Day 10 Table of Contents

• Review the midterm exam
• Causal-Comparative vs. Correlational vs. Experimental research
• Designing Causal-Comparative Research
• Interpretation of Scores
• Statistical symbols
• Closure

Definition and purpose: Causal-comparative research

"Causal-comparative, or ex post facto, research . . . attempts to determine the cause, or reason, for existing differences in the behavior or status of groups of individuals." Gay, p 321.

Used to explore causal relationships, not confirm them. Like correlational research, causal-comparative research is used to describe differences that already exist.

Although causal comparative attempts to show cause and effect, results can not be strongly stated. We may not know which variable was the cause, or a third variable caused both.

Causal comparative variables are not manipulated because they:

• can't (such as a person's gender);
• should not (such as whether a person smokes cigarettes); or
• are simply not manipulated (whether or not a person has a computer at home).

Designing Causal-Comparative Research

1. Statement of the research problem Speculate about the causes of the phenomena from previous research, theory, or observations	"Students who used computers at home before before attending first grade will have higher achievement scores at the end of elementary school than students without computers at home."
2. Define the group that possesses the characteristic you wish to study.	Look for the smallest homogeneous group that contains the critical variables.
3. Select comparison group	The characteristic or experience that differentiate the groups must be clearly and operationally defined (each group represents a different population). Control extraneous variables to help insure equality of groups: Pair-wise matching on members of the two groups Compare similar subgroups (ex. high, medium, and low) Factor analysis allows statistical comparison of independent and control variable together and in combination Statistically equate groups by covarying the variable under study
4. Collect data	Any kind of instrument will work if it is valid and reliable
5. Analyze data	Start with descriptive statistics Mean, SD Provide more in-depth analysis with inferential statistics t test to look for differences between the means of two groups; ANOVA to look for differences between the means for three or more groups or; chi square test, to compare group frequencies (if the event occurs more frequently in one group).

Interpretation of scores

Causal-comparative studies identify relationships that may lead to experimental studies.

Cause-effect relationships established through causal-comparative research are at best tenuous and tentative.

The only way to show true cause and effect is to conduct experimental research.

Causal-comparative research often occurs first because:

• Data may already exist
• They permit investigation of variables that cannot or should not be investigated experimentally (do juvenile delinquents have more aggressive behavior?)
• They provide guidance for experimental studies
• They are less costly than experimental research

Inferential statistical symbols

Inferential statistics help us make inferences from sample statistics to the population parameters.

Some of the following symbols to be familiar include:

Closure; Review and Assignments

Review questions: (To find the answers, click on the question mark icon)

• What is, and what is the purpose of, causal-comparative research?
  Why do researchers use causal-comparative research if it can't ensure cause and effect?
  How do you identify groups to be examined in correlational research?
  Explain the basic differences between correlational, causal-comparative, and experimental research.
  Explain the following statistical terms:

Before next week:

• Read 341-357 (in Chapter 11)
• Turn in Critique #2 during next class