Confounds
From PsychWiki - A Collaborative Psychology Wiki
Definition
One of the advantages of using the experimental method is that it allows a relatively unambiguous interpretation of the results. Because the researcher keeps all variables other than the independent variable constant, the researcher can conclude that a change in the independent variable is what caused a change in the dependent variable. Variables, other than the independent variable, that are not kept constant are called confounding variables.
A confounding Variable is an extraneous variable (a variable that is not a focus of the study) that is statistically related to (or correlated with) the independent variable. This means that as the independent variable changes, the confounding variable changes along with it. Furthermore, when confounding occurs, the effects on the independent variable and an uncontrolled variable are intertwined and the researcher cannot determine which of the variables is responsible for the observed effect (Cozby, 2008). Confounding variables add error to an experiment and compromise a study’s internal validity. A typical error associated with confounding is a type I error; an erroneous 'false positive' conclusion that the independent variable causes a change in the dependent variable (Triola, 2007). As such, a good experimental design will eliminate possible confounding that will result in alternative explanations
The NPR video portrays a study in which personal space violations and people’s responses to these invasions are examined in the virtual world Second Life. The video explains how Stanford psychologist, Nick Yee, studies social behavior by creating avatars and interacting with other people (via avatars) in Second Life. The video implies that the participants in Second Life are unaware that their interactions are experimental manipulations staged by psychologists studying human behavior. This study demonstrates that personal space violations (the independent variable), defined by extreme staring and invasion of personal space), elicit the same reactions (the dependent variable) in Second Life as they do in the real world.
Application
Is confounding involved in the experiment? While the parameters of the experiment are only vaguely defined, because of the nature of the experiment one can only assume that there is some confounding that may affect the results of the experiment. For instance, while the experiment is clever and the researchers offer precise definitions, controlling for extraneous variables in a “naturalistic setting” is challenging and in some cases impossible. Thus, it is difficult to establish a causal relationship between the independent variable and the dependent variable. Such is the case for Yee’s experiment in Second Life.
For example, the video implies that when personal space is violated, the participant averts their gaze and/or backs away in effort to break the tension. The problem here is that the research team assumes that a person/avatar moves because they feel uncomfortable. Because the participants are unaware that they are involved in an experiment and do not provide any information about their motives for moving, it is unfair to assume that a person/avatar moved because they felt uncomfortable. For instance, in the case of the girl who flew away, it is impossible to know if she logged out of Second Life or if she fled the scene because she was uncomfortable.
References
Cozby, P. (2008). Methods in behavioral research. Columbus, OH: McGraw Hill.
Triola, M. F. (2007). Elementary statistics. Reading, MA: Addison Wesley.
◄ Back to Stenstrom Lab Group page
