Control of Variables
Extraneous Variables
The key to an experiment is that an independent variable is manipulated to see how this affects the
dependent variable. The only thing that should affect the DV is the IV. Any other variables that
would interfere with the IV or DV should be controlled and removed. These additional, unwanted
variables are called extraneous variables (EV), and these should be identified at the start of the
experiment by the researcher, who should take steps to minimize their influence.
Many extraneous variables are straightforward to control such as the age of the participants and the
lighting in the lab etc. These are “nuisance variables” that do not vary systematically with the IV.
They do not confound the findings of the study and may just make it harder to detect a result.
Confounding Variables
These do change systematically with the IV. Personality can be a confounding variable as this would
become an independent variable- for example when doing an experiment for an energy drink, if all
the participants in the variation drinking the drink were extravert types and the control group with
water were introverts, this is a confounding variable. The problem was that extraversion varied
systematically with the IV and this alone could explain changes in the DV.
Demand Characteristics
Participants are not passive within experiments and are likely to be spending much of their time
trying to make sense of the new situation they find themselves in. Participant reactivity is a
significant extraneous variable in experimental research and one that is exceedingly difficult to
control. The demand characteristics (clues or cues) of the experimental situation may help the
participant to “second-guess” the experimenters’ intentions as well as the aims of the study.
Participants may also look for clues to tell them how they should behave in the experimental
situation. They may act in a way they think is expected and over-perform to please the experimenter
(the “please-U effect”) or they may underperform to sabotage the results of the study (the “screw-U
effect”). Either way, the participants behaviour is no longer natural and affects the result.
Investigator Effects
Participant reactivity also leads to investigator effects. For example, if we are expecting the energy
drink group to speak more than the water group- we may unknowingly in our unconscious behavior,
encourage a greater level of chattiness from the energy drink participants. The investigator effect
refers to any unwanted influence of the investigator on the research outcome.
Hugh Coolican (2006) points out, this can include expectancy effects and unconscious cues. It might
also refer to an action of the researcher that were related to the study’s design such as the selection
of participants, the materials, and the instructions.
Leading questions which are discussed in relation to eyewitness testimony are a good example of the
power of investigator effects.
Randomisation
This is an attempt to control investigator effects. Randomisation refers to the use of chance
wherever possible to reduce the researchers influence on the design of the investigation. The order
of the list of words to recall in a memory experiment should be randomly generated so that the
position of each word is not decided by the experimenter. In an experiment, where participants are
1
Extraneous Variables
The key to an experiment is that an independent variable is manipulated to see how this affects the
dependent variable. The only thing that should affect the DV is the IV. Any other variables that
would interfere with the IV or DV should be controlled and removed. These additional, unwanted
variables are called extraneous variables (EV), and these should be identified at the start of the
experiment by the researcher, who should take steps to minimize their influence.
Many extraneous variables are straightforward to control such as the age of the participants and the
lighting in the lab etc. These are “nuisance variables” that do not vary systematically with the IV.
They do not confound the findings of the study and may just make it harder to detect a result.
Confounding Variables
These do change systematically with the IV. Personality can be a confounding variable as this would
become an independent variable- for example when doing an experiment for an energy drink, if all
the participants in the variation drinking the drink were extravert types and the control group with
water were introverts, this is a confounding variable. The problem was that extraversion varied
systematically with the IV and this alone could explain changes in the DV.
Demand Characteristics
Participants are not passive within experiments and are likely to be spending much of their time
trying to make sense of the new situation they find themselves in. Participant reactivity is a
significant extraneous variable in experimental research and one that is exceedingly difficult to
control. The demand characteristics (clues or cues) of the experimental situation may help the
participant to “second-guess” the experimenters’ intentions as well as the aims of the study.
Participants may also look for clues to tell them how they should behave in the experimental
situation. They may act in a way they think is expected and over-perform to please the experimenter
(the “please-U effect”) or they may underperform to sabotage the results of the study (the “screw-U
effect”). Either way, the participants behaviour is no longer natural and affects the result.
Investigator Effects
Participant reactivity also leads to investigator effects. For example, if we are expecting the energy
drink group to speak more than the water group- we may unknowingly in our unconscious behavior,
encourage a greater level of chattiness from the energy drink participants. The investigator effect
refers to any unwanted influence of the investigator on the research outcome.
Hugh Coolican (2006) points out, this can include expectancy effects and unconscious cues. It might
also refer to an action of the researcher that were related to the study’s design such as the selection
of participants, the materials, and the instructions.
Leading questions which are discussed in relation to eyewitness testimony are a good example of the
power of investigator effects.
Randomisation
This is an attempt to control investigator effects. Randomisation refers to the use of chance
wherever possible to reduce the researchers influence on the design of the investigation. The order
of the list of words to recall in a memory experiment should be randomly generated so that the
position of each word is not decided by the experimenter. In an experiment, where participants are
1
