Reading article tablesKlandermans & Oegema 1987Klandermans & Oegema: Network EffectsKlandermans & Oegema: Deciding to GoReading regression tablesInterpreting Klandermans & Oegema regression: DV=intend to goKlandermans & Oegema: Deciding to GoReading Wood & Hughes TableWood & Hughes TableMcAdam High Risk ActivismReading article tablesKlandermans, Wood & Hughes, McAdam “High Risk” model.74.17.4Conditional probabilities.79Klandermans & Oegema 1987Klandermans & Oegema: Network Effects19%24%7%0%% ParticipateBothInformalFormalNoneLinksTable suggests that informal links are most important!!Tilt of movement to educated, leftist, men came through networks, NOT opinions for/against movementKlandermans & Oegema: Deciding to Go1. Left parties, educated are key. Overlap with expect friends to go2. Additional effect of believing Dutch government has effect3. Majority who said they would go did not (6/10): cited specific “reasons”Reading regression tables• Look for symbols about “significance”, usually *’s. (Check footnotes, occasionally non-significant results are *’d.) Significant = effect too large to be due to change. MORE SIGNIFICANT IS SMALLER P, p<.05 is significant, p<.001 is more significant.• Look at sign of coefficient: + or -, and meaning of variables. [In a few cases, the coefficients are odds ratios instead, which are above 1.0 if effect is positive and below 1.0 if effect is negative.]• Unstandardized “b” or “B” coefficients can be compared across equations for the same variables• Standardized $ (beta) coefficients tell you how “strong” each variable is compared to others in the same equation.Interpreting Klandermans & Oegema regression: DV=intend to goEquivalent to saying:LogOdds(Intend) = .04Age -.32Gender + .37Educ + 2.33Voting –7.53Numbers in parentheses are standard errors.Coefficients are significant when they are substantially larger than their standard error.Here, only education and voting for left parties are significant.Table SHOULD have labeled direction for gender, votingKlandermans & Oegema: Deciding to Go1. Left parties, educated are key. Overlap with expect friends to go2. Additional effect of believing Dutch government has effect3. Majority who said they would go did not (6/10): cited specific “reasons”Reading Wood & Hughes TableB is undstandardized regression coefficient; gives equation for attitude toward pornography.The beta ($)column gives standardized coefficients. You can use it to find which independent variables are strongest. Here, age, the send, then education.The t column is for a t-test, ratio of B to its standard error (not shown). The symbols show that all independent variables are significant.Wood & Hughes TableSignificant predictors are older, female (-male), -education, rural now & rural at 16, Southern manual or lower white collar occupation, came from low income family, not Black, Catholic & Conservative Protestant.McAdam High Risk