Read in the play 2 data ahead of time using other script. (DiggingCPC_5_play2)

Play 2 has 1922 unique participants’ worth of data.

Risk-Taking Behaviour in the Gain Domain, Based on Future Information

participants take the riskier option around 50% of the time in gain domains when the future is positive.
when the future is negative, participants take the riskier option less.

happyData1 <- readRDS("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/data/happyData/happyData1.rds")
happyData1_diag <- readRDS("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/data/happyData/happyData1_diag.rds")
happyData1_past <- readRDS("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/data/happyData/happyData1_past.rds")

library(readr)
library(NatParksPalettes)
library(ggplot2)
library(Hmisc) #for computing correlation stuff

## 
## Attaching package: 'Hmisc'

## The following objects are masked from 'package:base':
## 
##     format.pval, units

library(knitr) 
library(tidyverse, warn.conflict=F)

## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.4     ✔ stringr   1.5.1
## ✔ forcats   1.0.0     ✔ tibble    3.2.1
## ✔ lubridate 1.9.3     ✔ tidyr     1.3.1
## ✔ purrr     1.0.2

## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter()    masks stats::filter()
## ✖ dplyr::lag()       masks stats::lag()
## ✖ dplyr::src()       masks Hmisc::src()
## ✖ dplyr::summarize() masks Hmisc::summarize()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

library(corrplot)

## corrplot 0.95 loaded

library(dplyr)
library(MASS)

## 
## Attaching package: 'MASS'
## 
## The following object is masked from 'package:dplyr':
## 
##     select

library(sandwich)
library(lmtest)

## Loading required package: zoo
## 
## Attaching package: 'zoo'
## 
## The following objects are masked from 'package:base':
## 
##     as.Date, as.Date.numeric

library(broom)
library(dplyr)
library(gt)

## 
## Attaching package: 'gt'
## 
## The following object is masked from 'package:Hmisc':
## 
##     html

library(lme4)

## Loading required package: Matrix
## 
## Attaching package: 'Matrix'
## 
## The following objects are masked from 'package:tidyr':
## 
##     expand, pack, unpack

library(dunn.test)

source("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/poster_theme.R") #for figure themes

## Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
## ℹ Please use the `linewidth` argument instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.

source("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/functions/process_initquiz.R")
source("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/functions/process_digging_long.R")
source("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/functions/process_params.R")
source("~/Desktop/Rutledge Lab/ApathyDepression/updatedData/functions/softmax.R")

sample_size = length(unique(happyData1$userKey))

gamb_future <- happyData1 %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative")) %>% 
  filter(Trial=="Gain") %>% # selecting only gain domain
  group_by(userKey,Future) %>%
  summarise(total_choices = n(),
            risky_choices = sum(Choice == 2, na.rm = TRUE),
            percent_risky = (risky_choices / total_choices) * 100,
            .groups = 'drop')

summary_gamb_future <- gamb_future %>%
  group_by(Future) %>%
  summarise(mean_gambling = mean(percent_risky),
            sem_gambling = sd(percent_risky)/sqrt(n())) %>%
  na.omit()

fig1a <- ggplot(summary_gamb_future, aes(x=Future, y=mean_gambling, color=Future)) +
    geom_point(position = position_dodge(0.3), width = 0.3) +
    geom_errorbar(aes(ymin = mean_gambling - sem_gambling, ymax = mean_gambling + sem_gambling),
                  position = position_dodge(0.3), width = 0.1) + 
    geom_hline(yintercept = 50, linetype = "dotted", color = "black") +  
    scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
    labs(
      title = paste0("% Gambling based on Future (Gain Domain) (N=", sample_size, ")"),
      x = "Future",
      y = "% Risky Choices"
    ) +
    ylim(40,70) + 
    theme_minimal() +
    poster_theme

## Warning in geom_point(position = position_dodge(0.3), width = 0.3): Ignoring
## unknown parameters: `width`

print(fig1a)

# summary: increased gambling when the future is positive vs negative in gain domain.

Risk-Taking in the Gain Domain Based on Future Information and Anxiety Diagnosis

Participants without anxiety particularly show the above effect of taking the riskier option more in the gain domain when the future is positive vs negative.
Participants with anxiety show similar behaviour irrespective of future information.

sample_size = length(unique(happyData1_diag$userKey))

# Prepare data for plotting
gamb_future <- happyData1_diag %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative")) %>% 
  filter(Trial == "Gain") %>%  # selecting only gain domain
  group_by(userKey, Future, diagnosis_anxiety) %>%
  summarise(total_choices = n(),
            risky_choices = sum(Choice == 2, na.rm = TRUE),
            percent_risky = (risky_choices / total_choices) * 100,
            .groups = 'drop')

# Summarize the data for plotting
summary_gamb_future <- gamb_future %>%
  group_by(Future, diagnosis_anxiety) %>%
  summarise(mean_gambling = mean(percent_risky, na.rm = TRUE),
            sem_gambling = sd(percent_risky, na.rm = TRUE) / sqrt(n()),
            .groups = 'drop')

# Create the plot with facet_wrap for "anx" and "no anx" groups
fig1b <- ggplot(summary_gamb_future, aes(x = Future, y = mean_gambling, color = Future)) +
  geom_point(position = position_dodge(0.3)) +
  geom_errorbar(aes(ymin = mean_gambling - sem_gambling, ymax = mean_gambling + sem_gambling),
                position = position_dodge(0.3), width = 0.1) +
  geom_hline(yintercept = 50, linetype = "dotted", color = "black") +
  scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
  labs(
    title = paste0("% Gambling based on Future (Gain Domain) (N = ", sample_size, ")"),
    x = "Future",
    y = "% Risky Choices"
  ) +
  ylim(40, 70) +
  theme_minimal() +
  poster_theme + 
  facet_wrap(~ diagnosis_anxiety, labeller = labeller(diagnosis_anxiety = c("anx" = "Anxiety", "no anx" = "No Anxiety")))

# Display the plot
print(fig1b)

Risk-Taking Behaviour in Gain Domain Based on Future Information and Depression Diagnosis

Reversed patten to anxiety diagnosis, where participants with depression exhibit different risk-taking behaviour when the future is positive vs negative in the gain domain.
Participants with depression diagnosis take more risks when the future is positive vs negative
Participants without depression also do this, but there is less of a difference in the % of risk-taking based on future information.

sample_size = length(unique(happyData1_diag$userKey))

# Prepare data for plotting
gamb_future <- happyData1_diag %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative")) %>% 
  filter(Trial == "Gain") %>%  # selecting only gain domain
  group_by(userKey, Future, diagnosis_depression) %>%
  summarise(total_choices = n(),
            risky_choices = sum(Choice == 2, na.rm = TRUE),
            percent_risky = (risky_choices / total_choices) * 100,
            .groups = 'drop')

# Summarize the data for plotting
summary_gamb_future <- gamb_future %>%
  group_by(Future, diagnosis_depression) %>%
  summarise(mean_gambling = mean(percent_risky, na.rm = TRUE),
            sem_gambling = sd(percent_risky, na.rm = TRUE) / sqrt(n()),
            .groups = 'drop') %>%
  na.omit()

# Create the plot with facet_wrap for "dep" and "no dep" groups
fig1b2 <- ggplot(summary_gamb_future, aes(x = Future, y = mean_gambling, color = Future)) +
  geom_point(position = position_dodge(0.3)) +
  geom_errorbar(aes(ymin = mean_gambling - sem_gambling, ymax = mean_gambling + sem_gambling),
                position = position_dodge(0.3), width = 0.1) +
  geom_hline(yintercept = 50, linetype = "dotted", color = "black") +
  scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
  labs(
    title = paste0("% Gambling based on Future (Gain Domain) (N = ", sample_size, ")"),
    x = "Future",
    y = "% Risky Choices"
  ) +
  ylim(40, 70) +
  theme_minimal() +
  poster_theme + 
  facet_wrap(~ diagnosis_depression, labeller = labeller(diagnosis_depression = c("dep" = "Depression", "no dep" = "No Depression")))

# Display the plot
print(fig1b2)

Risk-Taking Behaviour in Gain Domain Based on Past Information and Anxiety Diagnosis

As with future information, participants without anxiety exhibit different behaviour based on the past island, gambling more on the current gain trial if they previously experienced a loss domain.
Participants with anxiety show similar behaviour based on the past island domains.

happyData1_past_diag <- happyData1_diag %>%
  arrange(userKey, TrialNumber) %>%
  group_by(userKey) %>%
  mutate(
    # Create an indicator for valid happiness ratings
    happyind = !is.na(zHappy),
    
    # Identify if the current trial is on the same island as the previous valid happiness rating
    prev_happyind_trial = lag(Trial, order_by = TrialNumber),  # Previous trial type (Gain/Loss)
    
    # Set PastIsland based on the previous valid happiness rating's trial type
    PastIsland = case_when(
      lag(happyind, order_by = TrialNumber) & prev_happyind_trial == "Gain" ~ 2,
      lag(happyind, order_by = TrialNumber) & prev_happyind_trial == "Loss" ~ 1,
      TRUE ~ NA_real_  # NA for the first island or no previous valid row
    )
  ) %>%
  ungroup()

happyData1_past_diag <- happyData1_past_diag %>%
  fill(PastIsland, .direction = "up")



sample_size = length(unique(happyData1_past_diag$userKey))

# Prepare data for plotting
gamb_past <- happyData1_past_diag %>%
  mutate(Past = case_when(PastIsland == 2 ~ "Positive",
                            PastIsland == 1 ~ "Negative")) %>% 
  filter(Trial == "Gain") %>%  # selecting only gain domain
  group_by(userKey, Past, diagnosis_anxiety) %>%
  summarise(total_choices = n(),
            risky_choices = sum(Choice == 2, na.rm = TRUE),
            percent_risky = (risky_choices / total_choices) * 100,
            .groups = 'drop')

# Summarize the data for plotting
summary_gamb_past <- gamb_past %>%
  group_by(Past, diagnosis_anxiety) %>%
  summarise(mean_gambling = mean(percent_risky, na.rm = TRUE),
            sem_gambling = sd(percent_risky, na.rm = TRUE) / sqrt(n()),
            .groups = 'drop') %>%
  na.omit()

# Create the plot with facet_wrap for "anx" and "no anx" groups
fig1c <- ggplot(summary_gamb_past, aes(x = Past, y = mean_gambling, color = Past)) +
  geom_point(position = position_dodge(0.3)) +
  geom_errorbar(aes(ymin = mean_gambling - sem_gambling, ymax = mean_gambling + sem_gambling),
                position = position_dodge(0.3), width = 0.1) +
  geom_hline(yintercept = 50, linetype = "dotted", color = "black") +
  scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
  labs(
    title = paste0("% Gambling based on Past (Gain Domain) (N = ", sample_size, ")"),
    x = "Past",
    y = "% Risky Choices"
  ) +
  ylim(40, 70) +
  theme_minimal() +
  poster_theme + 
  facet_wrap(~ diagnosis_anxiety, labeller = labeller(diagnosis_anxiety = c("anx" = "Anxiety", "no anx" = "No Anxiety")))

# Display the plot
print(fig1c)

Happiness Plots

Happiness with Future (in Loss Domain)

z-scored happiness shows decreased current happiness (in the loss domain) when the future is positive vs negative

sample_size = length(unique(happyData1$userKey))

happy_future <- happyData1 %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative")) %>% 
  filter(Trial=="Loss") %>% # selecting only gain domain
  group_by(userKey,Future) %>%
  summarise(happiness = mean(zHappy, na.rm=TRUE),
                        .groups = 'drop') %>%
  na.omit()


summary_happy_future <- happy_future %>%
  group_by(Future) %>%
  summarise(mean_happiness = mean(happiness),
            sem_happiness = sd(happiness)/sqrt(n())) %>%
  na.omit()

fig2a <- ggplot(summary_happy_future, aes(x=Future, y=mean_happiness, color=Future)) +
    geom_point(position = position_dodge(0.3), width = 0.3) +
    geom_errorbar(aes(ymin = mean_happiness - sem_happiness, ymax = mean_happiness + sem_happiness),
                  position = position_dodge(0.3), width = 0.1) + 
    geom_hline(yintercept = 0, linetype = "dotted", color = "black") +  
    scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
    labs(
      title = paste0("zHappy based on Future (Loss Domain) (N=", sample_size, ")"),
      x = "Future",
      y = "z-scored Happiness"
    ) +
    ylim(-0.3,0.3) + 
    theme_minimal() +
    poster_theme

## Warning in geom_point(position = position_dodge(0.3), width = 0.3): Ignoring
## unknown parameters: `width`

print(fig2a)

# summary: decreased happiness when future is positive in loss domain

Happiness based on Future Information and Past Experience in the Gain Domain

In the current gain domain, participants show increased happiness when the past island was a loss domain and when the future island is a loss domain
Happieness is similar regardless of whether the future is a loss or gain island if the past island was a gain domain.
So, if the past island was negative, future information seems to matter in affecting happiness.

sample_size = length(unique(happyData1_past$userKey))

happy_future_past <- happyData1_past %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative"),
         Past = case_when(PastIsland == 2 ~ "Positive",
                          PastIsland == 1 ~ "Negative")) %>% 
  filter(Trial=="Gain") %>% # selecting only gain domain
  group_by(userKey,Future, Past) %>%
  summarise(happiness = mean(zHappy, na.rm=TRUE),
                        .groups = 'drop') %>%
  na.omit()


summary_happy_future_past <- happy_future_past %>%
  group_by(Future, Past) %>%
  summarise(mean_happiness = mean(happiness),
            sem_happiness = sd(happiness)/sqrt(n())) %>%
  na.omit()

## `summarise()` has grouped output by 'Future'. You can override using the
## `.groups` argument.

fig2b <- ggplot(summary_happy_future_past, aes(x=Future, y=mean_happiness, color=Past)) +
    geom_point(position = position_dodge(0.3), width = 0.3) +
    geom_line(position = position_dodge(0.3), aes(group = Past)) +  # Add lines connecting points within groups
    geom_errorbar(aes(ymin = mean_happiness - sem_happiness, ymax = mean_happiness + sem_happiness),
                  position = position_dodge(0.3), width = 0.1) + 
    geom_hline(yintercept = 0, linetype = "dotted", color = "black") +  
    scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
    labs(
      title = paste0("zHappy based on Future (Gain Domain) (N=", sample_size, ")"),
      x = "Future",
      y = "z-scored Happiness"
    ) +
    ylim(-0.5,0.5) + 
    theme_minimal() +
    poster_theme

## Warning in geom_point(position = position_dodge(0.3), width = 0.3): Ignoring
## unknown parameters: `width`

print(fig2b)

# summary: increased gambling when the future is positive vs negative in gain domain.

Happiness based on Future Information and Anxiety Diagnosis in the Loss Domain

In the loss domain, people without anxiety show similar happiness levels regardless of whether the future is positive or negative.
People with anxiety are happier when the future is negative vs positive if they are currently in the loss domain.
Perhaps there is a preference for persistence - and participants with anxiety do not like having to switch policies when the island changes if they’re in a loss domain.

sample_size = length(unique(happyData1_diag$userKey))

happy_future <- happyData1_diag %>%
  mutate(Future = case_when(NextIsland == 2 ~ "Positive",
                            NextIsland == 1 ~ "Negative")) %>% 
  filter(Trial=="Loss") %>% # selecting only gain domain
  group_by(userKey,Future,diagnosis_anxiety) %>%
  summarise(happiness = mean(zHappy, na.rm=TRUE),
                        .groups = 'drop') %>%
  na.omit()


summary_happy_future <- happy_future %>%
  group_by(Future, diagnosis_anxiety) %>%
  summarise(mean_happiness = mean(happiness),
            sem_happiness = sd(happiness)/sqrt(n())) %>%
  na.omit()

## `summarise()` has grouped output by 'Future'. You can override using the
## `.groups` argument.

fig2c <- ggplot(summary_happy_future, aes(x=Future, y=mean_happiness, color=Future)) +
    geom_point(position = position_dodge(0.3), width = 0.3) +
    geom_errorbar(aes(ymin = mean_happiness - sem_happiness, ymax = mean_happiness + sem_happiness),
                  position = position_dodge(0.3), width = 0.1) + 
    geom_hline(yintercept = 0, linetype = "dotted", color = "black") +  
    scale_color_manual(values = c("Positive" = "#FFD700", "Negative" = "#FF6347")) +
    labs(
      title = paste0("zHappy based on Future (Loss Domain) (N=", sample_size, ")"),
      x = "Future",
      y = "z-scored Happiness"
    ) +
    ylim(-0.3,0.3) + 
    theme_minimal() +
    poster_theme +
facet_wrap(~ diagnosis_anxiety, labeller = labeller(diagnosis_anxiety = c("anx" = "Anxiety", "no anx" = "No Anxiety")))

## Warning in geom_point(position = position_dodge(0.3), width = 0.3): Ignoring
## unknown parameters: `width`

print(fig2c)

# summary: decreased happiness when future is positive in loss domain

Summary Plots Digging

Ellen Martin

2025-09-04

Risk-Taking Behaviour in the Gain Domain, Based on Future Information

Risk-Taking in the Gain Domain Based on Future Information and Anxiety Diagnosis

Risk-Taking Behaviour in Gain Domain Based on Future Information and Depression Diagnosis

Risk-Taking Behaviour in Gain Domain Based on Past Information and Anxiety Diagnosis

Happiness Plots

Happiness with Future (in Loss Domain)

Happiness based on Future Information and Past Experience in the Gain Domain

Happiness based on Future Information and Anxiety Diagnosis in the Loss Domain