PPA 415 Statistical Methods in Public Administration

1
PPA 415 Statistical Methods in Public
Administration

• Lecture 1c Research Design Summary

2
Research Questions

• Research questions can be descriptive.
• What is happening to a variable currently or what
  will happen in the future?
• Research questions can be causal.
• What factors are causing a variable to change?
• What factors influence the key variables in the
  research question?

3
Variables

• Concepts, variables, indicators.
• Causal direction.
• Independent variables (causes).
• Dependent variables (effects).

4
Hypotheses

• Hypotheses can be descriptive.
• Are arrests for DUI increasing over time?
• Hypotheses can also be causal.
• Do increases in DUI arrests reduce deaths in
  motor vehicle accidents?
• Does nearness to a presidential election increase
  the likelihood of granting a major disaster
  declaration?

5
Sampling

• Population (states, counties, IAEM members).
• Random samples (simple, systematic, cluster,
  random digit dialing).
• Snowball sampling
• Purposive or haphazard sampling.

6
Data Collection Procedures

• Surveys.
• Agency records and primary documents.
• Direct observation.
• Existing data sets (Census, BLS, ICPSR, etc.).
  www.fedstats.gov.
• Participant observation.
• Focus groups.
• Literature review.

7
Data Analysis Procedures Quantitative (What,
Where, When)

• Descriptive statistics (mean, median, mode,
  standard deviation).
• Inferential statistics.
• T-tests, analyses of variance, chi-square, cross
  tabulation.
• Measures of association.
• Bivariate.
• Regression-based analysis.
• Multivariate analysis.
• Multiple regression, analysis of covariance,
  logistic regression.
• Data reduction (factor analysis, scaling).

8
Data Analysis Procedures Qualitative (Why, How).

• Focus group summaries.
• Interview summaries.
• Historical analysis.
• Content analysis.
• Case studies.
• Ethnography.
• Participant observation.

9
Interpretation, Conclusions

• Looking for answers to research questions.
• Quantitative research often asks what, where, and
  when. Focuses on common elements across
  subjects. Makes inferences to whole population.
• Qualitative research asks why and how. Focuses
  on what is unique across subjects. Makes
  in-depth analysis of complexities of choice.