Results

library(tidyverse)
library(crackdownsphilanthropy)
library(rstan)
library(tidybayes)
library(broom)
library(ggstance)
library(ggdag)
library(ggraph)
library(patchwork)
library(pander)
library(scales)
library(janitor)
library(here)

source(here("analysis", "options.R"))

# Load data
results <- readRDS(here("data", "derived_data", "results_clean.rds"))
# qwraps2::lazyload_cache_dir("02_analysis_cache/html")

Causal pathway

Our theory and hypotheses are laid out in the causal pathway below. Our outcomes (Y: % likely to donate and amount donated) “listen to” or respond to C (crackdown), which is our main treatment or exposure. Funding (F) and issue (I) serve as heuristics for donation and influence a government’s decision to crack down on NGOs, hence the dual arrows to C and Y.

theory_dag <- dagify(Y ~ I + C + F,
                     C ~ I + F,
                     outcome = "Y",
                     exposure = "C",
                     labels = c("Y" = "Outcome (Likelihood of donation & amount donated)", 
                                "C" = "Anti-NGO legal crackdown", 
                                "I" = "NGO issue area", 
                                "F" = "NGO funding source")) %>% 
  tidy_dagitty(layout = "sugiyama")

plot_dag <- ggplot(theory_dag, aes(x = x, y = y, xend = xend, yend = yend)) +
  geom_dag_point(size = 6) +
  geom_dag_edges(start_cap = circle(4, "mm"),
                 end_cap = circle(4, "mm")) +
  geom_dag_text(size = pts(6), family = "Roboto Condensed", fontface = "bold") +
  geom_dag_label_repel(aes(label = label), size = pts(6), box.padding = 0.1, 
                       direction = "x", seed = 12345,
                       family = "Roboto Condensed", fontface = "plain") +
  labs(x = NULL) +
  theme_ngos() +
  theme(panel.grid = element_blank(),
        axis.title = element_blank(),
        axis.text = element_blank())

plot_dag %T>% 
  print() %T>%
  ggsave(., filename = here("analysis", "output", "figures", "causal-path.pdf"),
         width = 4.75, height = 2, units = "in", device = cairo_pdf) %>% 
  ggsave(., filename = here("analysis", "output", "figures", "causal-path.png"),
         width = 4.75, height = 2, units = "in", type = "cairo", dpi = 300)

Overview of data

Balance of experimental conditions

results %>% count(crackdown, issue, funding) %>% 
  rename(Crackdown = crackdown, Issue = issue, Funding = funding) %>% 
  janitor::adorn_totals(., where = "row") %T>% 
  pandoc.table() %>% 
  pandoc.table.return(caption = "Balance of experimental conditions {#tbl:experimental-conditions}") %>% 
  cat(file = here("analysis", "output", "tables", "tbl-experimental-conditions.md"))

Crackdown	Issue	Funding	n
No crackdown	Human rights	Government	68
No crackdown	Human rights	Private	64
No crackdown	Humanitarian assistance	Government	68
No crackdown	Humanitarian assistance	Private	65
Crackdown	Human rights	Government	65
Crackdown	Human rights	Private	65
Crackdown	Humanitarian assistance	Government	68
Crackdown	Humanitarian assistance	Private	68
Total	-	-	531

Descriptive statistics table

vars_to_summarize <- tribble(
  ~variable_name, ~clean_name,
  "donate_likely", "Likelihood of donation",
  "donate_likely_bin", "Likelihood of donation (binary)",
  "amount_donate", "Amount hypothetically donated ($)",
  "gender", "Gender",
  "age", "Age",
  "income", "Income",
  "education", "Education",
  "religiosity", "Frequency of attending religious services",
  "ideology", "Political views",
  "political_knowledge", "Frequency of following public affairs",
  "give_charity", "Frequency of charitable donations",
  "volunteer", "Volunteered in past 12 months",
  "favor_humanitarian", "Prior favorability towards humanitarian NGOs",
  "favor_humanitarian_bin", "Prior favorability towards humanitarian NGOs (binary)",
  "favor_human_rights", "Prior favorability towards human rights NGOs",
  "favor_human_rights_bin", "Prior favorability towards human rights NGOs (binary)",
  "favor_development", "Prior favorability towards development NGOs",
  "favor_development_bin", "Prior favorability towards development NGOs (binary)",
  "check2", "Attention check 2"
)

results_summary_stats <- results %>% 
  select(one_of(vars_to_summarize$variable_name)) %>% 
  summarise_all(~list(.)) %>% 
  pivot_longer(everything(), names_to = "variable_name", values_to = "value") %>% 
  mutate(N = value %>% map_int(~ length(na.omit(.))),
         summary = map2(.x = value, .y = variable_name, ~ md_summary_row(.x, .y))) %>% 
  left_join(vars_to_summarize, by = "variable_name") %>% 
  mutate(variable_name = factor(variable_name, 
                                levels = vars_to_summarize$variable_name, 
                                ordered = TRUE)) %>% 
  arrange(variable_name) %>% 
  select(-value, -variable_name) %>% 
  unnest(summary) %>% 
  select(Variable = clean_name, N, ` ` = spark, Details = summary)

results_summary_stats %>% 
  select(-N) %>% 
  pandoc.table.return(caption = "Descriptive statistics {#tbl:descriptive-stats}",
                      split.cell = 80, split.table = Inf) %T>% 
  cat(file = here("analysis", "output", "tables", "tbl-descriptive-stats.md")) %>%
  cat()

Descriptive statistics {#tbl:descriptive-stats}

Variable		Details
Likelihood of donation		Extremely unlikely (46; 8.66%) | Somewhat unlikely (110; 20.72%) | Neither likely nor unlikely (138; 25.99%) | Somewhat likely (192; 36.16%) | Extremely likely (45; 8.47%)
Likelihood of donation (binary)		Not likely (294; 55%) | Likely (237; 45%)
Amount hypothetically donated ($)		Median: 10 | Mean: 22.4 | Std. Dev.: 25.67
Gender		Female (291; 54.80%) | Male (237; 44.63%) | Other (1; 0.19%) | Prefer not to say (2; 0.38%)
Age		Under 18 (1; 0.19%) | 18 – 24 (43; 8.10%) | 25 – 34 (207; 38.98%) | 35 – 44 (129; 24.29%) | 45 – 54 (90; 16.95%) | 55 – 64 (44; 8.29%) | 65 – 74 (16; 3.01%) | 75 – 84 (1; 0.19%)
Income		Less than $10,000 (32; 6.03%) | $10,000 – $19,999 (43; 8.10%) | $20,000 – $29,999 (50; 9.42%) | $30,000 – $39,999 (78; 14.69%) | $40,000 – $49,999 (58; 10.92%) | $50,000 – $59,999 (55; 10.36%) | $60,000 – $69,999 (43; 8.10%) | $70,000 – $79,999 (38; 7.16%) | $80,000 – $89,999 (25; 4.71%) | $90,000 – $99,999 (30; 5.65%) | $100,000 – $149,999 (45; 8.47%) | More than $150,000 (21; 3.95%) | Prefer not to say (13; 2.45%)
Education		Less than high school (2; 0.38%) | High school graduate (47; 8.85%) | Some college (128; 24.11%) | 2 year degree (68; 12.81%) | 4 year degree (212; 39.92%) | Graduate or professional degree (66; 12.43%) | Doctorate (8; 1.51%)
Frequency of attending religious services		More than once a week (19; 3.58%) | Once a week (74; 13.94%) | Once or twice a month (50; 9.42%) | A few times a year (71; 13.37%) | Seldom (98; 18.46%) | Never (215; 40.49%) | Don’t know (4; 0.75%)
Political views		Strong liberal (76; 14.31%) | Liberal (150; 28.25%) | Independent, leaning liberal (82; 15.44%) | Independent (80; 15.07%) | Independent, leaning conservative (60; 11.30%) | Conservative (61; 11.49%) | Very conservative (22; 4.14%)
Frequency of following public affairs		Most of the time (217; 40.87%) | Some of the time (213; 40.11%) | Only now and then (84; 15.82%) | Hardly at all (17; 3.20%)
Frequency of charitable donations		Once a week (37; 6.97%) | Once a month (105; 19.77%) | Once every three months (105; 19.77%) | Once every six months (102; 19.21%) | Once a year (89; 16.76%) | Once every few years (56; 10.55%) | Never (37; 6.97%)
Volunteered in past 12 months		No (288; 54.2%) | Yes (243; 45.8%)
Prior favorability towards humanitarian NGOs		Very unfavorable (2; 0.38%) | Unfavorable (6; 1.13%) | Neutral (39; 7.34%) | Favorable (235; 44.26%) | Very favorable (249; 46.89%)
Prior favorability towards humanitarian NGOs (binary)		Not favorable (47; 9%) | Favorable (484; 91%)
Prior favorability towards human rights NGOs		Very unfavorable (5; 0.94%) | Unfavorable (12; 2.26%) | Neutral (61; 11.49%) | Favorable (226; 42.56%) | Very favorable (227; 42.75%)
Prior favorability towards human rights NGOs (binary)		Not favorable (78; 15%) | Favorable (453; 85%)
Prior favorability towards development NGOs		Very unfavorable (5; 0.94%) | Unfavorable (8; 1.51%) | Neutral (45; 8.47%) | Favorable (235; 44.26%) | Very favorable (238; 44.82%)
Prior favorability towards development NGOs (binary)		Not favorable (58; 11%) | Favorable (473; 89%)
Attention check 2		Correct (531; 100%)

Average likelihood and amount donated across conditions

conditions_summary <- bind_rows(group_by(results, crackdown, issue, funding) %>% nest(),
                                group_by(results, crackdown, issue) %>% nest(),
                                group_by(results, crackdown) %>% nest(),
                                results %>% nest(data = everything())) %>% 
  arrange(crackdown, issue, funding) %>% 
  mutate(summary = data %>%
           map(~ summarize(., pct_likely = table(donate_likely_bin)[["Likely"]] /
                             length(donate_likely_bin),
                           mean_donation = mean(amount_donate, na.rm = TRUE),
                           sd_donation = sd(amount_donate, na.rm = TRUE),
                           N = nrow(.)))) %>% 
  unnest(summary) %>% select(-data) %>% ungroup()

conditions_summary_clean <- conditions_summary %>% 
  mutate(funding = ifelse(is.na(funding) & !is.na(issue) , "*Total*", as.character(funding)),
         issue = ifelse(is.na(issue) & !is.na(crackdown), "*Total*", as.character(issue)),
         crackdown = ifelse(is.na(crackdown), "*Total*", as.character(crackdown))) %>% 
  group_by(crackdown) %>% 
  mutate(issue = replace(issue, duplicated(issue), NA)) %>% 
  ungroup() %>% 
  mutate(crackdown = replace(crackdown, duplicated(crackdown), NA)) %>% 
  mutate(pct_likely = percent(pct_likely)) %>% 
  rename(`Crackdown condition` = crackdown, `Issue condition` = issue,
         `Funding condition` = funding, `% likely to donate` = pct_likely,
         `Amount donated (mean)` = mean_donation, `Amount donated (sd)` = sd_donation)

conditions_summary_clean %T>% 
  pandoc.table() %>% 
  pandoc.table.return(caption = "Average likelihood and amount donated across experimental conditions {#tbl:avg-results}") %>% 
  cat(file = here("analysis", "output", "tables", "tbl-avg-results.md"))

Crackdown condition	Issue condition	Funding condition	% likely to donate	Amount donated (mean)	Amount donated (sd)	N
No crackdown	Human rights	Government	47.06%	22.4	24.8	68
		Private	39.06%	19.3	22.4	64
		Total	43.18%	20.9	23.6	132
	Humanitarian assistance	Government	44.12%	17.9	21.4	68
		Private	40.00%	21.9	26.9	65
		Total	42.11%	19.9	24.2	133
	Total		42.64%	20.4	23.9	265
Crackdown	Human rights	Government	29.23%	19.4	25.8	65
		Private	58.46%	28	26.4	65
		Total	43.85%	23.7	26.3	130
	Humanitarian assistance	Government	51.47%	30.4	32.7	68
		Private	47.06%	19.8	21.5	68
		Total	49.26%	25.1	28.1	136
	Total		46.62%	24.4	27.2	266
Total			44.63%	22.4	25.7	531

Visualize important variables

Likelihood of donation

donate_summary <- results %>% 
  count(donate_likely) %>% 
  mutate(perc = n / sum(n)) %>% 
  mutate(highlight = ifelse(donate_likely %in% c("Extremely likely", "Somewhat likely"), TRUE, FALSE))

plot_donate_summary <- ggplot(donate_summary, aes(x = n, y = donate_likely, 
                                                  fill = highlight)) +
  geom_barh(stat = "identity") +
  scale_x_continuous(sec.axis = sec_axis(~ . / sum(donate_summary$n),
                                         labels = percent_format(accuracy = 1))) +
  scale_fill_manual(values = ngo_cols("green", "blue", name = FALSE), guide = FALSE) +
  labs(x = NULL, y = NULL) +
  theme_ngos() +
  theme(panel.grid.major.y = element_blank())

plot_donate_summary %T>% 
  print() %T>%
  ggsave(., filename = here("analysis", "output", "figures", "donate_summary.pdf"),
         width = 9, height = 4, units = "in", device = cairo_pdf) %>% 
  ggsave(., filename = here("analysis", "output", "figures", "donate_summary.png"),
         width = 9, height = 4, units = "in", type = "cairo", dpi = 300)

Amount donated

plot_amount_summary <- ggplot(results, aes(x = amount_donate)) +
  geom_histogram(bins = 20, fill = ngo_cols("blue")) +
  scale_x_continuous(labels = dollar) +
  labs(x = NULL, y = NULL) +
  theme_ngos() +
  theme(panel.grid.major.x = element_blank(),
        panel.grid.minor.x = element_blank())

plot_amount_summary %T>% 
  print() %T>%
  ggsave(., filename = here("analysis", "output", "figures", "amount_summary.pdf"),
         width = 9, height = 2.75, units = "in", device = cairo_pdf) %>% 
  ggsave(., filename = here("analysis", "output", "figures", "amount_summary.png"),
         width = 9, height = 2.75, units = "in", type = "cairo", dpi = 300)

all_models <- tribble(
  ~title, ~df,
  "Crackdown", results,
  # Ha, cheat here by renaming the issue and funding columns to crackdown. 
  # All other tests are based on the crackdown column except these two, so
  # rather than build all sorts of convoluted arguments and functions, we just
  # rename these as issue and funding
  "Issue", mutate(results, crackdown = issue),
  "Funding", mutate(results, crackdown = funding),
  # Create all the nested conditions
  "Human rights | Crackdown", filter(results, issue == "Human rights"),
  "Humanitarian assistance | Crackdown", filter(results, issue != "Human rights"),
  "Government | Crackdown", filter(results, funding == "Government"),
  "Private | Crackdown", filter(results, funding == "Private"),
  "Human rights | Government | Crackdown", filter(results, issue == "Human rights", funding == "Government"),
  "Human rights | Private | Crackdown", filter(results, issue == "Human rights", funding == "Private"),
  "Humanitarian assistance | Government | Crackdown", filter(results, issue != "Human rights", funding == "Government"),
  "Humanitarian assistance | Private | Crackdown", filter(results, issue != "Human rights", funding == "Private")
) 

Treatment effects: Likelihood of donation

This time around, we’re not using interactionful regression models to calculate all these differences in groups. Instead we use real live Stan code to estimate the differences in group means and proportions!

Priors and models

We estimate the proportion of people responding that they’d be likely to donate to the organization with a binomial distribution, with a prior \(\theta\) distribution of \(\text{Beta}(5, 5)\). We build the following model in Stan:

\[ \begin{aligned} n_{\text{group 1, group 2}} &\sim \text{Binomial}(n_{\text{total in group}}, \theta_{\text{group}}) &\text{[likelihood]}\\ \text{Difference} &= n_{\text{group 2}} - n_{\text{group 1}} &\text{[difference in proportions]} \\ n &: \text{Number likely to donate} \\ \\ \theta_{\text{group 1, group 2}} &\sim \text{Beta}(5, 5) &\text{[prior prob. of being likely to donate]} \end{aligned} \]

likely_theta <- ggplot(data = tibble(x = c(0, 1)), aes(x = x)) +
  geom_area(stat = "function", fun = dbeta, args = list(shape1 = 5, shape2 = 5), 
            fill = "grey80", color = "black") +
  labs(x = expression(Probability ~ of ~ being ~ likely ~ to ~ donate ~ (theta)), y = "Density") +
  annotate(geom = "label", x = 0.5, y = 1, label = "Beta(5, 5)", size = pts(9)) +
  theme_ngos(base_size = 9, density = TRUE)
likely_theta

likely_theta %T>% 
  ggsave(., filename = here("analysis", "output", "figures", "prior-likely.pdf"),
         width = 2.75, height = 1.5, units = "in", device = cairo_pdf) %>%
  ggsave(., filename = here("analysis", "output", "figures", "prior-likely.png"),
         width = 2.75, height = 1.5, units = "in", type = "cairo", dpi = 300)

Differences

# This takes forever because of MCMC sampling, so this chunk is cached

# Sample from the compiled model
all_models_likely_run <- all_models %>% 
  # Simplify data for sending to Stan
  mutate(stan_data_counts = df %>% map(~ {
    df_counts <- .x %>% 
      count(crackdown, donate_likely_bin) %>% 
      group_by(crackdown) %>% 
      mutate(total = sum(n)) %>% 
      filter(donate_likely_bin == "Likely")
    
    return(list(
      n_total_1 = df_counts$total[1],
      n_total_2 = df_counts$total[2],
      n_likely_1 = df_counts$n[1],
      n_likely_2 = df_counts$n[2]
    ))
  })) %>% 
  # Run the actual model on the simplified data
  mutate(model = stan_data_counts %>% map(~ {
    donate_likely(n_total_1 = .x$n_total_1, n_total_2 = .x$n_total_2,
                  n_likely_1 = .x$n_likely_1, n_likely_2 = .x$n_likely_2,
                  control = list(max_treedepth = 15),
                  chains = CHAINS, iter = ITER, warmup = WARMUP, seed = BAYES_SEED)
  })) 

# Extract stuff
all_models_likely <- all_models_likely_run %>%
  # Extract posterior chains
  mutate(posterior_chains_long = map(model, ~ {
      .x %>% gather_draws(theta_1, theta_2, theta_diff, pct_change)
  })) %>% 
  mutate(posterior_chains_wide = map(model, ~ {
      .x %>% spread_draws(theta_1, theta_2, theta_diff, pct_change)
  })) %>% 
  # Get HDI medians in tidy form
  mutate(tidy = posterior_chains_long %>% map(~ {
    .x %>% median_hdci() %>% to_broom_names()
  }))

tidied_diffs_likely <- all_models_likely %>% 
  unnest(posterior_chains_long) %>% 
  filter(.variable == "theta_diff") %>% 
  mutate(category = case_when(
    str_count(title, "\\|") == 0 ~ "Level 1",
    str_count(title, "\\|") == 1 ~ "Level 2",
    str_count(title, "\\|") == 2 ~ "Level 3"
  )) %>% 
  mutate(title = str_remove(title, " \\| Crackdown"))

level1_likely <- tidied_diffs_likely %>% 
  filter(category == "Level 1") %>%
  mutate(title = recode(title, 
                        Crackdown = "Crackdown −\nNo crackdown",
                        Issue = "Humanitarian\nassistance −\nHuman rights",
                        Funding = "Private −\nGovernment\nfunding")) %>% 
  mutate(title = fct_inorder(title))

plot_diff_likely_a <- ggplot(level1_likely, aes(x = .value, y = fct_rev(title), fill = title)) +
  stat_halfeye(.width = c(0.8, 0.95)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  scale_fill_manual(values = ngo_cols(c("blue", "red", "orange"), name = FALSE), guide = FALSE) +
  labs(x = "Difference in donation likelihood", y = NULL, tag = "A") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

level2_likely <- tidied_diffs_likely %>% 
  filter(category == "Level 2") %>%
  mutate(condition = case_when(
    title %in% c("Human rights", "Humanitarian assistance") ~ "Issue",
    title %in% c("Government", "Private") ~ "Funding"
  )) %>% 
  mutate(facet_title = case_when(
    title %in% c("Human rights", "Humanitarian assistance") ~ paste(title, "issues"),
    title %in% c("Government", "Private") ~ paste(title, "funding")
  )) %>% 
  mutate(title = case_when(
    title %in% c("Humanitarian assistance", "Private") ~ "",
    TRUE ~ "Crackdown −\nNo crackdown")
  )

plot_diff_likely_b <- ggplot(filter(level2_likely, condition == "Issue"), 
                             aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("red", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  labs(x = NULL, y = NULL, tag = "B") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

plot_diff_likely_c <- ggplot(filter(level2_likely, condition == "Funding"), 
                             aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("orange", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  labs(x = "Difference in donation likelihood", y = NULL, tag = "C") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

level3_likely <- tidied_diffs_likely %>% 
  filter(category == "Level 3") %>%
  separate(title, c("issue", "funding"), sep = " \\| ") %>%
  mutate(issue = paste(issue, "issues"),
         funding = paste(funding, "funding")) %>% 
  mutate(facet_title = paste0(issue, "\n", funding)) %>% 
  mutate(title = case_when(
    funding == "Private funding" ~ "",
    TRUE ~ "Crackdown −\nNo crackdown")
  )

plot_diff_likely_d <- ggplot(level3_likely, aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("green", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = percent_format(accuracy = 1)) +
  labs(x = "Difference in donation likelihood", y = NULL, tag = "D",
       caption = "Point shows posterior median; thick black lines show 80% credible interval;\nthin black lines show 95% credible interval") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

plot_diffs_likely_all <- plot_diff_likely_a / 
  (plot_diff_likely_b / plot_diff_likely_c) / 
  plot_diff_likely_d

plot_diffs_likely_all

plot_diffs_likely_all %T>% 
  ggsave(., filename = here("analysis", "output", "figures", "likely-diffs.pdf"),
         width = 4.75, height = 5.5, units = "in", device = cairo_pdf) %>%
  ggsave(., filename = here("analysis", "output", "figures", "likely-diffs.png"),
         width = 4.75, height = 5.5, units = "in", type = "cairo", dpi = 300)

tbl_likely_tidy <- all_models_likely %>% 
  unnest(tidy) %>% 
  filter(term %in% c("theta_1", "theta_2", "theta_diff", "pct_change")) %>% 
  select(title, term, estimate) %>% 
  pivot_wider(names_from = term, values_from = estimate) %>%
  mutate(category = case_when(
    str_count(title, "\\|") == 0 ~ "Level 1",
    str_count(title, "\\|") == 1 ~ "Level 2",
    str_count(title, "\\|") == 2 ~ "Level 3"
  )) 

tbl_likely_probs <- all_models_likely %>% 
  unnest(posterior_chains_long) %>% 
  filter(.variable == "theta_diff") %>% 
  group_by(title) %>% 
  summarize(p.greater0 = mean(.value > 0),
            p.less0 = mean(.value < 0),
            p.diff.not0 = ifelse(median(.value) > 0, p.greater0, p.less0)) %>% 
  ungroup()

# Save combined table for later use in manuscript
tbl_likely_tidy %>% 
  left_join(tbl_likely_probs, by = "title") %>% 
  saveRDS(here("data", "derived_data", "results_models_likely.rds"))
  
tbl_likely_1 <- tbl_likely_tidy %>% 
  left_join(tbl_likely_probs, by = "title") %>% 
  filter(category == "Level 1") %>%
  mutate(title = recode(title, 
                        Crackdown = "Crackdown − No crackdown",
                        Issue = "*Humanitarian assistance − Human rights*",
                        Funding = "*Private − Government funding*")) %>% 
  mutate_at(vars(theta_1, theta_2, theta_diff, pct_change), 
            list(~percent_format(accuracy = 0.1)(.))) %>% 
  mutate(p.diff.not0 = as.character(round(p.diff.not0, 2))) %>%
  select(Frame = title, `% likely~Treatment~` = theta_2, `% likely~Control~` = theta_1,
         `$\\Delta$` = theta_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_likely_2 <- tbl_likely_tidy %>% 
  left_join(tbl_likely_probs, by = "title") %>% 
  filter(category == "Level 2") %>%
  mutate(title_clean = str_remove(title, " \\| Crackdown")) %>% 
  mutate(title_clean = case_when(
    title_clean %in% c("Human rights", "Humanitarian assistance") ~ paste(title_clean, "issues"),
    title_clean %in% c("Government", "Private") ~ paste(title_clean, "funding")
  )) %>% 
  mutate(title_clean = factor(title_clean, 
                              levels = c("Human rights issues", 
                                         "Humanitarian assistance issues", 
                                         "Government funding", "Private funding"))) %>% 
  arrange(title_clean) %>% 
  mutate_at(vars(theta_1, theta_2, theta_diff, pct_change), 
            list(~percent_format(accuracy = 0.1)(.))) %>% 
  mutate(p.diff.not0 = as.character(round(p.diff.not0, 2))) %>% 
  mutate(title_clean = as.character(title_clean)) %>% 
  select(`H~2a~ and H~3a~` = title_clean, 
         `% likely~Crackdown~` = theta_2, `% likely~No\\ crackdown~` = theta_1,
         `$\\Delta$` = theta_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_likely_3 <- tbl_likely_tidy %>% 
  left_join(tbl_likely_probs, by = "title") %>% 
  filter(category == "Level 3") %>%
  mutate(title_clean = str_remove(title, " \\| Crackdown")) %>% 
  separate(title_clean, c("Issue", "Funding"), sep = " \\| ") %>% 
  mutate(Issue = factor(Issue, levels = c("Human rights", "Humanitarian assistance")),
         Funding = factor(Funding, levels = c("Government", "Private"))) %>% 
  arrange(Issue, Funding) %>% 
  mutate_at(vars(theta_1, theta_2, theta_diff, pct_change), 
            list(~percent_format(accuracy = 0.1)(.))) %>% 
  mutate(p.diff.not0 = as.character(round(p.diff.not0, 2))) %>%
  mutate(Frame = paste0(Issue, " issues, ", Funding, " funding")) %>% 
  select(`H~2a~ and H~3a~ (nested)` = Frame,
         `% likely~Crackdown~` = theta_2, `% likely~No\\ crackdown~` = theta_1,
         `$\\Delta$` = theta_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_likely_2_header <- enframe(colnames(tbl_likely_2)) %>% 
  mutate(value = (value)) %>% 
  pivot_wider(names_from = name, values_from = value) %>% 
  set_names(colnames(tbl_likely_1))

tbl_likely_3_header <- enframe(colnames(tbl_likely_3)) %>% 
  mutate(value = (value)) %>% 
  pivot_wider(names_from = name, values_from = value) %>% 
  set_names(colnames(tbl_likely_1))

bind_rows(tbl_likely_1, 
          tbl_likely_2_header, 
          set_names(tbl_likely_2, colnames(tbl_likely_1)),
          tbl_likely_3_header,
          set_names(tbl_likely_3, colnames(tbl_likely_1))) %>% 
  rename(`H~1a~` = Frame) %>% 
  pandoc.table.return(caption = 'Likelihood of donation and differences in proportions in "crackdown" (treatment) and "no crackdown" (control) conditions; values represent posterior medians {#tbl:likely-diffs}',
                      justify = "lccccc") %T>%
  cat() %>%
  cat(file = here("analysis", "output", "tables", "tbl-likely-diffs.md"))

Likelihood of donation and differences in proportions in “crackdown” (treatment) and “no crackdown” (control) conditions; values represent posterior medians {#tbl:likely-diffs}

H1a	% likelyTreatment	% likelyControl	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Crackdown − No crackdown	46.7%	42.9%	3.8%	8.9%	0.82
Humanitarian assistance − Human rights	45.9%	43.7%	2.3%	5.4%	0.7
Private − Government funding	46.3%	43.4%	2.9%	6.8%	0.75
H2a and H3a	% likelyCrackdown	% likelyNo crackdown	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Human rights issues	44.3%	43.6%	0.6%	1.5%	0.54
Humanitarian assistance issues	49.4%	42.5%	6.8%	15.9%	0.88
Government funding	41.3%	45.8%	-4.6%	-10.0%	0.78
Private funding	52.4%	40.3%	12.1%	30.2%	0.98
H2a and H3a (nested)	% likelyCrackdown	% likelyNo crackdown	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Human rights issues, Government funding	32.0%	47.4%	-15.4%	-32.5%	0.97
Human rights issues, Private funding	57.4%	40.4%	16.9%	41.5%	0.98
Humanitarian assistance issues, Government funding	51.4%	44.7%	6.8%	15.2%	0.8
Humanitarian assistance issues, Private funding	47.5%	41.2%	6.1%	14.8%	0.78

Treatment effects: Amount donated

Priors and models

Following John Kruschke’s “Bayesian Estimation Supersedes the t-test (BEST)” procedure, we estimate means for each group with a t-distribution. We use the following priors for the distribution parameters:

\[ \begin{aligned} x_{\text{group 1, group 2}} &\sim \text{Student } t(\nu, \mu, \sigma) &\text{[likelihood]}\\ \text{Difference} &= x_{\text{group 2}} - x_{\text{group 1}} &\text{[difference in means]} \\ x &: \text{Mean amount donated} \\ \\ \nu &\sim \text{Exponential}(1 / 29) &\text{[prior normality]} \\ \mu_{\text{group 1, group 2}} &\sim \mathcal{N}(\bar{x}_{\text{group 1, group 2}}, 10) &\text{[prior donation mean per group]}\\ \sigma_{\text{group 1, group 2}} &\sim \text{Cauchy}(0, 1)&\text{[prior donation sd per group]} \end{aligned} \]

amount_nu <- ggplot(data = tibble(x = c(0, 200)), aes(x = x)) +
  geom_area(stat = "function", fun = dexp, args = list(rate = 1/29), 
            fill = "grey80", color = "black") +
  labs(x = expression(Normality ~ parameter ~ (nu)), y = "Density") +
  annotate(geom = "label", x = 100, y = 0.009, 
           label = "Exponential(1/29)", size = pts(9)) +
  theme_ngos(base_size = 9, density = TRUE)

amount_mu <- ggplot(data = tibble(x = c(0, 100)), aes(x = x)) +
  geom_area(stat = "function", fun = dnorm, args = list(mean = 50, sd = 10),
            fill = "grey80", color = "black") +
  scale_x_continuous(breaks = c(seq(0, 100, 25)),
                     labels = c("−$50", "−$25", "Group average", "+$25", "+$50")) +
  annotate(geom = "label", x = 50, y = 0.01, label = "N(bar(x), 10)", 
           parse = TRUE, size = pts(9)) +
  labs(x = expression(Average ~ donated ~ (mu)), y = NULL) +
  theme_ngos(base_size = 9, density = TRUE)

amount_sigma <- ggplot(data = tibble(x = c(0, 10)), aes(x = x)) +
  geom_area(stat = "function", fun = dcauchy, args = list(location = 0, scale = 1),
            fill = "grey80", color = "black") +
  scale_x_continuous(labels = dollar) +
  annotate(geom = "label", x = 5, y = 0.08, label = "Cauchy(0, 1)", size = pts(9)) +
  labs(x = expression(SD ~ donated ~ (sigma)), y = NULL) +
  theme_ngos(base_size = 9, density = TRUE)

amount_priors <- amount_nu + amount_mu + amount_sigma
amount_priors

amount_priors %T>% 
  ggsave(., filename = here("analysis", "output", "figures", "prior-amount.pdf"),
         width = 8, height = 1.5, units = "in", device = cairo_pdf) %>%
  ggsave(., filename = here("analysis", "output", "figures", "prior-amount.png"),
         width = 8, height = 1.5, units = "in", type = "cairo", dpi = 300)

Differences

# Sample from the compiled model
all_models_amount_run <- all_models %>% 
  # Simplify data for sending to Stan
  mutate(stan_data = df %>% map(~ {
    select(.x, amount = amount_donate, group = crackdown)
  })) %>% 
  # Run the actual model on the simplified data
  mutate(model = stan_data %>% map(~ {
    amount_donated_best(df = .x, chains = CHAINS, iter = ITER, 
                        warmup = WARMUP, seed = BAYES_SEED)
  })) 

# Extract stuff
all_models_amount <- all_models_amount_run %>% 
  # Extract posterior chains
  mutate(posterior_chains_long = map2(model, stan_data, ~ {
    .x %>% 
      recover_types(.y) %>% 
      gather_draws(mu[group], sigma[group], mu_diff, pct_change, cohen_d, cles, nu, log10nu)
  })) %>% 
  mutate(posterior_chains_wide = map2(model, stan_data, ~ {
    .x %>% 
      recover_types(.y) %>% 
      spread_draws(mu[group], sigma[group], mu_diff, pct_change, cohen_d, cles, nu, log10nu)
  })) %>% 
  # Get HDI medians in tidy form
  mutate(tidy = posterior_chains_long %>% map(~ {
    .x %>% median_hdci() %>% to_broom_names()
  }))

all_models_amount <- all_models_amount %>% 
  # The `group` and `tidy` columns in each of these nested models contain
  # ordered factors, but we can't combine them with unnest() directly because
  # they have different levels (crackdown, issue, and funding), so we need to
  # make them unordered factors before unnesting
  mutate(posterior_chains_long = map(
    posterior_chains_long, ~mutate(., group = factor(group, ordered = FALSE)))
  ) %>% 
  mutate(tidy = map(
    tidy, ~mutate(., group = factor(group, ordered = FALSE)))
  )

tidied_diffs_amount <- all_models_amount %>% 
  unnest(posterior_chains_long) %>% 
  filter(.variable == "mu_diff") %>% 
  mutate(category = case_when(
    str_count(title, "\\|") == 0 ~ "Level 1",
    str_count(title, "\\|") == 1 ~ "Level 2",
    str_count(title, "\\|") == 2 ~ "Level 3"
  )) %>% 
  mutate(title = str_remove(title, " \\| Crackdown"))

level1_amount <- tidied_diffs_amount %>% 
  filter(category == "Level 1") %>%
  mutate(title = recode(title, 
                        Crackdown = "Crackdown −\nNo crackdown",
                        Issue = "Humanitarian\nassistance −\nHuman rights",
                        Funding = "Private −\nGovernment\nfunding")) %>% 
  mutate(title = fct_inorder(title))

plot_diff_amount_a <- ggplot(level1_amount, aes(x = .value, y = fct_rev(title), fill = title)) +
  geom_vline(xintercept = 0) +
  stat_halfeye(.width = c(0.8, 0.95)) +
  scale_x_continuous(labels = dollar) +
  scale_fill_manual(values = ngo_cols(c("blue", "red", "orange"), name = FALSE), guide = FALSE) +
  labs(x = "Difference in amount donated", y = NULL, tag = "A") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

level2_amount <- tidied_diffs_amount %>% 
  filter(category == "Level 2") %>%
  mutate(condition = case_when(
    title %in% c("Human rights", "Humanitarian assistance") ~ "Issue",
    title %in% c("Government", "Private") ~ "Funding"
  )) %>% 
  mutate(facet_title = case_when(
    title %in% c("Human rights", "Humanitarian assistance") ~ paste(title, "issues"),
    title %in% c("Government", "Private") ~ paste(title, "funding")
  )) %>% 
  mutate(title = case_when(
    title %in% c("Humanitarian assistance", "Private") ~ "",
    TRUE ~ "Crackdown −\nNo crackdown")
  )

plot_diff_amount_b <- ggplot(filter(level2_amount, condition == "Issue"), 
                             aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("red", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = dollar) +
  labs(x = NULL, y = NULL, tag = "B") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

plot_diff_amount_c <- ggplot(filter(level2_amount, condition == "Funding"), 
                             aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("orange", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = dollar) +
  labs(x = "Difference in amount donated", y = NULL, tag = "C") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

level3_amount <- tidied_diffs_amount %>% 
  filter(category == "Level 3") %>%
  separate(title, c("issue", "funding"), sep = " \\| ") %>%
  mutate(issue = paste(issue, "issues"),
         funding = paste(funding, "funding")) %>% 
  mutate(facet_title = paste0(issue, "\n", funding)) %>% 
  mutate(title = case_when(
    funding == "Private funding" ~ "",
    TRUE ~ "Crackdown −\nNo crackdown")
  )

plot_diff_amount_d <- ggplot(level3_amount, aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95), fill = ngo_cols("green", name = FALSE)) +
  geom_vline(xintercept = 0) +
  scale_x_continuous(labels = dollar) +
  labs(x = "Difference in amount donated", y = NULL, tag = "D",
       caption = "Point shows posterior median; thick black lines show 80% credible interval;\nthin black lines show 95% credible interval") +
  facet_wrap(~ facet_title, scales = "free_y") +
  theme_ngos(base_size = 8) +
  theme(panel.grid.major.y = element_blank())

plot_diffs_amounts_all <- plot_diff_amount_a / 
  (plot_diff_amount_b / plot_diff_amount_c) / 
  plot_diff_amount_d

plot_diffs_amounts_all

plot_diffs_amounts_all %T>% 
  ggsave(., filename = here("analysis", "output", "figures", "amount-diffs.pdf"),
         width = 4.75, height = 5.5, units = "in", device = cairo_pdf) %>%
  ggsave(., filename = here("analysis", "output", "figures", "amount-diffs.png"),
         width = 4.75, height = 5.5, units = "in", type = "cairo", dpi = 300)

tbl_amounts_tidy <- all_models_amount %>% 
  unnest(tidy) %>% 
  filter(term %in% c("mu", "mu_diff", "pct_change")) %>% 
  group_by(title) %>% 
  mutate(group_id = 1:n()) %>% 
  mutate(term = case_when(
    term == "mu" ~ paste0(term, "_", group_id),
    TRUE ~ term
  )) %>% 
  select(title, term, estimate) %>% 
  pivot_wider(names_from = term, values_from = estimate) %>% 
  mutate(category = case_when(
    str_count(title, "\\|") == 0 ~ "Level 1",
    str_count(title, "\\|") == 1 ~ "Level 2",
    str_count(title, "\\|") == 2 ~ "Level 3"
  )) %>% 
  ungroup()

tbl_amounts_probs <- all_models_amount %>% 
  unnest(posterior_chains_long) %>% 
  filter(.variable == "mu_diff") %>% 
  group_by(title) %>% 
  summarize(p.greater0 = mean(.value > 0),
            p.less0 = mean(.value < 0),
            p.diff.not0 = ifelse(median(.value) > 0, p.greater0, p.less0)) %>% 
  ungroup()

# Save combined table for later use in manuscript
tbl_amounts_tidy %>% 
  left_join(tbl_amounts_probs, by = "title") %>% 
  saveRDS(here("data", "derived_data", "results_models_amount.rds"))

tbl_amount_1 <- tbl_amounts_tidy %>% 
  left_join(tbl_amounts_probs, by = "title") %>% 
  filter(category == "Level 1") %>%
  mutate(title = recode(title, 
                        Crackdown = "Crackdown − No crackdown",
                        Issue = "*Humanitarian assistance − Human rights*",
                        Funding = "*Private − Government funding*")) %>% 
  mutate_at(vars(mu_1, mu_2, mu_diff, p.diff.not0), list(~as.character(round(., 2)))) %>% 
  mutate(pct_change = percent_format(accuracy = 0.1)(pct_change)) %>%
  select(`Frame` = title, `Amount~Treatment~` = mu_2, `Amount~Control~` = mu_1,
         `$\\Delta$` = mu_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_amount_2 <- tbl_amounts_tidy %>% 
  left_join(tbl_amounts_probs, by = "title") %>% 
  filter(category == "Level 2") %>%
  mutate(title_clean = str_remove(title, " \\| Crackdown")) %>% 
  mutate(title_clean = case_when(
    title_clean %in% c("Human rights", "Humanitarian assistance") ~ paste(title_clean, "issues"),
    title_clean %in% c("Government", "Private") ~ paste(title_clean, "funding")
  )) %>% 
  mutate(title_clean = factor(title_clean, 
                              levels = c("Human rights issues", 
                                         "Humanitarian assistance issues", 
                                         "Government funding", "Private funding"))) %>% 
  arrange(title_clean) %>% 
  mutate_at(vars(mu_1, mu_2, mu_diff, p.diff.not0), list(~as.character(round(., 2)))) %>% 
  mutate(pct_change = percent_format(accuracy = 0.1)(pct_change)) %>%
  mutate(title_clean = as.character(title_clean)) %>% 
  select(`H~2b~ and H~3b~` = title_clean, `Amount~Crackdown~` = mu_2, `Amount~No\\ crackdown~` = mu_1,
         `$\\Delta$` = mu_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_amount_3 <- tbl_amounts_tidy %>% 
  left_join(tbl_amounts_probs, by = "title") %>% 
  filter(category == "Level 3") %>%
  mutate(title_clean = str_remove(title, " \\| Crackdown")) %>% 
  separate(title_clean, c("Issue", "Funding"), sep = " \\| ") %>% 
  mutate(Issue = factor(Issue, levels = c("Human rights", "Humanitarian assistance")),
         Funding = factor(Funding, levels = c("Government", "Private"))) %>% 
  arrange(Issue, Funding) %>% 
  mutate(Frame = paste0(Issue, " issues, ", Funding, " funding")) %>% 
  mutate_at(vars(mu_1, mu_2, mu_diff, p.diff.not0), list(~as.character(round(., 2)))) %>% 
  mutate(pct_change = percent_format(accuracy = 0.1)(pct_change)) %>%
  select(`H~2b~ and H~3b~ (nested)` = Frame,
         `Amount~Crackdown~` = mu_2, `Amount~No\\ crackdown~` = mu_1,
         `$\\Delta$` = mu_diff, `$\\%\\Delta$` = pct_change, 
         `$p(\\Delta \\neq 0)$` = p.diff.not0)

tbl_amount_2_header <- enframe(colnames(tbl_amount_2)) %>% 
  mutate(value = (value)) %>% 
  pivot_wider(names_from = name, values_from = value) %>% 
  set_names(colnames(tbl_amount_1))

tbl_amount_3_header <- enframe(colnames(tbl_amount_3)) %>% 
  mutate(value = (value)) %>% 
  pivot_wider(names_from = name, values_from = value) %>% 
  set_names(colnames(tbl_amount_1))

bind_rows(tbl_amount_1, 
          tbl_amount_2_header, 
          set_names(tbl_amount_2, colnames(tbl_amount_1)),
          tbl_amount_3_header,
          set_names(tbl_amount_3, colnames(tbl_amount_1))) %>% 
  rename(`H~1b~` = Frame) %>% 
  pandoc.table.return(caption = 'Mean values and differences in means for amount donated in "crackdown" (treatment) and "no crackdown" (control) conditions; values represent posterior medians {#tbl:amount-diffs}',
                      justify = "lccccc") %T>%
  cat() %>%
  cat(file = here("analysis", "output", "tables", "tbl-amount-diffs.md"))

Mean values and differences in means for amount donated in “crackdown” (treatment) and “no crackdown” (control) conditions; values represent posterior medians {#tbl:amount-diffs}

H1b	AmountTreatment	AmountControl	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Crackdown − No crackdown	16.35	12.96	3.39	26.3%	0.97
Humanitarian assistance − Human rights	14.01	14.84	-0.78	-5.3%	0.66
Private − Government funding	15.11	13.65	1.44	10.6%	0.78
H2b and H3b	AmountCrackdown	AmountNo crackdown	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Human rights issues	17.43	14.94	2.49	16.8%	0.83
Humanitarian assistance issues	16.01	11.74	4.27	36.5%	0.96
Government funding	13.87	12.26	1.64	13.5%	0.74
Private funding	18.97	14.33	4.55	31.6%	0.96
H2b and H3b (nested)	AmountCrackdown	AmountNo crackdown	\(\Delta\)	\(\%\Delta\)	\(p(\Delta \neq 0)\)
Human rights issues, Government funding	10.55	15.04	-4.49	-29.9%	0.9
Human rights issues, Private funding	23.71	14.5	9.09	62.7%	0.99
Humanitarian assistance issues, Government funding	21.42	11.97	9.18	76.4%	0.99
Humanitarian assistance issues, Private funding	15.73	15.79	-0.03	-0.2%	0.5

Effect size

# Effect size
# (μ₁ - μ₂) / sqrt( (σ₁² + σ₂²) / 2)
tidied_eff_size_amount <- all_models_amount %>% 
  unnest(posterior_chains_long) %>% 
  filter(.variable == "cohen_d") %>% 
  mutate(category = case_when(
    str_count(title, "\\|") == 0 ~ "Level 1",
    str_count(title, "\\|") == 1 ~ "Level 2",
    str_count(title, "\\|") == 2 ~ "Level 3"
  )) %>% 
  mutate(title = str_remove(title, " \\| Crackdown")) %>% 
  mutate(title = fct_inorder(title))

effect_sizes <- tribble(
  ~size, ~x_end,
  "Small", 0.2,
  "Medium", 0.5,
  "Large", 0.8
) %>% 
  mutate(x_start = -x_end)

ggplot(tidied_eff_size_amount, aes(x = .value, y = fct_rev(title))) +
  stat_halfeye(.width = c(0.8, 0.95)) +
  geom_vline(xintercept = 0, size = 1) +
  geom_vline(xintercept = c(0.2, -0.2), linetype = "dotted", size = 0.5) +
  geom_vline(xintercept = c(0.5, -0.5), linetype = "dotted", size = 0.5) +
  geom_vline(xintercept = c(0.5, -0.5), linetype = "dotted", size = 0.5) +
  coord_cartesian(xlim = c(-1, 1)) +
  labs(x = "Cohen's d (effect size)", y = NULL) +
  theme_ngos()

Original computing environment

Here’s what we used the last time we built this page

writeLines(readLines(file.path(Sys.getenv("HOME"), ".R/Makevars")))

## # http://dirk.eddelbuettel.com/blog/2017/11/27/#011_faster_package_installation_one
## VER=
## CCACHE=ccache
## CC=$(CCACHE) gcc$(VER)
## CXX=$(CCACHE) g++$(VER)
## CXX11=$(CCACHE) g++$(VER)
## CXX14=$(CCACHE) g++$(VER)
## FC=$(CCACHE) gfortran$(VER)
## F77=$(CCACHE) gfortran$(VER)
## 
## CXX14FLAGS=-O3 -march=native -mtune=native -fPIC

devtools::session_info()

## ─ Session info ───────────────────────────────────────────────────────────────
##  setting  value                       
##  version  R version 4.0.2 (2020-06-22)
##  os       macOS Catalina 10.15.6      
##  system   x86_64, darwin17.0          
##  ui       X11                         
##  language (EN)                        
##  collate  en_US.UTF-8                 
##  ctype    en_US.UTF-8                 
##  tz       America/New_York            
##  date     2020-10-01                  
## 
## ─ Packages ───────────────────────────────────────────────────────────────────
##  package                * version  date       lib source        
##  arrayhelpers             1.1-0    2020-02-04 [1] CRAN (R 4.0.0)
##  assertthat               0.2.1    2019-03-21 [1] CRAN (R 4.0.0)
##  backports                1.1.9    2020-08-24 [1] CRAN (R 4.0.2)
##  base64enc                0.1-3    2015-07-28 [1] CRAN (R 4.0.0)
##  blob                     1.2.1    2020-01-20 [1] CRAN (R 4.0.0)
##  boot                     1.3-25   2020-04-26 [1] CRAN (R 4.0.2)
##  broom                  * 0.7.0    2020-07-09 [1] CRAN (R 4.0.2)
##  callr                    3.4.3    2020-03-28 [1] CRAN (R 4.0.0)
##  cellranger               1.1.0    2016-07-27 [1] CRAN (R 4.0.0)
##  cli                      2.0.2    2020-02-28 [1] CRAN (R 4.0.0)
##  coda                     0.19-3   2019-07-05 [1] CRAN (R 4.0.0)
##  codetools                0.2-16   2018-12-24 [1] CRAN (R 4.0.2)
##  colorspace               1.4-1    2019-03-18 [1] CRAN (R 4.0.0)
##  crackdownsphilanthropy * 0.9      2020-10-01 [1] local         
##  crayon                   1.3.4    2017-09-16 [1] CRAN (R 4.0.0)
##  curl                     4.3      2019-12-02 [1] CRAN (R 4.0.0)
##  dagitty                  0.3-0    2020-07-21 [1] CRAN (R 4.0.2)
##  DBI                      1.1.0    2019-12-15 [1] CRAN (R 4.0.0)
##  dbplyr                   1.4.4    2020-05-27 [1] CRAN (R 4.0.2)
##  desc                     1.2.0    2018-05-01 [1] CRAN (R 4.0.0)
##  devtools                 2.3.1    2020-07-21 [1] CRAN (R 4.0.2)
##  digest                   0.6.25   2020-02-23 [1] CRAN (R 4.0.0)
##  distributional           0.2.0    2020-08-03 [1] CRAN (R 4.0.2)
##  dplyr                  * 1.0.2    2020-08-18 [1] CRAN (R 4.0.2)
##  ellipsis                 0.3.1    2020-05-15 [1] CRAN (R 4.0.0)
##  evaluate                 0.14     2019-05-28 [1] CRAN (R 4.0.0)
##  fansi                    0.4.1    2020-01-08 [1] CRAN (R 4.0.0)
##  farver                   2.0.3    2020-01-16 [1] CRAN (R 4.0.0)
##  forcats                * 0.5.0    2020-03-01 [1] CRAN (R 4.0.0)
##  fs                       1.5.0    2020-07-31 [1] CRAN (R 4.0.2)
##  generics                 0.0.2    2018-11-29 [1] CRAN (R 4.0.0)
##  ggdag                  * 0.2.2    2020-02-13 [1] CRAN (R 4.0.0)
##  ggdist                   2.2.0    2020-07-12 [1] CRAN (R 4.0.2)
##  ggforce                  0.3.2    2020-06-23 [1] CRAN (R 4.0.2)
##  ggplot2                * 3.3.2    2020-06-19 [1] CRAN (R 4.0.2)
##  ggraph                 * 2.0.3    2020-05-20 [1] CRAN (R 4.0.2)
##  ggrepel                  0.8.2    2020-03-08 [1] CRAN (R 4.0.0)
##  ggstance               * 0.3.4    2020-04-02 [1] CRAN (R 4.0.0)
##  glue                     1.4.2    2020-08-27 [1] CRAN (R 4.0.2)
##  graphlayouts             0.7.0    2020-04-25 [1] CRAN (R 4.0.0)
##  gridExtra                2.3      2017-09-09 [1] CRAN (R 4.0.0)
##  gtable                   0.3.0    2019-03-25 [1] CRAN (R 4.0.0)
##  haven                    2.3.1    2020-06-01 [1] CRAN (R 4.0.2)
##  here                   * 0.1      2017-05-28 [1] CRAN (R 4.0.0)
##  hms                      0.5.3    2020-01-08 [1] CRAN (R 4.0.0)
##  htmltools                0.5.0    2020-06-16 [1] CRAN (R 4.0.0)
##  httr                     1.4.2    2020-07-20 [1] CRAN (R 4.0.2)
##  igraph                   1.2.5    2020-03-19 [1] CRAN (R 4.0.0)
##  inline                   0.3.15   2018-05-18 [1] CRAN (R 4.0.0)
##  janitor                * 2.0.1    2020-04-12 [1] CRAN (R 4.0.0)
##  jsonlite                 1.7.0    2020-06-25 [1] CRAN (R 4.0.2)
##  knitr                    1.29     2020-06-23 [1] CRAN (R 4.0.2)
##  labeling                 0.3      2014-08-23 [1] CRAN (R 4.0.0)
##  lattice                  0.20-41  2020-04-02 [1] CRAN (R 4.0.2)
##  lifecycle                0.2.0    2020-03-06 [1] CRAN (R 4.0.0)
##  loo                      2.3.1    2020-07-14 [1] CRAN (R 4.0.2)
##  lubridate                1.7.9    2020-06-08 [1] CRAN (R 4.0.2)
##  magrittr                 1.5      2014-11-22 [1] CRAN (R 4.0.0)
##  MASS                     7.3-52   2020-08-18 [1] CRAN (R 4.0.2)
##  matrixStats              0.56.0   2020-03-13 [1] CRAN (R 4.0.0)
##  memoise                  1.1.0    2017-04-21 [1] CRAN (R 4.0.0)
##  modelr                   0.1.8    2020-05-19 [1] CRAN (R 4.0.2)
##  munsell                  0.5.0    2018-06-12 [1] CRAN (R 4.0.0)
##  pander                 * 0.6.3    2018-11-06 [1] CRAN (R 4.0.0)
##  patchwork              * 1.0.1    2020-06-22 [1] CRAN (R 4.0.2)
##  pillar                   1.4.6    2020-07-10 [1] CRAN (R 4.0.2)
##  pkgbuild                 1.1.0    2020-07-13 [1] CRAN (R 4.0.2)
##  pkgconfig                2.0.3    2019-09-22 [1] CRAN (R 4.0.0)
##  pkgload                  1.1.0    2020-05-29 [1] CRAN (R 4.0.2)
##  plyr                     1.8.6    2020-03-03 [1] CRAN (R 4.0.0)
##  polyclip                 1.10-0   2019-03-14 [1] CRAN (R 4.0.0)
##  prettyunits              1.1.1    2020-01-24 [1] CRAN (R 4.0.0)
##  processx                 3.4.3    2020-07-05 [1] CRAN (R 4.0.0)
##  ps                       1.3.4    2020-08-11 [1] CRAN (R 4.0.2)
##  purrr                  * 0.3.4    2020-04-17 [1] CRAN (R 4.0.0)
##  R6                       2.4.1    2019-11-12 [1] CRAN (R 4.0.0)
##  Rcpp                     1.0.5    2020-07-06 [1] CRAN (R 4.0.2)
##  RcppParallel             5.0.2    2020-06-24 [1] CRAN (R 4.0.2)
##  readr                  * 1.3.1    2018-12-21 [1] CRAN (R 4.0.0)
##  readxl                   1.3.1    2019-03-13 [1] CRAN (R 4.0.0)
##  remotes                  2.2.0    2020-07-21 [1] CRAN (R 4.0.2)
##  reprex                   0.3.0    2019-05-16 [1] CRAN (R 4.0.0)
##  rlang                    0.4.7    2020-07-09 [1] CRAN (R 4.0.2)
##  rmarkdown                2.3      2020-06-18 [1] CRAN (R 4.0.2)
##  rprojroot                1.3-2    2018-01-03 [1] CRAN (R 4.0.0)
##  rstan                  * 2.21.2   2020-07-27 [1] CRAN (R 4.0.2)
##  rstudioapi               0.11     2020-02-07 [1] CRAN (R 4.0.0)
##  rvest                    0.3.6    2020-07-25 [1] CRAN (R 4.0.2)
##  scales                 * 1.1.1    2020-05-11 [1] CRAN (R 4.0.0)
##  sessioninfo              1.1.1    2018-11-05 [1] CRAN (R 4.0.0)
##  snakecase                0.11.0   2019-05-25 [1] CRAN (R 4.0.0)
##  StanHeaders            * 2.21.0-6 2020-08-16 [1] CRAN (R 4.0.2)
##  stringi                  1.4.6    2020-02-17 [1] CRAN (R 4.0.0)
##  stringr                * 1.4.0    2019-02-10 [1] CRAN (R 4.0.0)
##  svUnit                   1.0.3    2020-04-20 [1] CRAN (R 4.0.0)
##  testthat                 2.3.2    2020-03-02 [1] CRAN (R 4.0.0)
##  tibble                 * 3.0.3    2020-07-10 [1] CRAN (R 4.0.2)
##  tidybayes              * 2.1.1    2020-06-19 [1] CRAN (R 4.0.2)
##  tidygraph                1.2.0    2020-05-12 [1] CRAN (R 4.0.0)
##  tidyr                  * 1.1.2    2020-08-27 [1] CRAN (R 4.0.2)
##  tidyselect               1.1.0    2020-05-11 [1] CRAN (R 4.0.0)
##  tidyverse              * 1.3.0    2019-11-21 [1] CRAN (R 4.0.0)
##  tweenr                   1.0.1    2018-12-14 [1] CRAN (R 4.0.0)
##  usethis                  1.6.1    2020-04-29 [1] CRAN (R 4.0.0)
##  V8                       3.2.0    2020-06-19 [1] CRAN (R 4.0.2)
##  vctrs                    0.3.4    2020-08-29 [1] CRAN (R 4.0.2)
##  viridis                  0.5.1    2018-03-29 [1] CRAN (R 4.0.0)
##  viridisLite              0.3.0    2018-02-01 [1] CRAN (R 4.0.0)
##  withr                    2.2.0    2020-04-20 [1] CRAN (R 4.0.0)
##  xfun                     0.16     2020-07-24 [1] CRAN (R 4.0.2)
##  xml2                     1.3.2    2020-04-23 [1] CRAN (R 4.0.0)
##  yaml                     2.2.1    2020-02-01 [1] CRAN (R 4.0.0)
## 
## [1] /Library/Frameworks/R.framework/Versions/4.0/Resources/library

Copyright © 2017–2020 Suparna Chaudhry and Andrew Heiss. Licensed under the MIT License. Code available on GitHub.