will show elder boy your post when he comes back home.
I also dunno what he wants to find out. Told me he did not hv prior experience designing a qnaire
DONE :=) Let kaypoh aunty give your son some feedback too, hope he doesn't mind! Like CKS, I'm unable to detect the thesis statement.
1) There should ideally be a null hypothesis from the outset that the research sets out to nullify or support. The qns don't seem to tie to the overarching theme of "food wastage". Qns seem to be designed to elicit consumers' responses on their behaviour/shopping preferences. There is also no apparent correlation to food retailers' practices.
2) It's good practice to classify the qns as it's easier to present/structure his findings on powerpoint slides or report later. E.g., he can have a heading "Product" for all qns pertaining to the quality of produce. "Price" which is not factored in by his team. "Place" for all the shopping venues. Not sure if "People" is pivotal to this study as we dunno his hypothesis and he didn't ask for Gender either (respondents are aunties/maids?). He may classify them differently so long as they dovetail to his thesis statement. Headings should correspond to his ppt deck/report paras with a Conclusion, e.g.: "We conclude that factors (a) Price and (b) Place contribute to consumers' perception of...<thesis statement>".
3) Flow of qns is impt if you want to influence your respondents to respond in a specific way. Q2 & Q3 probe respondents' preferred retailers for fruit/veggies and meat respectively. Then another qn on generic "produce" (fruit/veggies/meat?) comes in as Q4, so is his team assuming respondents' expectations for both fruit/veggie and meat are the same? Then it jumps to Q5 which asks abt "fruit/veggie" only, as though the team doesn't want to investigate how consumer determines the quality of "meat". Also, Q6 pertains to Frequency and shouldn't appear after Q5. This is because Q7 pertains to Quality and is intrinsically linked to Q5. Logically, Q7 should come after Q5.
4) Phrasing of qns. Q5 & Q7 sound like asking the same thing, i.e. how to determine the freshness of produce. But in fact, Q5 is asking for respondents' opinion of what constitutes Good Quality. So it's good to rephrase it as "What do you think are the qualities of fresh food produce?" Q5's Option (1) "Freshness" to me is somewhat redundant. Option (5) "Shape" is a bit iffy cos fruit/veggie's shape has little bearing on its freshness, no? He could have "Colour" (vibrant not dull) or "Fragrance" (aroma or pesticide smell) as alternatives too. Qn (7) on the other hand addresses consumers' preferred action of determining freshness, so "Smell" refers to the action of sniffing.
5) The problem with "Others". This is usually a headache for people compiling the findings. If sway sway he gets many "Others" as returns, his thesis will be sabo'ed. I had 1 market survey carried out last year that came back with many fence-sitting/non-committal "Others" and miscellaneous responses. It was a nightmare for us to finalise the graphs and report, and to get public buy-in.
This is why most questionnaires phrased their qns in Declarative form, rather than Interrogative form. Declarative is like "I press raw meat to decide if it's of good quality" (Yes/No?; Rate from 1-5;Never> Always). Interrogative is like "How do you decide if raw meat of good quality?" The former enables you to dictate how your subjects should respond to skew the findings to match your objectives.
6) Multiple options not recommended. Enabling respondents to pick 1 not more options is very wise for a JC1 project. Luckily he only allows 1 choice. Having more than 1 option per qn is harder to qualify and quantify empirically.
7) Constant test environment/variables. Since the 2 sets of questionnaire are dissimilar due to his team adding new variables like "Occupational Status", earlier returns should be voided or "doctored" retrospectively? A re-test is recommended to preserve the Integrity of the survey. Why? His team can no longer state "Nearly 70% of SAHMs indicated that they prefer to apply pressure on the fruit/veggie to check for freshness". This statement is inaccurate as someone in the 70% didn't indicate his/her occupational status in Test 1. But nebbermind lah...school project only...
8) Wider sampling size. If his team only manages to ask Higher Income Group respondents and they only shop at Organic/Gourmet Shops, he cannot conclusively say: "Some 90% respondents in the >$11,000 income bracket prefer to purchase from Organic shops". It's not representative. Or, if he only manages to get 2 respondents from such income bracket and only 1 of them shop at wet markets, he cannot split his pie and write: "Our study showed that 50% of higher-income consumers prefer to shop at wet markets". Therefore, the larger sampling size, the better.
(9) Outreach can be "controlled"(manipulated). Ask his team to split the job by stationing someone at 1 wet market, 1 supermarket, 1 gourmet shop, 1 wholesale, 1 Raffles Place MRT. This way, they could grab a more "even" distribution of profiles. If all team members' relatives are rich SAHMs, their findings will lack <$1,000 lower-income consumers' responses and he cannot assume a certain group tend to waste food, for example...Seriously, I still dunno what's he's trying to find out.
10) Identify limitations. After the findings are compiled, it's important to outline all the limitations of this survey: What was not tested? What couldn't be tested? Why wasn't it tested? What could've been done to make our results more reliable?
Just my 2 cents and for reference only...Please tell your elder boy NO need to re-do hor...JC1 project nia...