RESEARCH BRAINTEASER
Last week, you commented on a school research assignment in which you conducted a focus group to gather qualitative data to assist the school’s cafeteria in planning upcoming menus. This is a continuation of that same question.
After conducting your cafeteria menu focus group, you realize that you want even more data to substantiate what you learned in the focus group. You feel that some sort of survey would be applicable. You don’t have the money to mail a survey to staff and students nor do you have the help to telephone all staff and students.
Question
What type of survey method would be applicable for this research and why?
Subscribe to:
Post Comments (Atom)
An online survey would be best. this would allow me to reach a large number of people quickly. This would also eliminate the need for staff or mailing materials and is very low cost.
ReplyDeleteHi Correne -
ReplyDeleteWhile online surveys appear to be the most popular way to go (we learned last week that 88% of research firms are conducting online surveys), can you list off some advantages as well as some disadvantages to utilizing online surveys?
i would agree with correne on an online survey. also a self serviced survey, maybe at the door of the cafeteria might be beneficial as well.
ReplyDeleteAdvantages: Fast,convenient, low cost, large reach.
ReplyDeleteDisadvantages: No personal involved to probe for more questions, unsure who you are actually reaching.
Hi Frankie -
ReplyDeleteA self-administered survey is another good option, esp. doing so at the cafeteria entrance. Can you list off some advantages and disadvantages to a self-administered survey?
oh by the way, i have a new number so the old one doesn't work anymore
ReplyDeleteadvantages - people can easily access it right then and now. not costly
ReplyDeletedisadvantages - nobody can ask to elaborate on the questions further or answer any confusion the person might have.
You both are correct! Re: self-administered surveys, these are quickly becoming kiosk-based computer questionnaires with touch screens, which increase response rate and accuracy. So the days of pencil and paper may soon be over!
ReplyDeleteOkay, let's do a review of what we learned in class last week and read about in Chapter 5 before moving on to new and exciting research topics!
ReplyDeleteLast week in class, we discussed several types of survey errors, specifically sampling errors, sample design errors and measurement errors. Let’s review these error types.
What is a sampling error?
It is the difference between the sample and population. the solution to avoid this problem is to increase your sample size.
ReplyDeletesampling error are different samples and population information that has been gathered. its when the sample is small and doesn't cover everybody in the whole picture
ReplyDeleteRight. A sample error is the difference between the sample value (i.e. 10 people) and the true value of population (i.e. 100 people).
ReplyDeleteAnd you answered my next question, which is how can a sampling error be reduced? From what we learned in our textbooks, a sample error can never be eliminated (it's a byproduct of doing research), but it can be reduced by increasing the size of our sample.
We talked about three different kinds of sample design errors. Can you describe any of these types of sample design errors?
ReplyDeletepopulation error: incorrect sample like people who dont fit those you want to target.
ReplyDeletepopulation error (wrong target involvement)
ReplyDeleteframe error (true samples)
selection error (didn't follow correct sampling procedures)
Correne - Can you give me an example of a population error, something you may encounter in research?
ReplyDeleteFrankie - Can you give me an example of a frame error, something you may encounter in research?
is frame error the one involving selective people by being bias? like the example in class, a run down house wouldn't want to answer survey questions as much as a nice well kept house?
ReplyDeleteIf I wanted to get opinions from senior citizens on how the felt about current health care, I may encounter a population error if I utilized a kiosk-based computer survey in the mall. I would get opinions from people of all ages.
ReplyDeleteExactly Correne. A population error is an incorrect definition of the population to be selected.
ReplyDeleteEx: population defined as 20-25; later determined that an older population should be included, i.e. 50-75. If a difference between two populations exists = Bias.
Frankie - You are confusing a frame error with a selection error (which fits the example you provided).
ReplyDeleteA frame error is when you use an incomplete/inaccurate sample population that doesn't represent your true population.
Ex: using a telephone book for a telephone survey. You will encounter households that are not listed, people moved, wrong numbers, etc.
And for your notes, and research know-how, a selection error is when sampling procedures are incomplete or when sample selection procedures are not properly followed.
ReplyDeleteEx: door-to-door interviewer avoiding houses that look messy because they think resident would be unpleasant.
Lastly, the most common form of survey errors is called a measurement error, i.e. +/- 5%. What does the +/- indicate?
ReplyDeleteThat is the allotted margin of error that must also be considered.
ReplyDeleteits a "give or take" way of showing that the numbers shown are not accurate per say and it maybe be off by a few points.
ReplyDeleteRight. The +/- indicates the difference between the research information being sought and the research information obtained.
ReplyDeleteWhen designing a survey or questionnaire, why is learning about errors important to us as advertising researchers?
ReplyDeleteso that we can get the most accurate numbers or involvement from our target market as possible without leaving many people out. we cannot obtain everybody in out target market but we can narrow down our options to primarily get a focus on who we want our efforts to involve
ReplyDeleteBecause errors will effect the outcome of our information. This can result in wasted time and money. Worse can result in bad campaign
ReplyDeleteYou both present good ideas, especially as you state Correne, errors affect the outcome of the research.
ReplyDeleteI will urge you both to reread and become familiar with page 147 in your textbook, esp. as you go out and do your own research projects.
i like correnes answer. short and to the point and informative <3
ReplyDelete<3
ReplyDeleteSo this week, we are going to build on our knowledge of surveys (also called a questionnaire) and learn how to design and write the best questions to get the most impact out of our questionnaire.
ReplyDeleteFirst things first, what exactly is a questionnaire?
This comment has been removed by the author.
ReplyDeletea questionnaire, when you divide the words up, is part "question" and part "aire". overall its a list of questions to get people to answer, like a survey
ReplyDeleteIt is a list of questions designed to gather info on peoples opinions and thoughts.
ReplyDeleteExactly. A questionnaire is a set of questions designed to generate data to achieve the goals of a research project.
ReplyDeleteThe key when designing a questionnaire is that it is a standard set of questions asked so that ALL respondents get the same questions.
What happens if we make custom questionnaires to handout to certain respondents?
That could get confusing fast. You would have to make "respondent identification" part of the questionnaire.
ReplyDeleteThere are three criteria that we must answer when starting to design a questionnaire. If we can check these items off with a “yes”, then we’re probably on the right path.
ReplyDeleteThe first set of criteria for questionnaires is does it provide decision-making information. What does this mean?
making custom questionnaires would make it difficult to focus on one particular type of objective that is based on specific questions versus random ones. and i agree with correne again, "respondent identification" would be necessary to be more organized.
ReplyDeleteThat might mean do the questions motivate choice oriented thought rather than yes or no answers.
ReplyDeleteYou are a little ahead of the curve, Correne, but I can tell you've done your reading!
ReplyDeleteCriteria 1 in designing a questionnaire - decision-making information - simply means...
does the questionnaire meet the research problem and objectives. Will doing a questionnaire get you the research results you are looking for.
The second criteria for questionnaires is does it consider respondents. What does this mean?
ReplyDeleteisn't that a debate type of question? like an internal decision between two brands used before and the feels of the consumers experience towards each brand?
ReplyDeleteFrankie - I'm not sure what you mean. Are you referring to criteria 1 or 2?
ReplyDeleteThis may mean are the questions designed to consider who will be answering them. i.e. Questionnaires for teenagers would be worded in such a way to keep their attention and not talk over their heads.
ReplyDeletewell the respondents overall will decide how to answer the questions. for example, if the questionnaire was designed for 18-24 year olds of an upper class level and asked them specific questions they would be familiar with but a person that is of lower middle class was approached to answer it then they are left out in the specifics of the questions?
ReplyDeleteCriteria #2 we must keep in mind is the questionnaire topic and type of respondents, i.e. adults vs. children; therefore, use of different language.
ReplyDeleteAnother issue to keep in mind is your sample population’s ability to read/literacy and/or does the questionnaire contain difficult or technical language.
Also, are there language issues involved? Should you produce multiple versions of your questionnaire, say in English and Spanish?
The third criteria for questionnaires to check off on ou list is does the questionnaire meet editing and coding requirements.
ReplyDeleteIn your opinion, what is editing and coding as it pertains to questionnaire design?
i was referring to the first part
ReplyDeleteAny thoughts on editing and/or coding?
ReplyDeleteEditing and coding would ask if it had been through a review or editing process by someone else or even a test sample.
ReplyDeleteYou got the first part Correne re: editing.
ReplyDeleteEditing is the process of going through questionnaire as would a respondent to ensure questions are being asked in a logical manner. (Think about when you write a paper and you reread it to make sure it makes sense. Same concept.)
Coding is a little more complicated, and it involves grouping and assigning # codes to question responses. We can discuss more on Thursday. However, please see page 289, ex. 9.3 and page 298 for examples of coding.
Let's move over to Ethics.
ReplyDeleteRES. HOMEWORK REMINDER
-Chapter 9 for Thursday, Mar. 5th
-Chapter 5 questions #1, 5, 6 by Thurs. March 5th
-Frankie to present an Adv. Res. Current Events presentation on March 5th
-Part II Semester Res. Project: “Research Goals & Objectives” due March 5th
if we are referring to the questionnaire still, editing would be a skipping of questions directed towards how they answered the previous question.
ReplyDeleteThat's true Frankie. What you just described is called a "skip pattern". If a certain question is answered, the respondent moves to the next question. Or, if a certain question is not answered or the respondent answers "no" (instead of "yes"), then they are moved to another set of questions.
ReplyDelete