level adequacy. Source: S2.

Arkansas instructional materials | FY2025: $156.9 million, all funds. Recorded expenditure across Arkansas districts and open-enrollment charters. Accounting and commingled-fund limits apply. Not an unmet-needs inventory. Source: S3.

Sources: S1, S2, S3.

Figure 1. Resource context uses separate units and denominators. Historical teacher spending, national per-pupil expenditure and Arkansas materials spending do not estimate a Runstr-addressable shortfall. Text and data alternative.

The practical inference is to start with a school-approved need, not a national market-size calculation. A resource may be desirable but ineligible for purchase, already funded, difficult to maintain or unacceptable under school policy. A useful request would specify the item or experience, intended use, receiving institution and acceptance process. The research reviewed here does not supply that local inventory.

Digital access also shapes who could use the proposed service. In the NCES table derived from the 2023 American Community Survey, 92.4% of household-resident children ages 3-18 had internet with a computer, 5.4% had smartphone-only access and 2.3% lacked access. Independently rounded values total 100.1%. Smartphone-only access was 11.4% in the lowest family-income quarter and 1.0% in the highest; no access was 4.3% and 0.9%, respectively. The table does not measure dependable bandwidth, a private recording space or individual device ownership. National Center for Education Statistics; Census ACS, S7

Figure 2. Household access differs by income

U.S. household-resident children ages 3-18, 2023 ACS; lowest and highest family-income quarters. Point estimates only, without uncertainty intervals. These figures are not Arkansas-specific and do not measure bandwidth, privacy or individual device ownership.

Access category	Lowest income quarter	Highest income quarter
Smartphone-only access	11.4%	1.0%
No internet access	4.3%	0.9%

U.S. children ages 3-18 living in households, separately within each family-income quarter; group quarters excluded

Source reports standard errors; no error bars or inferential comparison shown. Standard errors are not transcribed into this point-estimate graphic.

None; published percentages, unchanged. Zero baseline; shared 0-12 percent scale.

Sources: S7.

Figure 2. U.S. household-resident children ages 3-18, 2023 ACS; lowest and highest family-income quarters. Point estimates only, without uncertainty intervals. These figures are not Arkansas-specific and do not measure bandwidth, privacy or individual device ownership. Text and data alternative.

Consequently, a polished desktop experience or a high national connection rate would be insufficient evidence of practical access. Arkansas-specific needs, household circumstances and school preferences remain unmeasured in this report. No interviews or first-person accounts have been collected for it.

4. Existing alternatives and the limits of a claimed gap

Runstr would enter a setting where schools and families can already communicate. Conversations, calls, tailored messages and school information-system alerts are meaningful alternatives. A school may improve the relevance or timing of an existing message without introducing another profile, content channel or paid relationship. The family-school partnership framework discussed below supports taking these comparators seriously. Karen L. Mapp and Paul J. Kuttner, S4

ParentSquare's own documentation describes SMS, email, app and voice delivery, two-way translation and student information-system integration. This establishes that channel choice and digital school-family communication are existing capabilities. It does not prove that every family is reached effectively or that the service improves educational outcomes. Runstr should not explain its purpose by suggesting schools lack communication products. ParentSquare, S14

Recurring support also exists. DonorsChoose documents monthly giving, including support directed toward projects, teachers or schools, cancellation and handling of unused credits. That is a charitable comparator, whereas Runstr proposes paid memberships. The comparison establishes an existing recurring-support workflow; it supplies no estimate of Runstr demand, renewal or net school resources. DonorsChoose Help Center, S15

The relevant choice depends on the problem a school or family actually has:

This comparison does not presume that separate tools are always preferable. Integration could reduce repeated effort, but it could also add account management, moderation and unfamiliar financial rules. The case for combining functions should come from observed user benefit. It cannot be inferred merely because the functions can share a profile.

5. What the evidence supports, contradicts and leaves uncertain

Communication and supportive relationships

Mapp and Kuttner's 2013 dual-capacity framework emphasizes the capabilities and trusting relationships that families and educators need for partnership. It offers a way to ask whether participation is reciprocal and useful. It is a conceptual framework with illustrative cases, not a randomized test of a platform or proof that more frequent updates create stronger relationships. Karen L. Mapp and Paul J. Kuttner, S4

Bergman and Chan’s January 2019 working paper reports a randomized academic-alert intervention in 2015-16 across 22 middle and high schools in one West Virginia district. Alerts reduced course failures and improved class attendance, but state test scores did not improve and middle-school effects were not significantly positive. Messages drew on existing academic information. A participant-authored Check-in asks a different question: whether selected personal context helps a recipient understand and support the participant. This report consistently uses the 2019 paper, rather than mixing it with the later journal version. Peter Bergman and Eric W. Chan, S5

Together, these sources justify investigating timely, understandable communication. They do not establish an optimal posting frequency, an improvement in relationship quality from Check-ins, or a reason to attach payment to access. More content could be redundant or unwanted. A recipient's ability to read, understand and respond matters as much as whether a message was sent.

CDC's analysis of the 2023 Youth Risk Behavior Survey associates school connectedness with protective patterns among high-school students. It is cross-sectional, relies on self-report and does not test a product intervention. It cannot support a claim that Runstr improves mental health. Nor does it justify inferring a child's mental state from a recording, transcript or missed update. Jorge V. Verlenden and colleagues, S6

Voice, text and the risk of unequal effort

Optional voice could help some people express context with less typing. That is an untested Runstr hypothesis. The retained evidence does not include a comparative child voice-versus-text trial establishing its benefit, and adult speech-system research cannot fill that gap.

Koenecke and colleagues' 2020 audit found average word error rates of 0.35 for Black speakers and 0.19 for White speakers across five studied commercial systems. The study used selected adult interview corpora and historical systems. It demonstrates that transcription errors can be unevenly distributed, not how any current Runstr provider performs with children, accents or languages. Allison Koenecke and nine coauthors, S8

For participants, the relevant cost includes having to correct a transcript or abandon an inaccurate recording. W3C guidance for prerecorded audio-only content calls for an equivalent alternative, subject to its stated exception for audio that is itself an alternative to text. Automatically producing words is not evidence that an equivalent, understandable alternative exists. W3C Web Accessibility Initiative, S9

Rewards: different questions produce different answers

The reward literature should not be flattened into either “rewards work” or “rewards harm.” Deci, Koestner and Ryan's 1999 synthesis of 128 experiments found that expected tangible rewards and several contingent reward arrangements reduced free-choice intrinsic motivation on average. Positive feedback showed a different pattern. Many tasks were brief, and the outcomes concerned interest and freely chosen activity after reward conditions, rather than school resource delivery. Edward L. Deci, S10

Cerasoli, Nicklin and Ford's 2014 meta-analysis examined performance across school, work and physical domains. Intrinsic motivation and incentives jointly predicted performance; intrinsic motivation explained more unique variance in quality, while incentives explained more in quantity. Mixed populations and correlational predictors limit causal interpretation. This does not negate the earlier review: producing more while incentives exist and choosing an activity afterward are different outcomes. Christopher P. Cerasoli, S11

There is also directly relevant adverse evidence from schools. Robinson and colleagues' 2019 accepted author paper reports a January-February 2016 randomized attendance-award experiment with 15,329 analyzed students in grades 6-12 across 14 California districts. Prospective awards did not improve attendance on average; retrospective awards increased subsequent absence. Eligibility and household exclusions constrain generalization. Symbolic attendance certificates differ from chosen physical prizes, but the result warns that recognition can communicate an unintended message. Carly D. Robinson, S12

A 2024 gamification review by Li, Hew and Du found a small positive average intrinsic-motivation effect, with substantial heterogeneity. Most included studies were in higher education, and none measured free-choice behavior. Bundled game elements are not equivalent to redeemable prizes connected to paid support. The review gives a reason to investigate particular designs, not to promise that adding points will improve motivation. Liuyufeng Li, S13

Figure 3. Reward studies measure different outcomes

Qualitative comparison of different populations, designs and outcomes. Findings cannot be combined into a quantitative Runstr forecast. Publication years are not necessarily intervention years.

Deci et al. (1999) | S10: Post-reward free-choice interest. 128 experiments, children and college students. Expected tangible and several contingent rewards reduced this outcome on average; positive feedback differed. Transfer limit: many brief tasks, not resource delivery.

Cerasoli et al. (2014) | S11: Performance quality and quantity. Meta-analysis across school, work and physical domains. Motivation explained more unique variance in quality; incentives more in quantity. Transfer limit: mixed populations and correlational predictors.

Robinson et al. (2019) | S12: Attendance after symbolic awards. Randomized field study, grades 6-12, 14 California districts; January-February 2016 intervention. Prospective awards did not improve average attendance; retrospective awards increased later absence. Transfer limit: eligibility exclusions; certificates differ from chosen prizes.

Li et al. (2024) | S13: Self-reported learning motivation. Systematic review and meta-analysis, mainly higher education. Small positive average effect with substantial heterogeneity. Transfer limit: no free-choice measures; bundled game elements do not isolate physical prizes.

Sources: S10, S11, S12, S13.

Figure 3. Qualitative comparison of different populations, designs and outcomes. Findings cannot be combined into a quantitative Runstr forecast. Publication years are not necessarily intervention years. Text and data alternative.

For Runstr, the unresolved question is whether participants find the activity meaningful and voluntary during and after any reward arrangement. Post counts alone would miss coerced participation, superficial updates and declining interest once rewards stop.

6. The product rationale, with each untested step made explicit

The following is EAG's interpretation of the evidence, not a tested causal chain.

A resource rationale. Historical personal spending and current institutional expenditure reports justify asking schools about specific needs. They do not establish the amount Runstr can meet. The proposed mechanism is a permitted membership relationship that contributes to a school purchase after applicable costs and approvals. Its missing evidence includes supporter demand, renewal, transaction and fulfillment costs, administration and institutional acceptance. Gross receipts would not be an adequate measure of resources delivered.

A relationship rationale. Communication research suggests that relevant information can matter in defined settings. Runstr proposes participant-authored Check-ins as a way to share context beyond an academic alert. The untested step is whether that context matters to the authorized recipient enough to sustain a useful relationship. Paying for a membership, reading a post and offering constructive support are distinct behaviors. None is a substitute measure for the others.

A participation rationale. Choice between text and voice could let people use a form that suits them. The mechanism fails if voice requires extensive correction, text is treated as second-class participation, or producing updates becomes a recurring obligation. A shorter composition time would be helpful only if the resulting content remains understandable and the participant retains control over the audience.

A motivation rationale. Points and prizes might make participation more salient. The evidence also supports concern about pressure and activity undertaken mainly to obtain a reward. A reasonable hypothesis is that voluntary, meaningful participation can coexist with bounded rewards, but the review cannot tell EAG which design achieves that. The personal-prize pathway must remain distinct from school-directed resources; a school's need must not become a reason to pressure a student to post.

An economic rationale. A recurring paid relationship could support planning if supporters renew and the permitted resource route is workable. Predictability itself remains unproven. Runstr needs to explain what the membership provides, how cancellation works and what resources it can and cannot promise. It must not borrow donation language or assumed tax treatment from charitable comparators.

Figure 4. Proposed paths require separate evidence

EAG conceptual model, not an evaluated causal chain or approved economic rule. Payment does not grant audience access. Institutional resources never pass through a student personal-prize balance; personal-prize eligibility remains unresolved.

Communication hypothesis. Participant chooses text or voice → Authorized Check-in and permitted audience → Recipient understands useful context. Unknown: usefulness, effort, voluntary participation and relationship value.

School-resource hypothesis. Paid membership → Costs and permitted allocation → Institutional acceptance → School-directed resource. Unknown: demand, renewal, total costs, allocation and accepted delivery.

Separate personal-prize question. Personal-prize eligibility remains unresolved → Separate approval and fulfillment rules. No automatic entitlement or economic conversion is established. This is not the school-resource route.

Sources: EAG product definition.

Figure 4. EAG conceptual model, not an evaluated causal chain or approved economic rule. Payment does not grant audience access. Institutional resources never pass through a student personal-prize balance; personal-prize eligibility remains unresolved. Text and data alternative.

The combined proposition is stronger only if these functions help one another in practice. If supporters value school updates but not student updates, a school-centered approach may be more appropriate. If communication is useful but memberships are unattractive, the communication finding remains valid while the economic model needs revision. If prizes drive volume without usefulness, increasing reward expenditure would not solve the underlying problem.

7. Design implications and illustrative concepts

These examples describe possible experiences, not implemented functionality or approved economic rules.

A participant might write two sentences about a team activity, save a draft and later choose to share it with the permitted audience. Another might record the same update, review the transcript and correct a name before sharing. Both should be able to understand who can receive it and recover work after an interrupted connection. Neither example requires a public audience or a paid relationship to prove the basic communication mechanism.

A permitted supporter might receive an update that explains what happened and offers an appropriate way to reply. A useful response could acknowledge the activity or ask a relevant question. Whether recipients actually find this worthwhile must be observed, rather than inferred from the presence of a reply button.

Separately, a school might identify an eligible resource and designate who can accept it. The resource's status should distinguish a request, an authorized purchase and receipt by the institution. A student's personal prize would require its own eligibility and fulfillment explanation. A single progress display should not imply that these different obligations are interchangeable.

The design priorities follow from the evidence and these distinctions: a complete text route, understandable audience controls, correctable audio alternatives, recoverable drafts and clear resource status. Low-capability phone and assistive-technology checks are needed to determine whether those priorities are achieved. These recommendations are design inferences from the access, transcription and accessibility evidence, not proof of conformance or improved participation. National Center for Education Statistics; Census ACS, S7 Allison Koenecke and nine coauthors, S8 W3C Web Accessibility Initiative, S9

8. Burden, equity, privacy and unintended consequences

The work of sustaining the service needs attention before scale. A brief Check-in can still involve deciding what to say, reviewing a transcript, checking an audience and handling replies. Educators may acquire moderation, support and purchasing duties.

The academic-alert experiment's low marginal messaging expense depended on existing information infrastructure; it does not estimate the full cost of Runstr content production or supervision. Peter Bergman and Eric W. Chan, S5

Supporter networks introduce another uncertainty. Profiles with larger or wealthier networks might attract more paid support. That is a distributional hypothesis, not an observed Runstr outcome. Aggregate revenue could rise while benefits remain concentrated. Evaluation should therefore ask which participants and schools receive resources, who does the associated work and who cannot readily recruit supporters.

Device access, language, disability and household privacy could also change the cost of participation. A person with a shared phone or no quiet place to record should not have to disclose those circumstances to use text. Declining to post should not automatically be treated as lack of motivation. These are reasons to provide choice and investigate barriers directly, rather than diagnose people from their activity.

Membership and familiarity do not establish permission to access a student's information. Institutional approval, guardian permission where required, account authorization and a chosen audience are separate decisions. FTC guidance limits a school's ability to consent on a parent's behalf to the educational context, without another commercial purpose. It also addresses collecting only information reasonably necessary for an activity. How that applies to a particular supporter service requires assessment of its actual data uses. This report makes no legal-compliance determination. Federal Trade Commission, S16

The Department of Education provides school-facing privacy and online-service review resources. This review inspected its resource index, not every linked guidance document, and cannot claim a full FERPA appraisal. District purchasing, records, acceptance and consent requirements remain a separate local inquiry. U.S. Department of Education, S17

Finally, a system intended to encourage participation could make students feel responsible for keeping supporters satisfied or obtaining resources for adults. The possibility warrants direct, age-appropriate questions about pressure. It should not be dismissed because participation is described as optional in product copy.

9. What to evaluate and what would change the decision

The first evaluation should answer whether the basic interaction is useful and understandable. Can participants complete a Check-in, recover from interruption and correctly explain its audience? Can permitted recipients understand it? Do participants prefer text, voice or neither? Record completion and correction effort alongside voluntary accounts of usefulness. Any future study involving participants requires its own consent and research arrangements; none has occurred for this report.

Later questions require separate measures:

These are proposed decision criteria, not completed results or numerical success thresholds. A later evaluation should choose comparators that match the problem: existing communication for relationship value, a simpler support route for resource delivery, and participation without prizes for reward effects. It should examine people who decline or leave, rather than describing only active users.

A successful early interaction would justify further investigation of that interaction. It would not establish learning gains, mental-health improvement or sustainable economics. If the combined model performs poorly, the appropriate response may be to remove or separate a function rather than add more incentives or more messaging.

10. Methods, limitations, references and corrections

This is a focused, question-led review for a proposed U.S. K-12 service with Arkansas context. It is not a systematic review or meta-analysis. Research assembled on September 19, 2026 prioritized original empirical papers, original research syntheses, official statistics and guidance, and first-party descriptions of alternatives. Full retrieved sources supported retained empirical findings; inaccessible leads and search snippets did not support numerical claims.

Evidence was interpreted according to population, geography, design, measured period and outcome. Descriptive need, causal findings, conceptual frameworks, service capabilities and EAG product definitions were kept separate. Correction checks were bounded. The absence of a visible correction notice is not proof that none exists. Historical figures were not treated as current prevalence, and effect sizes from different outcomes were not pooled into a Runstr forecast.

The review has no original interviews, new participant dataset, Runstr impact study or independent expert review. It does not establish current local unmet demand, district approval, willingness to pay, access preferences or safe and effective reward arrangements. AI assistance supported search, appraisal and drafting; EAG remains responsible for the interpretation and its disclosed product interest.

Correction history: September 19, 2026, first report revision. It uses the revised NCES spending source, the January 2019 Bergman-Chan working paper and the 2019 accepted Robinson author copy consistently. No prior public edition exists. Readers identifying an error can contact EAG with the section, disputed statement and supporting source.

References

All web sources below were retrieved September 19, 2026. Undated service and guidance pages are retrieval-date snapshots. Dates for measured data are distinguished from publication dates in the report.

• S1. Maura Spiegelman, NCES. Public School Teacher Spending on Classroom Supplies. 2021-03 revised.

• S2. U.S. Census Bureau. Public School Spending Per Pupil Reaches Historic High in 2024. 2026-05-07; revised 2026-05-12.

• S3. Arkansas Bureau of Legislative Research. 2026 Educational Adequacy Study: Resource Allocation, Part I. 2026-03-10.

• S4. Karen L. Mapp and Paul J. Kuttner. Partners in Education: A Dual Capacity-Building Framework for Family-School Partnerships. 2013.

• S5. Peter Bergman and Eric W. Chan. Leveraging Parents through Low-Cost Technology: The Impact of High-Frequency Information on Student Achievement. 2019-01 working paper; journal version 2021.

• S6. Jorge V. Verlenden and colleagues, CDC. Mental Health and Suicide Risk Among High School Students and Protective Factors: YRBS 2023. 2024-10-10.

• S7. National Center for Education Statistics; Census ACS. Digest of Education Statistics, Table 702.12. 2024-11 table prepared.

• S8. Allison Koenecke and nine coauthors. Racial disparities in automated speech recognition. 2020-03-23 online.

• S9. W3C Web Accessibility Initiative. Understanding SC 1.2.1: Audio-only and Video-only (Prerecorded). Undated; retrieved September 19, 2026.

• S10. Edward L. Deci, Richard Koestner and Richard M. Ryan. A Meta-Analytic Review of Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation. 1999-11.

• S11. Christopher P. Cerasoli, Jessica M. Nicklin and Michael T. Ford. Intrinsic Motivation and Extrinsic Incentives Jointly Predict Performance: A 40-Year Meta-Analysis. 2014-02-03 online.

• S12. Carly D. Robinson, Jana Gallus, Monica G. Lee and Todd Rogers. The Demotivating Effect (and Unintended Message) of Awards. 2019 accepted author copy; accepted 2019-03-18.

• S13. Liuyufeng Li, Khe Foon Hew and Jiahui Du. Gamification enhances student intrinsic motivation, perceptions of autonomy and relatedness, but minimal impact on competency: a meta-analysis and systematic review. 2024-01-16.

• S14. ParentSquare. ParentSquare Frequently Asked Questions. Undated; retrieved September 19, 2026.

• S15. DonorsChoose Help Center. How can I manage or cancel a monthly donation?. Undated; retrieved September 19, 2026.

• S16. Federal Trade Commission. Complying with COPPA: Frequently Asked Questions. Undated; retrieved September 19, 2026.

• S17. U.S. Department of Education, Student Privacy Policy Office. Privacy and Education Technology. Undated; retrieved September 19, 2026.

Product-definition source: Education Angel Group, Runstr product definition reviewed September 19, 2026. This definition establishes intended scope only; it is not empirical evidence or proof of availability.