Campus Recruitment¶

1. Dataset Summary¶

This dataset is about academic and employability factors influencing placement which happens at the end of MBA, before graduation. The students are recruited based on their past academic performance, choice of specialization, placement test score, as well as work experience. The dataset consists of several categorical features such as gender, school board, specialization, and degree type. The numerical features are the student's performance in secondary, higher secondary school, college, and MBA, and salary offered.

The features are:

• gender
• ssc_p = Secondary Education percentage- 10th Grade
• ssc_b = Board of Education- Central/ Others
• hsc_p = Higher Secondary Education percentage- 12th Grade
• hsc_b = Board of Education- Central/ Others
• hsc_s = Specialization in Higher Secondary Education
• degree_p = Degree Percentage
• degree_t = Under Graduation(Degree type)- Field of degree education
• workex = Work Experience
• etest_p = Employability test percentage ( conducted by college)
• specialisation = Post Graduation(MBA)- Specialization
• mba_p = MBA percentage
• status = Status of placement- Placed/Not placed
• salary = Salary offered by corporate to candidates

Out[4]:

	sl_no	gender	ssc_p	ssc_b	hsc_p	hsc_b	hsc_s	degree_p	degree_t	workex	etest_p	specialisation	mba_p	status	salary
0	1	M	67.00	Others	91.00	Others	Commerce	58.00	Sci&Tech	No	55.0	Mkt&HR	58.80	Placed	270000.0
1	2	M	79.33	Central	78.33	Others	Science	77.48	Sci&Tech	Yes	86.5	Mkt&Fin	66.28	Placed	200000.0
2	3	M	65.00	Central	68.00	Central	Arts	64.00	Comm&Mgmt	No	75.0	Mkt&Fin	57.80	Placed	250000.0
3	4	M	56.00	Central	52.00	Central	Science	52.00	Sci&Tech	No	66.0	Mkt&HR	59.43	Not Placed	NaN
4	5	M	85.80	Central	73.60	Central	Commerce	73.30	Comm&Mgmt	No	96.8	Mkt&Fin	55.50	Placed	425000.0

It appears that the features in our dataset are already clean and have no missing values, except for the salary column because those are available only for candidates who passed the placement.

Out[6]:

	features	data_type	nan_total	nan_pct	unique	values_ex
0	sl_no	int64	0	0.00	215	[48, 149]
1	gender	object	0	0.00	2	[M, F]
2	ssc_p	float64	0	0.00	103	[43.0, 69.6]
3	ssc_b	object	0	0.00	2	[Central, Others]
4	hsc_p	float64	0	0.00	97	[74.0, 71.98]
5	hsc_b	object	0	0.00	2	[Central, Others]
6	hsc_s	object	0	0.00	3	[Arts, Science, Commerce]
7	degree_p	float64	0	0.00	89	[85.0, 56.2]
8	degree_t	object	0	0.00	3	[Comm&Mgmt, Others, Sci&Tech]
9	workex	object	0	0.00	2	[No, Yes]
10	etest_p	float64	0	0.00	100	[71.2, 88.0]
11	specialisation	object	0	0.00	2	[Mkt&HR, Mkt&Fin]
12	mba_p	float64	0	0.00	205	[56.11, 64.27]
13	status	object	0	0.00	2	[Placed, Not Placed]
14	salary	float64	67	31.16	45	[216000.0, 255000.0]

2. Candidates Profiling¶

2.1 How many of the candidates passed the job placement?¶

Most of the candidates got placed. Almost 70% of the candidates were offered the job.

2.2 How was the difference in job placement based on gender?¶

There are more male candidates compared with female candidates. Male candidates made up more than 60% of total candidates.

More females were not getting placed compared with their male counterparts. As much as 36.8% of the female candidates were turned down from the job while only 28% of male candidates were turned down.

Out[60]:

status	% Not Placed	% Placed
gender
Female	36.842105	63.157895
Male	28.057554	71.942446

3. Candidate Education History during Secondary and Higher Secondary School¶

3.1 How did the candidates' performance in secondary and higher secondary school affect their placement?¶

Candidates who passed the job placement were reported as having higher performance during their secondary education on the 10th grade. Their average score was 71.7%, far higher than those who did not pass who had average score of only 57.5%.

Out[47]:

	ssc_p
status
Not Placed	57.544030
Placed	71.721486

Candidates who passed the job recruitment also had better higher secondary education performance, as shown by their average score of 69.9%, more than 10 points higher than those who did not pass.

Out[48]:

	hsc_p
status
Not Placed	58.395522
Placed	69.926554

It could be concluded that high performance during secondary and higher secondary education are an indicator of whether the candidate will pass for the job.

3.2 Did the secondary and higher secondary school board affect candidates' placement?¶

There are several school boards in India, each has different education focus thus might have differences in curriculum. For example, the most popular school board, Central Board of Secondary Education (CBSE), focuses on preparing students for Science and Math related subjects$^{1}$. Other boards put different emphasis on science, arts and language subjects. This dataset groups the school board into two: (1) Central or CBSE, and (2) Others which are school boards other than CBSE.

From the chart below, most of the candidates came from secondary and higher secondary education under CBSE. However, it seems that there the type of school board the students attended for their secondary and higher secondary school, since the ratio of students who got the job were similar for both categories, 67.2% for candidates with Central Board education and 70.7% for candidates not from Central Board education. We could do hypothesis testing to compare the passed candidate proportion difference between the Central and Others group.

Out[69]:

status	% Not Placed	% Placed
ssc_b
Central	32.758621	67.241379
Others	29.292929	70.707071

Out[76]:

status	% Not Placed	% Placed
hsc_b
Central	32.142857	67.857143
Others	30.534351	69.465649

Hypothesis testing

H0: $p_{passed, Central} = p_{passed, Others}$

(proportion of candidates who passed from Central board is the same with proportion of candidates who passed from Other boards)

H1: $p_{passed, Central} \neq p_{passed, Others}$

(proportion of candidates who passed from Central board is not the same with proportion of candidates who passed from Other boards)

import scipy.stats.distributions as dist

def two_proportions(column, sample1, sample2, value='Placed'):

    # Sample sizes
    n1 = len(df_train[df_train[column] == sample1])
    n2 = len(df_train[df_train[column] == sample2])

    y1 = len(df_train[(df_train[column] == sample1) & (df_train.status == value)])
    y2 = len(df_train[(df_train[column] == sample2) & (df_train.status == value)])

    # Estimates of the population proportions
    p1 = round(y1 / n1, 2)
    p2 = round(y2 / n2, 2)

    # Estimate of the combined population proportion
    phat = (y1 + y2) / (n1 + n2)

    # Estimate of the variance of the combined population proportion
    va = phat * (1 - phat)

    # Estimate of the standard error of the combined population proportion
    # standard error of estimate
    se = np.sqrt(va * (1 / n1 + 1 / n2))

    # Test statistic and its p-value
    test_stat = (p1 - p2) / se
    pvalue = 2*dist.norm.cdf(-np.abs(test_stat)) # two-tailed test!!

    # Print the test statistic its p-value
    print('alpha : 0.05')

    print("\nTest Statistic")
    print(round(test_stat, 2))

    print("\nP-Value")
    print(round(pvalue, 2))

Hypothesis testing on ssc_b (Secondary education school board):

two_proportions(column='ssc_b', sample1='Central', sample2='Others', value='Placed')

alpha : 0.05

Test Statistic
-0.63

P-Value
0.53

There is not enough evidence to reject H0, which means the proportion of candidates who passed from Central board is the same with proportion of candidates who passed from Others.

Hypothesis testing on hsc_b (Higher secondary education school board)

two_proportions(column='hsc_b', sample1='Central', sample2='Others', value='Placed')

alpha : 0.05

Test Statistic
-0.15

P-Value
0.88

There is not enough evidence to reject H0, which means the proportion of candidates who passed from Central board is the same with proportion of candidates who passed from Others.

3.3 Did the placement favors a certain specialization of candidates in high secondary school?¶

Most of the candidates had Commerce as their specialization in high secondary school, followed by Science, and Arts. Commerce and Science had the same proportions of candidates who passed the placement, which was around 69%.

Out[93]:

status	% Not Placed	% Placed
hsc_s
Arts	45.454545	54.545455
Commerce	30.088496	69.911504
Science	30.769231	69.230769

4. Candidate Education History during Undergraduate School¶

4.1 How did the candidates' performance in undergraduate affect their placement?¶

Candidates who passed the job placement were reported as having higher performance during their undergraduate study. Their average score was 68.7%, slightly higher than those who did not pass who had average score of only 61.1%.

Out[96]:

	degree_p
status
Not Placed	61.134179
Placed	68.740541

4.2 What was the most popular degree among the candidates and the candidates who passed the placement?¶

This dataset lists only three types of undergraduate degree, which are Science & Technology, Communication & Management, and others. As expected, most of the candidates had undergraduate degree in Communication & Management, followed by Science & Technology.

Candidates with undergraduate degree in Comm&Mgmt had similar chance of getting placed (70.3%) compared with those having degree in Sci&Tech (69.5%).

Out[99]:

status	% Not Placed	% Placed
degree_t
Comm&Mgmt	29.655172	70.344828
Others	54.545455	45.454545
Sci&Tech	30.508475	69.491525

5. Candidate Education and Work History during Post-Graduation and MBA School¶

5.1 How many of the candidates had work experience?¶

Most of the candidates did not have any work experience, but it is clear that having work experience would favor them in getting the job. About 86.5% of the candidates with work experience were accepted, a proportion far higher than those without work experience.

Out[111]:

status	% Not Placed	% Placed
workex
Experienced	13.513514	86.486486
Not Experienced	40.425532	59.574468

5.2 Did the candidates performance in MBA affect the placement?¶

Surprisingly, there was no difference in performance during MBA school between those who passed and those who did not.

5.3 What were the specializations taken by candidates during MBA?¶

There are two specialization types in this dataset, (1) Marketing & Finance, and (2) Marketing & Human Resources. Most of the candidates came from Mkt&Fin specialization. A higher portion of Mkt&Fin candidates passed the placement (79.2%), while only 55.8% of candidates from Mkt&HR passed. This might suggests that the job application is for a role related to Marketing & Finance field.

Out[114]:

status	% Not Placed	% Placed
specialisation
Mkt&Fin	20.833333	79.166667
Mkt&HR	44.210526	55.789474

6. Candidate Performance in Placement Test¶

The chart below shows that there was no difference of test placement score between candidates who passed and not. This suggests that the placement test played small role in deciding whether a candidate qualify for the job or not.

Out[115]:

	features	data_type	nan_total	nan_pct	unique	values_ex
0	sl_no	int64	0	0.00	215	[55, 174]
1	gender	object	0	0.00	2	[Female, Male]
2	ssc_p	float64	0	0.00	103	[48.0, 61.08]
3	ssc_b	object	0	0.00	2	[Central, Others]
4	hsc_p	float64	0	0.00	97	[67.2, 72.8]
5	hsc_b	object	0	0.00	2	[Others, Central]
6	hsc_s	object	0	0.00	3	[Commerce, Arts, Science]
7	degree_p	float64	0	0.00	89	[63.35, 54.38]
8	degree_t	object	0	0.00	3	[Sci&Tech, Others, Comm&Mgmt]
9	workex	object	0	0.00	2	[Not Experienced, Experienced]
10	etest_p	float64	0	0.00	100	[74.28, 93.4]
11	specialisation	object	0	0.00	2	[Mkt&HR, Mkt&Fin]
12	mba_p	float64	0	0.00	205	[64.34, 66.06]
13	status	object	0	0.00	2	[Not Placed, Placed]
14	salary	float64	67	31.16	45	[260000.0, 393000.0]

7. Conclusions and Suggestions¶

By far, the most obvious feature that decides candidate's placement is workex or whether a candidate has a work experience or not. On the other hand, this is only based on comparing candidate pass rate between the categorical features, but we have not compared the numerical features in more detail.

The next step for analysis is to include the categorical and numerical features and group them into three: (1) Candidate secondary & higher secondary education history, (2) Candidate undergraduate history, and (3) Candidate post-graduate and MBA history.