Table of contents
Published October 27, 2026.
Research question and decision use
The study asks how long a volunteer application takes to move from submission to a screening decision and then to a first scheduled shift. The unit is one applicant-role pair. The result can guide staffing and vendor coordination, but it does not measure the quality of a volunteer's work or the appropriateness of a screening standard.
Population and eligibility
Eligible applicants applied for a defined role during the window and consented to the required checks. Record the risk tier, application channel, vendor package, and final status. Exclude applicants who never consented, and count withdrawals by stage so the denominator remains interpretable.
Sampling protocol
Include every applicant for the selected roles during the window, then sample only if the volume exceeds the review capacity. If sampling, stratify by risk tier and application week, publish the seed, and report the unsampled counts by status. Never drop withdrawn applicants from the stage totals.
Evidence dictionary
Timestamps cover application received, consent received, packet sent, vendor result, reference completion, staff decision, and first shift scheduled. Record the source system and whether the stamp is automatic or manual. A missing timestamp is a finding, not an invitation to estimate from email memory.
Classification rubric
Each stage is within target, late, or blocked. Late means the target was exceeded without a documented exception. Blocked means the record waited on missing information or an unavailable approver. Rate evidence quality separately so a fast decision with no reference note does not look complete.
Reviewer calibration
Calibrate on cases that sat over a weekend, cases with a vendor delay, and cases reopened after a missing document. Reviewers should agree on which timestamp starts each stage. Disagreements about the clock are resolved before measuring the sample.
Analysis plan
Report median, upper quartile, and maximum days by stage, with counts at each step. Show business-hour and calendar-day versions side by side. Break results by risk tier and vendor package, but suppress cells small enough to identify an applicant.
Sensitivity and exception analysis
Recompute excluding holiday weeks and excluding roles that require an extra credential. Compare results when withdrawals are counted as completed at the withdrawal date. If the bottleneck ranking changes, describe the operational condition instead of asserting a single cause.
Privacy and role boundaries
Screening records can include identity and background information. Use a restricted extract, aggregate before sharing, and suppress small cells. Do not publish an example that could identify an applicant or a volunteer, and apply the retention schedule to the working file.
Limitations
The study describes elapsed time for one period and cannot prove why a stage was slow. Vendor load, staff leave, applicant responsiveness, and role complexity may all contribute. A fast handoff does not establish that the screening decision was correct or that the role was a good fit.
Replication package
Retain the application export, timestamp dictionary, target clock, vendor reference list, stage calculations, and suppression rules. A later cycle should reproduce the bottleneck ranking without access to the original sensitive records.
Conclusion
The useful output is a stage map with an owner and a target for each handoff, not one average that blends vendor time, reference time, and scheduling. Publish the stage that consumed the most time and the evidence behind that conclusion.