rm(list = ls()) library(plyr) library(dplyr) library(tidyr) library(RColorBrewer) library(ggplot2) library(Rmisc) library(cregg) library(stringi) library(forcats) library(hrbrthemes) library(viridis) library("PerformanceAnalytics") setwd("~/Box Sync/Journal of Asia and Africa Studies/RR_2") HK_only = read.csv("JAAS_supp_material_dataset.csv", header = TRUE) ################################## # Table 3: Logit regression ################################## HK_only$us_act_logit<-ifelse(HK_only$us_act_dummy=="yes",1,0) HK_only$us_act_online_logit<-ifelse(HK_only$us_act_online_dummy=="yes",1,0) HK_only$us_act_offline_logit<-ifelse(HK_only$us_act_offline_dummy=="yes",1,0) # Model(1) logit_1 = glm(us_act_logit ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(2) logit_2 = glm(us_act_online_logit ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(3) logit_3 = glm(us_act_offline_logit ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) ############################ #Table 4: OLS ############################ # Model(4) ols_4 = lm(US_act_total ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(5) ols_5 = lm(online ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(6) ols_6 = lm(offline ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) ####################################################################### #Appendix Table A1: Correlation Matrix of Transnational Ties Variables ####################################################################### corr_data = select(HK_only,level_comm_home_s,language_s,culture_s,HK_act_total_s,value_score_s,HK_id_s) corr_datafa=corr_data[complete.cases(corr_data),] res=cor(corr_datafa) trace("chart.Correlation", edit=T) chart.Correlation(corr_datafa, histogram=TRUE, pch=19) ##################################################### #Appendix Table B1: VIF test for muliticollinarity ##################################################### library(car) vif(logit_1) vif(ols_4) ##################################################### #Appendix Table C ##################################################### #Table C1 # Model(1) ols_c1.1 = lm(US_act_total ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(2) ols_c1.2 = lm(online ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(3) ols_c1.2 = lm(offline ~ level_comm_home_s+language_s+culture_s+ HK_act_total_s+value_score_s+HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) #Table C2: Logit regression on overall participation # Model(1) logit_c2.1 = glm(us_act_logit ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(2) logit_c2.2 = glm(us_act_logit ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(3) logit_c2.3 = glm(us_act_logit ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(4) logit_c2.4 = glm(us_act_logit ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(5) logit_c2.5 = glm(us_act_logit ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(6) logit_c2.6 = glm(us_act_logit ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) #Table C3: Logit regression on offline participation # Model(1) logit_c3.1 = glm(us_act_offline_logit ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(2) logit_c3.2 = glm(us_act_offline_logit ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(3) logit_c3.3 = glm(us_act_offline_logit ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(4) logit_c3.4 = glm(us_act_offline_logit ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(5) logit_c3.5 = glm(us_act_offline_logit ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(6) logit_c3.6 = glm(us_act_offline_logit ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) #Table C4: Logit regression on online participation # Model(1) logit_c4.1 = glm(us_act_online_logit ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(2) logit_c4.2 = glm(us_act_online_logit ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(3) logit_c4.3 = glm(us_act_online_logit ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(4) logit_c4.4 = glm(us_act_online_logit ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(5) logit_c4.5 = glm(us_act_online_logit ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) # Model(6) logit_c4.6 = glm(us_act_online_logit ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, family = binomial(),data = HK_only) #Table C5: OLS regression on overall participation # Model(1) ols_c5.1 = lm(US_act_total ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(2) ols_c5.2 = lm(US_act_total ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(3) ols_c5.3 = lm(US_act_total ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(4) ols_c5.4 = lm(US_act_total ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(5) ols_c5.5 = lm(US_act_total ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(6) ols_c5.6 = lm(US_act_total ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) #Table C6: OLS regression on offline participation # Model(1) ols_c6.1 = lm(offline ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(2) ols_c6.2 = lm(offline ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(3) ols_c6.3 = lm(offline ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(4) ols_c6.4 = lm(offline ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(5) ols_c6.5 = lm(offline ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(6) ols_c6.6 = lm(offline ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) #Table C7: OLS regression on online participation # Model(1) ols_c7.1 = lm(online ~ level_comm_home_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(2) ols_c7.2 = lm(online ~ language_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(3) ols_c7.3 = lm(online ~ culture_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(4) ols_c7.4 = lm(online ~ HK_act_total_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(5) ols_c7.5 = lm(online ~ value_score_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only) # Model(6) ols_c7.6 = lm(online ~ HK_id_s+ Q20a+age_s+edu_con_s+income_con_s+time_in_US_s+citizen+Q6i+Q20g, data = HK_only)