Campus Placement Analysis for Jain University
0.1 Introduction
0.1.1 Overview
The dataset consists of campus placement data for MBA Business Analytics students from Jain University in Bangalore, India. There are 215 observations and 15 variables containing the following information:
- placement status of the student
- salary offered to students who were placed
- MBA specialization and grade
- record of previous employment
- specialization and grade for secondary and post-secondary degrees
It should be noted that numerous factors beyond the scope allowed by data collection can influence why a student/canadidate is or is not hired. The variables inspected here do not account for the importance of interview skills and human interaction during a job interview. We also do not have any infomation on the employment test set up by the college, the list of employers and type of speicialization or academic background they were looking for in a candidate.
0.1.2 List of Variables
sl_no: Serial Number of the observations for each studentgender: Male = ‘M’; Female = ‘F’ssc_p: Secondary Education percentage (10th grade)ssc_b: Board of Education for Secondary (Central / Others)hsc_p: Higher Secondary Education percentage (12th grade)hsc_b: Board of Education for Higher Secondary (Central / Others)hsc_s: Specialization in Higher Secondary Education (Commerce / Science / Other)degree_p: Degree Percentagedegree_t: Under Graduation degree type (Field of education)workex: Record of previous work experience (Yes / No)etest_p: Employability test percentage (conducted by college)specialisation: Post Graduation (MBA) specialization (Mkt&Fn / Mkt &HR)mba_p: MBA percentagestatus: Status of placement (Placed / Not Placed)salary: Annual salary offered by corporate to candidates who were placed
0.2 Exploratory Data Analysis
The exploratory data analysis is divided into two sections. First, the relationship between status and all the relevant dependent variables, excluding sl_no and salary, are examined. In the second section, the influence on salary is inspected.
0.2.1 Influence on Status
0.2.1.1 What is the placement rate?
Figure 1: Bar plot showing campus placement rate for students.
prop.table(table(placement$status)) * 100##
## Not Placed Placed
## 31.16279 68.83721
Remark: Approximately 69% students were hired and 31% were not.
0.2.1.2 What is the status by work experience?
Hypothesis is that candidates with previous work experience are likely to get hired.
Figure 2: Bar plots showing placement status for students based on their work experience.
Remark: The bar plot on the left shows that more students did not have work experience. The proportion bar plot on the right shows that almost 60% of students without work experience were placed, whereas the placement rate for students with work experience is almost 90%. This result matches with the hypothesis.
0.2.1.3 What is the status by employment test score?
The employment test was conducted by the college. Hypothesis is that this test score positively influences the placement decision.
Figure 3: Histogram, density, and box plots to illustrate the relationship between placement status and employment test score.
Remark: The histogram and density plots for employment test scores with respect to status show that, in general, for higher test scores, a greater percentage of students were placed than not placed.
The boxplot shows the test score for students who were not placed in the interquartile range 60-76 (median = 67), and for those placed in the interquartile range 60-85 (median = 72). Therefore, it is certain that students with high employment test scores were preferred for placement, but in regards to those with scores 60-76 who were also placed, there must be other factors that influenced it.
0.2.1.4 How does the status depend on both employement test score and work experience?
Figure 4: Histogram, density, and box plots to illustrate the relationship between placement status, work experience, and employment test score.
Remark: From the histogram plots it is clear that a very small percentage of students with previous work experience were not placed, and more among these students had low employment test score than high. However, for the students who were placed, etest_p scores does not seem to have any effect on the placement decision.
In the box plot, we see that among the students who were not placed, the median etest score for those with work exprience were lower than those without. Among the students who were placed, we see the opposite, but the interquartile range of test scores for these students almost overlap. Comparing the status and employment test scores for students without work experience, we see that the group of students who were hired had a higher median score.
0.2.1.5 What is the effect of MBA test score on placement status?
Hypothesis is that more students with good MBA test score and work experience were hired.
Figure 5: Histogram, density, and box plots to illustrate the relationship between placement status and MBA test score.
Remark: First, looking at the box plots we see that the median MBA test scores for students with work experience are higher.
From the histogram, density, and box plots we see that regardless of the placement status, the MBA test score distribution for students without work experience is skewed right in comparison to test scores of those students with work experience. The median and third quartile scores are the highest for those students with work experience who were hired.
Comparing the width of the box plots for the Not Placed category we see that the number of students with work experience who were not hired is a lot smaller.
To summarize, the hypothesis holds only for the group with work experience and test scores above 65.
0.2.1.6 Did the MBA specialization play a role in the hiring process?
Figure 6: Bar plots to inspect the relationship between placement status and MBA specialization.
Remark: We see that more students specialized in Mkt&Fin than in Mkt&HR. In addition, we see that more candidates with Mkt&Fin specialization were hired. Based on the proportion bar plot, the ratio of receiving placement for the two MBA streams is ~80% : ~55%.
0.2.1.7 What is the effect of undergraduate degree stream?
Figure 7: Bar plots to inspect the relationship between placement status and undergraduate degree stream.
Remark: The first bar plot shows that there are a lot more students in Comm&Mgmt than in Sci&Tech and other streams combined. But from the proportion bar plot We see that an equal proportion of students from Comm&Mgmt and Sci&Tech were hired, and less than 50% of students from other streams were hired.
0.2.1.8 How does both undergraduate degree stream and degree test score reflect on status?
Figure 8: Histogram, density, and box plots to illustrate the relationship between placement status, undergraduate degree stream, and degree test score.
Remark: Here, the first three plots show histogram and density plots for degree test scores, where the degree scores are grouped according to degree streams and according to placement status within each stream. The box plot shows a different arrangement where the streams are grouped according to placement status.
According to the histograms and the density plot, across all streams, candidates with higher degree percentage were preferred. In the box plot for students in Comm&Mgmt who were not place, we see two outliers in the test score range of 75-80.
0.2.1.9 How does both higher secondary degree stream and test score influence the placement decision?
Figure 9: Histogram, density, and box plots to illustrate the relationship between placement status, higher secondar degree stream, and the test score.
Remark: Here, the first three plots show histogram and density plots for degree test scores, where the degree scores are grouped according to degree streams and according to placement status within each stream. The box plot shows a different arrangement where the streams are grouped according to placement status.
In the density plot, dividing the placement status according to the science, commerce, and arts streams, we see that students who were placed generally had higher hsc scores than those who were not placed.
From the box plot, comparison of the degree streams based on the placement status shows that the median scores for students who were hired were greater than those who were not. We also notice that for each degree stream, the lower quratile scores for students who were hired is greater than the upper quartile scores of those who were not.
0.2.1.10 Did the secondary test score have an effect?
Figure 10: Histogram, density, and box plots to illustrate the relationship between placement status and secondary school test score.
Remark: For secondary school education the streams were not provided, or they do not exist. Based on the density plot and box plot we see that candidates with higher ssc percentages were preferred for placement. Students who were hired have a significantly high ssc test score and the interquartile range of for the two groups do not overlap. The test scores for the group that was hired is skewed left, and for the group that was not hired the test scores are skewed right.
0.2.1.11 Were students from any particular secondary and higher secondary school board preferred?
Figure 11: Bar plots to show if the higher secondary or secondary school board had any influence in the placement decision.
Remark: There are two secondary and higher secondary school board groups, Central and Other school boards. There does not seem to be any visual evidence to suggest one school board group is preferred than the other.
Thus far, the visual analysis suggests that test scores from all secondary and post-secondary levels were taken into account when deciding which student/candidate to hire.
Now let us explore these test scores with respect to the MBA specializations to inspect if one group fared better than the other in terms of their work experience.
0.2.1.12 Is there a connection between MBA stream and work experience?
Figure 12: Bar plots to illustrate the relationship between MBA stream and work experience.
Remark: The bar graph on the left shows there are more students specializing in Mkt&Fin than in Mkt&HR. From the proportion bar graph on the right we see that more students in Mkt&Fin with work experience than those in the Mkt&HR specialization.
0.2.1.13 What is the relationship between MBA specialization, MBA score, status, and work experience?
Figure 13: Density and box plots to illustrate the relationship between placement status, work experience, and MBA specialization.
Remark: Overall, MBA test score distributions for students without work experience are skewed to the right in comparison to test score distributions for students who have work experience.
The box plots show that students with work experience generally have higher MBA test score. Among the students who were hired from Mkt&FIn, almost equal number of students are with and without work experience. Among the students who were hired form Mkt&HR, there are less students with work experience. From the Not Placed category we see that the percentage of students with work experience is much smaller. We also notice that the median test score, regardless of placement status, work experience, specialization, is in the range of 60-65. The third quartile score mark students with work experience are higher in general.
0.2.1.14 What is the relationship between MBA specialization, employement test score, and work experience?
Figure 14: Histogram, density, and box plots to illustrate the relationship between MBA specialization, employement test score, and work experience.
Remark: Based on the density curve and box plot, we see that employement test scores for students in Mkt&Fin is more uniformly distributed, with a median score of 75. Comapring in terms of work experience, the interquartile test score range for these students are almost the same.
For Mkt&HR students, the test scores are skewed to the right, with a lower median score of 65 and an upper quartile score of 80. This is in contrast with a higher median score of those students without work experience (at 67), although the upper quartile score is much lower at 75.
0.2.1.15 What is the relationship between MBA specialization, employement test score, status, and work experience?
Figure 15: Density and box plots to illustrate the relationship between MBA specialization, employement test score, placement status, and work experience.
Remark: Analyzing the previous plot with respect to placement status we see that for Mkt&Fin students who were hired, the median, lower and upper quartile employment scores are higher irrespective of having work experience. For students from Mkt&HR no such observable trend is seen.
Based on the density curve and box plot, we see that employement test scores for students in Mkt&Fin is more uniformly distributed, with a median score of 75. For Mkt&HR students, the test scores are skewed to the right, with a median score of 67.
In the box plot we see an outlier where a Mkt&HR student with work experience and high employment test score of 86 was not hired. Filtering for that particular criteria we get the following observation.
knitr::kable(placement %>%
filter(etest_p == 86 & status == "Not Placed" & workex == "Yes" & specialisation == "Mkt&HR"))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 110 | M | 52 | Central | 63 | Others | Science | 65 | Sci&Tech | Yes | 86 | Mkt&HR | 56.09 | Not Placed | NA |
Remark: From the table above we see that even though this particular student had a very high employment test score due as well as work experience, their scores for the remaining secondary and post-secondary tests show scores from 52-65 which are quite low. This suggests the possibility that having work experience and high employment test scores are not enough when other test scores are in the lower range.
0.2.1.16 What is the relationship between MBA specialization, undergraduate degree test score, and work experience?
Figure 16: Histogram, density, and box plots to illustrate the relationship between MBA specialization, undergraduate degree test score, and work experience.
Remark: The histogram and density plots show that more students with Mkt&HR had lower degree scores, and more students specializing in Mkt&Fin scored higher. Based on the box plot, median degree test score for Mkt&Fin is 67 and that for Mkt&HR is 65. Even though the median scores for both specializations are almost the same, the scores for Mkt&Fin students are skewed right and those for Mkt&HR are skewed left. Which means, more Mkt&Fin students scored in the second quartile (64 - 67), and more Mkt&HR students scored in the third quartile (65 - 69).
0.2.1.17 What is the relationship between MBA specialization, undergraduate degree test score, placement status, and work experience?
Figure 17: Density and box plots to illustrate the relationship between MBA specialization, undergraduate degree test score, placement status, and work experience.
Remark: From the box plots we see higher median and upper quartile degree test scores for those students who were hired.
Again, we are interstested in the two Mkt&Fin students without work experience and who were not hired with degree test scores that lie outside the upper inner fence (top left segment).
We are also interested in the Mkt&HR student who scored 55 (outside the lower inner fence) in the degree test, did not have work experience, and was hired.
Case 1: Mkt&Fin students without work experience and really high degree test scores who were not hired
knitr::kable(placement %>%
filter(degree_p >=74 & status == "Not Placed" & workex == "No" & specialisation == "Mkt&Fin"))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 7 | F | 46.0 | Others | 49.20 | Others | Commerce | 79 | Comm&Mgmt | No | 74.28 | Mkt&Fin | 53.29 | Not Placed | NA |
| 83 | M | 63.0 | Central | 67.00 | Central | Commerce | 74 | Comm&Mgmt | No | 82.00 | Mkt&Fin | 60.44 | Not Placed | NA |
| 166 | F | 63.3 | Central | 78.33 | Others | Commerce | 74 | Comm&Mgmt | No | 80.00 | Mkt&Fin | 74.56 | Not Placed | NA |
Remark: The table above shows there were three students in Mkt&Fin and without work experience that were not hired whose degree test scores are outside the upper inner fence. We see that MBA and secondar test scores for students with serial numbers 7 and 83 were extremely low. Student with serial number 166 had much better test scores in comparison, therefore, the reason behind why this student was not hired cannot be determined with the existing information.
Case 2: Mkt&HR student without work experience and extremely low degree test score who was hired
knitr::kable(placement %>%
filter(degree_p == 56 & status == "Placed" & workex == "No" & specialisation == "Mkt&HR"))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 173 | M | 73 | Others | 58 | Others | Commerce | 56 | Comm&Mgmt | No | 84 | Mkt&HR | 52.64 | Placed | 300000 |
| 177 | F | 59 | Central | 60 | Others | Commerce | 56 | Comm&Mgmt | No | 55 | Mkt&HR | 57.90 | Placed | 220000 |
Remark: Once again we see two students whose degree test scores were identified by a single outlier. The student with serial number 173 had high secondary and employement test scores but the student with serial number 177 had poor test scores throughout. It is worth noting that salary offered to student# 173 is higher than that offered to student# 177, indicating the possibility that test scores might influence salaries offered to those who are hired. Salary will be inspected in the second part of the exploratory data analysis.
0.2.1.18 What is the relationship between MBA specialization, higher secondary test score, status, and work experience?
Figure 18: Density and box plots to illustrate the relationship between MBA specialization, higher secondary test score, placement status, and work experience.
Remark: From the density and box plots we see that the higher secondary test score range, and therefore median scores, for students who were placed is higher. We also see an outlier where a candidate with a test score of 74.7 and work experience was not hired.
knitr::kable(placement %>%
filter(hsc_p >= 74 & status == "Not Placed" & workex == "Yes" & specialisation == "Mkt&HR"))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 156 | M | 51.57 | Others | 74.66 | Others | Commerce | 59.9 | Comm&Mgmt | Yes | 56.15 | Mkt&HR | 65.99 | Not Placed | NA |
Remark: We see that despite having work experience, and scoring 74.7 (74.66 rounded up) in higher secondary this student from Mkt&HR did not score well in the other tests. Secondary score tests are likely not be relevant if degree, MBA, and employment test scores are very low.
Side note: Let us check the other outlier test score from the box plot to establish a connection with salary as seen in the analysis of employment test score, specialization, work experience, and placement status.
knitr::kable(placement %>%
filter(hsc_p >= 97 & status == "Placed" & workex == "No" & specialisation == "Mkt&Fin"))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 25 | M | 76.5 | Others | 97.7 | Others | Science | 78.86 | Sci&Tech | No | 97.4 | Mkt&Fin | 74.01 | Placed | 360000 |
Remark: We see that this student scored well across all tests but did not have work experience, which could be the reason why they were not offered a higher salary. Salary will be inspected in the second part of the exploratory data analysis.
0.2.1.19 What is the relationship between MBA specialization, secondary test score, status, and work experience?
Figure 19: Density and box plots to illustrate the relationship between MBA specialization, secondary test score, placement status, and work experience.
Remark: From the density and box plots we see that the higher secondary test score range, and therefore median scores, for students who were placed is higher. We also see multiple outlier scores where a student with a test score of 49 was hired and students with test scores in 70-80 were not.
knitr::kable(placement %>%
filter(workex == "Yes") %>%
filter((ssc_p == 49 & status == "Placed" & specialisation == "Mkt&Fin") | (ssc_p == 74 & status == "Not Placed" & specialisation == "Mkt&HR")))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 42 | F | 74 | Others | 63.16 | Others | Commerce | 65 | Comm&Mgmt | Yes | 65 | Mkt&HR | 69.76 | Not Placed | NA |
| 154 | M | 49 | Others | 59.00 | Others | Science | 65 | Sci&Tech | Yes | 86 | Mkt&Fin | 62.48 | Placed | 340000 |
Remark: We see that despite having work experience, and scoring 74.7 (74.66 rounded up) in higher secondary this student from Mkt&HR did not score well in the other tests. Secondary score tests are likely not be relevant if degree, MBA, and employment test scores are very low.
Analyses focussed on MBA specialization and placemenet decision showed that more students specialized in Mkt&Fin than in Mkt&HR, and more students from Mkt&Fin were hired. In general, more students with work experience had better MBA test scores, as a result we found more students from Mkt&FIn had overall higher MBA test scores.
More students from Mkt&FIn also had higher employment test scores, regardless of having work experience. With respect to undergraudate degree test scores, we found that students who were hired had higher median and upper quartile test scores.
There were several cases of outliers when work experience and test scores based on students’ MBA specializations were inspected for placement decision. For example, we found that having work experience and really high employment test score does not gurantee placement if rest of the test scores are low. Sometimes students with really low post-secondary test scores and no work experience were hired, whereas studetns with good test records and work experience were not. In a select few of these cases it appeared that work experience and test scores of students who were hired played a role in the salaries offered.
Next, it is reasonable to ask if gender played any role in the hiring process.
0.2.1.20 Did gender play a role in the hiring process?
Figure 20: Bar plots to illustrate the relationship between gender and placement status.
Remark: The first bar plot on the left shows that there are less female stuents than male. The proportion bar plot on the right shows that around 60% of female students were placed and around 70% of male students were placed. At this point it is not evident if gender played a role in the placement process.
0.2.1.21 What is the work experience according to the gender?
Figure 21: Bar plots to illustrate the relationship between gender and work experience.
Remark: The proportion bar plot on the left shows that there are more male students with work experience than female.
From the proportion bar plot on the right we see that among those students with work experience, around 70% are male and 30% are female. In case of those students without work experience, around 62% are male and 38% are female.
0.2.1.22 Is there a preference for MBA specialization according to gender?
Figure 22: Bar plots to illustrate the relationship between gender and MBA specialization.
Remark: The proportion bar plot on the right shows that more female students specialized in Mkt&HR and more male students specialized in Mkt&Fin.
0.2.1.23 How did students from each gender perform on the MBA test? What is the relationship with respect to placement status?
Figure 23: First sets of density and bar plots show the relationship between gender and MBA test score. The second set of plots show the relationship of these variables with respect to placement status.
Remark: From the first two density and box plots we see that MBA test scores for male students are skewed to the right, and the scores for the female students is skewed slightly to the left (almost symmetrical). The median score for female students is approsimately 65 and that for male students is around 61. The box plot also shows some extremely high scores for male students as outliers from the trend observed. In conclusion, more female students performed better in the MBA test.
The second set of plots show the relationship between gender and MBA test score with respect to the placement status. We see overall the same trend in test scores for each placement decision category. We also notice that the median scores for both genders are roughly the same whether or not they were hired. We also see an outlier where a male student with test above 75 who was not hired.
0.2.1.24 How did students from each gender perform on the employment test? What is the relationship with respect to placement status?
Figure 24: First sets of density and bar plots show the relationship between gender and employment test score. The second set of plots show the relationship of these variables with respect to placement status.
Remark: In the first density plot for employement test scores for both genders we see multimodal distribution of the data. The blue density curve have two peaks, one around 60 and one around 85. The pink density curve have two peaks around 58 and 90, and a small plateau around 70. The prominent peaks for both curves overlap around the same regions, where the peaks around the lower scores (58 and 60) have higher density, and the peaks around the higher scores (85 and 90) have lower density. This corresponds to the possibility that students with work experience performed better at the employment test. The lower density for the higher scores is in agreement with the finding that fewer students have previous work experience. According to the box plot, median score for female students is around 69 and that for male students is around 72. Therefore, it can be concluded that more male students performed better in the employment test.
The second set of density and box plots show the relationship between gender and employment test score with respect to the placement status. We see that regardless of the gender, median test scores of students who were hired is higher than those who were not. Within the group of students who were hired, male students have a higher median score than females. Among those who were not hired, interquartile test score range and median test scores for students from both genders are approximately equal.
0.2.1.25 How did students from each gender perform on the undergraduate degree test? What is the relationship with respect to placement status?
Figure 25: First sets of density and bar plots show the relationship between gender and undergraduate degree test score. The second set of plots show the relationship of these variables with respect to placement status.
Remark: The first set of density and box plots show that the degree test score distribution for male and female students is slightly skewed to the right. Median score for female students is around 67, and median score for male students is 65. It can be concluded that more female students performed better in the degree test.
The second set of density and box plots show the relationship between gender and degree test score with respect to the placement status. We see that regardless of the gender, median test scores of students who were hired is higher than those who were not, and overall female students have higher median test score.
0.2.1.26 How did students from each gender perform on the higher secondary test? What is the relationship with respect to placement status?
Figure 26: First sets of density and bar plots show the relationship between gender and higher secondary test score. The second set of plots show the relationship of these variables with respect to placement status.
Remark: The first set of density plot and box plot show that the higher secondary test score for male and female students is almost identical. For both genders, the distributions are sligtly skewed to the right, the median scores are in the range 62-63, and the interquartile range is approximately 60-67.
The second set of density and box plots show that for both genders interquatile higher secondary test score range for students who were hired is higher than those who were not. We do notice that for students who were not hired the test scores are skewed left, but no high outlier scores in this group.
0.2.1.27 How did students from each gender perform on the secondary test? What is the relationship with respect to placement status?
Figure 27: First sets of density and bar plots show the relationship between gender and secondary test score. The second set of plots show the relationship of these variables with respect to placement status.
Remark: The first set of density plot and box plot show that secondary school test scores for male students are skewed slightly to the right with a median score of 65. The test score for female students is skewed to the left with a median score of 70. It can be concluded that more female students performed better in the secondary school test.
The second set of density and box plots show that with each gender, the interquartile secondary test score range for students who were hired is higher tha those who were not. We do not see any high outlier test scores among those students who were not hired.
Visual analysis of the test scores based on gender shows that for the most part female students performed better in the academic tests. Male students performed better in the employment test because of a larger number of male students than female had work experience.
Evidence suggests that post-secondary test scores and employment test scores are more important than secondary test scores. If a student scores well in the secondary exams and poorly in the post-secondary exams, they will not be hired in general, inspite of having work experience. However, there are exceptions to this which can depend on the interview skills. Background in Scince&Tech or Comm&Mgmt or MBA specialization are likely deciding factors depending on the position for which the candidate is being interviewed.
0.2.2 Influence on Salary
0.2.2.1 What is the numeric range of the salaries offered?
Figure 28: Histogram distribution of salary offered to students who were hired.
Remark: The salary offered ranges from roughly Rs.200,000 to Rs.950,000. And the median salary is around Rs.265,000.
0.2.2.2 How does the salary for hired students vary based on their gender?
Figure 29: Salary distribution based on gender.
Remark: From all three plots we see that the salaries for both genders are skewed right, similar to the overal salary distribution. We see that the lowest salary offered to both genders were Rs.200,000. For female students the median salary is around Rs.250,000 and the interquatile range is Rs.220,000 - Rs.300,000. For male students the median salary is Rs.275,000 and the interquartile range is Rs.225,000 - Rs.300,000. In the box plot we see outliers in the form of salary offered to a single female student at Rs.650,000. For male students, we see an extremely large number of salary beyond fourth quartile in the range from Rs.375,000 - Rs.925,000.
0.2.2.3 How does the salary for hired students vary based on their MBA specialization?
Figure 30: Salary distribution based on MBA specialization.
Remark: Based on the overlayed histogram and box plot we see that the median salary offered to students from Mkt&Fin are higher than that offered to students from Mkt&HR. We also see that really high salaries in the range Rs.500,000 - Rs.950,000 were offered to students in Mkt&Fin. Highest salary offered to students placed in Mkt&HR is approximately Rs.460,000.
0.2.2.4 Did students with work experience receive higher salary?
Figure 31: Salary distribution based on work experience.
Remark: From the density plot we see that a similar proportion of stduents with and without work experience received salaries between Rs.200,000 - Rs.450,000. However, salaries in the range of Rs.600,000 - Rs.800,000 were exclusively offered to those with work experience.
Note that when we compare this density plot with that from comparison of salary based on MBA specialization (remark for salary vs specialization), we see that the students in the higher salary range are excluseively from Mkt&Fin.
Upon comparison with the box plot for salary vs. gender (remark for salary vs. gender), we see that both male and female students belong to that salary bracket.
To summarize: The following plots visually summarize the observations on salary based on work experience, MBA specialization, and gender
Figure 32: Salary based on work experience, MBA specialization, and gender.
0.2.2.5 Did the undergraduate degree stream influence the salary offered?
Figure 33: Salary based on undergraduate degree stream.
Remark: Sci&Tech and Comm&Mgmt students received highest salaries. For all three streams the salary distributions are skewed to the right. Salaries above Rs.400,000 went to Sci&Tech and Comm&Mgmt students only, with the latter group receiving a larger number of salaries beyond the fourth quartile. We see that only three students received salaries over Rs.600,000.
0.2.2.6 Did the salary depend on the MBA test score? How did work experience influence it?
Figure 34: Salary based on MBA test score and speicializaiton, and work experience.
Remark: From the first scatter plot we see that regardless of the MBA test score, only students with work experience were offered salaries in the extremely high range. This implies that other factors are likely to have influenced this decision.
The second set of scatter plots show that all three students with work experience who received salary higher than Rs.600,000 are from Mkt&Fin, and one student is female and two are male.
0.2.2.7 Did the salary depend on the employment test score? How did work experience influence it?
Figure 35: Salary based on employment test score, MBA speicializaiton, and work experience.
Remark: The scatter plots show that the highest salaries were offered only to students who had work experience regardless of their empployment test score, especially for student # 151. This implies other factors are likely to have influenced this decision.
0.2.2.8 Did the salary depend on the undergraduate degree test score? How did work experience influence it?
Figure 36: Salary based on undergraduate degree test score, MBA speicializaiton, and work experience.
Remark: The scatter plot on the left shows for salaries less than Rs.500,000 an even distribution of salaries between students with and without work experience. Salaries larger than Rs.500,000 were offered specifically to students with work experience, regardless of their degree test score. It is reasonable to suggest that other factors have influenced this decision.
0.2.2.9 Did the salary depend on the higher secondary test score? How did work experience influence it?
Figure 37: Salary based on higher secondary test score, MBA speicializaiton, and work experience.
Remark: The scatter plot on the left shows for salaries less than Rs.500,000 an even distribution of salaries between students with and without work experience. Salaries larger than Rs.500,000 were offered specifically to students with work experience, regardless of their higher secondary test score. It is reasonable to suggest that other factors have influenced this decision.
0.2.2.10 Did the salary depend on the secondary test score? How did work experience influence it?
Figure 38: Salary based on secondary test score, MBA speicializaiton, and work experience.
Remark: The scatter plot on the left shows for salaries less than Rs.500,000 an even distribution of salaries between students with and without work experience. Salaries larger than Rs.500,000 were offered specifically to students with work experience, regardless of their secondary test score. It is reasonable to suggest that other factors have influenced this decision.
Thus far, we have seen one female and two male students specializing in Mkt&Fin with work expeirence, were offered salaries higher than Rs.600,000, irrespective of their secondary and post-secondary test scores and undergraduate degree stream. The following table shows details on these three students.
knitr::kable(placed %>%
filter(salary > 600000))| 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 120 | M | 60.8 | Central | 68.40 | Central | Commerce | 64.6 | Comm&Mgmt | Yes | 82.66 | Mkt&Fin | 64.34 | Placed | 940000 |
| 151 | M | 71.0 | Central | 58.66 | Central | Science | 58.0 | Sci&Tech | Yes | 56.00 | Mkt&Fin | 61.30 | Placed | 690000 |
| 178 | F | 73.0 | Central | 97.00 | Others | Commerce | 79.0 | Comm&Mgmt | Yes | 89.00 | Mkt&Fin | 70.81 | Placed | 650000 |
Remark: The table above lists the three students who were offered salary greater than Rs.600,000. We see that all three students have previous work experience and specialization in Mkt&Fin. Two of them are male and one female. One male and one female student are from Comm&Mgmt and the other male student is from Sci&Tech. The test scores did not seem to have speicifically influenced the salaries as we see test scores in the approximate range of 50-80.
0.3 Conclusion
The exploratory data analysis conducted on the Campus Placement data set was done to understand relationships between the variables related to the students’ academic and professional background, and whether these variables could help in creating a model. The data set contains variables that are both categorical and numeric, and these were inspected using histogram, bar, density, box-and-whiskers, and scatter plots.
We saw that more sudents specialized in Mkt&Fin than in Mkt&HR, and more students from Mkt&Fin were hired. This could potentially be due to the overall test scores or the demand for one specialization over another. Students with work experience had higher MBA test scores, and regardless of work experience, Mkt&Fin students had higher employment test scores than those in Mkt&HR. Outlier score analysis suggested that having work experience and high employment test score does not gurantee placement if rest of the exam test scores are low. Students who were hired had higher median and upper quartile mark for undergraduate degree test scores.
There were several cases of outliers when work experience and test scores were examined together with respect to MBA specialization. In these cases sometimes students with really low post-secondary test scores and no work experience were hired, whereas, students with good test record and work experience were not. In a select few of these cases it appeared that work experience and test scores of students who were hired played a role in the salaries offered.
Relationship between placement status and gender of students was inspected as well. There are more male students than female in the MBA program, among whom 70% of male students and 60% of female students were hired. There are more male students with work experience than female (M:70%, F:30%). More female students specialized in Mkt&HR and more male students specialized in Mkt&Fin. Overall, there was no visual evidence suggesting preference towards one gender over another.
Analyzing the salary offered to those students who were hired showed the salaries to be in the range of Rs.200,000 - Rs.950,000, where the median salary is Rs.265,000. The distribution of the salary is skewed right, and the same trend is true for either gender, where the median salary for male students is slighly higher than that of female students. Students who specialized in Mkt&Fin, who are mostly male and have work experience, received much higher salaries. These students are also exclusively from Comm&Mgmt and Sci&Tech streams. Salaries offered to students in other degree and higher secondary streams are much lower and skewed right. With respect to all secondary and post-secondary test scores, there does not seem to be a distinct relationship with the salary offered.