HR Analytics
Group Presentation
Course: DECISION MODELING & ANALYTICS (MGT 611)
Lecturer: DR. TING CAO
By: UJU OBI | YASAMAN MOHAMMADI | RENE DOWNIE
Order of Presentation
1) Brief overview of our dataset |
Descriptive Statistics
a) Viewing the variables together
| Data Visualization
2) Exploring the relationships between
pairs of variables | Predictive Statistics
3) Summary of our Observations &
Recommendations
Our Objectives
Monthly Income
★ Explore the relationship (Compensation)
between these key variables Job Satisfaction
★ Assess whether these variables
can be explained by changes in
the other variables
★ Identify any patterns in how
these variables are managed in
the organization.
Performance
About the HR Analytics Dataset
Number of Variables = 36
Types of Variables
Numerical Categorical
23 13
Raw sample size (n) = 1,481
After Cleaning:
Sample size (n) = 1,423
Numerical Variables
★ Age
★ Daily Rate ★ Performance Rating
★ Distance from Home ★ Relationship Satisfaction
★ Education ★ Standard Hours
★ Employee Count ★ Stock Option Level
★ Hourly Rate ★ Total Working Years
★ Job Involvement ★ Training Times Last Year
★ Job Level ★ Work-Life Balance
★ Job Satisfaction ★ Years at the Company
★ Monthly Income ★ Years in Current Role
★ Monthly Rate ★ Years Since Last Promotion
★ Num Companies Worked ★ Years With Current Manager
Categorical Variables
1. EmpID
2. Age Group
3. Attrition
4. Business Travel
5. Department
6. Education Field
7. Gender
8. Job Role
9. Marital Status
10. Salary Slab
11. OverTime
12. Over 18
13. Employee Number
DESCRIPTIVE
STATISTICS
General / Personal Data Related Statistics
Compensation Related Statistics
Employee Satisfaction & Engagement Statistics
DATA VISUALIZATION
Age Group & Gender
★ The sample is predominantly male,
hence why the males consistently
outnumber the females in all age
groups.
★ The ratio of males to females is 6:4.
★ Most employees fall within the 26-
35 age group. Nonetheless, the
mean, median and mode are in the
second largest group, 36-45.
Job Involvement & Marital Status
★ The vast majority (46%) of the sample is married.
★ The proportion of the job involvement levels are consistent across marital status.
★ This was consistent with the descriptive statistics generated for Job Involvement.
Satisfaction by Department &
Income
★ 3 and 4 are consistently the most frequently
observed ratings for Environment and Job
Satisfaction across all departments and income
groups.
Work-Life Balance & Overtime Work
Performance by Dept & Job Title
C
PREDICTIVE
STATISTICS
Performance (Y) & Job Satisfaction (β1)
Null & Alternative Hypotheses
When Y = β0 + β1X↿
H0: β1=0
HA: β1≠0
P value > σ
We failed to reject the null hypothesis. Job satisfaction is a poor predictor of performance at the 5%
significance level. This is because there is enough evidence to suggest that β1 = 0.
The Predictors of Performance Rating (Y)
All poor
H0: β1=0, β2=0….β15=0 predictors of
performance at
HA: β1≠0, β≠0…...β15≠0 the 0.05
significance level.
P values > σ
The Predictors of Job Satisfaction (Y)
All poor
predictors of job
H0: β1=0, β2=0….β14=0 satisfaction at the
HA: β1≠0, β≠0…...β14≠0 0.05 significance
level.
P values > σ
The Predictors of Monthly Income (Y)
Age (β1), Education (β2), Performance Rating (β3), Training Times Last Year (β4), Years at Company
(β5) and Job Level
Years at Company (β5), Years in Current Role (β6), Job Level (β7) (β7) are good
predictors of
Monthly Income at
the 0.05
Null & Alternative Hypotheses significance level.
H0: β1=0, β2=0, β3=0, …..β7=0
HA: β1≠0, β2≠0, β3≠0, …..β7≠0
P values > σ
The Predictors of Job Level (Y)
Yrs in Current Role (β1), Yrs at Company (β2), Yrs with Curr Manager (β3), Age (β4), Age (β4) and Years
Education (β5) at Company (β2)
are good predictors
of Job Level at the
0.05 significance
Null & Alternative Hypotheses level.
H0: β1=0, β2=0, β3=0, β4=0, β5=0
HA: β1≠0, β2≠0, β3≠0, β4≠0, β5≠0
P values > σ
Job Level & Monthly Income
Formulas
Job Level (Y) = β0 + β1(X1) + β2(X2)
When:
β1 is Age and β2 is Years at Company
Job Level (Y) =
(-0.17) + 0.0466 (X1) + 0.0744 (X2)
Monthly Income (Y) = β0 + β1(X1) +
β2(X2)
When:
β1 is Years at Company and β2 is Job
Level
Monthly Income (Y) =
(-1,861.23) + 7.64(X1) + 4,023.12 (X2)
OUR OBSERVATIONS &
RECOMMENDATIONS
Our Observations & Recommendations
★ No predictors for Job Satisfaction &
Performance
★ Additional investigation into job
satisfaction is required.
★ Performance measures need more
variety.
★ Seniority and years of service are the
main considerations for promotion
and compensation practices.
