Designing a Survey My Research

Designing a Survey My research question was essentially this: what items are best for serving at an elementary school fundraising luncheon? I also wanted to know what guests were willing to pay and whether I could get some degree of work out of them in addition to mere money. A needed to take into consideration a number of food-related factors. First, there are some foods which people cannot or will not eat.

For instance, Buddhists are opposed to eating any kind of animal flesh, so provision should be made for immigrant families from predominantly Buddhist countries. Similar considerations must be made for the dietary requirements of Hindus, Jews, and Moslems (approximately 31% of the local population outside Irvine, CA consists of Asians, many of whom may be Moslems, Hindus or Buddhists). Further, there are some people who prefer vegetarian foods because they view these as more healthful than any other.

While a strictly vegetarian cuisine would satisfy the "letter of the law" as far as keeping everyone fed with a clear conscience, it would not necessarily keep everyone happy. There are plenty of people who would be happy eating pork roast or sauteed beef tips; they won't be happy if everything is vegetarian.

Also, it is important to remember that in a population of approximately 300 students, there will be at least 300 parents, and if only half of them come to the fundraiser there is a good chance that three or four people will be allergic to at least some of the food we serve, so some consideration has to be made in that area as well. It is impossible to serve one dish that will cover every base. Variety is the answer, but variety alone becomes expensive.

If we offer French, Thai, Moroccan, German, Mexican, Japanese, Brazilian, Indian, Ethiopian, Italian and traditional American food we will find that most people go for only one or two varieties. This means that there will be a lot of leftovers (lost revenue) and very likely a lot of disgruntled people who didn't get their fill of their favorite cuisine. So to get at this information, I "cooked up" a cuisine questionnaire. I broke the meal down into Appetizer, Salad, Main Dish, Side Dish, Dessert, and Drink categories.

I offered survey subjects Asian and Mexican vegetarian/non-vegetarian food and named a few more typical selections for each category, such as pizza for the main course and ladyfingers for an appetizer. This offered subjects a good compromise between variety and practicality, and helped keep the survey sufficiently brief to allay subject fatigue. However, for each category I also allowed for a "fill-in-the-blank" open-form response. This indeed proved a wise measure as it yielded some surprising data as well as giving me ideas for inclusion in next year's survey.

Part II - Research Population Regarding the population, I would ideally like to poll the families of all 300 students. This was not practical - implementing the survey would have cost nearly as much as the luncheon itself. So, rather than try to measure the entire population, I chose a sample consisting of the 30 members of the local PTA, most of whom have or have had students in the school in question.

Although there is some range restriction in this sample because not all board members are parents of children currently in the elementary school, I consider their greater degree of expertise from prior projects more than compensates for this. Given that only 7.4% of the population around Irvine is of Hispanic/Latino background, I expected that very few people would respond favorably toward Mexican food as a choice.

However, the results from the pilot study were surprising: 50% marked Mexican food as an option in every instance where it occurred, and 80% chose to write-in "tortilla chips and salsa/queso" under the Appetizer category. I would not have gotten this information had there been no open-form item on the survey. Part III - Sampling Technique The sampling technique will be simple and straight-forward.

My intention is to distribute the survey to the thirty members of the local PTA at its next meeting with the request that they fill them out and return them to me by the meeting's adjournment.

Since I am also interested in getting volunteers to help out at the luncheon, I will announce this fact and ask that anyone willing to do so should write their contact info on the survey's reverse side; when they get to the part of the survey asking for volunteers, it will be the second time they are exposed to the idea of volunteering. Hopefully, multiple exposures will increase the positive response rate for that item.

If some members of the board are absent, I will gladly redistribute the extra surveys to whoever is there. The surveys are sufficiently brief it is unlikely any survey will be returned uncompleted. For items requiring a check mark, analysis will require a simple tally...25 subjects checked ladyfingers, 10 checked pizza rolls, only one checked vegetarian hors' doeuvres, and so forth. For the part requiring subjects to rank six items in order of importance, I can get meaningful results by summing the scores for each item across subjects.

Since sample size is limited to thirty observations, I feel comfortable with the write in question about price range; it is unlikely that there will be much variance in responses given the sample size and sample demographics. Further, giving subjects an opportunity to offer their own suggestions for price gives more descriptive feedback than giving them a "circle-one" list.

If they see $5 as part of a "circle-one" list, then they may infer that $5 is a reasonable price and immediately circle that when they would otherwise have responded with a higher figure in response to a write-in question. A circle-one question about pricing would have been implicitly leading, which should be avoided in any event but certainly so when it might have a negative impact on revenue.

Part IV The survey design has been sufficiently described above by way of explaining other aspects of the research that I can forego repeating the description here. The survey itself is attached as an appendix. Part V - Pilot Study In preparation for the actual survey's distribution to the PTA board, I conducted a pilot study and distributed the survey to ten strangers in a local grocery store (with the manager's permission). Subjects received the survey attached to a clip board and were provided with pencils for scoring.

The most surprising thing was that 8 out of 10 people surveyed wrote-in some version of "chips & salsa" as an appetizer. This is an item I would have included in the appetizer list as a choice had it occurred to me. Fortunately, other people have good ideas as well. A learned that in spite of the local population consisting mostly of "white" Caucasian or Asian people, Mexican food was selected 50% of the.