mod

code/cleaning_wrangling.R

@@ -7,6 +7,9 @@ data_party <- readRDS("_SharedFolder_article_spsa2024_gpt_party/data/expert_surv
 data_party$econ_distance <- abs(data_party$V4_Scale - data_party$econ_ideo_gpt_mean)
 data_party$sos_distance <- abs(data_party$V6_Scale - data_party$sos_ideo_gpt_mean)
 
+data_party$cat_econ_distance <- abs(data_party$econ_ideo_cat - data_party$econ_ideo_cat_gpt)
+data_party$cat_sos_distance <- abs(data_party$sos_ideo_cat - data_party$sos_ideo_cat_gpt)
+
 mean(data_party$V4_Scale - data_party$econ_ideo_gpt_mean, na.rm = T)
 
 data_party$Region_name <- NA 

code/t_test_mean_distance.R

@@ -1,6 +1,5 @@
 library(dplyr)
 library(ggplot2)
-library(patchwork)
 
 data_party <- readRDS("_SharedFolder_article_spsa2024_gpt_party/data/expert_survey/data_party.rds")
 
@@ -113,4 +112,79 @@ g2 <- ggplot(h1, aes(x = gps, y = gpt)) +
         axis.title.y = element_text(hjust = 0.5),
         axis.text.y = element_text(angle = 90))
 
-g2 + g1
+
+
+cat_econ_test <- t.test(data_party$econ_ideo_cat, data_party$econ_ideo_cat_gpt, paired = TRUE)
+cat_sos_test <- t.test(data_party$sos_ideo_cat, data_party$sos_ideo_cat_gpt, paired = TRUE)
+
+econ_t_value <- cat_econ_test$statistic
+sos_t_value <- cat_sos_test$statistic
+
+# Extract Mean Differences and Confidence Intervals
+results <- data.frame(
+  Category = c("Econ", "Sos"),
+  MeanDifference = c(econ_test$estimate, sos_test$estimate),
+  sd = c(econ_test$stderr[1], sos_test$stderr[1]),
+  LowerCI95 = c(econ_test$conf.int[1], sos_test$conf.int[1]),
+  UpperCI95 = c(econ_test$conf.int[2], sos_test$conf.int[2])
+) %>% 
+  mutate(LowerCI99 = MeanDifference - (sd * 2.807),
+         UpperCI99 = MeanDifference + (sd * 2.807))
+
+ggplot(results, aes(x = MeanDifference, y = Category)) +
+  geom_point(size = 3) +
+  geom_linerange(aes(xmin = LowerCI99, xmax = UpperCI99), linewidth = 0.7) +
+  geom_linerange(aes(xmin = LowerCI95, xmax = UpperCI95), linewidth = 2) +
+  clessnverse::theme_clean_light() +
+  labs(x = "\nMean Difference\n",
+       y = "\nIdeological Scale\n",
+       caption = "Thicker lines denote a 95% confidence interval.\nThinner lines indicate a 99% confidence interval.") +
+  scale_x_continuous(limits = c(-0.75, 0)) +
+  scale_y_discrete(labels = c("Sos" = "Social", "Econ" = "Economic")) +
+  geom_vline(xintercept = 0, linetype = "dotted") +
+  geom_text(aes(x = -0.34, y = "Econ", label = paste("t = ", round(econ_t_value, 2), sep = "")),
+            hjust = 0, size = 5) +
+  geom_text(aes(x = -0.1, y = "Sos", label = paste("t = ", round(sos_t_value, 2), sep = "")),
+            hjust = 0, size = 5) +
+  theme(plot.caption.position = "plot",
+        axis.title.x = element_text(hjust = 0.5, size = 20),
+        axis.title.y = element_text(hjust = 0.5, size = 20),
+        axis.text.x = element_text(size = 20),
+        axis.text.y = element_text(size = 20),
+        plot.caption = element_text(size = 20, hjust = 0))
+
+ggsave("_SharedFolder_article_spsa2024_gpt_party/graphs/paper/h1_ttest.png",
+       width = 8, height = 6)
+
+h1 <- data_party %>% 
+  select(ID_GPS,
+         scale_econ = V4_Scale,
+         scale_social = V6_Scale,
+         gpt_econ = econ_ideo_gpt_mean,
+         gpt_social = sos_ideo_gpt_mean) %>% 
+  pivot_longer(., cols = starts_with("scale"),
+               names_to = "scale",
+               names_prefix = "scale_",
+               values_to = "gps") %>% 
+  pivot_longer(., cols = starts_with("gpt"),
+               names_to = "gptscale",
+               names_prefix = "gpt_",
+               values_to = "gpt") %>% 
+  filter(scale == gptscale) %>% 
+  mutate(scale = ifelse(scale == "econ", "Economic", "Social"))
+
+g2 <- ggplot(h1, aes(x = gps, y = gpt)) +
+  facet_wrap(~scale) +
+  geom_jitter(alpha = 0.6, shape = 19,
+              width = 0.2, height = 0.2) +
+  geom_smooth(method = "lm",
+              color = "black",
+              alpha = 0.2) +
+  clessnverse::theme_clean_light() +
+  scale_x_continuous(breaks = c(1, 9), labels = c("Left", "Right")) +
+  scale_y_continuous(breaks = c(1, 9), labels = c("Left", "Right")) +
+  labs(x = "\nParty Alignment (GPS)\n",
+       y = "\nParty Alignment (GPT-4)\n") +
+  theme(axis.title.x = element_text(hjust = 0.5),
+        axis.title.y = element_text(hjust = 0.5),
+        axis.text.y = element_text(angle = 90))