Identifying the Population and Variables

By Dr. Elizabeth Pearman
Copyright 2009

            You now have a research problem and have probably asked a question but in order to ask a really good question, you need to be able to identify the population you want to use in the study. One thing to think of in identifying the population is to keep in mind where you might find this population and how will you have access to them. The population is the group of individuals you want to include in your study – the group you are proposing to study. They might be third grade students, teenagers in juvenile facilities, 10th grade students, boys and girls in preschool, Mothers, or maybe something bigger like studying differences in countries. But you population is the group you want to study.  You might want to put some specificity to identifying your potential population by being more specific so rather than Mothers or boys and girls you might specify an age group, or a typology, or a diagnosis. But the population addresses the individuals you want to study in your research. In order to ask a good question, you need to know who you want to study. Sometimes you also need to think about where you are going to find the individuals.
Independent and Dependent Variables
I know quite a few statisticians and people with doctorates with little understanding of independent (IV) and dependent (DV) variables.
The IV comes first in time or is antecedent to the DV.
For example – Is individual social skills counseling more effective than group social skills counseling for elementary school students? Based on what we have leaned about questions we might want to refine it a bit to establish who is the population. Elementary school students is a pretty big group so perhaps we decide to use only third grade students or some other grade. You might want to further refine this to third grade student with behavior problems, or below average or with disabilities or maybe gifted. So let’s decide on third grade students with disabilities. The independent variable is individual counseling or group counseling. Now we come to the DV – or what is the result of the two types of counseling for an example let’s use a social skills instrument – the School and Social Skills Survey since the counseling has been about. The diagram below should give you a good idea – there are a couple of examples. Do you see how the IV comes before or is antecedent to the DV?

Does high school driver education do any good?

How are dogmatism and political attitudes related?

In the third example, the IV and DV are not identified simply because it cannot be determined which is antecedent to the other.

            In a correlation study, you still specify the population but don’t really have IV or DV since you are asking about the relationship between two variables. In simple linear and multiple regression, the predictor variables are the IV and the variable being predicted is the DV. Such as, do age, height, and weight predict blood pressure in 45-50 year old males? The IVs are age, height, weight – since they are being used to predict blood pressure, so blood pressure is the DV.  You can use the same chart to figure out the IV and DVs in a simple or multiple regression.