A Tourist Satisfaction Survey with a Flawed Sample
A +2 Vocational Tourism student is interning with a Bhubaneswar travel agency. The project guide asks for a tourist satisfaction survey of 100 visitors at Konark. On the first two days the student stands near the ticket counter from 10:00 to 12:00, approaches only visitors who speak Odia or Hindi, and offers a ₹50 coupon for completing a five-point satisfaction form. The form asks ‘Are you satisfied with Konark?’ and one overall rating question. The student collects 62 responses; 48 are from families travelling in the same two buses. Most respondents tick 4 or 5 on the satisfaction scale, and the student plans to write that ‘90% of tourists are satisfied’. The agency supervisor wants a one-page finding for a client within three days, and the project guide asks the student to review the method before submission.
1. Identify the sampling, measurement and analysis problems in the survey. Explain how each could bias the 90% claim.
Read the worked answer
Sampling problems. The student uses a convenience sample: only people who pass the ticket counter between 10:00 and 12:00, only Odia/Hindi speakers, and only those willing to stop. Early-morning tour groups, foreign visitors, independent travellers and people with language barriers are excluded. The 48 respondents from two buses are a cluster, not 48 independent views; their answers may reflect the same tour guide or group mood. A coupon creates self-selection: people who want ₹50 may be more cooperative and more likely to give positive answers.
Measurement problems. ‘Are you satisfied with Konark?’ is a leading yes/no question. The five-point rating is a single global item and does not separate cleanliness, guides, ticketing, toilets or crowding. Offering a reward before completion can produce acquiescence bias - guests reciprocate with favourable ratings.
Analysis problems. The 90% claim is not supported. With 62 responses, most ticking 4 or 5, the student may be describing the respondents, but cannot generalise to ‘tourists at Konark’. The 62 are not a random 62% of visitors. Also, ‘90% satisfied’ is ambiguous: does it mean 90% of 62, or 90% of all visitors? The clustered bus groups and the small, biased sample make any precise percentage unreliable. The finding should be labelled indicative, not representative.
2. Design a feasible improved method to collect 100 usable responses in three days, including timing, selection, language and incentive.
Read the worked answer
Sampling frame and selection. Station the student at the main exit gate, not the ticket counter, and select every 5th visitor group after a random start. Collect across three shifts: 08:00–11:00, 11:00–14:00 and 14:00–17:00, including one weekend day. Aim for 120 approaches to obtain 100 usable responses and record refusals.
Language. Use a short form in Odia, Hindi, English and Bengali, with a trained interpreter available for other languages. Keep the questionnaire to one page: five service attributes rated 1–5, one overall rating, one open complaint question, and basic profile (domestic/foreign, group size, first visit).
Incentive. Avoid a cash coupon tied to completion. If an incentive is needed, give a small fixed souvenir or tea voucher after the form is completed, and state that it is a thank-you, not payment for a positive answer. This reduces reciprocity bias.
Field discipline. Pilot the form with 10 visitors, check that questions are understood, and do not survey people in the same bus group consecutively - take one or two from a group, then move to another group.
Analysis. Report the response rate, the number of refusals and the limitations. A sample of 100 at the exit gate in three days is still not a full probability sample of all Konark visitors, but it is far more defensible than 62 convenience responses.
3. What ethical and reporting issues must the student address, and how should the one-page finding be worded?
Read the worked answer
Ethical issues. The ₹50 coupon may be an undue inducement, especially for low-income visitors; participation must be voluntary and not conditional on a positive rating. The student should explain the purpose, keep responses anonymous, avoid photographing faces without consent, and not promise that complaints will be personally resolved. The agency must not use the data to make false claims to clients.
Reporting issues. Disclose the method, dates, times, location, language coverage, sample size and refusal rate. Do not say ‘90% of tourists are satisfied’. Instead write something like: ‘In an exit survey of 100 visitors at Konark on 14–16 June, 72% rated overall satisfaction 4 or 5 out of 5. The sample over-represents daytime domestic group visitors; results are indicative and not representative of all Konark visitors.’ Include the main complaint themes, even if they are uncomfortable for the client.
A one-page finding should have: purpose, method, three key numbers, two complaint themes, and one recommended action. This is honest, useful to the client, and academically defensible.