The Essay Explores The Relationship Between Credit Card Charges, Income, And Household Size. (70 Characters)

Descriptive Statistics for Income ($000s) and Amount Charged ($)

The analysis investigates the relation of credit card charges with Income and house hold size of 50 consumers. Table 1 presents the descriptive statistics of the Income ($000s) and amount charged on the credit card ($) of the client.

Table 1: Descriptive statistics for Income ($000s) and Amount Charged ($)

The average (standard deviation) income ($000s) of the consumers is $43.48 ($15). The analysis shows that the minimum and maximum income of the consumers is $21 and $67 respectively. Hence the range of income of the consumers is $46. The median income of the consumers is $42. Maximum number of consumers of the organzation has an income of $54 (000s).

The Average (Standard Deviation) of the amount charged by the client on credit cards is $3963.8 ($934). The analysis shows that the minimum and maximum amount charged for credit cards to the consumers is $1864 and $5678 respectively. Thus consumers are charged in the range of $3814 for credit card usage. $4090 is the median value charged on the credit card. Maximum number of consumers of the organization is charged $3890. (000s).

Table 2: Descriptive statistics for Household Size

Table 2 presents the house hold size of the consumers of the client. The analysis shows that most of the consumers (15) have a household size of 2. The maximum household size is 7, 3 consumers have the household size. The least household size is 1, there are 5 consumers which have the household size.

To investigate the relation between credit card charge by the client and income of the consumers a linear regression is carried out.

Table 4: Regression Coefficients

Table 4 presents the regression coefficients for the relation between credit card charge and income of the consumers. The regression equation is represented as

Amount charged ($) = 2204.24 + 40.470*Income ($000s)

Table 3 presents the regression statistics for the relation. The analysis shows that 39.79% of the variability in the amount charged on the credit card can be explained by the independent variable Income.

To investigate the relation between amounts charged on credit card by the client and the household size of the consumer a linear regression is carried out.

Table 6: Regression Coefficients

Table 6 presents the regression coefficients for the relation between credit card charge and household size of the consumers. The regression equation is represented as

Descriptive Statistics for Household Size

Amount charged ($) = 2581.644 + 40.157*Household size

Table 5 presents the regression statistics for the relation. The analysis shows that 56.68% of the variability in the amount charged on the credit card can be explained by the independent variable Household size. 

The comparison of the R² values shows that household size is a better predictor for amount charged on credit card.

To investigate the relation between income and household size of the consumers and amount charged on credit card by the client a multiple linear regression is carried done.

 Table 8: Regression Coefficients

Table 8 presents the regression coefficients for the relation between amount charged on credit card and income and household size of the consumers. The regression equation is represented as

Amount charged ($) = 1305.034 + 33.122*Income ($000s) + 356.340*Household size

Table 7 presents the regression statistics for the relation. The analysis shows that 82.54% of the variability in the amount charged on the credit card can be explained through the independent variables Income ($000s) and Household size.

The regression equation for amount charged on credit card

Amount charged ($) = 1305.034 + 33.122*Income ($000s) + 356.340*Household size

Thus the amount that would be charged on a consumer whose household size is 3 as well as has an income of $40,000 is

Amount charged ($) = 1305.03 + 33.122*40 + 356.340*3

= 3698.93 ≈ 3699

Independent variables in a regression equation are helpful for predicting the variability in the dependent variable. The higher the value of R² the better the model. Thus other independent variables which may help in developing a better model are

1. Maximum amount that the consumer spends with credit cards
2. The frequency of use of credit card by the consumer.

The variable Student ID is an identity variable. Thus the histogram for student ID cannot be created. 

Table 9: Descriptive Statistics

Activity 03

Table 10: Correlation between variables

p-value = 0.05 level of significance

The Karl-pearsons correlation between HI001 Final Exam and HI002 Final Exam is r = 0.049. Thus the correlation can be said to be positive, very weak and linear. In addition the correlation is not statistically significant p-value = 0.630 > 0.05, level of significance.

The correlation between HI001 Final Exam and HI003 Final Exam is r = 0.122. Thus the correlation can be said to be positive, weak and linear. In addition the correlation is not statistically significant p-value = 0.232 > 0.05, level of significance.

The correlation between HI002 Final Exam and HI003 Final Exam is r = 0.116. Thus the correlation can be said to be positive, weak and linear. In addition the correlation is not statistically significant p-value = 0.257 > 0.05, level of significance.

Linear Regression for Income and Credit Card Charges

The correlation between HI001 Assignment 01 and HI001 Assignment 02 is r = 0.659. Thus the correlation can be said to be positive, moderate and linear. In addition the correlation is statistically significant p-value < 0.001.

The correlation between HI002 Assignment 01 and HI002 Assignment 02 is r = 0.549. Thus the correlation can be said to be positive, moderate and linear. In addition the correlation is statistically significant p-value < 0.001.

The correlation between HI003 Assignment 01 and HI003 Assignment 02 is r = 0.520. Thus the correlation can be said to be positive, moderate and linear. In addition the correlation is statistically significant p-value < 0.001.

The correlation between HI001 Final Exam and HI001 Assignment 01 is r = 0.093. Thus the correlation can be said to be positive, very weak and linear. In addition the correlation is statistically not significant p-value =0.364 > 0.05, level of significance.

The correlation between HI002 Final Exam and HI002 Assignment 01 is r = 0.177. Thus the correlation can be said to be positive, weak and linear. In addition the correlation is statistically not significant p-value =0.081 > 0.05, level of significance.

The correlation between HI003 Final Exam and HI003 Assignment 01 is r = 0.197. Thus the correlation can be said to be positive, weak and linear. In addition the correlation is statistically not significant p-value =0.052 > 0.05, level of significance.

The correlation between HI003 Final Exam and HI003 Assignment 02 is r = 0.120. Thus the correlation can be said to be positive, weak and linear. In addition the correlation is statistically not significant p-value =0.239 > 0.05, level of significance.

Table 11: Depression Scores for individuals with reasonably good health

From the state of Florida individuals with reasonable good health had a depression score from 2 to 9. From the state of New York individuals with reasonable good health had a depression score from 4 to 13. From the state of North Carolina individuals with reasonable good health had a depression score from 3 to 12.

Table 12: Depression Scores for individuals with chronic health condition

Individuals with chronic health condition from Florida had a depression score from 9 to 21. Similarly individuals from New York had a depression score range of 9 to 24. Likewise individuals from North Carolina had a depression score in the range of 8 to 19.

Analysis of Variance is carried out to check for the relation between depression scores and geographical location

ANOVA 1 - Individuals with reasonably good health

Null Hypothesis: The average depression scores of individuals from Florida, New York and North Carolina are equal  

Alternate Hypothesis: The average depression scores of individuals from Florida, New York and North Carolina are not equal

Table 13: ANOVA for individuals with reasonably good health

The analysis of variance for individuals with reasonably good health shows that F(2,57) = 5.241. The p-value for the ANOVA = 0.008 < 0.05, level of significance. Thus, we reject the Null hypothesis.

Thus, it can be inferred that depressions scores of individuals with reasonably good health varies across geographical locations.

ANOVA 2 - Individuals with chronic health condition

Null Hypothesis: The average depression scores of individuals from Florida, New York and North Carolina are equal

Alternate Hypothesis: The average depression scores of individuals from Florida, New York and North Carolina are not equal 

Table 14: ANOVA for individuals with chronic health condition

The analysis of variance for individuals with chronic health conditions shows that F(2,57) = 0.714. The p-value for the ANOVA = 0.494 > 0.05, level of significance. Hence we do not reject the Null Hypothesis.

Thus, it can be inferred that depressions scores of individuals with chronic health conditions does not vary across geographical locations.

From the analysis of variance of depressions scores of individuals with good health and chronic health conditions it can be inferred: 

1. The depression scores of individuals with good health do not vary across geographical locations.
2. The depression scores of individuals with chronic health conditions vary across geographical locations.