We use cookies to enhance your experience on our website. By continuing to use our website, you are agreeing to our use of cookies. You can change your cookie settings at any time. Find out more

Chapter Summary

When psychologists discover a relationship between two variables, there are a number of possible interpretations. Most importantly, they cannot be sure that a third variable is not the source of that relationship unless they conduct an experiment. Experiments allow for the elimination of alternative explanations for research findings. In order to conduct an experiment, a researcher must manipulate the independent variable, hold constant all other variables, and measure the outcome on a second, dependent variable. This allows them to determine that changes in the independent variable directly cause changes in the dependent variable.

The independent variable oftentimes has varying levels. For example, looking at the relationship between heat and aggression, researchers might manipulate the temperature of a room with different levels (e.g., 20 degrees versus 30 degrees). There would then be two different experimental groups of people: those who participated in the study a room-temperature environment and those who participated in a hot environment.

Another way to introduce different groups within the study is by having a control condition. Control groups serve as a sort of baseline measure; they represent normal, everyday responses in the absence of intervention by an experimenter. Placebo groups are related to control groups. Placebo groups receive “treatment,” but this is a benign or ineffective version of the treatment provided to the experimental group.

In addition to controlling for alternative explanations for a relationship between two variables, the power of experiments is that they allow researchers to eliminate causal directions. If X is manipulated and produces a change in Y, they can say with some certainty that X causes Y, rather than Y causing X. Although this is highly beneficial, it is also important to note that experiments are not always possible or appropriate.

In independent sample designs participants in the study experience only one condition of the study. This can create concern around non-equivalent groups, a situation where relevant personal characteristics may be disproportionately represented in one of the conditions, creating an experimental confound. In order to ensure equivalent groups, researchers rely on strategies such as basic random allocation, pre-testing, and representative allocation.

Alternatively, repeated measures designs deal with the problem of non-equivalence by exposing all participants to all the conditions of the study. However, this strategy has its own shortcomings, such as concern over effects related to the order that the conditions are presented. These order effects can be addressed by counterbalancing the order of materials and/or allowing sufficient time between presentation of the different conditions.

Finally, matched pairs designs attempt to create equivalent groups by matching participants along some meaningful criterion. Then, one participant is assigned to one condition, while the other participant is assigned to the opposite condition. This ensures that existing differences along this criterion are equally represented in the conditions of our study.

Sometimes, small n designs are warranted in order to answer our research questions. These studies may have a handful of participants, or possible even just one. Oftentimes this research is conducted with very unique individuals, where a larger sample is not possible, as is the case with those suffering from rare psychological conditions.

Up to this point, discussion of psychological research has focused predominantly on studies with a single independent variable. However, researchers often time include more than one independent variable in their studies. These independent variables are referred to as factors, and studies with multiple independent variables are called multi-factorial designs. These studies allow us to look at the simultaneous impact of different independent variables on the dependent variable.


Additional Online Resources

Online video and accompanying quiz on independent and dependent variables by Nathan Lampson: http://www.sophia.org/tutorials/independentdependent-variables--3

“Is this causal claim justified?”: http://www.everydayresearchmethods.com/current-affairs/

Links to various public press articles that claim causality—or do they? http://jfmueller.faculty.noctrl.edu/100/correlation_or_causation.htm


Flashcards

Test your knowledge of the keywords and definitions in the chapter.

 


Interactive Quiz for Chapter 5

Instructions: For each question, click on the radio button beside your answer. When you have completed the entire quiz, click the “Submit my answers” button at the bottom of the page to receive your results.

Question 1:


a) placebo group
b) experimental group
c) control group
d) attention group

Question 2:


a) independent variable
b) dependent variable
c) unrelated variable
d) confounding variable

Question 3:


a) there are asymmetrical order effects
b) conditions are presented in random order
c) there is too much time between conditions
d) there is too little time between conditions

Question 4:


a) random allocation
b) representative allocation
c) pretesting
d) using an independent samples design

Question 5:


a) manipulation of an independent variable
b) measurement of a dependent variable
c) measurement of other participant variables
d) holding constant other variables

Question 6:


a) determining causality
b) greater certainty in findings
c) increased statistical significance
d) standardized procedures

Question 7:


a) complex experimental design
b) multi-factorial design
c) interactive experimental design
d) multi-level design

Question 8:


a) repeated measures design
b) between-groups design
c) unrelated groups design
d) mixed-groups design

Question 9:


a) the impact of order effects
b) losing more data when a participant drops out
c) requires extra time to complete
d) none of the above

Question 10:


a) placebo test
b) reactivity measure
c) baseline measure
d) difference value