---
title: "Health Algorithm Objectivity - Chapter 1"
author: "Matthew Leitao"
date: "`r Sys.Date()`"
output: html_document
---

<!--
Original analysis notebook for Study 1 (Chapter 1), as I ran it on the raw Qualtrics export.
Only two lines were changed for release: the local file paths on my computer were replaced
with "raw-data/..." placeholders. Everything else, including notes to myself and unfinished
exploratory chunks, is as written.

The raw .sav file is not released because it holds fields that could identify people.
To rerun the results from the public file, use study1-analysis.R and study1-ratings.csv instead.
-->

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```


```{r packages}
if(!require(haven)) {install.packages("haven")}
if(!require(tidyverse))  {install.packages("tidyverse")}
if(!require(gghalves))  {install.packages("gghalves")}
if(!require(psych))  {install.packages("psych")}
if(!require(effectsize))  {install.packages("effectsize")}
if(!require(lme4))  {install.packages("lme4")}
if(!require(MuMIn))  {install.packages("MuMIn")}
if(!require(sjPlot))  {install.packages("sjPlot")}
if(!require(sjmisc))  {install.packages("sjmisc")}
if(!require(EMAtools))  {install.packages("EMAtools")}
if(!require(papaja))  {install.packages("papaja")}
if(!require(lm.beta))  {install.packages("lm.beta")}
if(!require(viridis))  {install.packages("viridis")}
if(!require(ghibli))  {install.packages("ghibli")}
if(!require(report))  {install.packages("report")}
```

```{r}
Health_Algorithm_Objectivity_1 <- read_sav("raw-data/study1-raw.sav")  # raw file, not released; see the de-identified CSV
```


# Hypotheses

1.a.1 - People will perceive diagnostic medical decisions as requiring more quantitative reasoning than prescriptive medical decisions.

1.a.2 - People will perceive diagnostic medical decisions as requiring more quantitative reasoning than preventative medical decisions.

1.a.3 - People will perceive prescriptive medical decisions as requiring more quantitative reasoning than preventative medical decisions.

1.b.1 - People will perceive medical decisions about physical health as requiring more quantitative reasoning than decisions about mental health.

1.b.2 - People will perceive medical decisions about bodily injury as requiring more quantitative reasoning than decisions about acute sickness.

2.a.1 - People will perceive a doctor as being more capable of intuition-based decision-making than an algorithm

3.a. - People will judge doctors to be more trustworthy than medical artificial intelligence.


```{r}

describe(Health_Algorithm_Objectivity_1)
```

## Removing Preview
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1[-c(1,2,3,4,5,6,7), ]
```


```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  dplyr::rename(
    Int_Quant_Rec_Inj = Int_Quant_1,
    Int_Quant_Diag_Inj = Int_Quant_2,
    Int_Quant_Pres_Inj_Surgery = Int_Quant_3,
    Int_Quant_Pres_Inj_Med = Int_Quant_4,
    Int_Quant_Rec_Ment = Int_Quant_5,
    Int_Quant_Diag_Ment = Int_Quant_6,
    Int_Quant_Pres_Ment = Int_Quant_7,
    Int_Quant_Rec_AcuS = Int_Quant_8,
    Int_Quant_Diag_AcuS = Int_Quant_9,
    Int_Quant_Pres_AcuS = Int_Quant_10,
    Int_Quant_Rec_ChronS = Int_Quant_11,
    Int_Quant_Diag_ChronS = Int_Quant_12,
    Int_Quant_Pres_ChronS = Int_Quant_13,
    Int_Quant_Rec_Wellbeing = Int_Quant_14,
    Int_Quant_Med_Gen = Int_Quant_15,
    DM_QR_AI_Fitness = Q29_AI_Fitness
  )
```


## Creating Dummies

### Ethnicity Dummy Variable
```{r}
#Ethnicity Dummy Variables

Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    IsAsian = case_when(
      Ethnicity_1 == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsBlack = case_when(
      Ethnicity_2 == 1 ~ 1,
      TRUE ~ 0
    ),
    IsWhite = case_when(
      Ethnicity_3 == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsNativeAmerican = case_when(
      Ethnicity_5 == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsPacificIslander = case_when(
      Ethnicity_6 == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsOther = case_when(
      Ethnicity_8 == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsHispanic = case_when(
      Hispanic == 1 ~ 1,
      Hispanic == 0 ~ 0,
      TRUE ~ NA_real_
    )
  )

```


### Gender
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    IsMale = case_when(
      Gender == 1 ~ 1,
      TRUE ~ 0
    ),
    
    IsNonBinary = case_when(
      Gender == 4 ~ 1,
      TRUE ~ 0
    ),
    
    IsTrans = case_when(
      Transgender == 1 ~ 1,
      Transgender == 0 ~ 0,
      TRUE ~ NA_real_
    )
    
)
```


### Other Demographics
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    IsVisableMinority = case_when(
      Visible_Minority == 1 ~ 1,
      Visible_Minority == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    IsMinority = case_when(
      Minority == 1 ~ 1,
      Minority == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    EnglishFirst = case_when(
      First_Language == 1 ~ 1,
      First_Language == 2 ~ 2,
      TRUE ~ NA_real_
    ),
    
    NativeUSA = case_when(
      Native_USA == 1 ~ 1,
      Native_USA == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    ParentEd = case_when(
      Parent_College == 1 ~ 1,
      Parent_College == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    LearningDisability = case_when(
      Disablity_Learning == 1 ~ 1,
      Disablity_Learning == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    PhysicalDisability = case_when(
      Disablity_Physical == 1 ~ 1,
      Disablity_Physical == 0 ~ 0,
      TRUE ~ NA_real_
    ),
    
    MentalIllness = case_when(
      Disablity_Mental == 1 ~ 1,
      Disablity_Mental == 0 ~ 0,
      TRUE ~ NA_real_
    )
)
```


### Employment & Income
```{r}

Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    EmpFull = case_when(
      Employment == 1 ~ 1,
      TRUE ~ 0
    ),
    
    EmpPart = case_when(
      Employment == 2 ~ 1,
      TRUE ~ 0
    ),
    
    EmpSelf = case_when(
      Employment == 3 ~ 1,
      TRUE ~ 0
    ),
    
    EmpUn = case_when(
      Employment == 4 ~ 1,
      TRUE ~ 0
    ),
    
    EmpRet = case_when(
      Employment == 5 ~ 1,
      TRUE ~ 0
    ),
    
    EmpStud = case_when(
      Employment == 6 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_LessHigh = case_when(
      Education_Att == 1 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_LessHigh = case_when(
      Education_Att == 1 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_HighSchool = case_when(
      Education_Att == 2 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_LessColl = case_when(
      Education_Att == 3 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_College = case_when(
      Education_Att == 4 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_LessGrad = case_when(
      Education_Att == 5 ~ 1,
      TRUE ~ 0
    ),
    
    EduAtt_Graduate = case_when(
      Education_Att == 6 ~ 1,
      TRUE ~ 0
    )
  )
    

```


### Marriage Status
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    Single = case_when(
      Married == 1|2 ~ 1,
      TRUE ~ 0
    ))


Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    Single = case_when(
      Married == 1 ~ 1,
      TRUE ~ 0
    ),
    
    Marriage = case_when(
      Married == 2 ~ 1,
      TRUE ~ 0
    ),
    
    LivePartner = case_when(
      Married == 3 ~ 1,
      TRUE ~ 0
    ),
    
    Divorced = case_when(
      Married == 4 ~ 1,
      TRUE ~ 0
    )
  )

```





Demographics
```{r}
prop.table(table(Health_Algorithm_Objectivity_1_Clean$Gender))

describe(Health_Algorithm_Objectivity_1_Clean$Age)
```
```{r}
table(Health_Algorithm_Objectivity_1_Clean$Age)
```


# PreRegistered Analysis

Our primary analysis will be a series of paired samples t-tests between individual items and groups. 

1.	We will compare people’s perceptions of diagnostic decisions and prescriptive decisions.
2.	We will compare people’s perceptions of diagnostic decisions and preventative decisions.

3.	We will compare people’s perceptions of prescriptive decisions and preventative decisions.
4.	We will compare people’s perceptions of decisions about bodily injury and decisions about mental health.
5.	We will compare people’s perceptions of decisions about bodily injury and decisions about acute sickness. 
6.	We will compare people’s perceptions of the capability of doctors using intuition and medical artificial intelligence using intuition. 
7.	We will compare people’s perceptions of the capability of doctors using quantitative reasoning and medical artificial intelligence using quantitative reasoning.
8.	We will compare people’s reported trust in doctors and medical artificial intelligence.



## Correlation & Cronbachs Alpha Of Composites

### Diagonsis

Diagnostic Decisions:
1.	Diagnosing a Broken Bone
2.	Diagnosing Depression
3.	Diagnosing Strep Throat
4.	Diagnosing Diabetes

#### Correlation/Alpha
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_Ment, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_Ment, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS) %>%
  alpha()
```


### Prescriptive

Prescriptive Decisions:
1.	Prescribing Pain Medication for a Knee Injury
2.	Prescribing Anti-Depressants and Cognitive Behavioral Therapy for Depression
3.	Prescribing Anti-Biotics for Strep Throat
4.	Prescribing Insulin and a Change in Diet for Diabetes

```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Pres_Inj_Med, Int_Quant_Pres_Ment, Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Pres_Inj_Med, Int_Quant_Pres_Ment, Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS) %>%
  alpha()
```


### Preventative 

Preventative Decisions:
1.	Recommending Stretching to Prevent Injury
2.	Recommending Meditation and Exercise to Promote Mental Health
3.	Recommending Vitamin C + Zinc to Fight Off Infections
4.	Recommending a Strict Diet to Promote Better Physical Health

```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Rec_Inj, Int_Quant_Rec_Ment, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Rec_Inj, Int_Quant_Rec_Ment, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS) %>%
  alpha()
```

1.	Physical Health:
    a.	Bodily Injury:
             i.	Recommending Stretching to Prevent Injury
            ii.	Diagnosing a Broken Bone
           iii.	Prescribing Pain Medication for a Knee Injury
    b.	Acute Sickness:
             i.	Recommending Vitamin C + Zinc to Fight Off Infections
            ii.	Diagnosing Strep Throat
           iii.	Prescribing Anti-Biotics for Strep Throat
     c.	Chronic Sickness:
             i.	Recommending a Strict Diet to Promote Better Physical Health
            ii.	Diagnosing Diabetes
           iii.	Prescribing Insulin and a Change in Diet for Diabetes
2.	Mental Health:
     a.	Recommending Meditation and Exercise to Promote Mental Health
     b.	Diagnosing Depression
     c.	Prescribing Anti-Depressants and Cognitive Behavioral Therapy for Depression


### Physical Health
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS, Int_Quant_Rec_Inj, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS, Int_Quant_Pres_Inj_Med,  Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS, Int_Quant_Rec_Inj, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS, Int_Quant_Pres_Inj_Med,  Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS) %>%
  alpha()
```

### Mental Health
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Ment, Int_Quant_Rec_Ment, Int_Quant_Pres_Ment) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Ment, Int_Quant_Rec_Ment, Int_Quant_Pres_Ment) %>%
  alpha()
```


### Bodily Injury
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Rec_Inj, Int_Quant_Pres_Inj_Med) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Rec_Inj, Int_Quant_Pres_Inj_Med) %>%
  alpha()
```


### Acute Sickness
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_AcuS, Int_Quant_Rec_AcuS, Int_Quant_Pres_AcuS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_AcuS, Int_Quant_Rec_AcuS, Int_Quant_Pres_AcuS) %>%
  alpha()
```


### Chronic Sickness
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_ChronS, Int_Quant_Rec_ChronS, Int_Quant_Pres_ChronS) %>%
    corr.test()
```


```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_ChronS, Int_Quant_Rec_ChronS, Int_Quant_Pres_ChronS) %>%
  alpha()
```

### All
```{r}
Int_Quant_Cor_All <- Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS, Int_Quant_Rec_Inj, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS, Int_Quant_Pres_Inj_Med,  Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS,Int_Quant_Diag_Ment, Int_Quant_Rec_Ment, Int_Quant_Pres_Ment) %>% cor(use = "complete.obs")

Int_Quant_Cor_All
```

#### Exploritory Factor Analysis
```{r}
Int_Quant_EFA_Data <- Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS, Int_Quant_Rec_Inj, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS, Int_Quant_Pres_Inj_Med,  Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS,Int_Quant_Diag_Ment, Int_Quant_Rec_Ment, Int_Quant_Pres_Ment)
```


##### Scree Plot
```{r}

EFA_Fit <- fa(Int_Quant_EFA_Data,nfactors = ncol(Int_Quant_EFA_Data), rotate = "none")
n_factors <- length(EFA_Fit$e.values)

scree    <- data.frame(
  Factor_n =  as.factor(1:n_factors), 
  Eigenvalue = EFA_Fit$e.values)

ggplot(scree, aes(x = Factor_n, y = Eigenvalue, group = 1)) + 
  geom_point() + geom_line() +
  xlab("Number of factors") +
  ylab("Initial eigenvalue") +
  labs( title = "Scree Plot", 
        subtitle = "(Based on the unreduced correlation matrix)")
```
```{r}
scree
```


##### Analysis
```{r}
Int_Quant_FactorAnalysis <- fa(r = Int_Quant_Cor_All,nfactor = 3, rotate = "oblimin",  fm = "minchi")

Int_Quant_FactorAnalysis
```
##### Plot
```{r}
fa.diagram(Int_Quant_FactorAnalysis)
```

## Combine
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    Int_Quant_Diagnosis    = (Int_Quant_Diag_Inj + Int_Quant_Diag_Ment + Int_Quant_Diag_AcuS + Int_Quant_Diag_ChronS)/4,
    Int_Quant_Prescription  = (Int_Quant_Pres_Inj_Med + Int_Quant_Pres_Ment + Int_Quant_Pres_AcuS + Int_Quant_Pres_ChronS)/4,
    Int_Quant_Preventative  = (Int_Quant_Rec_Inj + Int_Quant_Rec_Ment + Int_Quant_Rec_AcuS + Int_Quant_Rec_ChronS)/4,
    Int_Quant_PhysHealth    = (Int_Quant_Diag_Inj + Int_Quant_Diag_AcuS + Int_Quant_Diag_ChronS + Int_Quant_Rec_Inj + Int_Quant_Rec_AcuS + Int_Quant_Rec_ChronS + Int_Quant_Pres_Inj_Med + Int_Quant_Pres_AcuS + Int_Quant_Pres_ChronS)/9,
    Int_Quant_MentHealth    = (Int_Quant_Diag_Ment + Int_Quant_Rec_Ment + Int_Quant_Pres_Ment)/3,
    Int_Quant_BodyInj       = (Int_Quant_Diag_Inj + Int_Quant_Rec_Inj + Int_Quant_Pres_Inj_Med)/3,
    Int_Quant_AcuteSic      = (Int_Quant_Diag_AcuS + Int_Quant_Rec_AcuS + Int_Quant_Pres_AcuS)/3,
    Int_Quant_ChronSic      = (Int_Quant_Diag_ChronS + Int_Quant_Rec_ChronS + Int_Quant_Pres_ChronS)/3
  )
    

```


## Descriptives Before Compositing
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(Int_Quant_Diag_Inj, Int_Quant_Diag_AcuS, Int_Quant_Diag_ChronS, Int_Quant_Rec_Inj, Int_Quant_Rec_AcuS, Int_Quant_Rec_ChronS, Int_Quant_Pres_Inj_Med, Int_Quant_Pres_AcuS, Int_Quant_Pres_ChronS, Int_Quant_Diag_Ment, Int_Quant_Rec_Ment, Int_Quant_Pres_Ment,Int_Quant_Pres_Inj_Surgery, Int_Quant_Rec_Wellbeing, Int_Quant_Med_Gen) %>% describe()
    
```


## Descriptives for Composites
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  select(
    Int_Quant_Diagnosis, Int_Quant_Prescription, Int_Quant_Preventative, Int_Quant_PhysHealth,
         Int_Quant_MentHealth, Int_Quant_BodyInj, Int_Quant_AcuteSic, Int_Quant_ChronSic, Int_Quant_Pres_Inj_Surgery, Int_Quant_Rec_Wellbeing, Int_Quant_Med_Gen
    ) %>%
  describe()
```

```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean%>%
  select(-Create_New_Field_or_Choose_From_Dropdown...)
```


# Saving Data Set
```{r}
write_sav(Health_Algorithm_Objectivity_1_Clean, "raw-data/study1-clean.sav")  # not released
```



To create the qualities of the decisions I have to dummy code categories to define the categories that I am evaluating.

# Analysis

### Hypothesis 1 - Not Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, paired = TRUE)
```
```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, paired = TRUE))
```




### Hypothesis 2 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE)
```


```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE)
```


```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Diagnosis, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE))
```


### Hypothesis 3 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE)
```


```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_Prescription, Health_Algorithm_Objectivity_1_Clean$Int_Quant_Preventative, paired = TRUE))
```



### Hypothesis 4 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_PhysHealth, Health_Algorithm_Objectivity_1_Clean$Int_Quant_MentHealth, paired = TRUE)
```

```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Int_Quant_PhysHealth, Health_Algorithm_Objectivity_1_Clean$Int_Quant_MentHealth, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_PhysHealth, Health_Algorithm_Objectivity_1_Clean$Int_Quant_MentHealth, paired = TRUE))
```


### Hypothesis 5 - Not Significant (Marginal)
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_BodyInj, Health_Algorithm_Objectivity_1_Clean$Int_Quant_AcuteSic, paired = TRUE)
```

```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Int_Quant_BodyInj, Health_Algorithm_Objectivity_1_Clean$Int_Quant_AcuteSic, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Int_Quant_BodyInj, Health_Algorithm_Objectivity_1_Clean$Int_Quant_AcuteSic, paired = TRUE))
```


## Decriptives For Decision Makers - Quantitative Reasoning
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
dplyr::select(
    DM_QR_Doctor_General, DM_QR_Nurse, DM_QR_Doctor_Remote, DM_QR_Therapist, DM_QR_PersonalTrainer, DM_QR_AI_General, DM_QR_AI_Diagonstic, DM_QR_AI_Chatbot, DM_QR_AI_Fitness, DM_QR_Doctor_AI, DM_QR_Lay
  ) %>%
  describe()

```

## Decriptives For Decision Makers - Intution
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
dplyr::select(
   DM_IN_Doctor_General, Decision_Maker_IN_Nurse, Decision_Maker_IN_Doctor_Remote, Decision_Maker_IN_Therapist, Decision_Maker_IN_PersonalTrainer, Decision_Maker_IN_AI_General, Decision_Maker_IN_AI_Diagonstic, Decision_Maker_IN_AI_Chatbot,Decision_Maker_IN_AI_Fitness,  Decision_Maker_IN_Doctor_AI, Decision_Maker_IN_Lay
  ) %>%
  describe()

```

## Decriptives For Decision Makers - Trust
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
dplyr::select(
   Decision_Maker_TR_Doctor_General, Decision_Maker_TR_Nurse, Decision_Maker_TR_Doctor_Remote, Decision_Maker_TR_Therapist, Decision_Maker_TR_PersonalTrainer, Decision_Maker_TR_AI_General, Decision_Maker_TR_AI_Diagonostic, Decision_Maker_TR_AI_Chatbot, Decision_Maker_TR_AI_Fitness, Decision_Maker_TR_Doctor_AI, Decision_Maker_TR_Lay
  ) %>%
  describe()

```

# Ranked Descriptives

## Decriptives For Decision Makers - Quantitative Reasoning
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
 dplyr::select(
    DM_QR_Doctor_General, DM_QR_Nurse, DM_QR_Doctor_Remote, DM_QR_Therapist, DM_QR_Doctor_AI, DM_QR_AI_General, DM_QR_PersonalTrainer, DM_QR_AI_Diagonstic, DM_QR_AI_Chatbot, DM_QR_AI_Fitness,  DM_QR_Lay
  ) %>%
  describe()

```

## Decriptives For Decision Makers - Intution
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  dplyr::select(
   DM_IN_Doctor_General, Decision_Maker_IN_Nurse, Decision_Maker_IN_Therapist, Decision_Maker_IN_Doctor_Remote,  Decision_Maker_IN_PersonalTrainer,  Decision_Maker_IN_Doctor_AI, Decision_Maker_IN_Lay, Decision_Maker_IN_AI_General, Decision_Maker_IN_AI_Diagonstic, Decision_Maker_IN_AI_Chatbot,Decision_Maker_IN_AI_Fitness
  ) %>%
  describe()

```

## Decriptives For Decision Makers - Trust
```{r}
Health_Algorithm_Objectivity_1_Clean %>%
  dplyr::select(
   Decision_Maker_TR_Doctor_General, Decision_Maker_TR_Nurse, Decision_Maker_TR_Doctor_Remote, Decision_Maker_TR_Therapist, Decision_Maker_TR_Doctor_AI, Decision_Maker_TR_PersonalTrainer, Decision_Maker_TR_AI_General, Decision_Maker_TR_AI_Diagonostic,  Decision_Maker_TR_Lay,  Decision_Maker_TR_AI_Fitness, Decision_Maker_TR_AI_Chatbot
  ) %>%
  describe()

```

### Hypothesis 6 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$DM_IN_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_AI_General, paired = TRUE)
```

```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$DM_IN_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_AI_General, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$DM_IN_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_AI_General, paired = TRUE))
```


### Hypothesis 6.2 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$DM_QR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$DM_QR_AI_General, paired = TRUE)
```


```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$DM_QR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$DM_QR_AI_General, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$DM_QR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$DM_QR_AI_General, paired = TRUE))
```



### Hypothesis 7 - Significant
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_AI_General, paired = TRUE)
```

```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_AI_General, paired = TRUE)
```

```{r}
report(t.test(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_AI_General, paired = TRUE))
```



#### Person As Effect Size
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    EffectDirection_TR_DocAiG  = case_when(
      Decision_Maker_TR_Doctor_General > Decision_Maker_TR_AI_General ~  1,
      Decision_Maker_TR_Doctor_General == Decision_Maker_TR_AI_General ~ 2,
      Decision_Maker_TR_Doctor_General < Decision_Maker_TR_AI_General ~ 3,
      TRUE ~ NA_real_
    )
  )
```


```{r}
Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_DocAiG <- recode_factor(Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_DocAiG, `1` = "Positive", `2` = "No Change", `3` = "Negative")

TRDocAiGCountTable <- table(Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_DocAiG)
TRDocAiGCountTable

# Get Percentage of Distribution
prop.table(TRDocAiGCountTable)
```





# Exploritory Analysis and Data Cleaning

## Creating Long Format Data

There are two types of long format data I want to analyse.

1. Qualities of the decision makers
2. Qualities of the decision.

### Reducing Data Frame
```{r}

Health_Algorithm_Objectivity_1_Tight_Variables <- c(

'ResponseId',


#Demographics
'Age', 'IsMale', 'IsNonBinary','IsTrans',
'IsAsian','IsBlack','IsHispanic','IsNativeAmerican','IsPacificIslander','IsOther','IsWhite','NativeUSA',
'IsMinority','IsVisableMinority',
'LearningDisability','MentalIllness','PhysicalDisability',
'Divorced','Single','LivePartner','Marriage',
'EduAtt_College','EduAtt_Graduate','EduAtt_HighSchool','EduAtt_LessColl','EduAtt_LessGrad','EduAtt_LessHigh','ParentEd',
'EmpFull','EmpPart','EmpRet','EmpSelf','EmpStud','EmpUn',
'EnglishFirst',
'Income',
#Politics
'Political_Econ','Political_Social','Religious',

# Medical Decisions
"Int_Quant_AcuteSic", "Int_Quant_BodyInj", "Int_Quant_ChronSic","Int_Quant_Diag_AcuS", "Int_Quant_Diag_ChronS", "Int_Quant_Diag_Inj", "Int_Quant_Diag_Ment","Int_Quant_Diagnosis","Int_Quant_Med_Gen","Int_Quant_MentHealth","Int_Quant_PhysHealth", "Int_Quant_Pres_AcuS","Int_Quant_Pres_ChronS", "Int_Quant_Pres_Inj_Med", "Int_Quant_Pres_Inj_Surgery", "Int_Quant_Pres_Ment", "Int_Quant_Prescription", "Int_Quant_Preventative","Int_Quant_Rec_AcuS","Int_Quant_Rec_ChronS","Int_Quant_Rec_Inj","Int_Quant_Rec_Ment", "Int_Quant_Rec_Wellbeing",

# Medical Decision Maker
#Intuition
"DM_IN_Doctor_General","Decision_Maker_IN_AI_Chatbot", "Decision_Maker_IN_AI_Diagonstic", "Decision_Maker_IN_AI_Fitness", "Decision_Maker_IN_AI_General","Decision_Maker_IN_Doctor_AI", "Decision_Maker_IN_Doctor_Remote","Decision_Maker_IN_Lay", "Decision_Maker_IN_Nurse", "Decision_Maker_IN_PersonalTrainer", "Decision_Maker_IN_Therapist", 


#Trust
"Decision_Maker_TR_Doctor_General","Decision_Maker_TR_AI_Chatbot","Decision_Maker_TR_AI_Diagonostic","Decision_Maker_TR_AI_Fitness","Decision_Maker_TR_AI_General", "Decision_Maker_TR_Doctor_AI",  "Decision_Maker_TR_Doctor_Remote", "Decision_Maker_TR_Lay", "Decision_Maker_TR_Nurse", "Decision_Maker_TR_PersonalTrainer", "Decision_Maker_TR_Therapist", 

# Quantitative Reasoning
"DM_QR_Doctor_General", "DM_QR_AI_Chatbot", "DM_QR_AI_Diagonstic", "DM_QR_AI_Fitness", "DM_QR_AI_General", "DM_QR_Doctor_AI",  "DM_QR_Doctor_Remote", "DM_QR_Lay", "DM_QR_Nurse", "DM_QR_PersonalTrainer", "DM_QR_Therapist", 


# Experience
"Exp_AI_ChatGTP", "Exp_AI_Fitness", "Exp_AI_General", "Exp_AI_Medical", "Exp_ME_Fitness", "Exp_ME_General", "Exp_ME_Hospital"

)

Health_Algorithm_Objectivity_1_Tight <- Health_Algorithm_Objectivity_1_Clean %>%
                dplyr::select(Health_Algorithm_Objectivity_1_Tight_Variables)



```

### Renaming
```{r}
Health_Algorithm_Objectivity_1_Tight <- Health_Algorithm_Objectivity_1_Tight %>%
  dplyr::rename(
    Int_Quant.AcuteSickness        = Int_Quant_AcuteSic,
    Int_Quant.BodilyInjury         = Int_Quant_BodyInj, 
    Int_Quant.ChronSickeness       = Int_Quant_ChronSic,
    Int_Quant.Diag_AcuteSickness   = Int_Quant_Diag_AcuS,
    Int_Quant.Diag_ChronSickness   = Int_Quant_Diag_ChronS,
    Int_Quant.Diag_Injury          = Int_Quant_Diag_Inj,
    Int_Quant.Diag_Mental          = Int_Quant_Diag_Ment,
    Int_Quant.Diagnosis            = Int_Quant_Diagnosis,
    Int_Quant.Med_General          = Int_Quant_Med_Gen,
    Int_Quant.MentHealth           = Int_Quant_MentHealth,
    Int_Quant.PhysHealth           = Int_Quant_PhysHealth,
    Int_Quant.Pres_AcuteSickness   = Int_Quant_Pres_AcuS,
    Int_Quant.Pres_ChronSickness   = Int_Quant_Pres_ChronS,
    Int_Quant.Pres_Injury_Medicine = Int_Quant_Pres_Inj_Med,
    Int_Quant.Pres_Inj_Surgery     = Int_Quant_Pres_Inj_Surgery,
    Int_Quant.Pres_Mental          = Int_Quant_Pres_Ment,
    Int_Quant.Prescription         = Int_Quant_Prescription,
    Int_Quant.Preventative         = Int_Quant_Preventative,
    Int_Quant.Rec_AcuteSickness    = Int_Quant_Rec_AcuS,
    Int_Quant.Rec_ChronSickness    = Int_Quant_Rec_ChronS,
    Int_Quant.Rec_Injury           = Int_Quant_Rec_Inj,
    Int_Quant.Rec_Mental           = Int_Quant_Rec_Ment,
    Int_Quant.Rec_Wellbeing        = Int_Quant_Rec_Wellbeing,
    
    
    DM_AI_Chatbot.Intuition       = Decision_Maker_IN_AI_Chatbot, 
    DM_AI_Diagonistic.Intuition   = Decision_Maker_IN_AI_Diagonstic,
    DM_AI_Fitness.Intuition       = Decision_Maker_IN_AI_Fitness,
    DM_AI_General.Intuition       = Decision_Maker_IN_AI_General,
    DM_Doctor_AI.Intuition        = Decision_Maker_IN_Doctor_AI,
    DM_Doctor_Remote.Intuition    = Decision_Maker_IN_Doctor_Remote,
    DM_Lay.Intuition              = Decision_Maker_IN_Lay,
    DM_Nurse.Intuition            = Decision_Maker_IN_Nurse,
    DM_Personal_Trainer.Intuition = Decision_Maker_IN_PersonalTrainer,
    DM_Therapist.Intuition        = Decision_Maker_IN_Therapist,
    DM_Doctor_General.Intuition   = DM_IN_Doctor_General,
    
    DM_AI_Chatbot.Trust       = Decision_Maker_TR_AI_Chatbot, 
    DM_AI_Diagonistic.Trust   = Decision_Maker_TR_AI_Diagonostic,
    DM_AI_Fitness.Trust       = Decision_Maker_TR_AI_Fitness,
    DM_AI_General.Trust       = Decision_Maker_TR_AI_General,
    DM_Doctor_AI.Trust        = Decision_Maker_TR_Doctor_AI,
    DM_Doctor_Remote.Trust    = Decision_Maker_TR_Doctor_Remote,
    DM_Lay.Trust              = Decision_Maker_TR_Lay,
    DM_Nurse.Trust            = Decision_Maker_TR_Nurse,
    DM_Personal_Trainer.Trust = Decision_Maker_TR_PersonalTrainer,
    DM_Therapist.Trust        = Decision_Maker_TR_Therapist,
    DM_Doctor_General.Trust   = Decision_Maker_TR_Doctor_General,
     
    DM_AI_Chatbot.Quantitative_Reasoning       = DM_QR_AI_Chatbot,
    DM_AI_Diagonistic.Quantitative_Reasoning   = DM_QR_AI_Diagonstic,
    DM_AI_Fitness.Quantitative_Reasoning       = DM_QR_AI_Fitness,
    DM_AI_General.Quantitative_Reasoning       = DM_QR_AI_General,
    DM_Doctor_AI.Quantitative_Reasoning        = DM_QR_Doctor_AI,
    DM_Doctor_Remote.Quantitative_Reasoning    = DM_QR_Doctor_Remote,
    DM_Lay.Quantitative_Reasoning              = DM_QR_Lay,
    DM_Nurse.Quantitative_Reasoning            = DM_QR_Nurse,
    DM_Personal_Trainer.Quantitative_Reasoning = DM_QR_PersonalTrainer,
    DM_Therapist.Quantitative_Reasoning        = DM_QR_Therapist,
    DM_Doctor_General.Quantitative_Reasoning   = DM_QR_Doctor_General
  )

Health_Algorithm_Objectivity_1_Tight <- dplyr::mutate(Health_Algorithm_Objectivity_1_Tight, ID = row_number())
```


### Pivoting Longer
#### By Decision
```{r}
Health_Algorithm_Objectivity_1_Longer_Decision <- Health_Algorithm_Objectivity_1_Tight %>%
  pivot_longer(
    cols = starts_with("Int_Quant"),
    names_to = c('Decision'),
    names_prefix = ("Int_Quant.")
  ) 

Health_Algorithm_Objectivity_1_Longer_Decision <- Health_Algorithm_Objectivity_1_Longer_Decision %>%
  dplyr::rename(
    Int_Quant = value
  )

```



#### By Decision-Maker
```{r}
Health_Algorithm_Objectivity_1_Longer <- Health_Algorithm_Objectivity_1_Tight %>%
  pivot_longer(
    cols = starts_with("DM"),
    names_to = c('Decision_Maker', '.value'), 
    names_sep = '\\.'
  )


```

```{r}
Health_Algorithm_Objectivity_1_Longer$Decision_Maker <- factor(Health_Algorithm_Objectivity_1_Longer$Decision_Maker, levels = c("DM_Doctor_General", "DM_Doctor_Remote", "DM_Lay", "DM_Nurse", "DM_Personal_Trainer", "DM_Therapist", "DM_AI_Chatbot", "DM_AI_Diagonistic", "DM_AI_Fitness", "DM_AI_General", "DM_Doctor_AI"))
```


```{r}
Health_Algorithm_Objectivity_1_Longer <- Health_Algorithm_Objectivity_1_Longer %>%
  mutate(
    IsAI = case_when(
    Decision_Maker == "DM_AI_Chatbot" | Decision_Maker == "DM_AI_Diagonistic" | Decision_Maker =="DM_AI_Fitness" | Decision_Maker == "DM_AI_General"| Decision_Maker == "DM_Doctor_AI" ~ 1,
    TRUE ~ 0
    )
    
  )

```

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision$Int_Quant.STD <- scale(Health_Algorithm_Objectivity_1_Longer_Decision$Int_Quant)
```


## Plots
Number Line Plots
I want to make plots so I can put the values of interest on a number line, kind of like grey 2007.
I should do this by taking the mean reported by each person and making a data frame out of them.

*Note - I need to change the names and perhaps find some way to highlight the compiled variables rather than making it look like the other variables. Also reconsider the colors.

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision$Decision <- 
  recode_factor(Health_Algorithm_Objectivity_1_Longer_Decision$Decision, 
         AcuteSickness        = "Acute Sickness Related",
         BodilyInjury         = "Bodily Injury Related",
         ChronSickeness       = "Chronic Sickness Related",
         Diagnosis            = "Diagnostic",
         Med_General          = "Medical General",
         MentHealth           = "Mental Health Related",
         PhysHealth           = "Physical Health Related",
         Prescription         = "Prescriptive",
         Preventative         = "Preventative",
         Rec_Mental           = 'Recommend Meditation',
         Rec_Injury           = 'Recommend Stretching',
         Rec_ChronS           = 'Recommend Strict Diet',
         Rec_AcuteSickness    = 'Recommend Vitimin C + Zinc',
         Pres_Inj_Surgery     = 'Prescribe Knee Surgery',
         Pres_Mental          = 'Prescribe Anti-Depressants + Therapy',
         Pres_Injury_Medicine = 'Prescribe Pain Medication',
         Rec_ChronSickness    = 'Recommend Strict Diet',
         Pres_ChronSickness   = 'Prescribe Insulin',
         Pres_AcuteSickness   = 'Prescribe Anti-Biotics',
         Diag_Mental          = 'Diagnose Depression',
         Diag_Injury          = 'Diagnose A Broken Bone',
         Diag_ChronSickness   = 'Diagnose Diabetes',
         Diag_AcuteSickness   = 'Diagnose Strep Throat',
         Rec_Wellbeing        = 'Recommend General Wellbeing Improvements'
         )

```

```{r}
unique(Health_Algorithm_Objectivity_1_Longer_Decision$Decision)
```
legend.background = element_blank(), legend.box.background = element_rect("transparent"), legend.text = element_text( size = 3),




```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  select(Decision) %>%
  group_by(Decision) %>% 
  summarise(n=n())
```



```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  select(Decision) %>%
  group_by(Decision) %>%
  filter(Decision == Decisions_Full) %>% summarise(n=n())
```


```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
  ggplot(aes(Int_Quant, Decision, color = Decision)) + geom_jitter() + stat_summary(fun.y=mean, geom="point", shape=18, size=6, color="black") + stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 1, size = 1.25, color = "black") + ggtitle("Medical Decision by Intuition/Quantiative Reasoning") + xlab("Intuition (0) versus Quantative Reasoning(100)") + ylab("Medical Decision") + theme_apa() +
  theme(panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "black"), axis.line = element_line(colour = "black"), axis.text = element_text(colour = "black", size = 12))


#ggsave("Medical Decisions Matrix Empty.png", width = 15, height = 11)
```

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
    filter(Decision == "Acute Sickness Related" | Decision ==  "Chronic Sickness Related" | Decision == "Bodily Injury Related" | Decision ==  "Mental Health Related" | Decision ==  "Physical Health Related") %>%
  ggplot(aes(Decision, Int_Quant,  color = Decision, fill = Decision)) + 
      geom_flat_violin(position = position_nudge(x = .2, y = 0), alpha = .6, adjust = 4) +
    geom_point(position=position_jitter(width = .15),size = 1, alpha = .90) +
       scale_fill_manual(values = c("#F18C8E", "#F0B7A4", "#F1D1B6", "#568EA6", "#305F72")) +
          scale_colour_manual(values = c("#F18C8E", "#F0B7A4", "#F1D1B6", "#568EA6", "#305F72")) +
  stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 0.35, size = 1, color = "black") +
 stat_summary(fun.y=mean, geom="point", shape=23, stroke = 1, size = 3, color="black", fill = "#FCBB6D") +  ggtitle("Domain of Medical Decision by Intuition/Quantiative Reasoning") + xlab("Medical Decision") + ylab("Intuition (0) versus Quantative Reasoning(100)") + theme_apa() +
  theme(panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "black"), axis.line = element_line(colour = "black"), axis.text = element_text(colour = "black", size = 12))


 ggsave("Medical Decisions Matrix Domain.png", width = 12, height = 9)
```

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
    filter(Decision == "Acute Sickness Related" | Decision ==  "Chronic Sickness Related" | Decision == "Bodily Injury Related" | Decision ==  "Mental Health Related" | Decision ==  "Physical Health Related") %>%
  ggplot(aes(Decision, Int_Quant,  color = Decision, fill = Decision)) + 
      geom_flat_violin(position = position_nudge(x = .2, y = 0), alpha = .6, adjust = 4) +
    geom_point(position=position_jitter(width = .15),size = 2.5, alpha = 1/2, color = "black") +
       scale_fill_manual(values = c("#F18C8E", "#F0B7A4", "#F1D1B6", "#568EA6", "#305F72")) +
          scale_colour_manual(values = c("#F18C8E", "#F0B7A4", "#F1D1B6", "#568EA6", "#305F72")) +
  stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 0.35, size = 1.5, stroke = 1.5, color = "white") +
 stat_summary(fun.y=mean, geom="point", shape=23, stroke = 1.5, size = 3, color="white") +  ggtitle("Domain of Medical Decision by Intuition/Quantiative Reasoning") + xlab("Medical Decision") + ylab("Intuition (0) versus Quantative Reasoning(100)") + theme_apa() + theme(plot.title = element_text(face = "bold", size = 16), axis.title = element_text(face = "bold", size = 14), panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "white"), axis.line = element_line(colour = "white"), axis.text = element_text(colour = "white", size = 12, face = "bold"))

 ggsave("Medical Decisions Matrix Domain Presentation.png", width = 12, height = 9)

```



```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  filter(Decision == "Diagnostic" | Decision == "Prescriptive" | Decision == "Preventative") %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
    ggplot(aes(Decision, Int_Quant,  color = Decision, fill = Decision)) + 
      geom_flat_violin(position = position_nudge(x = .2, y = 0), alpha = .6, adjust = 4, weight = 2) +
    geom_point(position=position_jitter(width = .15),size = 2.5, alpha = .90) +
       scale_fill_manual(values = c("#A2D4AB", "#3EACA8", "#547A82")) +
          scale_colour_manual(values = c("#A2D4AB", "#3EACA8", "#547A82"))+
  stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 0.35, size = 1, color = "black") + 
  stat_summary(fun.y=mean, geom="point", shape=23, stroke = 1, size = 3, color="black", fill = "#FCBB6D") + 
  ggtitle("Stage of Medical Decision by Intuition/Quantiative Reasoning") + xlab("Medical Decision" ) + ylab("Intuition (0) versus Quantative Reasoning(100)") + theme_apa() +
  theme(panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "black"), axis.line = element_line(colour = "black"), axis.text = element_text(colour = "black", size = 12))
 ggsave("Medical Decisions Matrix Stage.png", width = 9, height = 8)
```

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  filter(Decision == "Diagnostic" | Decision == "Prescriptive" | Decision == "Preventative") %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
    ggplot(aes(Decision, Int_Quant,  color = Decision, fill = Decision)) + 
      geom_flat_violin(position = position_nudge(x = .2, y = 0), alpha = .6, adjust = 4) +
    geom_point(position=position_jitter(width = .15),size = 2.5, alpha = 1/2, color = "black") +
       scale_fill_manual(values = c("#A2D4AB", "#3EACA8", "#547A82")) +
          scale_colour_manual(values = c("#A2D4AB", "#3EACA8", "#547A82"))+
stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 0.35, size = 1, stroke = 1, color = "white") +
 stat_summary(fun.y=mean, geom="point", shape=23, stroke = 1, size = 3, color="white") + 
ggtitle("Stage of Medical Decision by Intuition/Quantiative Reasoning") + xlab("Medical Decision") + ylab("Intuition (0) versus Quantative Reasoning(100)") +    theme_apa() + theme(plot.title = element_text(face = "bold", size = 16), axis.title = element_text(face = "bold", size = 14), panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "white"), axis.line = element_line(colour = "white"), axis.text = element_text(colour = "white", size = 12, face = "bold"))
  
 ggsave("Medical Decisions Matrix Stage Presentation.png", width = 9, height = 8)
```

```{r}
Health_Algorithm_Objectivity_1_Longer_Decision %>%
  filter(Decision == "Diagnose Strep Throat" | Decision == "Diagnose Diabetes" | Decision ==  "Diagnose A Broken Bone" | Decision == "Diagnose Depression" | Decision ==  "Medical General" | Decision == "Prescribe Anti-Biotics" | Decision ==  "Prescribe Insulin" | Decision ==  "Prescribe Pain Medication" | Decision ==  "Prescribe Knee Surgery" | Decision == "Prescribe Anti-Depressants + Therapy" | Decision ==  "Recommend Vitimin C + Zinc"| Decision ==  "Recommend Strict Diet" | Decision ==  "Recommend Stretching" | Decision == "Recommend Meditation" | Decision ==  "Recommend General Wellbeing Improvements") %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
  ggplot(aes(Int_Quant, Decision, color = Decision)) + geom_jitter(aes(size = .5)) +  stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 1, size = 2.5, color = "white") +
  scale_color_viridis(discrete = TRUE) +
  stat_summary(fun.y=mean, geom="point", shape=23, stroke = 2, size = 4, color="white", fill = "#FCBB6D") + ggtitle("Medical Judgment by Intuition/Quantiative Reasoning") + xlab("Intuition (0) versus Quantative Reasoning(100)") + ylab("Medical Decision") + theme_apa() + theme(plot.title = element_text(face = "bold", size = 16), axis.title = element_text(face = "bold", size = 14), panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position="none" , axis.ticks = element_line(colour = "white"), axis.line = element_line(colour = "white"), axis.text = element_text(colour = "white", size = 14, face = "bold"))


ggsave("Medical Decisions Matrix Empty.png", width = 15, height = 10)
```


```{r}
Decision_Mean_Table <- Health_Algorithm_Objectivity_1_Longer_Decision %>%
  mutate(Decision = fct_reorder(Decision, Int_Quant, .fun = "mean")) %>%
  group_by(Decision) %>%
  summarise(Mean = mean(Int_Quant, na.rm = TRUE))
            
write_csv(Decision_Mean_Table, "Decision Mean Table.csv")
```


## Algorithm Versus Lay Person Trust
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Lay, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_AI_General, paired = TRUE)
```

```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Lay, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_AI_General, paired = TRUE)
```

```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_Lay, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_AI_General, paired = TRUE)
```
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$DM_QR_AI_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_IN_AI_General, paired = TRUE)
```



#### Person As Effect Size
```{r}
Health_Algorithm_Objectivity_1_Clean <- Health_Algorithm_Objectivity_1_Clean %>%
  mutate(
    EffectDirection_TR_LayAlgG  = case_when(
      Decision_Maker_TR_Lay > Decision_Maker_TR_AI_General ~  1,
      Decision_Maker_TR_Lay == Decision_Maker_TR_AI_General ~ 2,
      Decision_Maker_TR_Lay < Decision_Maker_TR_AI_General ~ 3,
      TRUE ~ NA_real_
    )
  )
```


```{r}
Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_LayAlgG <- recode_factor(Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_LayAlgG, `1` = "Positive", `2` = "No Change", `3` = "Negative")

TRLayAlgGCountTable <- table(Health_Algorithm_Objectivity_1_Clean$EffectDirection_TR_LayAlgG)
TRLayAlgGCountTable

# Get Percentage of Distribution
prop.table(TRLayAlgGCountTable)
```



## Doctor versus Doctor AI
```{r}
t.test(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_AI, paired = TRUE)
```


```{r}
cohens_d(Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_General, Health_Algorithm_Objectivity_1_Clean$Decision_Maker_TR_Doctor_AI, paired = TRUE)
```


```{r}
summary(lm(Decision_Maker_TR_Doctor_General ~ DM_IN_Doctor_General + DM_QR_Doctor_General , data = Health_Algorithm_Objectivity_1_Clean))

lm.beta(lm(Decision_Maker_TR_Doctor_General ~ DM_IN_Doctor_General + DM_QR_Doctor_General , data = Health_Algorithm_Objectivity_1_Clean))


```

```{r}
summary(lm(Decision_Maker_TR_AI_General ~ Decision_Maker_IN_AI_General + DM_QR_AI_General , data = Health_Algorithm_Objectivity_1_Clean))

lm.beta(lm(Decision_Maker_TR_AI_General ~ Decision_Maker_IN_AI_General + DM_QR_AI_General , data = Health_Algorithm_Objectivity_1_Clean))


```
```{r}
summary(lm(Decision_Maker_TR_Lay ~ Decision_Maker_IN_Lay + DM_QR_Lay, data = Health_Algorithm_Objectivity_1_Clean))

lm.beta(lm(Decision_Maker_TR_Lay ~ Decision_Maker_IN_Lay + DM_QR_Lay, data = Health_Algorithm_Objectivity_1_Clean))


```


### Correlations

#### Doctor
```{r}
Health_Algorithm_Objectivity_1_Longer %>%
  filter(Decision_Maker == "DM_Doctor_General") %>%
  dplyr::select(Trust, Quantitative_Reasoning, Intuition) %>%
  corr.test()
```
#### AI General
```{r}
Health_Algorithm_Objectivity_1_Longer %>%
  filter(Decision_Maker == "DM_AI_General") %>%
  dplyr::select(Trust, Quantitative_Reasoning, Intuition) %>%
  corr.test()
```

#### Lay
```{r}
Health_Algorithm_Objectivity_1_Longer %>%
  filter(Decision_Maker == "DM_Lay") %>%
  dplyr::select(Trust, Quantitative_Reasoning, Intuition) %>%
  corr.test()
```





## Means By Group
```{r}
Health_Algorithm_Objectivity_1_Longer$Intuition <- Health_Algorithm_Objectivity_1_Longer$Intuition -1
Health_Algorithm_Objectivity_1_Longer$Quantitative_Reasoning <- Health_Algorithm_Objectivity_1_Longer$Quantitative_Reasoning -1

Health_Algorithm_Decision_Maker_Mean_Table <- Health_Algorithm_Objectivity_1_Longer %>%
  group_by(Decision_Maker) %>%
 summarise(Trust = mean(Trust, na.rm = TRUE),  Quantiative_Reasoning = mean(Quantitative_Reasoning, na.rm = TRUE),  Intuition_M = mean(Intuition, na.rm = TRUE), QR_CI = (sd(Quantitative_Reasoning, na.rm = TRUE)/sqrt(n()))*1.96, IN_CI = (sd(Intuition, na.rm = TRUE)/sqrt(n()))*1.96)


Health_Algorithm_Decision_Maker_Mean_Table 
```

####dddd
*Note - I need to visualize this better, if I can put error bars around each side of the circle that would be ideal. Maybe I can use symbols instead of putting the words. I will also have to change the names.

```{r}
Health_Algorithm_Decision_Maker_Mean_Table$Decision_Maker <- as.factor(Health_Algorithm_Decision_Maker_Mean_Table$Decision_Maker)

Health_Algorithm_Decision_Maker_Mean_Table$Decision_Maker  <- 
  recode_factor(Health_Algorithm_Decision_Maker_Mean_Table$Decision_Maker , 
         DM_Doctor_General = 'General Doctor',
         DM_Nurse          = 'Nurse',
         DM_Doctor_Remote  = 'Remote Doctor',
         DM_Therapist    = 'Therapist',
         DM_Doctor_AI   = 'Doctor w/ AI',
         DM_Personal_Trainer    = 'Personal Trainer',
         DM_AI_General = 'General AI',
         DM_AI_Diagonistic   = 'Diagnostic AI',
         DM_Lay   = 'Lay Individual',
         DM_AI_Fitness    = 'Fitness Related AI',
         DM_AI_Chatbot = 'AI Therapy Chatbot'
         )


```


, colour = "white"
```{r}
ggplot(Health_Algorithm_Decision_Maker_Mean_Table, aes(Intuition_M, Quantiative_Reasoning, label = Decision_Maker , fill = Decision_Maker, colour = Decision_Maker)) + 
  geom_point(size = 4, stroke = 2, aes(shape = Decision_Maker)) +
  scale_shape_manual(values=c(12,2,13,4,5,6,7,8,9,10,11)) +
  geom_errorbar(aes(ymin=Quantiative_Reasoning-QR_CI, ymax=Quantiative_Reasoning+QR_CI), size = 1.5) +
  geom_errorbar(aes(xmin=Intuition_M-IN_CI, xmax=Intuition_M+IN_CI), size = 1.5) + 
  coord_cartesian(xlim = c(3.5, 7.5), ylim = c(3.5,7.5)) + ggtitle("Medical Decision Maker by Intuition and Quantiative Reasoning") + xlab("Intuition") + ylab("Quantitative Reasoning") + 
  scale_color_viridis(discrete = TRUE) +
  theme_apa()  +  theme(plot.title = element_text(face = "bold", size = 16), axis.title = element_text(face = "bold", size = 14), panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.background = element_blank(), legend.box.background = element_rect("transparent", colour = NA), legend.title = element_text(colour = "black", face = "bold"), legend.text = element_text(colour = "white"),  axis.text = element_text(colour = "white", face = "bold"), axis.ticks = element_line(colour = "white"), axis.line = element_line(colour = "white")) 

 ggsave("Decision Maker Matrix Empty White Presentation.png", width = 8.5, height = 7)
```
#### CCCC
*Note have to rename the vales and change the intervals to a much more understandable values. If possible add statements below the x axis to show how people answered.
```{r}
Health_Algorithm_Objectivity_1_Longer$Decision_Maker <- as.factor(Health_Algorithm_Objectivity_1_Longer$Decision_Maker)

Health_Algorithm_Objectivity_1_Longer$Decision_Maker  <- 
  recode_factor(Health_Algorithm_Objectivity_1_Longer$Decision_Maker , 
         DM_Doctor_General = 'General Doctor',
         DM_Nurse          = 'Nurse',
         DM_Doctor_Remote  = 'Remote Doctor',
         DM_Therapist    = 'Therapist',
         DM_Doctor_AI   = 'Doctor w/ AI',
         DM_Personal_Trainer    = 'Personal Trainer',
         DM_AI_General = 'General AI',
         DM_AI_Diagonistic   = 'Diagnostic AI',
         DM_Lay   = 'Lay Individual',
         DM_AI_Fitness    = 'Fitness Related AI',
         DM_AI_Chatbot = 'AI Therapy Chatbot'
         )

Health_Algorithm_Objectivity_1_Longer$Trust <- Health_Algorithm_Objectivity_1_Longer$Trust - 1
```

```{r}

Health_Algorithm_Objectivity_1_Longer %>%
  mutate(Decision_Maker = fct_reorder(Decision_Maker, Trust, .fun = "mean")) %>%
  ggplot(aes(Trust, Decision_Maker, color = Decision_Maker)) + geom_jitter() +  stat_summary(geom = "errorbar", fun.data = mean_cl_boot, width = 1, size = 1.25, color = "black") + stat_summary(fun.y=mean, geom="point", shape=23, stroke = 1, size = 3, color="black", fill = "#FCBB6D") +
    scale_color_viridis(discrete = TRUE) +
  ggtitle("Trust by Medical Decision-Maker") + xlab("Trust") + ylab("Medical Decision-Maker") +  theme_apa()   + theme(plot.title = element_text(face = "bold", size = 16), axis.title = element_text(face = "bold", size = 14), panel.background = element_rect("transparent"), plot.background = element_rect("transparent", color = NA), legend.position = "none",  axis.text = element_text(colour = "white", face = "bold"), axis.ticks = element_line(colour = "white"), axis.line = element_line(colour = "white")) 


ggsave("Medical Decision Maker Trust Matrix Empty White Presentation.png", width = 8, height = 6)
```

## Linear Mixed Effect Model - Decision Maker

### Creating Person-Mean Centered and Person Mean Variables
```{r}
Health_Algorithm_Objectivity_1_Longer <- Health_Algorithm_Objectivity_1_Longer %>%
  mutate(
    Quantitative_Reasoning_PM = pmean(ResponseId, Quantitative_Reasoning),
    Quantitative_Reasoning_PMC = pcenter(ResponseId, Quantitative_Reasoning),
    Intuition_PM = pmean(ResponseId, Intuition),
    Intuition_PMC = pcenter(ResponseId, Intuition), 
    
  )

Health_Algorithm_Objectivity_1_Longer %>%
  select(ResponseId, Quantitative_Reasoning_PM, Quantitative_Reasoning_PMC, Intuition_PM, Intuition_PMC) %>%
  describe()


```



8### Base
```{r}
HealthAlgorithm_LMER_1 <- lmer(Trust ~ Quantitative_Reasoning + Intuition + IsAI + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_1)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_1)
```

```{r}
tab_model(HealthAlgorithm_LMER_1, show.std = TRUE)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_1)
```

#### Decomposed Model
```{r}
HealthAlgorithm_LMER_1D <- lmer(Trust ~ Quantitative_Reasoning_PM + Quantitative_Reasoning_PMC + Intuition_PM + Intuition_PMC + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_1D)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_1D)
```

```{r}
tab_model(HealthAlgorithm_LMER_1D, show.std = TRUE)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_1D)
```

### Base Difference Between AI and Humans
```{r}
HealthAlgorithm_LMER_IsAIOnly <- lmer(Trust ~ IsAI + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_IsAIOnly)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_IsAIOnly)
```

```{r}
tab_model(HealthAlgorithm_LMER_IsAIOnly, show.std = TRUE)
```


### Cov
```{r}
HealthAlgorithm_LMER_2 <- lmer(Trust ~ Quantitative_Reasoning + Intuition + IsAI + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_2)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_2)
```

```{r}
tab_model(HealthAlgorithm_LMER_2, show.std = TRUE)
```

```{r}
anova(HealthAlgorithm_LMER_1, HealthAlgorithm_LMER_2)
```

#### Decomposed Model
```{r}
HealthAlgorithm_LMER_2D <- lmer(Trust ~ Quantitative_Reasoning_PM + Quantitative_Reasoning_PMC + Intuition_PM + Intuition_PMC + IsAI + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_2D)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_2D)
```

```{r}
tab_model(HealthAlgorithm_LMER_2D, show.std = TRUE)
```

```{r}
report(HealthAlgorithm_LMER_2D)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_2D)
```

```{r}
anova(HealthAlgorithm_LMER_1D, HealthAlgorithm_LMER_2D)
```

### Interaction
```{r}
HealthAlgorithm_LMER_3 <- lmer(Trust ~ Quantitative_Reasoning*IsAI + Intuition*IsAI + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_3)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_3)
```

```{r}
anova(HealthAlgorithm_LMER_2, HealthAlgorithm_LMER_3)
```


```{r}
tab_model(HealthAlgorithm_LMER_3, show.std = TRUE)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_3)
```

#### Decomposed Model
```{r}
HealthAlgorithm_LMER_3D <- lmer(Trust ~ Quantitative_Reasoning_GM* IsAI + Quantitative_Reasoning_PMC* IsAI + Intuition_GM * IsAI + Intuition_PMC * IsAI + (1+  Intuition_PMC + Quantitative_Reasoning_PMC|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_3D)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_3D)
```

```{r}
tab_model(HealthAlgorithm_LMER_3D, show.std = TRUE)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_3D)
```


```{r}
anova(HealthAlgorithm_LMER_1D, HealthAlgorithm_LMER_2D, HealthAlgorithm_LMER_3D)
```

### Individual Models/Moderation

#### AI
```{r}
HealthAlgorithm_LMER_AI <- lmer(Trust ~ Quantitative_Reasoning + Intuition + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer, subset = IsAI == 1)

summary(HealthAlgorithm_LMER_AI)

```


```{r}
r.squaredGLMM(HealthAlgorithm_LMER_AI)
```



#### Human
```{r}
HealthAlgorithm_LMER_Hum <- lmer(Trust ~ Quantitative_Reasoning + Intuition + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer, subset = IsAI == 0)

summary(HealthAlgorithm_LMER_Hum)

```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_Hum)
```

## Combo Plot
```{r}
ggplot(data = Health_Algorithm_Objectivity_1_Longer,

       aes(x = Trust,

           y = Intuition ,

           col = as.factor(IsAI))) +

  geom_smooth(method = lm, se = F, fullrange = F, lty = 1, size = 2) +
  #Base
  ## Quantiative Reasoning
  #geom_abline(slope = 0.3724, intercept = 	1.225, colour = "black", size = 3, linetype = 4) +
  
  ## Intuition
  #geom_abline(slope = 0.4371, intercept = 	1.225, colour = "grey", size = 3, linetype = 4) +
  
  # AI
  ## Quantiative Reasoning
  geom_abline(slope = 0.27275, intercept = 	1.65439, colour = "red4", size = 3, linetype = 4) +
  
  ## Intuition
  geom_abline(slope = 0.42999, intercept = 	1.65439, colour = "orange", size = 3, linetype = 4) +
  
  # Hum
  ## Quantitative Reasoning
  geom_abline(slope = 0.4452, intercept = 	1.514, colour = "blue4", size = 3, linetype = 4) +
  
  ## Intuition
  geom_abline(slope = 0.3531, intercept = 	1.514, colour = "green", size = 3, linetype = 4) +
  
  xlab("Quantiative Reasoning/Intution") + ylab("Trust") +

  coord_cartesian(xlim = c(0,10),  ylim = c(0,10)) +
  
  scale_x_continuous(breaks=seq(0,10,by=1)) +

  scale_y_continuous(breaks=seq(0,10,by=1)) +
  
  theme_apa() +

  theme(
    panel.background = element_rect(fill='transparent'), #transparent panel bg
    plot.background = element_rect(fill='transparent', color=NA), #transparent plot bg
    panel.grid.major = element_blank(), #remove major gridlines
    panel.grid.minor = element_blank(), #remove minor gridlines
    legend.background = element_rect(fill='transparent'), #transparent legend bg
    legend.box.background = element_rect(fill='transparent') #transparent legend panel
  ) +
  
  theme(text = element_text(family = "serif")) +

  scale_color_manual(values = c("000000","00000"),

                     labels = c("AI", "Human"),

                     name = "Pandemic Happiness")


```

```{r}
ggplot(data = Health_Algorithm_Objectivity_1_Longer,

       aes(x = Trust,

           y = Intuition ,

           col = as.factor(IsAI))) +

  geom_smooth(method = lm, se = F, fullrange = F, lty = 1, size = 2) +
  # AI
  ## Quantiative Reasoning
  geom_abline(slope = 0.27275, intercept = 	1.65439, colour = "gray", size = 3) +
  
  ## Intuition
  geom_abline(slope = 0.42999, intercept = 	1.65439, colour = "blue", size = 3) +
  
  xlab("Quantiative Reasoning/Intution") + ylab("Trust") +

  coord_cartesian(xlim = c(0,10),  ylim = c(0,10)) +
  
  scale_x_continuous(breaks=seq(0,10,by=1)) +

  scale_y_continuous(breaks=seq(0,10,by=1)) +
  
  theme_apa() +

  theme(
    panel.background = element_rect(fill='transparent'), #transparent panel bg
    plot.background = element_rect(fill='transparent', color=NA), #transparent plot bg
    panel.grid.major = element_blank(), #remove major gridlines
    panel.grid.minor = element_blank(), #remove minor gridlines
    legend.background = element_rect(fill='transparent'), #transparent legend bg
    legend.box.background = element_rect(fill='transparent') #transparent legend panel
  ) +
  
  theme(text = element_text(family = "serif")) +

  scale_color_manual(values = c("000000","00000"),

                     labels = c("Quantitative Reasoning - Gray", "Intuition - Blue"),

                     name = "Artificial Intellgence")


```


```{r message=TRUE, warning=FALSE}
ggplot(data = Health_Algorithm_Objectivity_1_Longer,

       aes(x = Trust,

           y = Intuition ,

           col = as.factor(IsAI))) +

  geom_smooth(method = lm, se = F, fullrange = F, lty = 1, size = 2) +

  # Hum
  ## Quantitative Reasoning
  geom_abline(slope = 0.4452, intercept = 	1.514, colour = "gray", size = 3) +
  
  ## Intuition
  geom_abline(slope = 0.3531, intercept = 	1.514, colour = "blue", size = 3) +
  
  xlab("Quantiative Reasoning/Intution") + ylab("Trust") +

  coord_cartesian(xlim = c(0,10),  ylim = c(0,10)) +
  
  scale_x_continuous(breaks=seq(0,10,by=1)) +

  scale_y_continuous(breaks=seq(0,10,by=1)) +
  
  theme_apa() +

  theme(
    panel.background = element_rect(fill='transparent'), #transparent panel bg
    plot.background = element_rect(fill='transparent', color=NA), #transparent plot bg
    panel.grid.major = element_blank(), #remove major gridlines
    panel.grid.minor = element_blank(), #remove minor gridlines
    legend.background = element_rect(fill='transparent'), #transparent legend bg
    legend.box.background = element_rect(fill='transparent') #transparent legend panel
  ) +
  
  theme(text = element_text(family = "serif")) +

  scale_color_manual(values = c("000000","00000"),

                     labels = c("Quantitative Reasoning - Gray", "Intuition - Blue"),

                     name = "Human Expert")


```

## Within/Between Person Correlations
```{r}

Health_Algorithm_Objectivity_1_Longer$Intuition <- as.numeric(Health_Algorithm_Objectivity_1_Longer$Intuition)

Health_Algorithm_Objectivity_1_Longer$Quantitative_Reasoning <- as.numeric(Health_Algorithm_Objectivity_1_Longer$Quantitative_Reasoning)

Health_Algorithm_Objectivity_1_Longer$Trust <- as.numeric(Health_Algorithm_Objectivity_1_Longer$Trust)


W_B_Corr <- Health_Algorithm_Objectivity_1_Longer %>%
 dplyr::select(ID, Trust, Intuition, Quantitative_Reasoning, IsMale) %>%
  statsBy(group = c("ID"))


W_B_Corr$rbg
W_B_Corr$pbg


W_B_Corr$rwg
W_B_Corr$pwg
```

#### Is AI
```{r}

W_B_Corr_AI <- Health_Algorithm_Objectivity_1_Longer %>%
  filter(IsAI == 1) %>%
 dplyr::select(ID, Trust, Intuition, Quantitative_Reasoning) %>%
  statsBy(group = c("ID"))

W_B_Corr_AI$rbg
W_B_Corr_AI$pbg

W_B_Corr_AI$rwg
W_B_Corr_AI$pwg

```

#### Is Hum
```{r}

W_B_Corr_Hum <- Health_Algorithm_Objectivity_1_Longer %>%
  filter(IsAI == 0) %>%
 dplyr::select(ID, Trust, Intuition, Quantitative_Reasoning) %>%
  statsBy(group = c("ID"))

W_B_Corr_Hum$rbg
W_B_Corr_Hum$pbg

W_B_Corr_Hum$rwg
W_B_Corr_Hum$pwg

```



```{r}
HealthAlgorithm_LMER_1 <- lmer(Trust ~ Quantitative_Reasoning + Intuition + (1|ResponseId), data = Health_Algorithm_Objectivity_1_Longer)

summary(HealthAlgorithm_LMER_1)
```

```{r}
r.squaredGLMM(HealthAlgorithm_LMER_1)
```

```{r}
tab_model(HealthAlgorithm_LMER_1, show.std = TRUE)
```


```{r}
lm.beta.lmer(HealthAlgorithm_LMER_1)
```