Statistical Analysis of Gender Earnings Using Stata and T-test

This report presents a statistical analysis of gender-based earnings using the Stata software. The study utilizes data from the Job Corps experimental evolution project, focusing on variables such as gender, treatment group, ethnicity, and earnings. The analysis begins with data transformation, addressing missing and zero values, and creating a new variable named 'earncen'. A t-test is conducted to compare the earnings between male and female groups. The results reveal a significant difference in earnings, with males earning more than females. The t-test yielded a t-statistic of 11.028 and a p-value of 0, leading to the rejection of the null hypothesis. The report concludes that there is a statistically significant difference in earnings between genders, supported by the Stata analysis and t-test results.

Running head: STATISTICAL METHODS USING STATA
Statistical Methods Using STATA
Name of the Student:
Name of the University:
Author Note:

1STATISTICAL METHODS USING STATA
Table of Contents
Introduction................................................................................................................................2
Details of the dataset..................................................................................................................2
Data transformation....................................................................................................................2
Data Analysis.............................................................................................................................2
Results........................................................................................................................................3
Conclusion..................................................................................................................................4
Reference....................................................................................................................................5
Appendix....................................................................................................................................6
Appendix 1: Details of the dataset.........................................................................................6
Appendix 2: Summary statistics of total self-reported earnings and also for male and female. 7
Appendix 3: T-test result............................................................................................................8

2STATISTICAL METHODS USING STATA
Introduction
Data analytics is potentially able to forecast with the help of relevant data and
statistical software. Here, the data that is used to prepare the report, was used in experimental
evolution of the Job Corps, conducted by the Mathematical Policy Research under contract to
the U.S. Department of Labour. The report tries to find the difference of earning across
gender of the individual (Geiger, 2017, Schochet, 2018 & Wandner, 2017).
Details of the dataset
The data set includes too many important variables as it was used in a crucial research
work. However, the important variables that are considered here are gender which is named
identification number (mprid), treatment group (treatment), gender (female), ethnicity
(race_eth), age group, educational group (educ_gr), earnings in the year prior to random
assignment (earn_yr) and self-reported earnings per week in the 4th year after the random
assignment (earny4).
Data transformation
The missing values and the zero values of earny4 are dropped. Treatment group is
selected for the analysis and the control group is dropped. “earny4” is transformed into
“earncen”. The values of “earncen” is equal to “earny4” where all the values were less than
300 and the remaining values of “earncen” were fixed at 300.
Data Analysis
To check the difference in earning of the male and female groups, t-test is conducted.
Before t-test the details of “earncen” is calculated for male and female.

3STATISTICAL METHODS USING STATA
Results
The below table presents the mean value of self-reported weekly earning, same for the
male and female. The mean value of earning is 200.566, the mean vale of earning for male is
214.408 and the mean vale of earning for female is 176.938 (Schopohl, 2019).
Table 1: Summary statistics of “earncen”, “earncen for male” and “earncen for female”
Obs 3354
Sum of Wgt. 3354
Mean 200.566
Std. Dev. 96.6637
Variance 9343.86
Skewness -0.5209
Kurtosis 1.91629
Earncen
Obs 2115
Sum of Wgt. 2115
Mean 214.408
Std. Dev. 92.9412
Variance 8638.06
Skewness -0.7412
Kurtosis 2.2243
Earncen (Male)
Obs 1239
Sum of Wgt. 1239
Mean 176.938
Std. Dev. 98.3392
Variance 9670.6
Skewness -0.1856
Kurtosis 1.70763
Earncen (Female)
The t-test result is presented in the below table. The hypothesis are mentioned below:
Null Hypothesis: The difference between average earning of male and average earning of
female is 0.
Alternative Hypothesis: The difference between average earning of male and average earning
of female is not 0.
Table 2: T-test result
Group Obs mean std. error std. dev
0 2115 214.4078 2.0209 92.9412
1 1239 176.9384 2.7938 98.3392
diff 37.46937 3.397659
diff mean(0)-mean(1) t 11.028
diff !=0 Pr(t) 0
diff >0 Pr(t) 0
diff <0 Pr(t) 1

4STATISTICAL METHODS USING STATA
The t-stat is 11.028 and the p-value is 0. The p-value is less than 0.05 (general
significance level) for which the null hypothesis is rejected and alternative hypothesis is
accepted. This implies that there exists a difference between average earning of male and
average earning of male. The p-value is 0 for difference between average earning of male and
average earning of female which indicates that the difference is greater than 0 and significant
at 5% significance level. This implies that the alternative hypothesis which says that the
average earning of male is greater than the average earning of female will be accepted.
Conclusion
With the help of STAT, a statistical software, a t-test result shows that there exists a
difference between earnings of male and female and the earning of male is greater than the
female. The average earning of male is 214.408 and the average earning of female is 176.938.

5STATISTICAL METHODS USING STATA
Reference
Geiger, R.L., 2017. Research and relevant knowledge: American research universities since
World War II. Routledge.
Schochet, P., 2018. National Job Corps Study: 20-Year Follow-Up Study Using Tax Data.
Schopohl, L., 2019. STATA Guide for Introductory Econometrics for Finance.
Wandner, S.A. ed., 2017. Lessons Learned from Public Workforce Program Experiments.
WE Upjohn Institute.

6STATISTICAL METHODS USING STATA
Appendix
Appendix 1: Details of the dataset.

7STATISTICAL METHODS USING STATA
Appendix 2: Summary statistics of total self-reported earnings and also for male and
female.

8STATISTICAL METHODS USING STATA
Appendix 3: T-test result