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, union
library(plotly)
## Loading required package: ggplot2
## 
## Attaching package: 'plotly'
## The following object is masked from 'package:ggplot2':
## 
##     last_plot
## The following object is masked from 'package:stats':
## 
##     filter
## The following object is masked from 'package:graphics':
## 
##     layout
hr <- read_csv('https://raw.githubusercontent.com/aiplanethub/Datasets/refs/heads/master/HR_comma_sep.csv')
## Rows: 14999 Columns: 10
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (2): Department, salary
## dbl (8): satisfaction_level, last_evaluation, number_project, average_montly...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
hr1 <- hr %>% 
  mutate(Employee_Status = (ifelse(left == 0 , 'Stayed' , 'Left')))


t.test(hr1$average_montly_hours ~ hr1$Employee_Status)
## 
##  Welch Two Sample t-test
## 
## data:  hr1$average_montly_hours by hr1$Employee_Status
## t = 7.5323, df = 4875.1, p-value = 5.907e-14
## alternative hypothesis: true difference in means between group Left and group Stayed is not equal to 0
## 95 percent confidence interval:
##   6.183384 10.534631
## sample estimates:
##   mean in group Left mean in group Stayed 
##             207.4192             199.0602

The p-value is very small so there is a significant difference between means where employees that left work at least 6 hours more.

Descriptive: Employees that left, on average, work more hours, at least 3% more.

Prescriptive: To reduce employee attribution, average monthly hours can be reduced by 3%, for those that work longer hours.

plot_ly(hr1,
        x = ~Employee_Status ,
        y = ~average_montly_hours , 
        type = 'box' ,
        color = ~Employee_Status, 
        colors = c('#1a1aff' , '#ff3333'))%>%
  layout(title = 'Employees that left, on average, work more hours, at least 3% more')
t.test(hr1$satisfaction_level ~ hr1$Employee_Status)
## 
##  Welch Two Sample t-test
## 
## data:  hr1$satisfaction_level by hr1$Employee_Status
## t = -46.636, df = 5167, p-value < 2.2e-16
## alternative hypothesis: true difference in means between group Left and group Stayed is not equal to 0
## 95 percent confidence interval:
##  -0.2362417 -0.2171815
## sample estimates:
##   mean in group Left mean in group Stayed 
##            0.4400980            0.6668096

The p-value is very small so there is a significant difference between means where employees that left had an average satisfaction level .2 lower than ones who stayed.

Descriptive: Employees that left had a 67% lower satisfaction level.

Prescriptive: To maintain employee retention, managers need to increase satisfaction levels by 67%.

plot_ly(hr1,
        x = ~Employee_Status ,
        y = ~satisfaction_level , 
        type = 'box' ,
        color = ~Employee_Status, 
        colors = c('#1a1aff' , '#ff3333'))%>%
  layout(title = 'Employees that left had a 67% lower satisfaction level')
t.test(hr1$last_evaluation ~ hr1$Employee_Status)
## 
##  Welch Two Sample t-test
## 
## data:  hr1$last_evaluation by hr1$Employee_Status
## t = 0.72534, df = 5154.9, p-value = 0.4683
## alternative hypothesis: true difference in means between group Left and group Stayed is not equal to 0
## 95 percent confidence interval:
##  -0.004493874  0.009772224
## sample estimates:
##   mean in group Left mean in group Stayed 
##            0.7181126            0.7154734

The p-value is large so there is a non-significant difference between means where employees that left had very similar last evaluation scores to ones who stayed.

Descriptive: There was virtually no difference between last evaluation score for employees who left or stayed.

Prescriptive: Nothing needs to be done regarding last evaluation score.

plot_ly(hr1,
        x = ~Employee_Status ,
        y = ~last_evaluation , 
        type = 'box' ,
        color = ~Employee_Status, 
        colors = c('#1a1aff' , '#ff3333'))%>%
  layout(title = 'There was virtually no difference between average last evaluation
# score for employees who left or stayed')
t.test(hr1$number_project ~ hr1$Employee_Status)
## 
##  Welch Two Sample t-test
## 
## data:  hr1$number_project by hr1$Employee_Status
## t = 2.1663, df = 4236.5, p-value = 0.03034
## alternative hypothesis: true difference in means between group Left and group Stayed is not equal to 0
## 95 percent confidence interval:
##  0.006540119 0.131136535
## sample estimates:
##   mean in group Left mean in group Stayed 
##             3.855503             3.786664

The p-value is large so there is a non-significant difference between means where employees that left had a very similar average number of projects to ones who stayed.

Descriptive: Employees that left, on average, worked on more projects, about 2% more.

Prescriptive: To reduce employee attribution, number of projects can be reduced by 2%, for those who work on more projects.

plot_ly(hr1,
        x = ~Employee_Status ,
        y = ~number_project , 
        type = 'box' ,
        color = ~Employee_Status, 
        colors = c('#1a1aff' , '#ff3333'))%>%
  layout(title = 'Employees that left, on average, worked on more projects, about 2% more')