I was a Economics and Psychology major at college, but want to change to a statistics graduate program. I don't have a very strong math background and I think it's a major weakness in my application, so I try to stress the quantitative work that I've done in the past in my Statement of Purpose.
Here's my draft. Any suggestion is greatly appreciated. Thank you.
My country, China, has experienced many terrible disasters in the recent years, such as the major earthquake in Sichuan and the chemical factory explosion in Tianjin. Also, there are typhoons frequent the east coast and drought occur in Northern China every year, not to mention the extremely severe air pollution. The Chinese government is often criticized by the citizens and media as not responding to these disasters effectively and efficiently enough. As a person cares a lot about social impact, I want to use my knowledge and skills in statistics to engage in the Civic Technology area, help the government and organizations to better identify and solve the problems, long-term and short-term, as a data scientist, and therefore making the society a better place. To achieve this goal, I decide to pursue a master's degree in Statistics to gain more expertise.
Trained as an economist and psychologist at college, I have a good understanding of scientific research methods learnt from both textbook and practice. I've conducted analyses on various areas including the financial market, the labor market and the consumer market, and learnt several data analysis tools such as Excel, SPSS and STATA through my work as a data analyst for my professors. I enjoy finding the hidden information and patterns behind a large amount of data and it's always exciting to see the results come out after a long period of research design, data collection and model set-up. Even if the results contradict my hypothesis, though disappointing at first, it represents a great opportunity to improve my study design and statistical analysis skills.
As a curious person, I like to dig deeper into the problems to find the potential explanations and implications of the observed results. For example, in one of the labor market studies that I've conducted, initially I found that the wage gap between the Black and the White in the US labor market today remains roughly the same as it was in the 1990s after controlling for premarket academic skills. But does it mean that our Affirmative Action and many other efforts to eliminate the racial wage inequality had failed? And are there any explanations other than discrimination based on prejudice that can explain this difference? In order to probe further into the questions, I divided the labor market into more specific categories and found that in fact the trends of racial wage inequality in the last 20 years were very different for different job categories. What's more, a statistical model of discrimination as well as a model based on market competitiveness and employees' productivity could also explain the existence of wage gap, yet both of them have nothing to do with prejudice. Without these further analyses, I could have easily drawn intuitive, but wrong, conclusions. These research experiences also make me a better critical thinker. I realized that many opinions we see on TV or magazines that are based on scientific studies might not be as scientific or valid as the authors claimed. And I'll always consider the alternative explanations and come to the conclusions carefully.
However, as I work on more and more projects, I start to realize the limitations of my current statistics knowledge and data analysis skills to solve the increasingly complex problems today. For instance, I once worked at a health management start-up and we were developing a new medical tourism product. In order to find the customers' real interests, I designed the survey and conducted many interviews. However, the results obtained from these traditional market research methods were largely useless because they were so scattered that it was hard to extract any valuable business intelligence from them and we didn't know whether the results represent the customers' true feelings. What's more, such traditional ways of doing market research were very time-consuming and energy-consuming. It would have been much more effective and efficient if I could studied people's actual consumption behaviors or even use the data from media and social networks where people are more likely to express their true needs and feelings. But I was unable to study such a large amount of unstructured data with my current knowledge on statistics and data science. Thereafter, I learnt Python, SQL and MATLAB in my spare time and I'm currently taking an online course about machine learning on Coursera in order to be better at data analysis. I often impress others as a diligent and fast learner. I took the game theory class, the highest level course offered at the mathematics department, as a freshman while all of my classmates are juniors and seniors. I didn't learn linear algebra which was extensively used in the class. But I learnt the matrix theory by myself and got one of the three A's out of 19 students at the end of the semester. The professor, who was also the department chair, wanted me to be a math major and his advisee.
Although I'm confident that I can learn data analytical tools well, I believe it's important to learn statistics at a higher level systematically and that's why I think pursuing a higher degree in statistics is necessary. After all, the tools and algorithms are constantly changing as the world changes. Only when I have a solid understanding of the statistical ideas behind the methodologies can I be more adaptive and effective in facing the new challenges. (I'll continue this paragraph by arguing why I want to go to a specific program and why there is a good fit for me)
Upon obtaining my master's degree, I will decide whether I want to get further into research and pursue a PhD degree or go to work, depending on my research area and achievements during the program. In the longer term, as a statistician, I want to use my knowledge to help organizations and cities become more prepared and effective in response to shocks and challenges, whether they are physical, social or economic.
Here's my draft. Any suggestion is greatly appreciated. Thank you.
My country, China, has experienced many terrible disasters in the recent years, such as the major earthquake in Sichuan and the chemical factory explosion in Tianjin. Also, there are typhoons frequent the east coast and drought occur in Northern China every year, not to mention the extremely severe air pollution. The Chinese government is often criticized by the citizens and media as not responding to these disasters effectively and efficiently enough. As a person cares a lot about social impact, I want to use my knowledge and skills in statistics to engage in the Civic Technology area, help the government and organizations to better identify and solve the problems, long-term and short-term, as a data scientist, and therefore making the society a better place. To achieve this goal, I decide to pursue a master's degree in Statistics to gain more expertise.
Trained as an economist and psychologist at college, I have a good understanding of scientific research methods learnt from both textbook and practice. I've conducted analyses on various areas including the financial market, the labor market and the consumer market, and learnt several data analysis tools such as Excel, SPSS and STATA through my work as a data analyst for my professors. I enjoy finding the hidden information and patterns behind a large amount of data and it's always exciting to see the results come out after a long period of research design, data collection and model set-up. Even if the results contradict my hypothesis, though disappointing at first, it represents a great opportunity to improve my study design and statistical analysis skills.
As a curious person, I like to dig deeper into the problems to find the potential explanations and implications of the observed results. For example, in one of the labor market studies that I've conducted, initially I found that the wage gap between the Black and the White in the US labor market today remains roughly the same as it was in the 1990s after controlling for premarket academic skills. But does it mean that our Affirmative Action and many other efforts to eliminate the racial wage inequality had failed? And are there any explanations other than discrimination based on prejudice that can explain this difference? In order to probe further into the questions, I divided the labor market into more specific categories and found that in fact the trends of racial wage inequality in the last 20 years were very different for different job categories. What's more, a statistical model of discrimination as well as a model based on market competitiveness and employees' productivity could also explain the existence of wage gap, yet both of them have nothing to do with prejudice. Without these further analyses, I could have easily drawn intuitive, but wrong, conclusions. These research experiences also make me a better critical thinker. I realized that many opinions we see on TV or magazines that are based on scientific studies might not be as scientific or valid as the authors claimed. And I'll always consider the alternative explanations and come to the conclusions carefully.
However, as I work on more and more projects, I start to realize the limitations of my current statistics knowledge and data analysis skills to solve the increasingly complex problems today. For instance, I once worked at a health management start-up and we were developing a new medical tourism product. In order to find the customers' real interests, I designed the survey and conducted many interviews. However, the results obtained from these traditional market research methods were largely useless because they were so scattered that it was hard to extract any valuable business intelligence from them and we didn't know whether the results represent the customers' true feelings. What's more, such traditional ways of doing market research were very time-consuming and energy-consuming. It would have been much more effective and efficient if I could studied people's actual consumption behaviors or even use the data from media and social networks where people are more likely to express their true needs and feelings. But I was unable to study such a large amount of unstructured data with my current knowledge on statistics and data science. Thereafter, I learnt Python, SQL and MATLAB in my spare time and I'm currently taking an online course about machine learning on Coursera in order to be better at data analysis. I often impress others as a diligent and fast learner. I took the game theory class, the highest level course offered at the mathematics department, as a freshman while all of my classmates are juniors and seniors. I didn't learn linear algebra which was extensively used in the class. But I learnt the matrix theory by myself and got one of the three A's out of 19 students at the end of the semester. The professor, who was also the department chair, wanted me to be a math major and his advisee.
Although I'm confident that I can learn data analytical tools well, I believe it's important to learn statistics at a higher level systematically and that's why I think pursuing a higher degree in statistics is necessary. After all, the tools and algorithms are constantly changing as the world changes. Only when I have a solid understanding of the statistical ideas behind the methodologies can I be more adaptive and effective in facing the new challenges. (I'll continue this paragraph by arguing why I want to go to a specific program and why there is a good fit for me)
Upon obtaining my master's degree, I will decide whether I want to get further into research and pursue a PhD degree or go to work, depending on my research area and achievements during the program. In the longer term, as a statistician, I want to use my knowledge to help organizations and cities become more prepared and effective in response to shocks and challenges, whether they are physical, social or economic.
