How do you assess the digital equity gaps in a 1:1 device program before expansion in 2027?

The outcome you should expect

A serious equity assessment does not end with a percentage of students who "have a device." It ends with a segmented map of who can actually do the assigned work at home, on what quality of connection, with what adult support, and how fast a broken machine gets replaced. That map is what makes an expansion decision defensible.

Expect the assessment to change your expansion plan, not just validate it. Districts that do this honestly usually discover three uncomfortable things. First, the device-to-student ratio is healthier than the *working*-device-to-student ratio, because a meaningful share of the fleet is out of warranty, battery-dead, or sitting in a repair queue. Second, home connectivity is far more bimodal than the survey headline suggests — a household can answer "yes, we have internet" while sharing a single hotspot across four people on a throttled plan. Third, the support gap is often the largest single driver of lost instructional time, and it is invisible in every dashboard that counts hardware.

The concrete deliverable should be a short set of artifacts, not a 60-page report nobody reads:

• A gap register. One row per identified gap, with the affected population size, the school(s) where it concentrates, the estimated instructional-hours cost, and the intervention that would close it. Keep it under 20 rows or it stops being a decision tool.
• A readiness score by school. Composite of device health, connectivity, literacy, and support responsiveness. This is what tells you which buildings are ready to absorb more devices and which would just accumulate more broken ones.
• A cost-to-close estimate. For each gap, what it would cost to close it versus what expansion costs. This is the comparison that usually gets skipped, and it is the one that determines whether expansion is the right spend at all.
• A baseline you can re-run. The assessment is worthless as a one-time exercise. Define the instrument and the query set so the same measurement can run again in twelve months and produce a comparable number.

The honest outcome is sometimes "do not expand yet." A district with a 6% device-failure backlog, a 20% under-connected population, and a 9-day mean repair time will not improve outcomes by adding grades to the program. It will dilute an already-strained support function. Expansion should be earned by demonstrated capacity, and the assessment is how you demonstrate it.

There is a governance benefit too. Board members, funders, and state reporting bodies increasingly ask for evidence of equitable access rather than distribution counts. A district that can produce a segmented readiness map answers that question in one slide. A district that can only produce a device count spends the whole meeting defending a number that does not mean what people think it means.

Finally, expect the assessment to surface adjacent problems you were not looking for. Digital equity work tends to expose weak asset management, inconsistent LMS usage across departments, unclear acceptable-use enforcement, and family-communication channels that reach only the families who were already reachable. Those are not scope creep — they are the reasons the gaps persist. Note them, rank them, and decide which ones block expansion and which ones are simply worth fixing.

What drives that outcome

Four measurable domains drive digital equity, and they interact. Treating them as independent checkboxes is the most common analytical failure, because the households that struggle on one dimension usually struggle on several, and the compounding is what produces the outcome gap.

Device functionality. Count working devices, not deployed devices. A device is functional if it boots, holds a charge through a school day, runs the current OS build, and is not awaiting parts. Pull this from your MDM: last check-in date, OS version distribution, battery health where the platform exposes it, and open ticket status. Devices that have not checked in for 30+ days are a distinct category — they are neither functional nor recoverable without a human intervention, and they should be reported separately rather than buried in the deployed count.

Connectivity quality. Move past the binary. The useful measurements are downstream bandwidth during instructional hours, upstream bandwidth (the constraint for video submissions and live participation), data caps and throttling, and concurrent household load. A single hotspot serving three students and a working adult behaves nothing like a wired connection with the same nominal speed. Where you can, collect actual speed-test data from the device rather than self-report, since households routinely overstate connection quality — sometimes from optimism, sometimes because they do not want to appear to be asking for help.

Digital literacy — student and caregiver. Split these. A student can navigate the LMS fluently while the caregiver cannot reset a password, locate an assignment, or interpret a grade portal. That combination produces a specific failure pattern: the student is fine until something breaks, then loses days. Assess caregiver literacy through task-based questions ("have you been able to check assignments in the last two weeks?") rather than confidence self-ratings, which correlate poorly with actual capability.

Support access. Measure mean time to repair, mean time to replacement, the distance and hours of the repair location, whether a loaner is available immediately, and whether the ticket process requires English, a working email, or a portal login the family cannot access. This last item defeats an otherwise well-funded program more often than budget does.

The compounding matters more than any single domain. A household with weak connectivity *and* low caregiver literacy *and* a two-week repair cycle is not experiencing three small problems; it is experiencing near-total exclusion from any assignment that assumes home access. Your analysis should count students who fail on two or more dimensions as a separate, highlighted population, because that group drives most of the measurable outcome difference and is the group an expansion will underserve first.

Instrument design drives the quality of everything downstream. Keep the household survey under 12 questions, offer it in every language your district serves, and distribute it through more than one channel — paper backpack copy, SMS, and the enrollment portal at minimum. Response rates on portal-only distribution skew hard toward already-connected families, which is exactly the bias that makes an equity assessment useless. If your response rate is below 60% overall, or below 40% in any single school, treat the results as directional and follow up with phone sampling before you make a resourcing decision on them.

Benchmarks and realistic ranges

Treat every number below as a planning range to calibrate against, not a target handed down from a standard. Verify against your own baseline and your state's reporting requirements.

Device fleet health. In a mature program, expect 3–8% of the fleet to be in some non-functional state at any given moment — in repair, awaiting parts, lost, or unreturned. Above 10% and your support function is under-resourced relative to fleet size; expansion will make that worse, not better. Annual attrition (loss, theft, unrecoverable damage) commonly runs 2–5% depending on grade band, with the lower elementary and high school ends typically higher than middle grades. Budget replacement against attrition plus the refresh cycle, not against refresh alone.

Refresh cycle. Most districts run a 4–5 year refresh. If you are expanding while a large cohort of the existing fleet is entering year 4, you are stacking a replacement wave on top of an expansion purchase. Model the two together; that combined-year spike is the single most common budget surprise in 1:1 programs.

Repair turnaround. A same-day loaner with a 3–5 day repair turnaround is a reasonable operating target. Every day beyond that converts directly into lost instructional access for a student who is, statistically, already more likely to be in a high-need group. If your mean time to repair exceeds 7–10 days, fix that before you expand — it is cheaper and it moves outcomes more.

Connectivity. Video conferencing and typical cloud coursework want meaningful upstream, not just downstream. Households sharing a mobile hotspot across multiple simultaneous users will experience effective per-user throughput far below the plan's advertised speed, especially during evening peak. When you survey, ask specifically about the number of people using the connection during homework hours — it is the variable that most changes the interpretation of a stated speed.

Survey response rates. Multi-channel distribution with a paper option typically outperforms digital-only by a wide margin in exactly the populations you most need to hear from. Set an internal floor (60% district-wide, 40% per school) and treat anything below it as a signal to sample by phone rather than a reason to publish the number anyway.

Cost framing. Expansion cost is device unit cost plus support headcount plus insurance plus network capacity plus the software licensing that scales per seat. The common underestimate is support headcount: a district that runs one technician per several hundred devices at current scale often assumes that ratio holds through expansion, when in practice ticket volume rises with both device count *and* the newness of the deployed population. New users generate more tickets per device than experienced ones for roughly the first semester.

Gap sizing. When you segment, do not report a single district-wide gap percentage. Report the distribution. A district at 12% under-connected on average might have one school at 3% and another at 34%; the average is the least useful number you could publish, and it will lead you to spread expansion evenly across buildings with wildly different readiness.

Adjacent programs worth benchmarking against. Public library hotspot lending programs, community-college device loan programs, and municipal broadband pilots in your area all produce data about household connectivity in the same geography. Ask for it. A neighboring institution's connectivity map is often the cheapest external validation you can get for your own survey findings, and it can catch a sampling bias you would otherwise miss entirely.

Risks, edge cases, and failure modes

Survey bias is the dominant risk. Families with the least reliable access are the least likely to complete a digital survey, so a portal-only instrument systematically undercounts the exact gap you are trying to measure. This produces a comfortable number and a wrong decision. Mitigate with multi-channel distribution, translation, and phone follow-up on non-respondents in the schools with the lowest response rates.

Self-reported connectivity overstates reality. "We have internet" covers everything from symmetric fiber to a throttled phone plan. Where policy allows and privacy is protected, supplement self-report with device-side telemetry — connection type, observed throughput during instructional hours, frequency of disconnects. Be explicit and transparent about what is collected and why; a program that quietly instruments home connections will lose family trust, and trust is the thing that makes the next survey work.

Privacy and surveillance overreach. There is a real line between measuring access and monitoring students at home. Collect the minimum necessary, aggregate before reporting, avoid content-level monitoring outside school hours, and publish your data practices in plain language. Consult counsel on student-privacy obligations in your jurisdiction before you deploy any home-network telemetry. Getting this wrong does not just create legal exposure — it poisons participation in every subsequent assessment.

Counting devices instead of access. The most common failure. A 1:1 ratio on paper coexists comfortably with a large population that cannot complete homework. If your reporting stops at distribution counts, you will expand into the gap rather than closing it.

Ignoring the support gap. Support capacity is the constraint that quietly determines whether expansion helps. Adding devices to a district with a saturated help desk lengthens repair queues for everyone, including students who were previously well-served. The equity effect can be negative even though the access number went up.

Homogenizing the population. Students in temporary housing, students in foster care, migrant families, and students with disabilities each have distinct access patterns that a district-average masks. Students with disabilities may need specific assistive software or hardware that a standard fleet image does not include — a device that technically works but cannot run the student's required tools is a functional gap, not a solved one. Segment for these groups explicitly.

Assuming the gap is static. Household connectivity changes with moves, job loss, plan changes, and the expiration of subsidy programs. A subsidy that ends between your assessment and your expansion can move a household from connected to disconnected without anything in your data changing. Re-measure at least annually, and treat any known subsidy expiration as a scheduled re-measurement trigger.

Language and format access to the support path. If the only way to report a broken device is an English-language web form requiring a login, families who cannot use it will simply stop using the device. The failure shows up as low engagement, which gets misread as low motivation. Provide a phone line, an in-person option, and translation.

Scope creep into a research project. The opposite failure. An assessment that takes nine months to produce a perfect dataset has cost a year of remediation. Time-box it: 6–10 weeks from instrument design to gap register is achievable and sufficient for a resourcing decision. Precision beyond that rarely changes what you would do.

Vendor-supplied assessments. If a hardware vendor offers to run your equity assessment, read the instrument. Assessments designed by a party that benefits from expansion tend to measure device counts well and support capacity poorly. Use their data, own the analysis.

A practical rollout plan

Run the assessment as a time-boxed project with a named owner and a decision date. Without a decision date it becomes an ongoing data-collection habit that never produces a resourcing change.

Weeks 1–2: scope and instrument design. Define the four domains, decide what you will measure from telemetry versus survey, and write the household instrument. Get it translated. Run it past three or four families for comprehension before it goes out — this catches more problems than another internal review round. Simultaneously, pull your MDM baseline so the device-health picture is ready before survey data lands.

Weeks 3–5: collection. Distribute through every channel simultaneously rather than sequentially; sequential rollouts stretch the timeline and depress the total response. Track response rate by school daily, not weekly, so you can trigger phone follow-up in low-response buildings while there is still time. Pull ticket-system data for the trailing 12 months in parallel — mean time to repair, ticket volume by school, reopen rate, and the share of tickets originating from a non-digital channel.

Weeks 6–7: analysis and segmentation. Build the composite readiness score. Segment by school, grade band, and the subgroups your district reports on. Flag the multi-dimensional gap population — students failing on two or more domains — separately. Build the gap register and cost-to-close estimates.

Weeks 8–10: decision and remediation sequencing. Present the gap register alongside the expansion proposal, not after it. The decision is not binary; the useful output is usually a sequenced plan where some buildings expand immediately, some expand after a support-capacity increase, and some get remediation instead of expansion this cycle.

Sequencing remediation. Order interventions by cost-to-close divided by population affected, but override that ordering for anything that blocks the support path. Fixing a monolingual help desk is cheap and unblocks everything downstream; it should jump the queue regardless of what the ratio says.

Governance. Assign the gap register an owner who is not the same person who owns the expansion budget. That separation keeps the assessment honest. Report the readiness score to the board on the same cadence as the device count, and eventually instead of it.

Adjacent moves that pay off. While the instrument is in the field, it costs almost nothing to add questions about printer access, quiet study space, and whether the student has a consistent place to charge a device overnight. These are outside the narrow digital-equity frame but they show up in the same households and they change what interventions actually work. Similarly, coordinate with your public library and any municipal broadband effort — shared data and shared hotspot inventory are usually available for the asking, and a coordinated map beats two partial ones.

Related questions

How often should the assessment be re-run?

Annually at minimum, with an off-cycle re-measurement triggered by any known subsidy expiration, a major refresh wave, or a boundary change. Keep the instrument stable between runs so the numbers are comparable; changing the questions resets your trend line.

Can telemetry replace the household survey?

No. Telemetry measures device state and observed connection behavior well, but it cannot measure caregiver literacy, support-path friction, or whether a household has a usable place to work. Use telemetry to validate the survey, not to replace it.

What if the assessment says do not expand?

Present the sequenced alternative rather than a flat no. Some buildings will be ready; fund those. For the rest, propose a remediation budget with a re-measurement date, so the expansion is deferred with a defined path back rather than cancelled.

How do you handle families who decline to participate?

Sample non-respondents by phone to estimate the direction and size of the bias, then report the assessment with an explicit confidence caveat. Never impute a non-respondent as connected — that is the assumption that produces the comfortable, wrong number.

FAQ

What is the single most under-measured factor in a 1:1 program?

Support access. Mean time to repair, loaner availability, and the language and format of the ticket path determine whether a device stays usable after the first failure. Districts measure hardware inventory carefully and support responsiveness almost never, which is why the support gap is usually the largest recoverable source of lost instructional time.

Should the assessment cover staff as well as students?

Yes, at least lightly. Inconsistent teacher use of the LMS produces an access gap that looks identical to a household gap from the student's side. If assignments live in three different systems depending on the teacher, a family with limited literacy or limited time is effectively excluded regardless of their bandwidth.

How do you assess digital equity without over-collecting student data?

Aggregate before you analyze, collect the minimum fields that answer the question, avoid content-level monitoring outside school hours, set a retention limit, and publish what you collect in plain language. Consult counsel on your jurisdiction's student-privacy obligations before deploying any home-network telemetry.

Is a composite readiness score better than reporting the four domains separately?

Report both. The composite is what makes buildings comparable and drives sequencing decisions; the component scores are what tell you which intervention to fund. A composite alone hides whether a low score is a connectivity problem or a support problem, and those have completely different price tags.

What is a realistic timeline for the whole assessment?

Six to ten weeks from instrument design to gap register, assuming you already have MDM and ticket data available. Longer than that and you are trading remediation time for precision you will not use. Time-box it against a named decision date.

Does expansion always mean more devices?

No. Expansion can mean more grade levels, more take-home privileges, more hours of access, or more capable devices for students whose assistive-software needs the current fleet image cannot meet. Define which expansion you mean before assessing, because the gaps that block each one are different.

Sources

• https://nces.ed.gov/ — National Center for Education Statistics
• https://www.fcc.gov/ — Federal Communications Commission
• https://www.ntia.gov/ — National Telecommunications and Information Administration
• https://www.pewresearch.org/internet/ — Pew Research Center, Internet & Technology
• https://tech.ed.gov/ — U.S. Department of Education, Office of Educational Technology
• https://www.iste.org/ — International Society for Technology in Education
• https://www.cosn.org/ — Consortium for School Networking
• https://www.imls.gov/ — Institute of Museum and Library Services
• https://studentprivacy.ed.gov/ — U.S. Department of Education, Student Privacy Policy Office
• https://www.census.gov/topics/population/computer-internet-use.html — U.S. Census Bureau, Computer and Internet Use

Related on PULSE

• How do you size a help desk for a growing device fleet?
• What belongs in an annual technology refresh budget?
• How do you measure instructional time lost to device downtime?
• What does a defensible program readiness score look like?
• How do you design a household survey that reaches non-responders?
• How do you sequence a rollout across buildings with uneven capacity?