To evaluate a scholarship, connect the award with the outcomes it is intended to support: for example, continued study, completion or reduced financial pressure. Plan appropriate follow-up and report what is known, what is missing and what the evidence can establish. Award counts and disbursements remain useful delivery measures; they do not by themselves show a change in recipients’ circumstances.
This lesson is for scholarship administrators, program teams and evaluators. Bring the purpose of one award and its current collection process. You will build a recipient follow-up plan, define one outcome and prepare a report that separates delivery, observed results and unanswered questions.
Start with the scholarship’s purpose
A completion grant, a living-cost award and a professional-development scholarship may aim at different changes. Select outcomes that match the award rather than requiring every recipient to report GPA, employment, debt and confidence.
Write the expected change, who it concerns and when it could reasonably be observed. Include how the recipient views success. A change in study mode, a transfer or a temporary pause may have context that a simple “retained/not retained” label misses.
Separate selection from evaluation
Application information can provide a starting point when it is appropriate and collected for a clearly explained purpose. But adding evaluation questions is not cost-free: it takes time and may request sensitive information that is unnecessary for selection.
Decide which information is required to assess eligibility and which belongs in voluntary or post-award evaluation. Explain whether responses affect the award and who will see them. Avoid exposing personal circumstances to reviewers who do not need that information.
If non-recipients are part of an evaluation design, establish a suitable purpose and collection process. Funded and unfunded applicants may differ because of the selection process. Comparing them does not automatically isolate the scholarship’s effect.
Connect the person, award and observation
Keep the student’s identity distinct from each application, award, payment and academic observation. A person may receive several awards or change institutions. A new award should not erase an earlier record, and a new term should not overwrite the previous term’s result.
| Record | Useful context |
|---|---|
| Recipient | Stable identifier and appropriately maintained contact details |
| Award | Program, cycle, decision, amount and conditions |
| Disbursement | Payment date, amount and source record |
| Observation | Outcome definition, period, source and review status |
| Follow-up | When it is due, response status and permitted next action |
Use the relevant person and award relationships to join these records. Do not place all values in one row or sum the same award once for every term response.
Define a small, usable collection plan
For an award intended to support continued study, define what continuation includes: remaining at the same institution, an approved transfer, a different study load or another agreed status. Identify the checkpoint and eligible group. Keep completion, withdrawal, pause and unknown status distinguishable.
Across institutions, agree the few shared fields necessary for reporting. Local programs may ask different questions about support needs. Map compatible terms through a versioned dictionary rather than imposing one complete survey. Do not aggregate different academic calendars or grading systems without a defensible rule.
Reuse stable registration details. Confirm changing contact or study information at an appropriate interval, retaining the dates relevant to each observation. Collect academic or financial detail only when it serves the agreed purpose and access is appropriate.
Work through a continuation example
A fictional program has 40 recipients due for a next-term check. Thirty respond: 24 meet the agreed continuation definition and six do not. Ten have unknown status.
| Measure | Result | Interpretation |
|---|---|---|
| Response coverage | 30/40 = 75% | One quarter of due records remain unknown. |
| Continuation among respondents | 24/30 = 80% | Describes the responding group. |
| Known continuing share of the cohort | 24/40 = 60% | Does not classify the ten unknowns as withdrawals. |
If all ten unknowns continued, the cohort share would be 34/40, or 85%. The 60–85% range describes the unresolved statuses under this example’s definitions; it is not a confidence interval and does not estimate the scholarship’s causal effect.
Record a useful next action: confirm the ten missing statuses through the agreed process and review reasons for non-continuation where recipients choose to share them. Do not report 80% as an established rate for all 40 recipients.
Use reflections to understand reported experiences
A recipient may describe transport costs, paid work, caring responsibilities or a change in study plans. Code these accounts consistently and preserve the relevant source passages. Report how many people provided usable comments and whether several themes can apply to one response.
These accounts explain participants’ reported experiences. They may help design a better support response, but they do not prove that a theme caused withdrawal or establish its prevalence among non-respondents. Avoid inventing a reason when one was not given.
Check quotation permissions and disclosure risk before using a story. In a small award program, removing a name may still leave someone identifiable through their institution, course and circumstances.
Plan later outcomes when they serve a real question
If the award aims to support graduation or later employment, define when follow-up becomes due and what evidence is practical to obtain. Some recipients will not yet have reached that stage. Keep “not yet due” separate from missing and from an observed outcome.
A persistent identifier helps connect the records; it does not ensure recipients respond. Use an appropriate contact plan, clear purpose and reasonable burden. Reassess whether each later question is still useful to the program and recipients.
Prepare a balanced report
Show awards and disbursements, the outcome definition, eligible population, response coverage, observed results and limitations. Explain changes in the collection method or recipient mix before comparing cycles. Baseline-to-follow-up change is useful descriptive evidence but does not alone establish that the award caused it.
Do not apply red/amber/green labels to students based on incomplete outcomes. If the team uses a workflow status, make it describe the evidence or agreed follow-up action. A student’s circumstances require context, not an automatic score.
Configure the recurring process in Sopact
Connect recipient, award and dated evidence records. Use the approved dictionary to support comparable analysis while retaining local questions. Review candidate themes, missing statuses and conflicting observations before approving the reporting view.
The goal is to maintain the evidence through the award’s life, reducing repeated matching and assembly. Test the calculation on a small sample, review permissions and keep a dated reporting version. Human reviewers remain responsible for the interpretation and decisions.
Practice: design one follow-up
- Write the award’s purpose and one outcome definition.
- Identify which information can be reused and which must be collected later.
- Choose the eligible group, timing and response statuses.
- Draft a short question and an appropriate local follow-up.
- Use the fictional table to write a clear board-report finding without hiding unknowns.
Frequently asked questions
Must baseline data be collected from every applicant?
No. Choose the collection point from the evaluation purpose, burden and permitted use. Reuse suitable application data, but do not require unnecessary sensitive information for a later analysis.
Can we compare recipients with unsuccessful applicants?
That may support a carefully designed evaluation, but selection differences must be addressed. A simple comparison does not establish the award’s effect.
Can a small program evaluate outcomes?
Yes. It can describe known outcomes and learn from participants. Small numbers affect interpretation and disclosure, and do not justify stronger causal claims.