Section F - Group 6
Section F - Group 6
Group-6
Note: Entire team discusssed and collobarated to reach the final conclusion
Contributions
Defined and created new fields from existing data-set, Created pivot tables for Ethnicity and Education by Ethnicity, Worked o
Created Box plots for Limit, Income, Rating and Credit usage
Created Summary statistics for Limit, Income, Rating and Credit usage
Created Histograms for Limit, Income, Rating and Credit usage
Created pivot tables and bar charts for Marital Status by Gender and Rating by Gender, Worked on formatting
Created Pie-charts and Bar charts for Ethnicity and Education by Ethnicity
Count Ethinicity
199 Caucasian
102 Asian
99 African American
Income Limit Rating
Male
48.25%
Female
51.75%
This chart isn't available in your version of Excel. This chart isn't available in your v
Editing this shape or saving this workbook into a different file format will Editing this shape or saving this w
permanently break the chart. permanently break the chart.
Editing this shape or saving this workbook into a different file format will 12
permanently break the chart.
10
0
African American
This chart isn't available in your
Credit Balance
12 Editing this shape or saving this
permanently break the chart.
10
6 Married
0
Male Female Total
Result
This chart isn't available in your version of Excel. This chart isn't available in your
Editing this shape or saving this workbook into a different file format will Editing this shape or saving this
permanently break the chart. permanently break the chart.
12
Ethnic Diversity
10
African 8
American
24.75%
Caucasian 6
49.75%
Asian 4
25.50%
2
0
African American Asian Caucasian To
rt isn't available in your version of Excel. This chart isn't available in your version of E
his shape or saving this workbook into a different file format will Editing this shape or saving this workbook i
ntly break the chart. permanently break the chart.
Total Total
12
10
8
Age Group
6
0
merican Asian African American Asian Cau
rt isn't available in your version of Excel. This chart isn't available in your version of E
his shape or saving this workbook into a different file format will Editing this shape or saving this workbook i
ntly break the chart. permanently break the chart.
this shape or saving this workbook into a different file format will
ntly break the chart.
Credit Usage (CU) 72.31%
73.33%
74.36%
74.62%
75.04%
75.39%
75.40%
76.43%
Total
Gender
2. A majority of the sample is Caucasian (49.75%), followed by 25.5% Asians and 24.75% African
Americans.
3. Using the 'balance' and the 'limit' from the given data, we have defined a new quantitative variable - 'Cr
Usage'. On an average, married males have a higher credit usage than unmarried males, while it is vice ver
for females, where unmarried users have a higher credit usage.
4. On an average, males have a higher credit rating than females in Caucasian and African American group
while females have a higher credit rating in Asain group.
5. We defined 4 education brackets, starting from Less than 10 years of education, to more than 20 years of
education, with an interval of 5. In these, people falling in the bracket of 10-15 years have a higher credit
usage on an average.
6. We defined the 7 age brackets, starting with 20-30, till 80-90. In these, people falling in the 70-80 age
bracket had the highest credit usage on an average.
7. The mean income across the sample is 45.22, with a standard deviation of 35.24. A skewness of 1.74
suggests that the income is positively skewed. This is validated by the Box and Whiskers plot, which also
shows some outliers.
8. The mean rating across the sample is 354.94, with a standard devaition of 154.72. A skewness of 0.87
suggests that the ratings are positively skewed. This is validated by the Box and Whiskers plot, which also
shows some outliers.
9. The mean limit across the sample is 4735.6, with a standard deviation of 2308.2. A skewness of 0.84
suggests that the data is positively skewed. This is validated by the Box and Whiskers plot, which also sho
some outliers.
10. The average credit usage in the sample is 91.04%. A skewness of -0.13 suggests that the data is negativ
skewed. This is validated by the Box and Whiskers plot. Unlike other quantitative variables, credit usage
doesn't have any outliers.
11. Using all the above points, we can conclude that "Credit usage" data point can be a good input to predi
default rate among different groups. We can futher work on above conclusions and find out the reasons for
high/low credit usage using techniques like surveys and questionnaires.
d 24.75% African
. A skewness of 1.74
hiskers plot, which also
2. A skewness of 0.87
hiskers plot, which also
. A skewness of 0.84
kers plot, which also shows