| Title: Homework 8 |
| Author: Brandon Flores |
| Date: Nov. 3rd, 2021 |
When observing the age difference between transmen and transwomen it is with 95% confidence that transmen are about 5 to 16 years younger than transwomen. With this we can reject the null hypothesis. This is a statistically significant relationship (pvalue = 0.0002697) at the .001 level.
Observing the difference between the outputs of excel and R, the outputs were very similar with a critical T of 3.734642 and a p-value of 0.000137. Although these two numbers slightly differed from RStudio (more so for the pvalue) the interpretation of them are all the same.
This can also be said for the outputs of both programs concerning the confidence intervals both being able to be interpreted by being between about 5 to 16 with 95% confidence.
The biggest difference between the two programs of outputs is that in R all of the outputs are negative (besides the pvalue & df) and in excel they are positive. When I tried to mimic the negative output of R in excel my pvalue became extremely different showing about a 0.999 level being the wrong output.
Finally the difference can also be shown in the degrees of freedom with my df by hand being 138 compared to the model in R being at about 129 when using the FALSE function.
Even with these known differences, when calculated properly the outputs of both programs are strikingly similar and the same in terms of interpretation.
According to the data, cismen are older than ciswomen by about 1.94 years. When analyzing the t-test for these groups the relationship between them is statistically significant at the .001 level. But due to the extremely large sample size of the model (about 55,000 and a df at about 47,000) this statistical relationship is not practical in the real world. Due to the size of the sample within the model it makes the relationship between the two variables very likely to be statically related to one another.
library(tidyverse)
## -- Attaching packages --------------------------------------- tidyverse 1.3.1 --
## v ggplot2 3.3.5 v purrr 0.3.4
## v tibble 3.1.4 v dplyr 1.0.7
## v tidyr 1.1.3 v stringr 1.4.0
## v readr 2.0.1 v forcats 0.5.1
## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
library(dplyr)
library(rvest)
##
## Attaching package: 'rvest'
## The following object is masked from 'package:readr':
##
## guess_encoding
library(httr)
library(purrr)
library(stringr)
library(janitor)
##
## Attaching package: 'janitor'
## The following objects are masked from 'package:stats':
##
## chisq.test, fisher.test
library(plyr)
## ------------------------------------------------------------------------------
## You have loaded plyr after dplyr - this is likely to cause problems.
## If you need functions from both plyr and dplyr, please load plyr first, then dplyr:
## library(plyr); library(dplyr)
## ------------------------------------------------------------------------------
##
## Attaching package: 'plyr'
## The following objects are masked from 'package:dplyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarize
## The following object is masked from 'package:purrr':
##
## compact
library(ggplot2)
library(Rmisc)
## Loading required package: lattice
pulse39 <-read.csv("C:\\Users\\BTP\\Downloads\\pulse2021_puf_39.csv")
pulse39$subgroup <-paste(pulse39 $EGENID_BIRTH,
pulse39$GENID_DESCRIBE, sep = "")
pulse39 %>%
tabyl(subgroup)
## subgroup n percent
## 1-99 417 0.007307584
## 11 22652 0.396957802
## 12 66 0.001156596
## 13 60 0.001051451
## 14 263 0.004608860
## 2-99 544 0.009533156
## 21 73 0.001279265
## 22 32522 0.569921492
## 23 79 0.001384410
## 24 388 0.006799383
pulse39$subgroupcat <-car::Recode(pulse39$ subgroup,
recodes=" '11' = 'cisman'; '22' = 'ciswoman'; '23' = 'transman'; '13' = 'transwoman'; else=NA",
as.factor=T)
pulse39 %>%
tabyl(subgroupcat)
## subgroupcat n percent valid_percent
## cisman 22652 0.396957802 0.409523982
## ciswoman 32522 0.569921492 0.587963047
## transman 79 0.001384410 0.001428236
## transwoman 60 0.001051451 0.001084736
## <NA> 1751 0.030684845 NA
pulse39 <- transform(pulse39, age=2021-TBIRTH_YEAR)
pulse39$trans <- car::Recode(pulse39$ subgroup,
recodes=" '23' = 'transman'; '13' = 'transwoman';
else=NA", as.factor=T)
pulse39 %>%
tabyl(trans)
## trans n percent valid_percent
## transman 79 0.001384410 0.5683453
## transwoman 60 0.001051451 0.4316547
## <NA> 56925 0.997564139 NA
pulse39$cis <-car::Recode(pulse39$ subgroup,
recodes=" '11' = 'cisman'; '22' = 'ciswoman'; else=NA", as.factor=T)
pulse39 %>%
tabyl(cis)
## cis n percent valid_percent
## cisman 22652 0.39695780 0.4105557
## ciswoman 32522 0.56992149 0.5894443
## <NA> 1890 0.03312071 NA
pulse39 %>%
group_by(trans) %>%
summarise_at(vars(age), list(name = mean))
## # A tibble: 3 x 2
## trans name
## <fct> <dbl>
## 1 transman 33.6
## 2 transwoman 44.0
## 3 <NA> 53.9
pulse39 %>%
group_by(trans) %>%
summarise_at(vars(age), list(name = sd))
## # A tibble: 3 x 2
## trans name
## <fct> <dbl>
## 1 transman 16.6
## 2 transwoman 16.0
## 3 <NA> 15.9
pulse39 %>%
group_by(cis) %>%
summarise_at(vars(age), list(name = mean))
## # A tibble: 3 x 2
## cis name
## <fct> <dbl>
## 1 cisman 55.0
## 2 ciswoman 53.1
## 3 <NA> 52.9
pulse39 %>%
group_by(cis) %>%
summarise_at(vars(age), list(name = sd))
## # A tibble: 3 x 2
## cis name
## <fct> <dbl>
## 1 cisman 16.3
## 2 ciswoman 15.5
## 3 <NA> 18.2
t.test (age ~ trans, var.equals=FALSE, data = pulse39)
##
## Welch Two Sample t-test
##
## data: age by trans
## t = -3.7457, df = 129.47, p-value = 0.0002697
## alternative hypothesis: true difference in means between group transman and group transwoman is not equal to 0
## 95 percent confidence interval:
## -15.971303 -4.930807
## sample estimates:
## mean in group transman mean in group transwoman
## 33.58228 44.03333
t.test (age ~ cis, var.equals=FALSE, data = pulse39)
##
## Welch Two Sample t-test
##
## data: age by cis
## t = 14.056, df = 47056, p-value < 2.2e-16
## alternative hypothesis: true difference in means between group cisman and group ciswoman is not equal to 0
## 95 percent confidence interval:
## 1.672666 2.214754
## sample estimates:
## mean in group cisman mean in group ciswoman
## 55.04441 53.10070