experimental final format.Rmd

---
title: "Chi Squared Analysis Against Gender"
author: "Jessica Spencer"
date: "8/18/2021"
output: pdf_document
---

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

source("cleaning_for_numerics.R")
library(ggplot2)

make_barchart = function(tab,title){
  prop_tab = ftable(round(prop.table(tab,2)*100, 3))
  ggplot(data=screencast, aes(x=Reason,y=Percentage,fill=factor(Type))) +
  geom_bar(position="dodge",stat="identity") + 
  coord_flip() +
  ggtitle("Replace Title")
}

get_sig = function(df){
  sig = c()
  insig = c()
  for(col in 1:ncol(df)){
    table = table(df[,col],df$bucketed_gender)
    #print(table)
    #rownames(table)=c("Yes","No")
    #title = paste(colnames(df)[col],"with Gender")
    cont_tab = ftable(table)
    chi = chisq.test(table,simulate.p.value=TRUE)
    
    if(chi$p.value<0.05){
      #add to list of significant results
      sig = append(sig,col)
    }
  }
  return(sig)
}

make_table_chi_df = function(df){
  for(col in 1:ncol(df)){
    table = table(df[,col],df$bucketed_gender)
    title = paste(colnames(df)[col],"with Gender")
    make_table_chi(table, title)
  }
}

make_table_chi = function(tab,title){
  cont_tab = ftable(tab)
  prop_tab = ftable(round(prop.table(tab,2)*100, 3))
  chi = chisq.test(tab,simulate.p.value=TRUE)
  #if chi^2 is sig, then print the results
  if(chi$p.value < 0.05){
    print(title)
    print("Contingency Table")
    print(cont_tab)
    print("Results in Percent of Participants")
    print(prop_tab)
    print("Chi-Squared Test")
    print(chisq.test(tab))
  }
  #else{
    #print(paste(title," - No significant relationship was found"))
    
  #} 
  
}

master_chi_instrument = function(colStart,colEnd,df,gender_column){
  for(i in colStart:colEnd) {  
   # print(names(df)[i])# for-loop over columns
  #  print((length(unique(df[,i]))>1))
    if(length(unique(df[,i]))>1){
      g= table(df[,i],df$bucketed_gender)
      rownames(g) = c("No","Yes")
      title = paste( "Q: Do you play ", names(df)[i] ," at a professional level?")
      val = make_table_chi(g, title)
    }
  }
}
```

## Methods
A Chi-Squared test was run against "bucketed gender" and all other independent variables.  Bucketed gender was created by combining transgender and nonbinary or gender non-conforming participants into one category labelled "transgender and nonbinary". This was done so that there would be enought datapoints in each category for a chi-squared test to be reliably performed.  Here are the counts for each category of bucketed gender:
```{r}
table(clean_dat$bucketed_gender)
```

## Significant Results

The following variables were shown to have a significant relationship with gender.

```{r}
sig = get_sig(subset_cd)
sig_subset = subset_cd[,sig]
make_table_chi_df(sig_subset)
```
#Instrument Choice with Gender

The following shows the instruments,that have some significant relationship with gender.Each of these instruments was tested individually against gender :

piano, keyboard, flute,  clarinet, piccolo, oboe, trumpet, flugelhorn, saxaphone-write in, trombone, tuba, french horn, violin, viola, cello, acoustic bass, electric bass, acoustic guitar, electric guitar, harp, drums, percussion

```{r}
#For Instruments
master_chi_instrument(colStart=31,colEnd=53,df=poi,gender_column="bucketed_gender")
```

#Instrument Families

the following groupings were made and tested against bucketed gender. All were significant:

Strings : violin, viola, acoustic bass, cello  


Brass : trumpet, flugelhorn, trombone, tuba, frenchhorn


Woodwinds: flute, clarinet, piccolo, oboe, saxophone  


Rhythm: acoustic guitar, electric guitar, electric bass, drums, percussion  


piano or keyboard  



```{r}
master_chi_instrument(colStart=54,colEnd=58,df=poi,gender_column="bucketed_gender")
```