mccs.Rmd

---
title: "Study 3 - Masculine Crime Capital Scale (MCCS)"
author: "Timothy J. Luke & Pär Stern"
date: "December 19, 2018"
output: html_document
---

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

library(tidyverse)
library(lme4)
library(lmerTest)
library(broom)
library(remef)

mccs <- read_csv("./data/Study3_SMNI_MCCS_raw.csv")

mccs <- mccs %>% 
  mutate(
    Winning4_R = case_when(
      Winning4_R == 1 ~ 4,
      Winning4_R == 2 ~ 3,
      Winning4_R == 3 ~ 2,
      Winning4_R == 4 ~ 1
    ),
    Emotion2_R = case_when(
      Emotion2_R == 1 ~ 4,
      Emotion2_R == 2 ~ 3,
      Emotion2_R == 3 ~ 2,
      Emotion2_R == 4 ~ 1
    ),
    Emotion5_R = case_when(
      Emotion5_R == 1 ~ 4,
      Emotion5_R == 2 ~ 3,
      Emotion5_R == 3 ~ 2,
      Emotion5_R == 4 ~ 1
    ),
    Risk4_R = case_when(
      Risk4_R == 1 ~ 4,
      Risk4_R == 2 ~ 3,
      Risk4_R == 3 ~ 2,
      Risk4_R == 4 ~ 1
    ),
    Risk5_R = case_when(
      Risk5_R == 1 ~ 4,
      Risk5_R == 2 ~ 3,
      Risk5_R == 3 ~ 2,
      Risk5_R == 4 ~ 1
    ),
    Violence2_R = case_when(
      Violence2_R == 1 ~ 4,
      Violence2_R == 2 ~ 3,
      Violence2_R == 3 ~ 2,
      Violence2_R == 4 ~ 1
    ),
    Violence5_R = case_when(
      Violence5_R == 1 ~ 4,
      Violence5_R == 2 ~ 3,
      Violence5_R == 3 ~ 2,
      Violence5_R == 4 ~ 1
    ),
    Playboy4_R = case_when(
      Playboy4_R == 1 ~ 4,
      Playboy4_R == 2 ~ 3,
      Playboy4_R == 3 ~ 2,
      Playboy4_R == 4 ~ 1
    ),
    Playboy5_R = case_when(
      Playboy5_R == 1 ~ 4,
      Playboy5_R == 2 ~ 3,
      Playboy5_R == 3 ~ 2,
      Playboy5_R == 4 ~ 1
    ),
    SelfReliance5_R = case_when(
      SelfReliance5_R == 1 ~ 4,
      SelfReliance5_R == 2 ~ 3,
      SelfReliance5_R == 3 ~ 2,
      SelfReliance5_R == 4 ~ 1
    ),
    Hetero5_R = case_when(
      Hetero5_R == 1 ~ 4,
      Hetero5_R == 2 ~ 3,
      Hetero5_R == 3 ~ 2,
      Hetero5_R == 4 ~ 1
    ),
    WorkPrimacy1_R = case_when(
      WorkPrimacy1_R == 1 ~ 4,
      WorkPrimacy1_R == 2 ~ 3,
      WorkPrimacy1_R == 3 ~ 2,
      WorkPrimacy1_R == 4 ~ 1
    ),
    WorkPrimacy3_R = case_when(
      WorkPrimacy3_R == 1 ~ 4,
      WorkPrimacy3_R == 2 ~ 3,
      WorkPrimacy3_R == 3 ~ 2,
      WorkPrimacy3_R == 4 ~ 1
    ),
    PoW4_R = case_when(
      PoW4_R == 1 ~ 4,
      PoW4_R == 2 ~ 3,
      PoW4_R == 3 ~ 2,
      PoW4_R == 4 ~ 1
    ),
    winning = (Winning1 + Winning2 + Winning3 + Winning4_R + Winning5)/5,
    emotion = (Emotion1 + Emotion2_R + Emotion3 + Emotion4 + Emotion5_R)/5,
    risk = (Risk1 + Risk2 + Risk3 + Risk4_R + Risk5_R)/5,
    violence = (Violence1 + Violence2_R + Violence3 + Violence4 + Violence5_R)/5,
    playboy = (Playboy1 + Playboy2 + Playboy3 + Playboy4_R + Playboy5_R)/5,
    self_reliance = (SelfReliance1 + SelfReliance2 + SelfReliance3 + SelfReliance4 + SelfReliance5_R)/5,
    hetero = (Hetero1 + Hetero2 + Hetero3 + Hetero4 + Hetero5_R)/5,
    work_primacy = (WorkPrimacy1_R + WorkPrimacy2 + WorkPrimacy3_R + WorkPrimacy4 + WorkPrimacy5)/5,
    pow = (PoW1 + PoW2 + PoW3 + PoW4_R + PoW5)/5,
    smni = (winning + emotion + risk + violence + playboy + self_reliance + hetero + work_primacy + pow)/9
  )

mccs_long <- mccs %>%
  gather(key = "key", value = "capital", names(mccs[53:168])) %>% 
  extract(key, c("crime", "rating_type"), "(.*)_(.*)") %>% 
  filter(!is.na(capital)) %>% 
  mutate(
    rating_type = case_when(
      rating_type == 1 ~ "personal",
      rating_type == 2 ~ "peer",
      rating_type == 3 ~ "society"
    )
  )

mci_40 <- read_csv("./data/MCI_40_mean_ratings.csv") %>% 
  mutate(
    crime_masculinity = mean
  )

mci_vals <- mci_40 %>% 
  select(crime, crime_masculinity)

mccs_long <- mccs_long %>% 
  left_join(mci_vals, by = "crime") %>% 
  mutate(
    crime_masculinity_mc = scale(crime_masculinity, scale = FALSE),
    smni_mc = scale(smni, scale = FALSE)
  )

mccs_personal <- mccs_long %>% 
  filter(rating_type == "personal")

mccs_peer <- mccs_long %>% 
  filter(rating_type == "peer")

mccs_society <- mccs_long %>% 
  filter(rating_type == "society")

```

## Descriptives

### Sample information

```{r}
mccs %>%
  summarise(
    Male = sum(gender == "Male"),
    Female = sum(gender == "Female"),
    "Mean Age" = round(mean(age, na.rm = TRUE), 2),
    "SD Age" = round(sd(age, na.rm = TRUE), 2),
    "Median Age" = round(median(age, na.rm = TRUE), 2)
  ) %>% 
  knitr::kable()
```

```{r}
race_table <- table(mccs$ethnicty) %>% 
  broom::tidy() %>% 
  rename(
    Ethnicity = Var1,
    Frequency = n
  ) %>% 
  arrange(desc(Frequency))

knitr::kable(race_table)
```


### Average captial ratings

```{r}

capital_table <- mccs_long %>% 
  group_by(rating_type) %>% 
  summarise(
    mean = mean(capital, na.rm = TRUE),
    sd = sd(capital, na.rm = TRUE)
  )


knitr::kable(capital_table)
```

### Average SMNI scores

```{r}

smni_table <- mccs %>% 
  summarise(
    mean = mean(smni, na.rm = TRUE),
    sd = sd(smni, na.rm = TRUE)
  )


knitr::kable(smni_table)
```

## Hypothesis test

```{r mlm 1}

peer_model <- lmer(capital ~ crime_masculinity_mc + smni_mc  + crime_masculinity_mc*smni_mc + (1|ID), data = mccs_peer)

summary(peer_model)

```

### Visualizations

#### Effect of crime masculinity on social capital (other effects partialed out)
```{r}
partial_capital <- remef(peer_model, fix = c("smni_mc", "crime_masculinity_mc:smni_mc"), ran = "all")

mccs_peer %>% 
  filter(!is.na(smni)) %>% 
  ggplot(.,
         aes(
           y = partial_capital,
           x = crime_masculinity_mc 
         )) +
  geom_point(
    alpha = .3
  ) +
  geom_smooth(
    method = "lm",
    color = "red"
  ) +
  labs(
    x = "Crime masculinity (MCI-40 ratings, mean centered)",
    y = "Social capital (partial residuals)"
    ) +
  theme_classic()


```

#### Effect of SMNI on social capital (other effects partialed out)
```{r}
partial_capital_2 <- remef(peer_model, fix = c("crime_masculinity_mc", "crime_masculinity_mc:smni_mc"), ran = "all")

mccs_peer %>% 
  filter(!is.na(smni_mc)) %>%
  ggplot(.,
         aes(
           y = partial_capital_2,
           x = smni_mc 
         )) +
  geom_point(
    alpha = .3
  ) +
  geom_smooth(
    method = "lm",
    color = "red"
  ) +
  labs(
    x = "Subscription to masculine norms (mean centered)",
    y = "Social capital (partial residuals)"
  ) +
  theme_classic()
```

### Supplmental model (outlier SMNI values removed)

```{r}
peer_model_no_outliers <- lmer(capital ~ crime_masculinity_mc + smni_mc  + crime_masculinity_mc*smni_mc + (1|ID), data = filter(mccs_peer, smni_mc < 1))

summary(peer_model_no_outliers)
```


## Exploratory models

### Personal ratings

```{r}

personal_model <- lmer(capital ~ crime_masculinity_mc + smni_mc + crime_masculinity_mc*smni_mc + (1|ID), data = mccs_personal)

summary(personal_model)

```

### Society ratings

```{r}

society_model <- lmer(capital ~ crime_masculinity_mc + smni_mc + crime_masculinity_mc*smni_mc + (1|ID), data = mccs_society)

summary(society_model)

```

### Comprehensive model

```{r}
comp_model <- lmer(capital ~ crime_masculinity_mc + smni_mc + rating_type + rating_type*crime_masculinity_mc + smni_mc*rating_type + crime_masculinity_mc*smni_mc + crime_masculinity_mc*smni_mc*rating_type +(1|ID), data = mccs_long)

summary(comp_model)
```


## SMNI scale diagnostics

```{r}

psych::alpha(mccs[8:12])

```

```{r}

psych::alpha(mccs[13:17])

```

```{r}

psych::alpha(mccs[18:22])

```

```{r}

psych::alpha(mccs[23:27])

```

```{r}

psych::alpha(mccs[28:32])

```

```{r}

psych::alpha(mccs[33:37])

```

```{r}

psych::alpha(mccs[38:42])

```

```{r}

psych::alpha(mccs[43:47])

```

```{r}

psych::alpha(mccs[48:52])

```

### Factor analysis

```{r}
factanal(~., data = mccs[8:52], factors = 9, na.action = na.exclude)
```