Harris Bank Case Study

...

Dependent Variable

SALNOW – Current Salary for the employee

Independent Variable

GENDER - Male and Female employees

Independent Variable

BANKEXP – Time spent working at Harris Bank

Independent Variable

AGE - Age in months of the employee

Independent Variable

WORKEXP – Months of work experience

4.2 Relationships between variables

4.2.1 Relationship between Salary and Gender

Table 1 Relationship between Salary and Gender

SALNOW (USD)

GENDER

SALNOW (USD)

1

GENDER

-0.389439162

1

Based on Table 1, there are differences in salary in term of gender. Males and Females employees in Harris Bank have different salaries.

4.2.2 Relationship between Salary and Bank Experiences

[pic 2]

Figure 1 Relationship between Salary and Bank Experiences

Based on Figure 1, there is no linear relationship observed between salary and bank experience. Bank Harris employees’ salaries are not determined by the time they spent working in Harris Bank.

4.2.3 Relationship between Salary and Age

[pic 3]

Figure 2 Relationship between Salary and Age

Based on Figure 2, there is no linear relationship observed between salary and age. Bank Harris employees’ salaries are not determined by their age.

4.2.4 Relationship between Salary and Education

[pic 4]

Figure 3 Relationship between Salary and Education

Based on Figure 3, there is a positive linear relationship observed between salary and years in school (EDUC). The salaries are Harris Bank employees are determined by the years they spent in school.

4.2.5 Relationship between Salary and Work Experience

[pic 5]

Figure 4 Relationship between Salary and Work Experience

Based on Figure 4, there is no linear relationship observed between salary and work experience. Bank Harris employees’ salaries are not determined by their months of work experience.

4.4 Multiple Regression Model

Upon identifying the relationships of the variables, a model is constructed as below.

Y = β0 + β1 X1+ β2 X2+ β3 X3+ β4 X4+ β5 X5 + [pic 6]

SALNOW = β0 + β1 (GENDER) + β2 (BANKEXP) + β3 (AGE) + β4 (EDUC) + β5 (WORKEXP) + [pic 7]

4.5 Multiple Regression Equation

A multiple regression equation is constructed based on the model in 4.4.

= b0 + b1x1 + b2x2 + b3x3 + b4x4 + b5x5[pic 8]

= b0 + b1 (GENDER) + b2 (BANKEXP) + b3 (AGE) + b4 (EDUC) + b5 (WORKEXP)[pic 9]

4.6 Multiple Regression Analysis Result and Elaboration

[pic 10]

Figure 5 Multiple Regression Testing (T = 80)

The correlation of r = 0.7050 suggested that the overall model is generally showing moderately strong positive linear relationship for gender, bank experience, age, years in school and work experience.

Coefficient of Determination (R Square) shows that only 49.52% of the variation of gender and education can explain the salary, and 50.48% is affected by other variation which are not analyzed part of this model.

The Significance F Test in Figure 5 is less than 0.05. Hence, on average there is at least one independent variable which is significant in the model.

4.7 Two-Tailed Hypothesis Test

The Hypothesis two-tailed test is conducted as the Significance F Test did not tell the significance of the respective independent variables. We perform the hypothesis test on the slope parameters, at 5% level of significance. For , reject if the p-value 0.05.[pic 11][pic 12][pic 13]

[pic 14]

p-value = 0.004011, so we reject at 5% level of significance. [pic 15][pic 16]

Therefore, GENDER is a significant variable in the model. Gender affects salary.

[pic 17]

p-value > 0.6450, so we do not reject at 5% level of significance. [pic 18][pic 19]

Therefore, BANKEXP (months) is not a significant variable in the model. Bank experiences do not affect salary.

[pic 20]

p-value > 0.9540, so we do not reject at 5% level of significance. [pic 21][pic 22]

Therefore, AGE (months) is not a significant variable in the model. Age does not affect salary.

[pic 23]

p-value > 0.00000000519732, so we do not reject at 5% level of significance. [pic 24][pic 25]

Therefore, EDUC (Years in School) is a significant variable in the model. Education affect salary.

[pic 26]

p-value > 0.858477012, so we reject at 5% level of significance. [pic 27][pic 28]

Therefore,