Variable Control

Variables are commonly thought of aspects of research. They are. However, writers of all kinds deal with variables, especially fiction writers. I address both fiction and research variables.

Constants in Fiction

A story contains many moving parts. Things change all the time, for example, characters’ ages, interests, and relationships; story location; plot via twists. No matter what variables there are, and no matter how much those change, a writer needs to remember the story’s constants. For instance, if a character or groups of characters move into a location at a certain time, that must be remembered. An author most likely would not consider having the character move to, say, Arkansas five months ago in the beginning of the story only to refer to the character as having been in Arkansas for the last year a few pages down the line.

Variables and Control

Variables, on the other hand, can change at any time. If a character despises a workmate when we first meet him, when he decides he can stand the colleague in the next chapter, we can accept that. Acceptance is not without conditions, though. This is where control plays the game. The reader needs to understand how the character came to the change of mind. Even if the change is critical to the plot, it makes for poor storyline to leave variables to their own devices, so to speak. The reason the character had a change of heart could be summed up in one or two sentences or it may be a gradual shift shown to the reader rather than told. No matter what, control your variables! Take note of them and attend to them.

Research Constants

In research, specifically educational research, constants are anything about your sample that do not vary. In a study on third graders, the students’ grade level is constant regardless of the rest of the study. Sometimes there may be multiple constants such as age, gender, and socioeconomic status (SES). That might occur in a study regarding 12-year-old girls of low SES and their attitudes toward going to college. The dependent variable is their attitudes.

Variables and Control

Gaining control of research variables includes moving step-by-step through the proposed methodology and determining if any variables are not being measured, and, if so, whether that is desirable, acceptable, or a problem. You will also need to continually revisit your research questions to ensure your variables are being properly addressed.

There are different types of variables: independent, dependent, moderator, and extraneous. They play different roles in your design and study. The independent variable is something that will have an affect on dependent variable, such as a curriculum, a program, a lecture, a group classification if comparisons are made. The independent variable is applied to a situation, a certain type of grouping, or something manipulated in the study.

Dependent variables are those characteristics that vary. If your study is on children of all ages, age is a dependent variable. If you want to determine how many teachers at a specific site respond to using a new curriculum, the variables are as follows:

Curriculum                       Independent

Gender                            Dependent

Age                                 Dependent

Years taught                    Dependent

Grade level/subject          Dependent

School site                       Constant

When analyzing data, if you have kept track of all data for all variables, there is much you may do with that information. You may want to see if men ages 24-35 more positively responded to the curriculum than men ages 36-54. That may be valuable information. The data can also be split by gender as well as by years taught or grade level. There may be no differences between the curriculum response and most or any of the variables.

If, in the same study, you find a difference between how men and women responded to the curriculum, gender becomes a moderator variable.

Finally, extraneous variables are those which cannot be controlled. It is critical you decide how you will address these variables in your design as well as in your analysis. You can only control what you do with these variables since they themselves cannot be controlled.

Deciding on your variables has much to do with your research questions. But knowing your variables and controlling for them is key.

See also: http://www.gifted.uconn.edu/siegle/research/variables/variablenotes.htm