Samples, Variables and Survey Response Rates

[Rudy Renteria]

 

How can you tell the difference betwwen dependent and indepent variable? Also, what is a good method for sampling?  What is an acceptable response rate for surveys? I’m just trying to wrap my mind around this methodology in research….

[Eric Schwartz]

I usually think the simplest way to describe the difference between independent and dependent variables is just to think about a sentence. The dependent variable is the direct object. The independent variable is the subject. What is doing the affecting? What is being affected? The condition of the dependent variable DEPENDS on the independent variable or variables. What is acceptable response rate for surveys? Depends on the survey. How is it being administered. A mail survey could have a response rate as low as 2 percent. Face-to-face surveys could give you a response rate of 40 percent easily. What is a good method for sampling? Again – it depends on your research.

[Abiodun .A. Joachim]

Renteria

Schwartz gave an illuminating reply.Let me give you an example to illuminate what he is trying to put across. For example your growth is dependent on your diet . That is if you eat poor food that will tell on your growth. In this case your diet is independent and your growth dependent onyour diet, you can affect or determine your growth by eating well. Again the nature of your research and your objective will determine the best sampling methods. For example in other to allow a realistic pursuit of a particular data set in line with research objective one may advise purposive sampling method. I suppose your supervisor will be quite heplful in helping you on this or you may need to disclose more of what you are doing for others to meaningfully help

[George Kariuki]

dependent variable relies on the response of the independent variable. e.g when comparing the effects of temperature variation on growth rate, temp. is the independent variable while growth rate is the dependent variable as it relies on temp.

[Kristian Karlson]

There is no inherent difference between dependent and independent variables. Whether variables take one form or the other depends on your theory. The best method for sampling is simple, random sampling. Each individual has the same probability of being in the sample. This opens for a lot of nice properties (e.g., generalization). Acceptable response rates can’t be defined in general. It depends on your research interest and how many variables, you are interested in. Making reasonable assumptions about what kind of effect you want to find, you may make so-called power analyses (calculating the power of your tests).

Kristian

[Author]

Posts