Causal-comparative Research

1
Causal-comparative Research

• To determine the cause for, or consequences of,
  existing differences in groups of individuals
• Also referred to as ex post facto research
  (Latin for after the fact) retrospection

2
Cause for / Consequence of

• Cause for inquiry teaching method vs. lecture
  teaching method (difference in performance)
• Consequences of vegetarian diet vs.
  non-vegetarian diet (difference in mental health)
• Any other examples?

3
Questions

1. How is a causal-comparative study similar to a
   descriptive study?
2. How is a causal-comparative study similar to a
   correlational study?
3. How is a causal-comparative study different from
   an experimental study?

4
When to use a causal-comparative design?

1. When it is unethical to manipulate an independent
   variable (e.g. diet)
2. When the independent variable CANNOT be
   manipulated (e.g. sex, ethnicity, etc.)
3. When the independent variable not been changed
   due to ignorance or negligence (e.g. teaching
   methods)

5
Did you know?

• Causal-comparative studies are most common in the
  field of medicine and sociology
• Why?

6
Which of the following questions would lend
themselves well to causal-comparative research?

1. How many students were enrolled in PSYC101 this
   semester?
2. Which subject do high school students like least?
3. How do elementary school teachers teach phonics?
4. Are two-year-old girls more aggressive than
   two-year-old boys?
5. How might Jimmy Thomas be helped to read?
6. Is teacher enthusiasm related to student success
   in academic classes?
7. What is the best way to teach arithmetic?
8. Do female students perform better in literature
   classes than male students?
9. Does sleep (amount of time) affect academic
   performance of students at college?

7
Weaknesses

• Lack randomizationso what?
• Inability to manipulate an independent
  variableso what?

8
Threats to internal validity

• Location
• Instrumentation
• Loss of subjects

9
Data analysis

1. Construct frequency polygons (to compare data
   graphically)
2. Calculate means and standard deviations
3. Statistical testing t-test or ANOVA
4. Results of causal-comparative studies should
   always be interpreted with cautionthey do not
   prove cause and effect!