Wednesday, September 24, 2008

Problems with Data Gathering from Surveys

In the section on "Data Gathering" (page 448 of textbook), one of the methods recommended for gathering data is from Surveys. The textbook however warns that the method may not be accurate for various reasons, such as explaining the cause of perceptions, the social context under which a person holds an opinion, and the problem of quantifying an experience that may not be quantifiable.

One limitation of Surveys that was not discussed is that the wording of the question posed to participants may be interpreted in multiple ways, causing respondents to give answers that may not be the true intention of the questioner. For example, the survey question, "Do you own your home?" may be answered "yes" by a person because he/she does "own" the place, but that omits the fact that there is a mortgage on the house that will most likely not be repaid, and the person is likely to lose the house any moment. So the question is ambiguous - it does not distinguish between: (1) those who own their home after all the mortgage has been paid, (2) those who own their home but mortgage is likely to be paid in a few years, and (3) those who own their home now but are likely to have it foreclosed.

1 comment:

cathyblog08 said...

I completely agree with your comments. I would also add that many times, when taking a survey I have to pick very definite answers. And sometimes the way I want to answer the question is not one of the options. And also many time the answer I'd like to give is just "it depends". Surveys do not allow participants to expand on their answer or have anyway to ask for clarification. I believe they are great as an initial step to gather general data, but if a company is really serious about finding information then other complementary methods should be used as well.