
By Sunil Saxena
Reporters rarely use a questionnaire. It is more of an academic tool.
The reason for this is simple. Questionnaires use multiple choice or Yes/No questions to obtain information. The information is therefore limited, and lacks detailed nuances. Yes, there are some questions for which more detailed information is sought. But these are few.
The other limitation of a questionnaire is that the reporter or the blogger does not get to meet the respondent. He is therefore unable to form any opinion about the respondent or judge the veracity of the information provided by him.
Reporting requires considerable skills, one of which is the ability to pose supplementary questions to get to the truth. These supplementary queries not only help reporters / bloggers establish whether a respondent is lying but sometimes also produce answers which can change the question trail and the focus of the story.
This is why reporters / bloggers prefer to meet respondents or subjects in person and seek first-hand information.
Another reason questionnaires are not the preferred mode of obtaining information is because several respondents are cagey about answering questions in writing. They are especially reluctant to give their name or other identifying information.
Things changed with the advent of web-based Google and other survey forms. Reporters / bloggers can today obtain information from many actors who were part of an event using a Google form. They can do it from the comfort of their office, and in quick time.
# My Experiments with AI
The goal of this experiment was not to examine or assess the utility of information obtained through a questionnaire. Instead, the goal was to explore the ability of AI to generate a questionnaire that reporters, bloggers and even researchers could use.
It was also a continuation of my earlier experiment where I had asked ChatGPT 4.0 to generate a set of questions that can be posed to street vendors and small traders for a feature story on their lives / business after the Maha Kumbh had ended. The questions suggested by AI were published on Medium.com and can be browsed here.
I now resubmitted these questions to ChatGPT 4.o and asked it to generate a questionnaire.
Prompt: Please structure the questions that you had suggested into a questionnaire format. The questionnaire should be simple, clear, and easy to comprehend. It should provide both quantitative and qualitative insights.
# Response of ChatGPT 4.o
Post-Kumbh Mela Vendor & Small Trader Questionnaire
(For Journalistic Field Use — Structured for Ease of Response)
Section 1: Personal Information
- Name:
- Age:
- Hometown/City:
- How long have you been in this business?
- Less than 1 year
- 1–5 years
- 5–10 years
- More than 10 years
- Have you participated in previous Kumbh Melas?
- Yes
- No
- If yes, which years? ______________
Section 2: Business During Kumbh Mela
- What did you sell during Kumbh Mela?
- Food & beverages
- Religious items (rudraksha, idols, malas, etc.)
- Clothes & accessories
- Handicrafts & souvenirs
- Other (please specify) ______________
2. How much did your sales increase during Kumbh compared to normal days?
- Doubled
- Tripled
- More than tripled
- No significant increase
3. What was your best-selling item, and why do you think it was so popular?
- Item: ______________
- Reason: ______________
4. Did you take a loan or invest heavily to stock up for Kumbh?
- Yes, took a loan
- Yes, invested savings
- No, continued business as usual
5. What kind of customers did you attract the most? (Check all that apply)
- Local residents
- Pilgrims from other Indian states
- Foreign tourists
- Sadhus & ascetics
Section 3: Post-Kumbh Challenges
- How have your daily sales changed after Kumbh ended?
- Dropped by 50%
- Dropped by 75%
- Almost no customers now
2. What happens to your unsold stock?
- Can sell it elsewhere
- Returning it to suppliers
- Facing a complete loss
3. Are you planning to stay in Prayagraj or move elsewhere for work?
- Staying in Prayagraj
- Moving to another pilgrimage city
- Unsure
4. Have you faced any issues with authorities regarding stall removal or relocation?
- Yes, forced to shut down
- No, but sales are low
- No issues so far
5. How are you managing daily expenses now that the crowd is gone?
- Using savings from Kumbh earnings
- Looking for another business/job
- Struggling to make ends meet
Section 4: Future Plans & Adaptation
- Do you follow a religious tourism circuit, moving to other pilgrimage sites or fairs?
- Yes (specify upcoming event/location) ______________
- No, I stay in Prayagraj
2. Are you considering changing your business model or products after Kumbh?
- Yes, shifting to a different business
- No, waiting for the next big event
- Not sure yet
3. Have you received any government or NGO support for small traders post-Kumbh?
- Yes
- No
- Not aware of any schemes
4. If you could request one thing from the authorities to help traders like you, what would it be?
…………………………………………………………………………………………
Section 5: Personal Stories & Reflections
- What was the most memorable experience you had during Kumbh?
…………………………………………………………………………………………
2. Did the earnings from Kumbh help you achieve any personal or family goals? (e.g., education, healthcare, home repairs, etc.)
- Yes (specify) ______________
- No, earnings were not enough
3. If you had to describe Kumbh Mela in one word, what would it be, and why?
- Word: ______________
- Reason: ______________
4. What does Kumbh Mela mean to you — not just as a business opportunity but as a cultural or spiritual event?
…………………………………………………………………………………………
Closing Notes:
- Any additional comments or concerns? ______________
- Would you like to share your contact details for future follow-ups? (Optional)
- Yes, my phone number is ______________
- No
# My Take: Reporters / Bloggers / Researchers can depend upon ChatGPT 4.o to create a professional questionnaire. Based on their needs, they can modify it.
The process is quick and fast and saves considerable time.
Leave a Reply