Channels, Fall 2017

Page 105 Channels • 2017 • Volume 2 • Number 1 Appendix B: Linear Regression Germany Because there was a very strong, positive linear correlation between inflows of asylum seekers and terrorism incidents for Germany, a linear regression was run in order to determine whether the correlation would continue into 2016 if numbers increased. In addition, an f test was run to determine whether the differences between variables are not due to chance. If the f value exceeds the critical value, then we can say that the differences between inflows of asylum seekers and terrorism incidents are not due to chance. Relevant information for interpreting results are provided underneath the appropriate table. Lastly, further descriptive statistics are provided to round out the linear regression analysis. Model Summary b Model R R Square Adjusted R Square Std. Error of the Estimate 1 .949 a .901 .894 3.857 a. Predictors: (Constant), Inflows Asylum Seekers By Thousands b. Dependent Variable: Terrorism Incidents ANOVA a Model Sum of Squares df Mean Square F Sig. 1 Regression 1889.502 1 1889.502 127.027 .000 b Residual 208.248 14 14.875 Total 2097.750 15 a. Dependent Variable: Terrorism Incidents b. Predictors: (Constant), Inflows Asylum Seekers By Thousands p <.05 df = 14 Critical Value = 4.60 F = 127 Predicted Number of Terrorism Incidents (if asylum inflows are 500,000) - 52 95% confidence interval for mean number of Terrorism Incidents – b/w 43 and 61 95% prediction interval for mean number of Terrorism Incidents – b/w 40 and 64 Predicted Number of Terrorism Incidents (if asylum inflows are 600,000) – 63 95% confidence interval for mean number of Terrorism Incidents – b/w 52 and 74