Week 10
12-1 PM: in MA resource room
1-3 PM: in S 2.81
Week 10 (part 1) Bivariate analyses: Association at Interval-Ratio Level, Correlation, Regression
Week 10 (part 2) Multivariate analyses: Direct, Intervening, Spurious Relationships and Interaction
Homework:
A) study of core reading: Healey (2005), Chapter 15, 16 and 17
B) working with SPSS, make exercises 9 and 10 (see Einspruch 2005)
C) two SPSS replication exercises, see below. Please print your output and bring it to class to discuss it.
First SPSS exercise week 10 (bivariate association at Interval-Ratio Level, Correlation, Regression)
Replication on the basis of literature about relationship between economic development and democracy. Your research question is: why are some countries more democratic than other? Answer this question, while using your knowledge of Healey’s Chapter 1-14, Chapter 15 and 17. Use the Global Indicators Data. Save all your commands in the syntax.
Question 1: What is your most important explanation for the fact that some countries are more democratic than others?
Question 2: Formulate the null and alternative research hypotheses on the basis of the research question (see above).
Question 3:
Use only IR-variables. What is the dependent and what is the independent variable? Can you reject your null hypotheses? Are the differences significant? Use only IR-variables, choose the appropriate techniques, follow the 5-step-model,
and correctly interpret the results. Answer your research question, and describe the results as you would do in a formal research report. Use also the answers to the following questions in your research report.
- Describe the regression line. What do you see? Use the Graphs and Scatter and Simple command to create the regression line.
- Report the formula of the association between X and Y. Use the Analyze and Regression and Linear command to calculate the formula.
- What is the pattern and direction (positive or negative) of the association between X and Y? Interpret ‘b’ in regression formula.
- How strong is the relationship between X and Y? Is this association statistically significant? Use the correct measures of association to answer this question. Use the Analyze and Correlate and Bivariate command.
- What is the explained variance? Use the Analyze and Regression and Linear command to find out. What does this result mean?
Second SPSS exercise week 10 (Multivariate analyses: Direct, Intervening, Spurious Relationships and Interaction)
Replication on the basis of literature about relationship between economic development and democracy. Your research question is: why are some countries more democratic than other? Answer this question, while using your knowledge of Healey’s Chapter 1-14, but specifically Chapter 15, 16 and 17. Use the Global Indicators Data. Save all your commands in the syntax.
Question 1: Choose three interesting interval-ratio variables related to the research question (see above). What are the independent variable, dependent variable, and control variable?
Question 2: Take each bivariate relationship one at a time and describe the three characteristics of each bivariate relationship: the existence, strength, significance, and pattern/ direction. Use any of the appropriate measurements of association to analyze and describe each bivariate relationship. Describe the results in non-technical language. Use the five steps described in the handouts of week 9, and use the commands described in the first SPSS exercise (see above).
Question 3: How well is the dependent variable explained by the other variables? Pay attention to the three characteristics of a multivariate relationship: existence, strength, and pattern/direction. Describe the results in non-technical language. Use also the answers to the following questions in your research report.
- Report the formula of the association between the variables. Use the Analyze and Regression and Linear command to calculate the formula.
- What is the pattern and direction (positive or negative) of the association between the variables? Interpret the ‘b’s’ in regression formula.
- How strong is the relationship between the variables? Is this association statistically significant? Interpret the beta’s to answer this question.
- What is the explained variance? Use the Analyze and Regression and Linear command to find out. What does this result mean?
- Is the relationship between your dependent and independent variable affected by your control variable? In other words, is there a direct, spurious, intervening or interaction effect?