Experimental Designs
Imagine you were conducting research on the relationship between academic performance (e.g., better grades) and different levels of loudness of music (interval scale) while studying.
- How would you design the study using a correlational design?
- How would you design the study using a quasi-experimental design?
- How would you design the study using an experimental design?
Grading CriteriaMaximum PointsQuality of initial posting, including fulfillment of assignment instructions16Quality of responses to classmates12frequency of responses to classmates4Reference to supporting readings and other materials4Language and grammar4Total:40
Class,
A few issues to keep in mind here: In order to perform a correlational design you would need *quantitative* variables: if when one variable increases, the other also increases, then the correlation between the two is positive: That is, they vary together. If when one variable increases, the other decreases, then the correlation between the two is negative: They vary together, but in opposite directions. If the variables are not related–that is, if when one variable increases, you don’t have any information about the behavior of the other variable, then you have a zero correlation. Here
http://www.statsoft.com/textbook/basic-statistics/#Correlations
is one resource that folks might find helpful for some additional information on using and interpreting correlations.
Also, for experimental designs, you’re trying to assess whether or not there is a causal relationship between two variables. For instance, does a change in X cause a change in Y? In our experiment, we would want to manipulate the independent variable X (if X is music, we would specify different types of music), and then randomly assign participants to the different levels of X (types of music), and then see if there is a change in the dependent variable Y, whilst holding all other possible variables constant. Here
http://www.socialresearchmethods.net/kb/desexper.php
is another resource that folks might find useful as you think more about experiments in this context. Hope this helps!