Control of variables strategy refers to the method used when experimenting with a system that has many variables.  To avoid confounded experiments, all variables but those under consideration must be controlled.

I. Definition

In a scientific inquiry process involving many variables, the relationship between the variables needs to be determined.  To do so, we form a hypothesis and test it experimentally.  When designing these experiments, it is important to design controlled experiments rather than confounded experiments.  This means we have to control all other variables in order to analyze the relationship between key variables without interference. For example, when considering the relationship between age and frequency of delinquent activity, the variable of gender has been treated as a control variable.

II. Simplified examples of control of variables

1. Question ID: 20100221100

Shown are drawings of three strings hanging from a bar.  The three strings have metal weights attached to their ends.  String 1 and String 3 are the same length.  String 2 is shorter.  A 10 unit weight is attached to the end of String 1.  A 10 unit weight is also attached to the end of String 2.  A 5 unit weight is attached to the end of String 3.  The strings (and attached weights) can be swung back and forth, and the time it takes to make a swing can be timed.

Suppose you want to find out whether the length of the string has an effect on the time it takes to swing back and forth. Which strings would you use to find out?

1. only one string
2. all three strings
3. 2 and 3
4. 1 and 3
5. 1 and 2

Note: In this problem, there are two variables that may influence the time it takes to swing back and forth: the length of the string and the mass of the attached weight. Students are asked to determine the relationship between the length of the string and the time it takes to swing back and forth, so the size of the weight needs to be controlled (held constant).  The weights attached to the end of string 1 and 2 are the same, but the lengths of these two strings are different. They can be chosen to test the relationship between length and swing time.

2. Question ID: 20100341101

Twenty fruit flies are placed in each of four glass tubes. The tubes are sealed. Tubes I and II are partially covered with black paper; Tubes III and IV are not covered. The tubes are placed as shown. Then they are exposed to red light for five minutes. The number of flies in the uncovered part of each tube is shown in the drawing.

This experiment shows that flies respond to (move to or away from):

1. red light but not gravity
2. gravity but not red light
3. both red light and gravity
4. neither red light nor gravity

Note: In this problem, there are two variables that would affect the distribution of the fruit flies—gravity and red light. To investigate the relationship between red light and the distribution of fruit flies, gravity is treated as a controlled variable. Tubes II and IV are compared since gravity is not having an affect on those tubes.  The two tubes have a very similar distribution of fruit flies, so we know red light did not have an impact.  To test the relationship between gravity and the distribution of fruit flies, red light is the controlled variable.  By comparing Tubes I and II (or III and IV), we see that gravity does have an impact on the distribution.

III. Review of control of variables in research

Control of variables is a necessary strategy in designing unconfounded experiments and in determining whether a given experiment is controlled or confounded.

Control of variables strategy is used in creating and conducting experiments.  Because variables interact with each other, the experimenter needs to make inferences in order to appropriately control variables and interpret the results (Chinn, C. A., & Hmelo-Silver, C. E., 2002). Usually, experimenters focus on the effect of a single variable of interest (Kuhn, D., & Dean, D. 2005).

Control of variables strategy is used in a logical sense to distinguish controlled and confounded experiments, which is necessary in determining whether an experiment can lead to a conclusive result. The logical aspects of control of variables include the ability to make appropriate inferences from the outcomes of unconfounded experiments and to understand the inherent indeterminacy of confounded experiments. In short, control of variables is the fundamental idea underlying the design of unconfounded experiments from which valid, causal, inferences can be made (Chen & Klahr, 1999).

Control of variables is also used in control theory. Control variables are those in a control system.  Consider an experiment where the experimenter is trying to determine the relationship between a chemical’s concentration and the rate of a chemical reaction.  Reaction rate is the dependent variable while chemical concentration is the independent variable.  Everything else that could change the reaction rate must be controlled (held constant) so that only the effects of the concentration are seen.  Variables that need to be controlled in this case include temperature, amount and type of catalyst, surface area of any solid chemicals, and pressure.  If these factors are not controlled, the experiment becomes confounded and loses its validity.

IV. Importance of control of variables

1. The importance of control of variables in learning

Control of variables is one of the National Research Council’s (1996) aspects of “design[ing] and conduct[ing] a scientific investigation” (p. 145).  Several sub-skills are identified in this category including ‘‘systematic observation, making accurate measurements, and identifying and controlling variables’’ (p. 145).

Control of variables is a component of scientific inquiry, which is broadly understood to mean skill in discovering or constructing knowledge for oneself (Dean, D., & Kuhn, D. 2007). It is also one of several types of procedural knowledge—or “process skills”—that are deemed central to early science instruction (Klahr, D., & Nigam, M. 2004).

2. The importance of control of variables in society

An experimental result can be affected by different kinds of variables, so if an experimenter wants to determine which variable directly influences an outcome, all other variables must be controlled while the variable of interest is changed.  For example, if a city planner wants to find out whether temperature affects the comfort levels of a city, all variables other than temperature, such as humidity and cloud cover, must be held constant while temperature is varied.

V. References

http://en.wikipedia.org/wiki/Control_variable

Chinn, C.A. & Hmelo-Silver, C.E. 2002. Authentic Inquiry: Introduction to the Special Section. Science Education, 86, 171-174.

Klahr, D. & Nigam, M. 2004. The equivalence of learning paths in early science instruction: Effects of direct instruction and discovery learning. Psychological Science, 15, 661–667.

Kuhn, D. & Dean, D. 2005. Is developing scientific thinking all about learning to control variables? Psychological Science, 16, 866–870.

Toth, Klahr, & Chen. 2000. Bridging Research and Practice: A Cognitively Based Classroom Intervention for Teaching Experimentation Skills to Elementary School Children. Cognition and Instruction, 18, 423–459.

Chen & Klahr. 1999. All other things being equal: Aquisition and transfer of the control of variables strategy. Child Development, Vol. 70, No. 5 (Sep. – Oct., 1999), pp. 1098-1120.

National Research Council, 1996.

Dean, D., & Kuhn, D. 2007. Direct instruction vs. discovery: The long view. Science Education. DOI: 10.1002/ sce.20194. 1-15.