320 likes | 529 Views
Crosstabs. Types of relationships. Linear Spurious Intervening Interaction effects. Spurious effects. Does past participation affect current participation?. Before September 2004, had you ever written a letter to some member of the local government? yes, no.
E N D
Types of relationships • Linear • Spurious • Intervening • Interaction effects
Does past participation affect current participation? • Before September 2004, had you ever written a letter to some member of the local government? yes, no
But, is this spurious? Z Participation before Violence Participation after Violence What could be Z?
What is the theoretical argument that this relationship is spurious? (What is the common cause?)
Z could be self efficacy • Self efficacy causes both past and current political participation. • So, the idea is that past participation is not the only cause of current political participation. • So, we need to control for self efficacy.
…if self efficacy is low Pearson r = .10 syntax temporary. select if capabyou < 3. crosstabs tables = polpartbesc by letterloc/cells = col/stats = corr.
So, what is the conclusion? • Past participation only has an impact when self efficacy is high. • Controlling for self efficacy, the effect of past participation on current participation is mitigated. • Therefore, the effect is partially spurious.
Intervening relationship Political interest Self efficacy Participation
Alternate cause Political interest Self efficacy Participation
So, let’s look at the relationship between political interest and participation
Political interest and political participation Pearson R = .29
So, now what do we need to do to see if political interest has an effect on participation BECAUSE it causes self efficacy? • In other words, how do we find out that the only way interest has an impact on participation is because interest causes self efficacy?
Crosstabs, controlling for self efficacy • If the relationship between political interest and participation is mitigated (lower), controlling for self efficacy, then it is perhaps spurious or intervening • If the relationship is maintained even after controlling for self efficacy, then we say we have alternate causes
Political interest and participation when self efficacy is high Pearson r = .27
Political interest and participation when self efficacy is low Pearson r = .28
Conclusion • Political interest has a direct effect on participation, controlling for self efficacy
So, does controlling for political interest mitigate the effect of self efficacy on political participation?
Linear relationship between self efficacy and political participation
Correlation between self efficacy and political participation
Self efficacy and participation when political interest is low
Self efficacy and participation when political interest is high
Conclusion • Empirical analysis shows us that it is neither spurious or intervening, since the relationship is not mitigated • Note that, when the relationship is mitigated only THEORY can determine whether it is spurious or intervening