LAST CLASS

1
LAST CLASS!

• Data Analysis and Interpretation and Review
  Session

2
Quasi-Experiments

• Involve some form of intervention and a
  comparison group, but they lack the control found
  in true experiments.
• No randomization
• Common form of quasi-experiment
• One-group pretest-posttest design

3
Nonequivalent Control Group Design

• Two conditions
• A group similar to the treatment group serves
  as a comparison group.
• You obtain pretest and posttest scores from
  individuals in both the treatment group and the
  comparison group.

4
Additive Effect of Selection and Maturation
Pretest
Posttest
5
Additive Effect of Selection and History

• The Local History Effect one group of third
  graders may have gotten new books, or may have
  gone on a special field trip that really inspired
  them to get more involved in learning.

6
Other Possible Threats to Internal Validity not
Controlled for in the Nonequivalent Control Group
Design

• Observer bias
• Contamination
• Hawthorne Effects

7
What about external validity?

• Nursing home example
• Plants and happiness.
• How do we know if these findings have external
  validity?

8
Program Evaluation

• Provide information to administrators of
  organizations regarding what services to provide
  to whom and how to best provide them.
• Needs
• Process
• Outcome
• Efficiency

9
Is this stuff science?

• Yes, Applied research is linked to basic
  research!
• By conducting applied research (program
  evaluation, etc) you observe phenomena in your
  environment and can then scientifically (using
  basic research) investigate those phenomena to
  determine more precisely what is happening.
• Applied Research ?? Basic Research

10
Chapter 12 Data Analysis Interpretation

• Get to know your data
• Summarize the data
• Confirm findings in the data

11
Getting to you know your data

• Clean the data
• Reverse code appropriate items
• Recode where necessary
• Check for data entry errors
• Exploratory Data Analysis (EDA)
• Check for appropriate sample sizes in sub-samples
  (e.g., males only)
• Check for missing data
• Frequencies and Descriptives

12
Summarizing the data

• Mean, median, mode, range, and standard deviation
• Standard error of the mean
• The standard deviation of the theoretical
  sampling distribution of means
• Measures of effect size
• Cohens d
• Eta squared

13
Confirming findings.

• Confidence intervals
• How is a confidence interval related to a margin
  of error???