Categorical Data Analysis
library(readr)
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)NHIS_Data <- read_csv("Downloads/NHIS Data.csv")##
## ── Column specification ────────────────────────────────────────────────────────
## cols(
## .default = col_double(),
## Demo_Race = col_logical(),
## Demo_Hispanic = col_character(),
## Demo_RaceEthnicity = col_character(),
## Demo_Region = col_character(),
## Demo_sex_C = col_character(),
## Demo_sexorien_C = col_logical(),
## Demo_agerange_C = col_character(),
## Demo_marital_C = col_character(),
## Demo_hourswrk_C = col_character(),
## MentalHealth_MentalIllnessK6_C = col_character(),
## MentalHealth_depressionmeds_B = col_logical(),
## Health_SelfRatedHealth_C = col_character(),
## Health_diagnosed_STD5yr_B = col_logical(),
## Health_BirthControlNow_B = col_logical(),
## Health_EverHavePrediabetes_B = col_logical(),
## Health_HIVAidsRisk_C = col_character(),
## Health_BMI_C = col_character(),
## Health_UsualPlaceHealthcare_C = col_character(),
## Health_AbnormalPapPast3yr_B = col_logical(),
## Behav_CigsPerDay_C = col_character()
## # ... with 1 more columns
## )
## ℹ Use `spec()` for the full column specifications.## Warning: 683386 parsing failures.
## row col expected actual file
## 68557 Demo_Race 1/0/T/F/TRUE/FALSE Black or African American 'Downloads/NHIS Data.csv'
## 68558 Demo_Race 1/0/T/F/TRUE/FALSE Asian 'Downloads/NHIS Data.csv'
## 68559 Demo_Race 1/0/T/F/TRUE/FALSE American Indian or Alaskan Native 'Downloads/NHIS Data.csv'
## 68560 Demo_Race 1/0/T/F/TRUE/FALSE White 'Downloads/NHIS Data.csv'
## 68561 Demo_Race 1/0/T/F/TRUE/FALSE White 'Downloads/NHIS Data.csv'
## ..... ......... .................. ................................. .........................
## See problems(...) for more details.I hypothesize that there is a relationship between Race/Ethnicity and Mental Health Illness.
Crosstab (Null Vs Actual)
table(NHIS_Data $ Demo_RaceEthnicity) ##
## American Indian or Alaskan Native Hispanic
## 1557
## American Indian or Alaskan Native Not Hispanic
## 3136
## Asian Hispanic
## 1035
## Asian Not Hispanic
## 24733
## Black or African American (Hispanic Identity Unknown)
## 2
## Black or African American Hispanic
## 2725
## Black or African American Not Hispanic
## 77661
## Hispanic (Race Identity Unknown)
## 11100
## Multiple Races (Hispanic Identity Unknown)
## 1
## Multiple Races Hispanic
## 1588
## Multiple Races Not Hispanic
## 6810
## Not Hispanic (Race Identity Unknown)
## 58278
## Other Race Hispanic
## 4807
## Other Race Not Hispanic
## 151
## White (Hispanic Identity Unknown)
## 16
## White Hispanic
## 81356
## White Not Hispanic
## 344526table(NHIS_Data $ MentalHealth_MentalIllnessK6_C)##
## Low Risk MMD SMI
## 487109 97837 21633Actual Distribution
table(NHIS_Data $ Demo_RaceEthnicity, NHIS_Data $ MentalHealth_MentalIllnessK6_C)##
## Low Risk MMD SMI
## American Indian or Alaskan Native Hispanic 1130 330 78
## American Indian or Alaskan Native Not Hispanic 2221 626 197
## Asian Hispanic 810 161 48
## Asian Not Hispanic 20541 3031 450
## Black or African American (Hispanic Identity Unknown) 1 1 0
## Black or African American Hispanic 2022 517 124
## Black or African American Not Hispanic 59796 13026 2945
## Hispanic (Race Identity Unknown) 8409 1944 583
## Multiple Races (Hispanic Identity Unknown) 1 0 0
## Multiple Races Hispanic 1078 368 108
## Multiple Races Not Hispanic 4591 1605 447
## Not Hispanic (Race Identity Unknown) 45845 9589 1977
## Other Race Hispanic 3580 894 252
## Other Race Not Hispanic 120 23 4
## White (Hispanic Identity Unknown) 11 3 1
## White Hispanic 64133 12557 3215
## White Not Hispanic 272783 53152 11202- The null hypothesis is that Race/Ethnicity and Mental Health Illness are independent to one another.
Row or Column %
table(NHIS_Data $ Demo_RaceEthnicity, NHIS_Data $ MentalHealth_MentalIllnessK6_C) %>%
prop.table(1)##
## Low Risk MMD
## American Indian or Alaskan Native Hispanic 0.73472042 0.21456437
## American Indian or Alaskan Native Not Hispanic 0.72963206 0.20565046
## Asian Hispanic 0.79489696 0.15799804
## Asian Not Hispanic 0.85509117 0.12617601
## Black or African American (Hispanic Identity Unknown) 0.50000000 0.50000000
## Black or African American Hispanic 0.75929403 0.19414195
## Black or African American Not Hispanic 0.78920902 0.17192181
## Hispanic (Race Identity Unknown) 0.76892831 0.17776152
## Multiple Races (Hispanic Identity Unknown) 1.00000000 0.00000000
## Multiple Races Hispanic 0.69369369 0.23680824
## Multiple Races Not Hispanic 0.69110342 0.24160771
## Not Hispanic (Race Identity Unknown) 0.79854035 0.16702374
## Other Race Hispanic 0.75751164 0.18916631
## Other Race Not Hispanic 0.81632653 0.15646259
## White (Hispanic Identity Unknown) 0.73333333 0.20000000
## White Hispanic 0.80261561 0.15714911
## White Not Hispanic 0.80911618 0.15765698
##
## SMI
## American Indian or Alaskan Native Hispanic 0.05071521
## American Indian or Alaskan Native Not Hispanic 0.06471748
## Asian Hispanic 0.04710500
## Asian Not Hispanic 0.01873283
## Black or African American (Hispanic Identity Unknown) 0.00000000
## Black or African American Hispanic 0.04656403
## Black or African American Not Hispanic 0.03886916
## Hispanic (Race Identity Unknown) 0.05331017
## Multiple Races (Hispanic Identity Unknown) 0.00000000
## Multiple Races Hispanic 0.06949807
## Multiple Races Not Hispanic 0.06728888
## Not Hispanic (Race Identity Unknown) 0.03443591
## Other Race Hispanic 0.05332205
## Other Race Not Hispanic 0.02721088
## White (Hispanic Identity Unknown) 0.06666667
## White Hispanic 0.04023528
## White Not Hispanic 0.03322685Multiple Races (Hispanic Identity Unknown) have the highest percentage of Low Risk cases of mental illness: 100%
Black or African American’s (Hispanic Identity Unknown) have the lowest percentage of Low Risk mental illness: 5%
Black or African American’s (Hispanic Identity Unknown) have the highest percentage of Moderate Mental Distress: 5%
Multiple Races (Hispanic Identity Unknown) have the lowest percentage of Moderate Mental Distress: 0%
White, American Indian or Alaskan Native, Multiple Races Hispanic,and Multiple Races have the highest percentage of Serious Mental Illness: 6%
Black or African American’s have the lowest percentage of Serious Mental Illness: 0%
NHIS_Data %>%
group_by(Demo_RaceEthnicity,MentalHealth_MentalIllnessK6_C) %>%
summarize(n=n()) %>%
mutate(percent=n/sum(n)) %>%
ggplot()+
geom_col(aes(x= Demo_RaceEthnicity, y= percent, fill= MentalHealth_MentalIllnessK6_C))## `summarise()` has grouped output by 'Demo_RaceEthnicity'. You can override using the `.groups` argument.
Chi-Square Statistical Test
- data: Demo_RaceEthnicity and MentalHealth_MentalIllnessK6_C
chisq.test(NHIS_Data $ Demo_RaceEthnicity, NHIS_Data $ MentalHealth_MentalIllnessK6_C)## Warning in chisq.test(NHIS_Data$Demo_RaceEthnicity,
## NHIS_Data$MentalHealth_MentalIllnessK6_C): Chi-squared approximation may be
## incorrect##
## Pearson's Chi-squared test
##
## data: NHIS_Data$Demo_RaceEthnicity and NHIS_Data$MentalHealth_MentalIllnessK6_C
## X-squared = 1825.8, df = 32, p-value < 2.2e-16- The results from the chi-square test are in favor of me rejecting the null hypothesis. There is a significant relationship, statistically speaking, between Demographic Race/Ethnicity and Mental Health Illness because the p-value is less than .05. These variables are not independent of one another.