Benchmarks / SurveyBench / Why Riverside Food Network is losing volunteers / Qwen4 Max

Measured by Spring Prompt

Qwen4 Max: Why Riverside Food Network is losing volunteers

The decision: how to spend next year's £45,000 for volunteer support: reimburse volunteers' travel costs, fund a part-time volunteer coordinator with online shift booking, or do both at a smaller scale. The brief and the data →

Verdict
✗ Not sound
Research score
90 of 100
Analysis rating
1,373
Head to head, this task
won 9 of 15

Why it is not sound

The survey

  • Answer bands that overlap or leave gaps

The analysis

  • Reported a material finding the data does not support

Findings the judges found unsupported

  • The operations lead's inference does not hold: travel cost is the most-mentioned irritation, but it is not what people say drove them out. Scheduling is. Spend should not be ranked by how often a difficulty is mentioned in a prompted list.Q3 provides respondents' reported main reasons, not established causal effects. Calling it causal evidence and saying it identifies what actually ended volunteering overstates what this observational survey establishes.
  • Travel cost is background friction that everyone experiences, not something that distinguishes leavers from stayers — so reimbursing it is a retention/faithfulness measure, not a win-back lever.The insignificant active–lapsed difference does not establish that reimbursement is a retention measure rather than a win-back lever. Neither effect was measured, and some lapsed respondents explicitly identify travel cost as their main reason for stopping.
  • Lapsed volunteers leave with a cluster of problems, not one, and the cluster is communication about the rota. Fixing notification is the cheapest, highest-coverage action available and needs no software purchase.Notification may be a sensible immediate improvement, but there are no comparative costs or respondent-level overlaps establishing that it is the cheapest, highest-coverage intervention.
  • Answer to the chair: this survey cannot tell you how many lapsed volunteers will come back. It can tell you that booking is the stronger stated lever, and that any headcount you publish would be false precision.The assumed 20–50% intention-to-action conversion has no supplied evidential basis. Applying it to the unweighted, self-selected return-intention result produces an unsupported estimate of 80–200 actual returners, despite calling it illustrative.
  • Answer to the volunteer manager: the under-25 travel signal is real and directionally strong, but it rests on 11 people. Fund a targeted, capped travel pilot for young and low-income volunteers now; do not commit the budget to an under-25 scheme on this base, and do not make it age-exclusive.Calling the under-25 signal 'real' and using it to justify immediate youth-targeted spending goes beyond an explicitly uncertain result from 11 lapsed respondents. The evidence supports further investigation and a needs-based hardship pilot, not an established age-specific effect.
  • About a quarter of the losses are outside the reach of either option, and a further slice needs site matching rather than cash or software. Do not build a business case that assumes the whole 1,000 is recoverable.Selecting 'my circumstances changed' does not establish that these volunteers are unreachable by either intervention. Changed work, study or health circumstances could sometimes be accommodated by easier scheduling or travel support.

Scorecard

Numbers 12 of 12 right

QuestionIts answerTrue
What percentage of active volunteers who responded chose 'The cost of travelling to the site' as a difficulty in Q2?✓37.2%37.2%
What percentage of lapsed volunteers who responded chose 'The cost of travelling to the site' as a difficulty in Q2?✓40.2%40.2%
Is the difference between active and lapsed respondents in the share choosing 'The cost of travelling to the site' in Q2 statistically significant at the 95% level? (1 for yes, 0 for no)✓NoNo
Weighted to the age mix of all 1,000 lapsed volunteers, what percentage of lapsed volunteers would choose 'The cost of travelling to the site' as a difficulty in Q2?✓47.4%47.4%
What percentage of lapsed respondents gave a scheduling reason as their main reason for stopping in Q3 (either of its first two options: shifts changed at short notice, or hard to arrange shifts)?✓42.9%42.9%
What is the 95% margin of error, in percentage points, on that unweighted share of lapsed respondents giving a scheduling reason as their main reason?✓9.2 points9.17 points
On average, how many difficulties did each respondent choose in Q2 (counting every option except 'None of these', which counts as zero), across all respondents?✓1.551.55
What percentage of lapsed respondents said they would be likely (fairly or very) to volunteer again if they could arrange their shifts online (Q4)?✓40.2%40.2%
How many of the open-text answers mention shift scheduling (booking or swapping shifts, or last-minute changes or cancellations)?✓3734
What percentage of the active volunteers invited completed the survey?✓52.4%52.4%
What percentage of the lapsed volunteers invited completed the survey?✓11.2%11.2%
If the lapsed respondents were representative, about how many of the 1,000 lapsed volunteers would give a scheduling reason (either of the first two Q3 options) as their main reason, using the unweighted share?✓429429

Traps in the data 8 of 8 handled

  • ✓Travel cost is the most-cited difficulty overall, but it is cited about as often by active volunteers (37%) as by lapsed ones (40%), so it does not explain who left. Short-notice shift changes (14% active against 52% lapsed), no way to book or swap shifts and not being told about changes differ sharply.
  • ✓The operations lead believes volunteers leave because of travel costs. Among lapsed respondents, 43% give a scheduling reason as their main reason against 12% travel cost.
  • ✓Q2 is multi-select: respondents chose 1.6 difficulties on average, so columns add up to more than 100%.
  • ✓52% of active volunteers responded but only 11% of lapsed ones, so the lapsed base is small (112) and those who replied may differ from the 89% who did not (for example, still feeling some attachment).
  • ✓Volunteers aged 65 or over are 43% of respondents but about 26% of volunteers in the database; younger volunteers cite travel cost more. Weighted to the lapsed age mix, travel cost among lapsed volunteers rises from 40% to 47%; a scheduling main reason goes from 43% to 41%, still well ahead of travel cost (21%).
  • ✓Lapsed under-25s (base 11) mostly give travel cost as their main reason (8 of 11).
  • ✓The travel-cost gap between lapsed and active respondents (3.0 points) is not significant at 95%.
  • ✓40% of lapsed respondents say they would be likely to volunteer again with online booking: a stated intention about a hypothetical, from a small and self-selected base.

The analysis

Recommendation

Do both, but asymmetrically — and do not spend the £45,000 on universal travel reimbursement. The evidence supports buying coordination first and travel support second, targeted. Illustrative split, to be confirmed by costings: ~£18–22k for a part-time volunteer coordinator (roughly 15–20 hours/week including on-costs), ~£2–3k for a low-cost online booking tool with automated SMS/email change alerts and a phone fallback, ~£18–20k ring-fenced for a capped, targeted travel fund, ~£2–3k for evaluation and contingency. Three points drive this. (1) The causal question (Q3) says scheduling ended volunteering for 42.9% of lapsed respondents versus 12.5% for travel; travel cost is statistically indistinguishable between actives and lapsed (37.2% vs 40.2%, p≈0.60), so it is a retention irritant rather than a win-back lever. (2) The largest lapsed complaints — shifts changed at short notice (51.8%), no easy way to arrange shifts (42.0%), not told in good time (41.1%) — are partly free to fix: publish rotas four weeks ahead, freeze them, set a 48-hour notice rule with one authoritative alert channel, stop over-recruiting sessions, and let volunteers book across all nine sites so they can move to their nearest. Appoint the coordinator to own that discipline; software alone will not deliver it, as V076, V292, V133 and V019 show. (3) Travel money should be aimed where it plausibly changes behaviour rather than spread thinly: under-25s (72.7% of under-25 lapsed name cost as the main reason, but on only 11 people) and volunteers on low or fixed incomes (V234, V237, V189), at the worst-connected sites, at a capped rate (e.g. up to £4–5 per shift, max two shifts/week, simple claim). Age-exclusive targeting is not justified by an n=11 cell and would exclude pensioners who describe the same squeeze; use age plus means/need as the eligibility rule. Alongside this, do the free or near-free items now: negotiate parking bays or a drop-off point (V118, V289 — cash cannot create spaces), a 20-minute induction walkthrough (V130), and thanking volunteers by name (V165, V144). Finally, tell the chair plainly that no defensible returner headcount exists from this survey: 40.2% stated likelihood on an 11.2% lapsed response rate is an intention signal, worth perhaps 80–200 actual returners on conventional conversion assumptions, and about a quarter of losses (23.2% 'circumstances changed') are unreachable by either option. Commit the first tranche now, stage-gate the rest at month 9 on measured extra shifts per pound.

What would change it

Six things would flip or reshape this recommendation. (1) A follow-up of a random sample of lapsed non-respondents showing travel cost as the main reason for stopping in, say, 35–40%+ rather than 12.5%, or scheduling below ~25% — that would reverse the priority order and justify travel-led spend. (2) Costings showing coordinator plus tool above ~£30k, leaving a token travel fund: then buy the tool plus a few hours of paid admin and put the balance into travel, rather than a coordinator we cannot afford. (3) Database shift volumes and a claim-rate estimate showing universal reimbursement costs under ~£15–20k: universal would then be affordable and the targeting argument weakens. (4) An under-25 travel-main share holding near 70% on a base of 60+ young lapsed volunteers: that would justify a much larger ring-fenced young-volunteer travel scheme, and possibly a site or minibus/lift-share solution instead of cash. (5) Pilot behaviour: if fewer than ~10% of those who said 'likely' actually take a shift after booking goes live, while the travel pilot generates extra shifts at a lower cost per shift, move the balance towards travel in year two. (6) Evidence that the 65+ group (43% of respondents, 30% of active population) cannot or will not use online booking without a phone route — that changes the tool specification and the coordinator's job description, not the priority, but it must be designed in from the start.

Findings

  1. The operations lead's inference does not hold: travel cost is the most-mentioned irritation, but it is not what people say drove them out. Scheduling is. Spend should not be ranked by how often a difficulty is mentioned in a prompted list.

    Medium confidenceBase 112

    Q3 (lapsed, base 112, forced single main reason): shifts changed at short notice 29.5% + hard to arrange shifts 13.4% = 42.9% (48 people, ±9.2pp at 95%) versus travel cost 11.6% + journey time 0.9% = 12.5% (14 people). Yet Q2 travel cost is the top tick for both actives (37.2%, 80/215) and lapsed (40.2%, 45/112). Q2 asks what has 'ever made volunteering difficult' (multi-select, 1.55 ticks per head); Q3 asks what actually ended the volunteering. Only Q3 is causal evidence, and it points at the rota.

  2. Travel cost is background friction that everyone experiences, not something that distinguishes leavers from stayers — so reimbursing it is a retention/faithfulness measure, not a win-back lever.

    High confidenceBase 327

    37.2% of actives (80/215) vs 40.2% of lapsed (45/112) chose travel cost in Q2: a 3.0pp gap, z≈0.53, p≈0.60, 95% CI roughly −8pp to +14pp. Not significant, and the survey could not have detected any gap smaller than about 11pp. By contrast the scheduling items separate the groups sharply: short notice 14.4% active vs 51.8% lapsed; no easy way to arrange shifts 11.6% vs 42.0%; not told in good time 7.4% vs 41.1%.

  3. Lapsed volunteers leave with a cluster of problems, not one, and the cluster is communication about the rota. Fixing notification is the cheapest, highest-coverage action available and needs no software purchase.

    High confidenceBase 327

    Lapsed respondents chose 2.33 difficulties each on average vs 1.15 for actives (Q2 sums: 233.2% vs 114.7%; overall 1.55). Only 2.7% of lapsed chose 'None of these' vs 28.4% of actives. The three rota/communication items are 51.8%, 42.0% and 41.1% among lapsed. Open text: 37 of 109 answers (34%) mention booking/swapping or short-notice change; several describe finding out by word of mouth (V009, V055, V134, V253, V308).

  4. Answer to the chair: this survey cannot tell you how many lapsed volunteers will come back. It can tell you that booking is the stronger stated lever, and that any headcount you publish would be false precision.

    Low confidenceBase 112

    Q4 net likely 40.2% (45/112) vs Q5 travel reimbursed 26.8% (30/112). Among those whose main reason was scheduling (base 48): 60.4% likely to return with online booking, 27.1% with travel money. Among travel-main (base 14): 57.1% with travel money, 14.3% with booking. Scaling the unweighted 40.2% to 1,000 lapsed gives ~400 'likely' intentions; at a conventional 20–50% intention-to-action conversion that is roughly 80–200 actual returners — an illustrative range, not a forecast, on an 11.2% response rate. The Q4/Q5 splits by main reason are also partly tautological (people who left over scheduling say scheduling would bring them back), so they are consistency checks, not independent proof.

  5. Answer to the volunteer manager: the under-25 travel signal is real and directionally strong, but it rests on 11 people. Fund a targeted, capped travel pilot for young and low-income volunteers now; do not commit the budget to an under-25 scheme on this base, and do not make it age-exclusive.

    Low confidenceBase 11

    Under-25 lapsed (base 11): 72.7% (8 people) name travel cost as the main reason for stopping (Q3), 81.8% (9) tick it in Q2 — versus 2.3% of lapsed 65+ (1/44) in Q3. Wilson 95% interval on 72.7% with n=11 is roughly 44%–90%; one respondent moves the figure 9.1pp. The same 11 also report not being told about changes (54.5%) and short-notice changes (45.5%), so travel is not their only problem. Cost pressure is not confined to the young: pensioners describe fares as a chunk of a fixed income (V234, V237, V189, V070). Under-25s are 22% of the lapsed population but 9.8% of lapsed respondents.

  6. The age skew materially changes the travel number, so treat any unweighted population claim about travel as understated.

    High confidenceBase 112

    Weighting lapsed Q2 travel cost to the age mix of all 1,000 lapsed volunteers (22/30/23/25%) gives 47.4%, versus 40.2% unweighted. Completes are 43.1% aged 65+ and 8.3% under 25, against 25% and 22% in the lapsed population and 30% and 18% among actives. Weighting fixes age composition only; it cannot fix self-selection within age bands.

  7. About a quarter of the losses are outside the reach of either option, and a further slice needs site matching rather than cash or software. Do not build a business case that assumes the whole 1,000 is recoverable.

    Medium confidenceBase 112

    Q3 'my circumstances changed' 23.2% (26/112) — work, study, health, moving. Nine open answers are about journey time or distance and ask for a nearer site or a different placement (V008, V025, V029, V062, V099, V228, V230, V304, V317); Q2 journey time is 17.1% overall and does not differ by segment (16.7% active vs 17.9% lapsed). Lapsed respondents are more than twice as likely to work across sites equally (10.7% vs 4.2%, Q6), and V286 asks to see all nine sites' shifts in one place.

  8. Much of the scheduling damage is rota discipline, not the absence of an app. Buying booking software without a coordinator who owns published rotas and notice rules will disappoint.

    Medium confidenceBase 109

    Open text describes rotas changing after publication and sessions cancelled for over-recruitment: V076 ('turn up only to be told they've got enough people today'), V292 ('a rota that doesn't change after it's been sent out'), V133 (month ahead and then stuck to), V161 ('fixed shifts that stay fixed'), V215, V019 (turned up to a locked door), V058 (8am text for a 9:30 start). Feature requests are also specific and cheap: recurring bookings (V102, V120), self-service swap (V104, V112, V210, V239), confirmation of a place (V225, V262), cross-site visibility (V286).

  9. Universal reimbursement risks spending the whole £45,000 subsidising shifts that already happen. Affordability hinges on two numbers we do not have: annual shift volume and claim rate.

    Low confidenceBase 327

    Open-text per-trip costs cluster at £4–£12 (V070 £2 each way; V053 £3.80 tram; V237 £5.20 day ticket; V034 £7 parking; V189 ~£9 train; V182 £12 fuel; V299 14 miles each way). At £4–£5 per shift, £45,000 buys roughly 9,000–11,000 shifts — of the same order as a plausible existing volume (410 actives × ~2 shifts/month ≈ 9,800 shifts/year). At 40–50% claim rates the cost falls to ~£18–25k, but 80 of 215 actives already cite cost, so most of the spend would go to retained volunteers. Retention value is real; win-back value is not evidenced.

  10. There is a low-cost tail worth doing regardless of the £45,000 decision, and one trap: some 'travel cost' complaints are actually a shortage of parking spaces, which cash cannot fix.

    Medium confidenceBase 327

    Q2: not enough training 11.9% (39/327), not feeling thanked or valued 9.8% (32/327), physical access 7.0% (23/327). Open text: induction walkthrough and date-checking rules (V130); thanks by name (V165), recognition (V144); parking supply and drop-off while carrying boxes (V118, V289 — six spaces, van blocked) as distinct from parking charges (V034, V060, V095, V158).

Caveats

  • Lapsed response is the weak point: 112 completes from 1,000 invited (11.2%), against 215 from 410 actives (52.4%). The lapsed column is a self-selected 11% — plausibly people with a live grievance or a remaining attachment to Riverside. Every lapsed figure, including the 42.9% scheduling share and the 429 projection, is conditional on that selection. Direction of bias is unknown, so do not treat 429 as a count of recoverable volunteers.
  • Margins of error are wide where it matters: lapsed base 112 gives ±9.2pp on the 42.9% scheduling share (≈337–521 people if representative); active base 215 gives ±6.5pp on 37.2%. The under-25 lapsed cell is 11 people (one person = 9.1pp); lapsed travel-main is 14 (one person = 7.1pp). No conclusion should hinge on those two cells alone.
  • Age composition is badly skewed: 43.1% of completes are 65+ and 8.3% under 25, versus 25%/22% in the lapsed population and 30%/18% among actives. Weighting lapsed travel cost to the population age mix moves it from 40.2% to 47.4%. Tables are unweighted; weighting only corrects age, not non-response within age.
  • The 'Total' column is not a population estimate: it is 66% active / 34% lapsed, while the volunteer population is 29% active / 71% lapsed. Use the segment columns, never Total, for population claims.
  • Invitation was by email and text from the database. Volunteers with stale contact details — likely the youngest and most recently moved — had less chance of being reached. No bounce or delivery data was supplied, so this bias cannot be sized.
  • Q2 is a prompted 'ever a difficulty' multi-select (1.55 ticks per respondent; 2.33 for lapsed). It measures prevalence and salience, not cause, and it cannot rank spending priorities. Q3 forces one main reason, which understates multi-causal exits — the 12.5% travel figure means 'travel was the single biggest factor', not 'travel mattered to only 12.5%'.
  • Q4 and Q5 are hypothetical intentions with no price, cap, effort or trade-off attached, asked in fixed order (booking before travel), which may anchor responses. 'Paid back' implies full reimbursement, not the capped scheme that is actually affordable. Stated likelihood routinely over-states behaviour; the paired booking-vs-travel difference cannot be tested from these marginal tables.
  • The Q4/Q5 breakdowns by main reason are partly circular: people who left because of scheduling saying scheduling would bring them back is consistency, not corroboration. Bases 48 and 14 are small.
  • 'Lapsed' here means no shift in 3 months but at least one in 12 months, and S2 screened out anyone longer ago. These are recently lapsed, possibly seasonal, volunteers. We have no evidence at all about people who left years ago, or about people who never joined. Ten lapsed open-text respondents say nothing is wrong, which sits oddly with only 2.7% of lapsed choosing 'None of these' in Q2 — some lapsing is probably life rhythm, not dissatisfaction.
  • Open text was optional: 109 of 327 (33.3%) answered and 38 of those said nothing needed changing. Theme counts describe the people who wrote, not prevalence among volunteers.
  • No site-level conclusions are safe: 9 sites with roughly 20–43 respondents each, and some answers name a different site from the respondent's Q6 site (V034 Hollins Park/Northgate, V095 Millbrook/Eastfield, V289 Bridge End/Eastfield), indicating multi-site working or recall noise.
  • The survey contains no costs, shift volumes, claim rates, coordinator salary quotes or software quotes. Every £ figure in this report is illustrative arithmetic and must be replaced with real costings before the budget is committed.

Next steps

  • decisionAppoint the part-time volunteer coordinator and, in their first month, publish and enforce a rota standard across all 9 sites: rotas released 4 weeks ahead, frozen once published, minimum 48 hours' notice for any change, one authoritative alert channel (SMS/email) to every affected volunteer, no over-recruiting of sessions, and a named person accountable for each change.
  • experimentRun a 3-site pilot (choose one high-lapsed, one typical, one well-connected site) of a low-cost online booking tool with self-service swap, recurring bookings, booking confirmation, cross-site visibility and automatic change alerts, with a phone/text fallback maintained by the coordinator.
  • experimentRun a capped, targeted travel fund pilot in parallel at 3 different sites (under-25s plus volunteers on low/fixed incomes or with no direct transport route; e.g. up to £5 per shift, two shifts per week, one-page claim), with matched comparison sites and a monthly cost cap.
  • researchCommission a short telephone follow-up with a random sample of ~200 lapsed non-respondents, deliberately boosting under-25s to at least 60 completes, asking the Q3 main-reason question plus one open 'why did you stop?' and whether a capped travel payment or self-service booking would change anything.
  • researchBuild the cost model from the volunteer database and supplier quotes: shifts per year by site and age, current fill and cancellation rates, realistic reimbursement claim rates, coordinator salary/on-costs at 15/20/25 hours, and three booking-tool quotes including SMS costs.
  • researchInterview 12–15 recently lapsed volunteers (6 under 25, 6 on low or fixed incomes, spread across sites) about what would actually get them back: capped versus full reimbursement, lift-share versus cash, recurring versus ad-hoc shifts, and what would make them trust the rota again.
  • decisionDo the free and near-free fixes now: enable booking across all nine sites and offer nearest-site transfers; negotiate parking bays or a drop-off point at the sites with charged or scarce parking; introduce a 20-minute induction walkthrough with date-checking rules; thank volunteers by name at every shift.
  • monitorStand up a monthly volunteer-support dashboard from the database: active volunteer count, lapsed reactivations (first shift back after 3+ months), shifts filled versus published, changes inside 48 hours, self-service booking share, travel claims (£ and shifts), all cut by age band and site.
  • decisionSet a trustee stage-gate at month 9: release the remaining travel and coordinator funding only on measured cost per additional shift and reactivation numbers from the two pilots, and re-run a short pulse survey (with phone chasing of lapsed and under-25 volunteers) to re-test the Q2/Q3 pattern.

Open-text themes it coded

Self-service booking, swapping and recurring shifts (online rota) 20Short-notice changes/cancellations and news not passed on (rota stability and notification) 20Cost of getting to the site: fares, fuel and parking charges 21Journey time, distance and wanting a nearer site or placement 9Parking supply and physical access at the site (not charges) 2Recognition and being thanked 2Induction training and knowing the sorting rules 1Nothing to change — works fine as it is 38

The survey it planned

2 screening questions and 10 questions, as the model wrote them.

  1. S1

    Have you done at least one volunteering shift with Riverside Food Network in the past 12 months? A shift means sorting donations or helping to hand out food parcels at one of our distribution sites.

    One answer
    • Yes
    • No
    • I'm not sure

    Continues if Yes; I'm not sure

  2. S2

    When was your most recent shift with Riverside Food Network?

    One answer
    • In the last 4 weeks
    • 1 to 3 months ago (within the last 3 months)
    • More than 3 months ago, up to 6 months ago
    • More than 6 months ago, up to 12 months ago
    • More than 12 months ago
    • I have never completed a shift

    Continues if In the last 4 weeks; 1 to 3 months ago (within the last 3 months); More than 3 months ago, up to 6 months ago; More than 6 months ago, up to 12 months ago

  3. Q1

    Which Riverside Food Network distribution site did you volunteer at most in the past 12 months?

    One answer
    • [Site 1 name]
    • [Site 2 name]
    • [Site 3 name]
    • [Site 4 name]
    • [Site 5 name]
    • [Site 6 name]
    • [Site 7 name]
    • [Site 8 name]
    • [Site 9 name]
    • I volunteered at several sites equally
    • I can't remember the site name
  4. Q2

    How old are you?

    One answer
    • Under 25
    • 25–34
    • 35–49
    • 50–64
    • 65–74
    • 75 or over
    • Prefer not to say
  5. Q3

    Think about the shifts you did in the past 12 months. On a typical shift, how much did you spend in total getting to the site and back — bus or train fares, or fuel and parking? Please give your best estimate in pounds (for example 3.50). Enter 0 if travelling cost you nothing.

    Number
  6. Q4

    Thinking about the shifts you did in the past 12 months, which of these made volunteering difficult or less appealing for you? Select all that apply.

    Any that apply
    • Bus or train fares to and from the site
    • Fuel or parking costs
    • No suitable bus or train at the times shifts start or finish
    • Having to phone or message the site lead to book or swap a shift
    • Not being able to find out which shifts were available
    • Shifts changing or being cancelled at short notice
    • The site closing or cutting its hours because too few volunteers turned up
    • Shift times not fitting around work, study or caring
    • The physical demands (lifting, standing, sorting)
    • Not feeling supported or thanked by the site lead or the team
    • Shifts feeling disorganised, or not having enough to do
    • My own health or a disability
    • Something else
    • None of these — volunteering was not difficult for me
  7. Q5

    You haven't done a shift in the last 3 months. Which of these played a part in you stopping? Select all that apply.

    Any that apply
    • The money I was spending on travel added up
    • Getting to the site was difficult (distance, transport, shift times)
    • Phoning or messaging the site lead to book or swap shifts was too much hassle
    • I couldn't easily find out about shifts, or I wasn't asked to help
    • Work or study commitments changed
    • Caring responsibilities changed
    • My health changed
    • I moved away from the area
    • I didn't feel supported, thanked or included
    • I didn't enjoy the tasks or the atmosphere, or shifts felt disorganised
    • I had a bad experience with someone at Riverside
    • I only ever meant to help for a short time
    • Something else
    • I don't know / can't remember

    Routing Shown only if S2 is: More than 3 months ago, up to 6 months ago; More than 6 months ago, up to 12 months ago

  8. Q6

    Of the things you selected, which was the single biggest reason you stopped volunteering with Riverside Food Network?

    One answer
    • The money I was spending on travel added up
    • Getting to the site was difficult (distance, transport, shift times)
    • Phoning or messaging the site lead to book or swap shifts was too much hassle
    • I couldn't easily find out about shifts, or I wasn't asked to help
    • Work or study commitments changed
    • Caring responsibilities changed
    • My health changed
    • I moved away from the area
    • I didn't feel supported, thanked or included
    • I didn't enjoy the tasks or the atmosphere, or shifts felt disorganised
    • I had a bad experience with someone at Riverside
    • I only ever meant to help for a short time
    • Something else
    • I don't know / can't remember

    Routing Shown only if S2 is: More than 3 months ago, up to 6 months ago; More than 6 months ago, up to 12 months ago

    Judges: assumptive

    The routing establishes only when the respondent last volunteered, not whether they have stopped. The question nevertheless requires a biggest reason for stopping.

    Q6 says 'Of the things you selected' but is shown to every lapsed respondent, whatever they answered at Q5. Answers are not forced, so some will have skipped Q5. Others will have chosen 'I don't know / can't remember' there. Q6 assumes these people selected reasons, and the full list is shown again rather than only the Q5 selections.

  9. Q7

    Riverside Food Network has not decided to do this, but imagine that next year it introduced online shift booking: you could see the shifts available at your site on your phone or computer, book one yourself, and swap with another volunteer without phoning or messaging the site lead. Everything else about volunteering stays the same, including paying your own travel costs. How likely would you be to do at least one shift in the 3 months after that change?

    Scale

    Scale 1-5: Very unlikely to Very likely

    Routing Shown only if S2 is: More than 3 months ago, up to 6 months ago; More than 6 months ago, up to 12 months ago

  10. Q8

    Now imagine a different change: next year Riverside Food Network paid back the travel costs you actually spend getting to and from your shift (bus or train fares, or fuel and parking), claimed back after each shift. Everything else about volunteering stays the same, including arranging shifts by phoning or messaging your site lead. How likely would you be to do at least one shift in the 3 months after that change?

    Scale

    Scale 1-5: Very unlikely to Very likely

    Routing Shown only if S2 is: More than 3 months ago, up to 6 months ago; More than 6 months ago, up to 12 months ago

  11. Q9

    Riverside Food Network has about £45,000 for volunteer support next year. It could pay back volunteers' travel costs, or fund a part-time volunteer coordinator with online shift booking, or do a smaller amount of each. Nothing has been decided yet. Which would make the biggest difference to whether you volunteer with Riverside in the next year?

    One answer
    • Paying back my travel costs
    • A part-time volunteer coordinator with online shift booking
    • A smaller amount of both
    • Neither would make a difference to me
    • Not sure
  12. Q10

    Is there anything else the trustees should know about what makes volunteering easy or hard for you, or about what would help you volunteer (again)? Please don't name individual people. This question is optional.

    Open text
Sample plan

FRAME AND INVITATION: census of the volunteer database — everyone with at least one shift in the last 12 months (410 active = shift in last 3 months; ~1,000 lapsed = last shift 3–12 months ago) = ~1,410 people. Personalised single-use links sent by email and SMS (SMS for the ~lapsed records with no valid email), with two reminders at day 4 and day 10 and a short telephone top-up call to a random 200 lapsed non-responders in week 3. Unconditional £5 voucher (or charity donation choice) offered to all completers to lift lapsed response and reduce non-response bias; funded from the research budget, not the £45,000. Before launch, append admin data from the site rotas to each record (main site, number of shifts in last 12 months, first-shift date, last-shift date) so these do not need to be asked and can be used for weighting and non-response checks. The nine site names in Q1 must be replaced with the exact names used on the rotas before the survey goes out; nothing else in the instrument is a placeholder.

COMPLETES AND QUOTAS: target 330 completes (expected yield: ~185 of 410 active at ~45% response, ~150 of 1,000 lapsed at ~15%). Soft quotas: minimum 150 active and 150 lapsed; per-site floor of 30 completes, remainder allocated proportionally to each site's 12-month roster; monitor daily and send site-specific reminders to lagging sites. Weight completes back to frame totals on status × site (and age if held on the database); report unweighted base for every cell and flag/merge any cell with n<30 so individuals at small sites cannot be identified.

PRECISION — THE BRIEFED TARGET IS NOT ACHIEVABLE AND MUST BE RENEGOTIATED BEFORE LAUNCH: ±7pp at 95% confidence for a worst-case proportion needs n≈196 per site. With 330 completes spread over 9 sites (~37 per site) the site margin is about ±16pp (worse at small sites), although the whole-sample margin is about ±5.4pp and the lapsed-vs-active contrast is around ±8pp per group. Even applying the finite-population correction (average site roster ≈157 of the 1,410), ±7pp per site needs roughly 70–130 completes per site depending on roster size, i.e. ~800 completes ≈ 57% of the whole frame — not credible for a lapsed group. Options for trustees, in order of recommendation: (a) keep 330 completes, hold ±7pp only for whole-sample and lapsed-vs-active estimates, and report site figures with explicit confidence intervals as directional, pooling the 9 sites into 3 groups (by roster size and transport catchment) for ~±10pp; (b) extend fieldwork and telephone top-up to ~550–600 completes for about ±10pp per site; (c) commission site-level qualitative follow-up instead of site-level survey precision. Do not present site percentages to trustees without intervals.

FIELDWORK QUALITY AND NEUTRALITY: mobile-first, one question per screen, no forced answers except the screener; pilot with 8–10 volunteers (mix of active and lapsed) and confirm median completion ≤5 minutes (lapsed see 10 questions, active see 6). Randomise the order of the Q4 and Q5 option lists (anchoring 'Something else', 'None of these', 'I don't know' at the bottom), carry the displayed Q5 order into Q6, and randomise whether Q7 or Q8 is shown first. The invitation and introduction say only that trustees are deciding how to spend the volunteer-support budget; the operations lead's travel-cost hypothesis is not mentioned anywhere, cost appears alongside booking, information, support and life-circumstance items, both counterfactuals explicitly hold the other condition constant, and Q9 includes 'neither' and 'not sure'. No dedicated attention item is spent from the 10-question budget; instead exclude completes under 120 seconds, straight-lined scales with identical open-text gibberish, and duplicate database IDs, and report the number excluded. Stated intentions are not behaviour: all return-likelihood figures will be reported as intentions with a stated-behaviour discount, not as forecasts.