Academic Success Report

Jeekeshen Chinnappen

0308104/110107264

Academic Success Report

By Jeekeshen Chinnappen Taylor’s ID: 0308104 UniSA ID: 110107264

1

Jeekeshen Chinnappen

0308104/110107264

Contents Executive Summary....................................................................................................................................... 3 Introduction .................................................................................................................................................. 4 Hypothesis..................................................................................................................................................... 5 Methodologies .............................................................................................................................................. 6 Results & Discussion ..................................................................................................................................... 7 Limitations of Analysis ................................................................................................................................ 13 Conclusion ................................................................................................................................................... 14 Appendices.................................................................................................................................................. 15 References .................................................................................................................................................. 17

2

Jeekeshen Chinnappen

0308104/110107264

Executive Summary

Objectives 

To determine the whether people with higher education levels have higher wages



To estimate the difference between people with postgraduate/undergraduate degrees and those with the lowest level of education.



To determine if there is an increasing returns to education.

Results After in-depth analysis, we concluded that people with higher levels of education definitely have a greater hourly wage for both genders. Postgraduate/Undergraduate degrees get 6.0 times more than those with the lowest level of education for males and 10.5 times more for females and it has been proved quantitatively that there is an increasing return to education.

3

Jeekeshen Chinnappen

0308104/110107264

Introduction Research Questions: The following are the research questions we would like to investigate on: 

“Is there an empirical evidence to support the assertion that people who have higher education levels have higher wages?”



“What is the wage difference between people with postgraduate/Undergraduate degrees and those with the lowest level of education?”



“Are there increasing returns to education?”

4

Jeekeshen Chinnappen

0308104/110107264

Hypothesis Human capital theory suggests that the level of education of people is directly related to their working skills and capability to profit in the workforce, meaning that people with higher education gets higher wages. (Becker, 1964; Schultz, 1975). According to Pew Research (2014), the wage difference between degrees and those without any degrees is increasing. Employees with bachelor degrees earned $45,000 annually compared to High School graduates who makes only $28,000. As per the Australia National University (2006), there is an increasing return to education assuming a competitive economy.

5

Jeekeshen Chinnappen

0308104/110107264

Methodologies 

Data Analysis

Analysis of the data to estimate the ratio of Males to Females, White to Nonwhites and Married to Single to get a look and feel of the data. 

Histogram

Histograms have been used to determine skewness of data for average wages by gender. 

Correlation

Correlation coefficients have been used to determine any possibility for multicollinearity that might affect the sample of data, noting that correlation will only lie between -1 and 1. 

Simple Regression

Simple regression has been used to identify the relationships between the dependent variable and independent variable with aim to estimate the slope and intercept for each line. 

Multiple Regression

Multiple regression has been used to identify multiple relationship between one dependent variable onto many independent variables which will help to analyze which of the explanatory variables are more crucial.

6

Jeekeshen Chinnappen

0308104/110107264

Results & Discussion Basic Data Analysis As per the data displayed below we can say that our data is comprised of 49.41% Males and 50.59% Females where 48.80% are married and 51.20% are single.

Histogram As per the histogram for Males below we can conclude that the ‘Number of Males’ to ‘Average wage’ is positively skewed.

As per the histogram for Females below we can conclude that the ‘Number of Females’ to ‘Average wage’ is positively skewed.

7

Jeekeshen Chinnappen

0308104/110107264

Correlation There is positive correlation between average wages for females and their level of education. We can determine from here that females with higher level of education tend to have a higher hourly wage. The Average wage for males is also positively correlated in terms of education but less than that of females.

Simple Regression

When we regress Y = Average Wage on X = Education, the following regression model is obtained: Y = 1.145X -5.430 where β = 1.145 and α = -5.430. The column labelled “Coefficient” represents the OLS (Ordinary Least Square) estimate, β = 1.145, indicating that increasing level of education by one grade is associated with an increase of 1.145% in average wage for females. The “Lower 95%” and “Upper 95%” convey the lower and upper bounds of the 95% confidence interval for β is [1.00002, 1.29000]. Thus we are 95% confident that the marginal effect of education level on average wage for Females lies between 1.00002% and 1.29000%.

8

Jeekeshen Chinnappen

0308104/110107264

LNY = Average Wage, X = Education LNY = 0.1556X -0.1457 β = 0.1556 indicates that increasing level of education by one grade is associated with an increase of 0.1556% in Average Wage for females. We are 95% confident that the marginal effect of education level on average wage for Females lies between 0.1348% and 0.1764%.

9

Jeekeshen Chinnappen

0308104/110107264

Y = Average Wage, X = Education Y = 1.1449X -2.6217 β = 1.1449 indicates that increasing level of education by one grade is associated with an increase of 1.1449% in Average Wage for males. We are 95% confident that the marginal effect of education level on LN average wage for Females lies between 0.8018% and 1.4879%.

Y = LN Average Wage, X = Education LNY = 0.1387X + 0.3567 β = 0.1387 indicates that increasing level of education by one grade is associated with an increase of 0.1387% in LN Average Wage for males. We are 95% confident that the marginal effect of education level on LN average wage for Females lies between 0.1181% and 0.1592%.

10

Jeekeshen Chinnappen

0308104/110107264

Note: The hypothesis test of β = 0 at the 5% of level of significance can be rejected for all of the above simple regressions since zero does not lie in the confidence interval for β and since Pvalue is much less than 0.05.

Multiple Regression Y = LN Average Wage for Males, X = Education LNY = -1.0410D1 - 0.5146D2 – 0.1349D3 + 0.1817D4 – 0.0148D5 + 0.4021D6 + 0.02465X1 – 0.0390D8 + 1.2006 + e The hypothesis test of β = 0 at the 5% of level of significance can be rejected for No Diploma, High School/Diploma/Vocational, marital status and age since their P-value is less than 0.05.

11

Jeekeshen Chinnappen

0308104/110107264

Y = LN Average Wage for Females, X = Education LNY = -1.7754D1 – 1.4386D2 – 0.8646D3 – 0.3518D4 – 0.05126D5 + 0.09674D6 + 0.02185X1 – 0.01632D8 + 1.9679 + e The hypothesis test of β = 0 at the 5% of level of significance can be rejected for No Diploma, High School/Diploma/Vocational and age since their P-value is less than 0.05.

12

Jeekeshen Chinnappen

0308104/110107264

Limitations of Analysis The amount of healthy people of 69.74% compared to the unhealthy ones with 30.26% and the amount of white people of 85.74% compared to nonwhite of 14.26% could easily result in a biased primary data since it should be 50:50 to be able to get a good set of data to be analyzed. There is also a correlation between the explanatory variables like the ages of people and their education level which proves the presence of multicollinearity. And also there are respondents which did not reply to all of their data which led to missing data during data collection. This has reduced the accuracy of the whole analysis due to reduced number of observations. Since the residuals are not so evenly distributed along the line, Heteroskedasticity is indicated.

Residuals For Males 5 4 3 2 1

Residuals

0 -1 0

0.5

1

1.5

2

2.5

3

3.5

-2 -3 -4

Residuals For Females 3 2 1 0 -1 0

0.5

1

1.5

2

2.5

3

3.5

Residuals

-2 -3 -4 -5

13

Jeekeshen Chinnappen

0308104/110107264

Conclusion We can conclude that there is a positive correlation between the level of education and average hourly wages, supported by the PC report which means there is empirical evidence to support that people with higher level of education gets higher wages. We have concluded that those with a postgraduate/undergraduate obtained 6.0 times more than the lowest level of education for males and 10.5 times more for females compared to the PC report where it is about 1.9 times in general. The rise in β coefficients from ‘No Diploma’ to ‘Doctorate’ shows that the higher the level of education, the greater is the hourly wages which means that there is an increasing returns to education which is supported by the PC report. People could actually move out of disadvantages and get a higher hourly wage by having a greater educational level.

Words: 1100

14

Jeekeshen Chinnappen

0308104/110107264

Appendices Y = LN Average Wage for Males, X = Education LNY = -1.0410D1 - 0.5146D2 – 0.1349D3 + 0.1817D4 – 0.0148D5 + 0.4021D6 + 0.02465X1 – 0.0390D8 + 1.2006 When Bachelor Degree/Associate Degree_D3 = 1 and all the other variables = 0 LN Y = 1.2006 - 0.1349 – 0 = 1.0657 Y = e^ 1.0657 = 2.90287 When No Diploma_D1 and High School/Diploma/Vocational_D2 = 1 and all the other variables = 0 LN Y = -1.0410 – 0.5146 + 1.2006 = -0.724 Y = e^ -0.724 = 0.484809 Ratio Males for Bachelor Degree/Associate Degree_D3 to No Diploma_D1 and High School/Diploma/Vocational_D2 = 2.90287/0.484809 = 5.9877 Y = LN Average Wage for Females, X = Education LNY = -1.7754D1 – 1.4386D2 – 0.8646D3 – 0.3518D4 – 0.05126D5 + 0.09674D6 + 0.02185X1 – 0.01632D8 + 1.9679 When Bachelor Degree/Associate Degree_D3 = 1 and all the other variables = 0 LN Y = 1.9679 - 0.8646 – 0 = 1.1033 Y = e^ 1.1033 = 3.014096 When No Diploma_D1 and High School/Diploma/Vocational_D2 = 1 and all the other variables = 0 LN Y = -1.7754 – 1.4386 + 1.9679 = -1.2461

15

Jeekeshen Chinnappen

0308104/110107264

Y = e^ -1.2461 = 0.28762 Ratio Females for Bachelor Degree/Associate Degree_D3 to No Diploma_D1 and High School/Diploma/Vocational_D2 = 3.014096/0.28762 = 10.479

16

Jeekeshen Chinnappen

0308104/110107264

References Booth, A, Coles, M. (2006) ‘Increasing Returns to Education: Theory and Evidence’, Australian National University, .

Fengliang, L, Xiaohao, D, Morgan, WJ. (2009) ‘Higher Education and the starting wages of graduates’, International Journal of Educational Development, Vol. 29, pp. 374-381, .

Pew Research. (2014) ‘The Rising Cost of Not Going to College’, Pew Research Social & Demographic Trends, viewed on 15th April 2014, .

17