Tutorial of Group-to-Person Generalizability Problem (McManus et al. 2021)

Load Data/Packages

Experiment 1 data from McManus et al. (2021) should be loaded into R: https://osf.io/kv6sw/. Additionally, analyses from that experiment should be conducted using the .Rmd file from the paper’s OSF page: https://osf.io/6mtg9/. Then, this script can be run.

options(scipen = 99) # removes scientific notation

Reorganizing Data

# Select E1a-b long datasets, select variables of interest, widen data, create new vars

E1ab_wide <- E1_all_long  %>%
  filter(BSs_cond != "CUZ vs SIB") %>% # removes dataset which compares judgments of helping family to other family
  select(ResponseId,
         BSs_cond, Relation, `Choice Context`,
         moral) %>% # selects only relevant vars
  pivot_wider(names_from = c(`Choice Context`, Relation), values_from = moral) %>% # creates dataset with four within-subj conditions per participant (similar to how data would be structured in SPSS) 
  mutate(NoChoice_Moral_Diff = (`No Choice_Distant` - `No Choice_Close`)) %>%
  mutate(Choice_Moral_Diff = (Choice_Distant - Choice_Close)) %>% # creates person-level difference scores for each simple effect of interest
  mutate(Int_Moral = (NoChoice_Moral_Diff - Choice_Moral_Diff)) # creates person-level interaction score

Creating All Possible 2x2 Person-Level Patterns

E1ab_wide <- E1ab_wide %>%
  mutate(`2x2_Pattern` = case_when(
    (NoChoice_Moral_Diff == 0 & Choice_Moral_Diff == 0) ~ "Zero, Zero, Zero",
    (NoChoice_Moral_Diff == 0 & Choice_Moral_Diff < 0) ~ "Zero, Neg, Pos",
    (NoChoice_Moral_Diff == 0 & Choice_Moral_Diff > 0) ~ "Zero, Pos, Neg",
    (NoChoice_Moral_Diff < 0 & Choice_Moral_Diff == 0) ~ "Neg, Zero, Neg",
    (NoChoice_Moral_Diff < 0 & Choice_Moral_Diff < 0 & NoChoice_Moral_Diff == Choice_Moral_Diff) ~ "Neg, Neg, Zero",
    (NoChoice_Moral_Diff < 0 & Choice_Moral_Diff > 0) ~ "Neg, Pos, Neg",
    (NoChoice_Moral_Diff < 0 & Choice_Moral_Diff < 0 & NoChoice_Moral_Diff > Choice_Moral_Diff) ~ "Neg, Neg, Pos",
    (NoChoice_Moral_Diff < 0 & Choice_Moral_Diff < 0 & NoChoice_Moral_Diff < Choice_Moral_Diff) ~ "Neg, Neg, Neg",
    (NoChoice_Moral_Diff > 0 & Choice_Moral_Diff == 0) ~ "Pos, Zero, Pos",
    (NoChoice_Moral_Diff > 0 & Choice_Moral_Diff < 0) ~ "Pos, Neg, Pos", # predicted effect
    (NoChoice_Moral_Diff > 0 & Choice_Moral_Diff > 0 & NoChoice_Moral_Diff == Choice_Moral_Diff) ~ "Pos, Pos, Zero",
    (NoChoice_Moral_Diff > 0 & Choice_Moral_Diff > 0 & NoChoice_Moral_Diff < Choice_Moral_Diff) ~ "Pos, Pos, Neg",
    (NoChoice_Moral_Diff > 0 & Choice_Moral_Diff > 0 & NoChoice_Moral_Diff > Choice_Moral_Diff) ~ "Pos, Pos, Pos"))

# make variable a factor
E1ab_wide$`2x2_Pattern` <- as.factor(E1ab_wide$`2x2_Pattern`)

Plotting Persons-as-Effect-Sizes

E1a (STR vs SIB)

print(E1a_MMY_Moral_Person <- ggplot(data = E1ab_wide %>% filter(BSs_cond == "STR vs SIB"), aes(x = `2x2_Pattern`, 
                                                  fill = `2x2_Pattern`)) +
  geom_bar(position = "dodge", size = 0.5) +
        scale_y_continuous(labels = function(x) paste0(round(x/203*100), "%"), breaks = seq(0,101.5,20.3)) +
        coord_cartesian(ylim = c(1,101.5)) +
        scale_x_discrete(drop = FALSE) +
        scale_fill_manual(drop = FALSE, values = c(
                                     "lightgrey", 
                                     "azure4",
                                     "lightgrey",
                                     "lightgrey",
                                     "lightgrey",
                                     "black", # group-level pattern
                                     "lightgrey",
                                     "azure4",
                                     "lightgrey",
                                     "azure4",
                                     "azure4",
                                     "lightgrey",
                                     "lightgrey"))+
        theme_classic() +
        theme(legend.position = "none") + 
        xlab("\n2x2 Direction (Simple Effects = Stranger - Sibling; Interaction = No Choice Simple - Choice Simple)") +
        ylab("Perentage of Participants\n") +  
        theme(axis.title.x = element_text(size = 18), 
              axis.title.y = element_text(size = 20),
              axis.text.x = element_text(color = "black", size = 13, angle = 60, vjust = 0.5, hjust=0.5),
              axis.text.y = element_text(color = "black", size = 20),
              strip.text.x = element_text(color = "black", size = 20),
              legend.title = element_text(color = "black", size = 18),
              legend.text = element_text(color = "black", size = 16)))

ggsave("E1a_MMY_Moral_Person.png")

Saving 14 x 9 in image

E1b (STR vs CUZ) - Tutorial Figure

print(E1b_MMY_Moral_Person <- ggplot(data = E1ab_wide %>% filter(BSs_cond == "STR vs CUZ"), aes(x = `2x2_Pattern`, 
                                                  fill = `2x2_Pattern`)) +
  geom_bar(position = "dodge", size = 0.5) +
        scale_y_continuous(labels = function(x) paste0(round(x/203*100), "%"), breaks = seq(0,101.5,20.3)) +
        coord_cartesian(ylim = c(1,101.5)) +
        scale_x_discrete(drop = FALSE) +
        scale_fill_manual(drop = FALSE, values = c(
                                     "lightgrey", 
                                     "azure4",
                                     "lightgrey",
                                     "lightgrey",
                                     "lightgrey",
                                     "black", # group-level pattern
                                     "lightgrey",
                                     "azure4",
                                     "lightgrey",
                                     "azure4",
                                     "azure4",
                                     "lightgrey",
                                     "lightgrey"))+
        theme_classic() +
        theme(legend.position = "none") + 
        xlab("\n2x2 Direction (Simple Effects = Stranger - Cousin; Interaction = No Choice Simple - Choice Simple)") +
        ylab("Perentage of Participants\n") +  
        theme(axis.title.x = element_text(size = 18), 
              axis.title.y = element_text(size = 20),
              axis.text.x = element_text(color = "black", size = 13, angle = 60, vjust = 0.5, hjust=0.5),
              axis.text.y = element_text(color = "black", size = 20),
              strip.text.x = element_text(color = "black", size = 20),
              legend.title = element_text(color = "black", size = 18),
              legend.text = element_text(color = "black", size = 16)))

ggsave("E1b_MMY_Moral_Person.png")

Saving 14 x 9 in image

Tabulating Raw Frequencies

E1ab_wide %>% 
  group_by(BSs_cond, `2x2_Pattern`) %>% 
  dplyr::summarize(`2x2_pattern` = n())

`summarise()` regrouping output by 'BSs_cond' (override with `.groups` argument)