Assignment-1

Assignment_01

Submitted By: Jyotirmoy Sen Gupta

: Visually shows the Relationship Between Satisfaction Level and Last

Graphical Presentation_1

The provided graph shows the relationship between Satisfaction Level and Average Last Evaluation
Score.

X-axis (Satisfaction Level): Represents the levels of satisfaction ranging from 0.09 to 1.0.

Y-axis (Mean Last Evaluation): Represents the average last evaluation score, ranging from 0.50 to 0.90.

Data Representation: The graph uses a line plot to show the trend of mean evaluation scores across
satisfaction levels. Each satisfaction level is associated with its corresponding mean last evaluation
score.
Assignment-1

Detailed Observations:

High Mean Evaluations at Low and High Satisfaction Levels:

1. At satisfaction level 0.09, the mean last evaluation score is very high (0.87).
2. Similarly, for satisfaction levels between 0.8 and 1.0, the mean evaluation scores are
consistently high (around 0.76–0.89).
3. This suggests that both extremely dissatisfied and highly satisfied individuals tend to receive
high last evaluation scores.

Low Mean Evaluations at Mid Satisfaction Levels:

1. Satisfaction levels between 0.30 and 0.50 show the lowest mean evaluation scores, dropping as
low as 0.53.
2. This indicates that individuals with moderate satisfaction levels tend to perform poorly in their
evaluations.

Increasing Trend Beyond 0.50:

After the lowest point around 0.50 satisfaction, the mean evaluation score begins to rise steadily,
reaching higher values as satisfaction increases.

Recommendations for Further Analysis:

1. Segmentation: Investigate whether specific job roles, departments, or teams contribute to these
trends.
2. Statistical Modeling: Apply regression analysis to quantify the relationship and identify the
inflection points.
3. Qualitative Analysis: Conduct surveys or interviews to understand why moderately satisfied
employees perform poorly.
Assignment-1

Graphical Presentation_2

Relationship Between Satisfaction Level and Last Evaluation

700.00
600.00
500.00
400.00
300.00
200.00
100.00
0.00
0.09
0.12
0.15
0.18
0.21
0.24
0.27

0.33
0.36
0.39
0.42
0.45
0.48
0.51
0.54
0.57

0.62
0.65
0.68
0.71
0.74
0.77

0.83
0.86
0.89
0.92
0.95
0.98
0.3

0.6

0.8
SATISFACTION_LEVEL

The graph displays the Relationship Between Satisfaction Level and the Number of Last
Evaluations

X-axis (satisfaction_level_1): Satisfaction levels ranging from 0.09 to 1.0.

Data Representation:

A line plot is used to track changes in the number of last evaluations as satisfaction levels vary. A
prominent spike is observed, with otherwise stable values across the range.

Detailed Observations:

Significant Spike at Satisfaction Level ~0.57:

1. There is a massive spike in the number of last evaluations, reaching approximately 800 at
satisfaction level 0.57.
2. This anomaly is highly distinct and may indicate a specific cohort or an event tied to this
satisfaction level.

Fluctuations at Lower Satisfaction Levels:

1. At very low satisfaction levels (0.09 to ~0.3), there are small but noticeable variations in the
number of evaluations.
2. The count remains relatively low (between 100 and 200 evaluations).

Stability at Higher Satisfaction Levels:

1. Beyond the satisfaction level of ~0.6, the number of evaluations stabilizes around 200–300, with
minor fluctuations.
2. This trend continues consistently toward the maximum satisfaction level (1.0).

Spike at 0.57: The sharp increase could indicate:

1. A large group of employees at this satisfaction level were evaluated during a particular period.
2. A specific event or intervention targeted at individuals with moderate satisfaction levels,
prompting higher evaluations.
3. Further investigation is needed to determine whether this represents a data anomaly, a
reporting issue, or a genuine trend.

Lower Evaluations at Extreme Satisfaction Levels:

Both very low (<0.2) and very high (>0.8) satisfaction levels have relatively fewer evaluations. This could
indicate:

1. A smaller population of employees in these satisfaction brackets.

General Stability Across Most Satisfaction Levels:

1. Aside from the spike, the number of evaluations remains relatively consistent across the range
of satisfaction levels.
2. This may suggest that evaluations are not strictly tied to satisfaction levels for the majority of
employees.

Recommendations for Further Investigation:

Analyze the Spike: Investigate why satisfaction level 0.57 shows such a significant increase. Check for
correlations with other variables such as department, job role, or time of evaluation.

Examine Evaluation Frequency by Satisfaction Group: Compare the frequency and outcomes of
evaluations for employees in extreme (low/high) satisfaction groups versus moderate satisfaction
groups.

Action Plan for High and Low Satisfaction Groups: If the low numbers at extreme satisfaction levels
indicate neglect, consider targeted strategies to engage and evaluate these employees.
Assignment-1

: Visually display the Frequency of Employees Based on Their Tenure In The

: Have you noticed any significant changes in satisfaction between different

Based on the problem statement, null and alternative hypothesis are as follows:

Null Hypothesis: There is no difference in satisfaction between different salary levels

Alternative hypothesis: There is at least one difference in satisfaction score among different salary levels
(at least one pair of salary level has significant difference)

Methodology: First we will check normality assumptions through Kolmogorov-Smirnov Test which
hypothesis are as follows:

Normality Assumptions Test:

Null Hypothesis: Variables come from normally distributed population

Alternative hypothesis: Variables don’t come from normally distributed population.

Here, the P Value of Kolmogorov-Smirnov Test is <0.001 which is less than 0.05. So, the null hypothesis is
rejected. So, it can be concluded that the normality assumption test is violated.

Since the normality test has been violated, non-parametric test will be carried out under Kruskal Walis
test.

Nonparametric Tests
Assignment-1

Results: Here, the P Value of Kruskal Wallis Test is <0.001 which is less than 0.05. So, the null hypothesis
that there is no difference in satisfaction between different salary levels, is rejected. From the pairwise
comparison of employee’s salary level it was found that there is difference between satisfaction score in
medium-high pair which P value is 0.107 which is more than 0.05.
Assignment-1

: Compare the employee's most recent performance evaluation score between

Based on the problem statement, null and alternative hypothesis are as follows:

Null Hypothesis: There is no difference in performance evaluation score between different departments

Alternative hypothesis: There is at least one difference in performance evaluation score between different
departments (at least one pair of department has significant difference)

Methodology: First, we will check normality assumptions through Shapiro-Wilk Test which hypothesis
are as follows:

Normality Assumptions Test:

Null Hypothesis: Variables come from normally distributed population

Alternative hypothesis: Variables don’t come from normally distributed population.

Here, the P Value of Shapiro-Wilk Test is <0.001 which is less than 0.05. So, the null hypothesis is
rejected. So, it can be concluded that the normality assumption test is violated.

Since the normality test has been violated, non-parametric test will be carried out under Kruskal Walis
test.
Assignment-1

Nonparametric Tests

Results: Here, the P Value of Kruskal Wallis Test is 0.090 which is greater than 0.05. So, we are failed to
reject the null hypothesis and can conclude that there is no difference in performance evaluation score
between different departments.
Assignment-1

:. Are the average monthly work hours higher in the sales department compared
to the technical department? (Test it using the proper approach.)

Based on the problem statement, null and alternative hypothesis are as follows:

Null Hypothesis: There is no difference in average monthly work hours in the sales department compared
to the technical department.

Alternative hypothesis: Average monthly work hours is higher in the sales department compared to the
technical department.

Methodology: First, we will check normality assumptions through Shapiro-Wilk Test which hypothesis
are as follows:

Normality Assumptions Test:

Null Hypothesis: Variables come from normally distributed population

Alternative hypothesis: Variables don’t come from normally distributed population.

Here, the P Value of Shapiro-Wilk Test is <0.001 which is less than 0.05. So, the null hypothesis is
rejected. So, it can be concluded that the normality assumption test is violated.

Since the normality test has been violated, non-parametric test will be carried out Mann-Whitney U Test

Nonparametric Tests

Results: Here, the P Value of Mann-Whitney U Test is 0.245 which is greater than 0.05. So, we are failed to
reject the null hypothesis and can conclude that there is no difference in average monthly work hours in
the sales department compared to the technical department.