|
|
Non-experimental
|
Experimental
|
|
Correlational
|
Differential
|
Quasi-experimental
|
True
Experiment
|
|
Manipulation
|
No
manipulation of variables.
|
No
manipulation
of variables.
|
Manipulation of the independent variable
|
Manipulation of the independent variable
|
|
Subjects
|
Subjects are NOT assigned to groups. Usually, there is only ONE group of
subjects.
However, subjects are Randomly SELECTED for
participation.
|
Subjects cannot be randomly assigned to groups. The groups of subjects
differ on some PRE- EXISTING
variable (ex: gender)
Subjects should still be randomly selected for participation
|
Subjects are NOT randomly assigned to control and experimental
groups because it is logistically difficult (e.g., comparing 3rd
period and 5th period AP psych classes after each class has be "treated"
differently.) But, there are control & experimental groups in
this type of design....just no random assignment.
If possible, they should be randomly selected for
participation.
|
Subjects are randomly assigned to control and experimental
groups.
(Ex: control group gets regular
teaching and the experimental group gets new teaching method)
If possible, they should be randomly selected for participation.
|
|
Variables
|
Two variables (X and Y) are measured
and the STRENGTH and DIRECTION of the RELATIONSHIP is determined.
(Ex: measuring GPA and depression
level)
|
Subjects are divided into groups based
on a pre-existing variable (X)
(such as sex, religion, etc.) and compared on some other variable (Y)
(i.e., IQ, self-esteem, depression, anxiety, etc.).
|
Subjects are in pre-formed groups.
But, unlike correlational and differential research, an independent variable (IV) is manipulated and the groups are measured & compared on a dependent variable (DV). (Ex: Using
one teaching technique with 3rd period and a new technique with
5th period. Then the two classes would be compared on
final grades (the DV) to see if a statistically
significant difference existed)
|
The Independent variable (IV) is manipulated and the dependent
variable (DV) is measured. The groups’ scores
on the dependent variable are
then COMPARED to determine if a STATISTICALLY SIGNIFICANT DIFFERENCE
EXISTS.
|
|
Statistics
|
Pearson product-moment, correlation
(Pearson’s r)
|
Chi-square, t-test, ANOVA, point-biseral correlation
|
Chi-square, t-test, ANOVA
|
Chi-square, t-test, ANOVA
|
|
Conclusions
|
Variable X co-varies with variable Y
(i.e., there is a relationship between
the two variables.) Cause and effect cannot be proven.
|
Differences in variable X may be RELATED to the differences in
variable Y, but cause and effect cannot
be proven.
|
While we may be able to draw some causal conclusions, we can’t do
it with as much confidence as if we had used a TRUE experimental design.
(This is due to lack of random assignment and other controls).
|
Changes in the IV CAUSED changes in the DV. We can be most confident when we have
controlled for as many threats to internal validity as possible.
|