Showing posts with label Research. Show all posts
Showing posts with label Research. Show all posts

Thursday, December 20, 2012

Some updates about CHNS newest biomarker data

[Forwarded from the UNC CHNS data center]


Dear CHNS data users,

Since we released the biomarker data collected in 2009, many users are excited to explore the data and of course have some questions. Assuming you may have  the same question, I make a response to the mailing list instead of emailing one by one.

To better understand the biomarker dataset, you may need to review the codebook first, which is available at http://www.cpc.unc.edu/projects/china/data/datasets/C10BIOMARKER.pdf.


In general, each biomarker has a set of 8 variables. Take Y48_1 (blood glucose) as an example:

Y48_1: Blood glucose measured in mmol/L in the field
Y48_1CD: Whether or not glucose levels are within normal range (.=missing ; 0=normal; 1=abnormal high but within linearity; 2=abnormal high and  beyond linearity; 99=status unknown because age/gender is missing)
Y48_1CS: Whether or not glucose data are available for the participants(.=missing; 1=assay available)
Y48_1SD: At which level the glucose value is (.=missing; 0=within +/- 3  standard deviations; 2=above 4 SD; 3=blow -3 SD)
Y48_1MG: Blood glucose measured in mg/dL in the field
Y48_1MG_CD: same as Y48_1CD
Y48_1MG_CS: same as Y48_1CS
Y48_1MG_SD: same as Y48_1SD

Some biomarkers were measured twice: one in local hospitals in the field (labeled as Field) immediately after blood samples were drawn; and one in the central lab in China-Japan Friendship Hospital (labeled as CJFH) after frozen samples were shipped to Beijing and unthawed. Generally speaking, biomarkers measured in the field were named with a prefix Y, and biomarkers measured in the central lab were named in a way that people can easily recognize them, e.g. glucose, urea, HDL, etc.

In analysis, if you use glucose, we recommend that you use glucose measured  in the field as blood glucose may depreciate when frozen and unthawed. For other biomarkers, we prefer to use those measured in the central lab as they were measured by the same people with the same machine.

Thank you for your interest in our study.
Happy Holidays.


Shufa Du
Research Assistant Professor
123 W. Franklin Street
University Square East, CB# 8120
Chapel Hill, NC 27516


Wednesday, May 26, 2010

In Search of a Right Topic

Recently, in different scenarios I have revealed some level of anxieties to my friends who were very kind and willing to listen to my complaints about hurdles here and there toward getting my dissertation done.

Prof. Don Davis of Columbia University has an interesting and very insightful little essay for Econ Ph.D students on how to start the Ph.D dissertation research. He listed out all the important strategies to keep in mind for a student to discover his/her own "right topic". Although my previous 4+ years of experiences in studying and conducting research almost coincidentally yet strictly followed his advices, such as "read theory and empirics as much, find research frontier, go to weekly seminar in my interested field, read NBER working papers and best journals selectively, Talk! Talk! and Write! and Write!", I definitely also made some big mistakes in terms of allocating time and resources on some wrong directions, which were warned by Don. I spent WAY too much time attending classes in not just all econ subjects, but also in Sociology, Law, Education, Psychology, etc. My motivation was, to truly understand the comparative advantages and disadvantages of both quantitative and qualitative researches methods and to learn as much as knowledge about institutions that backs up our economic phenomena.

So now, on the third day of the summer of 2010 where I am dedicated to a daily routine of conducting research, what strategies should I take to pick up on what I am doing and make the type of progress that I am expecting efficiently? In other words, Don gave a lot of great ideas to students to "start" their very first research project, but how do I "resume" upon what I have done before and have outstanding achievements?

I would like to go back to the very 2nd question Don suggested students to ask---How do I know if I have an interesting topic? In my last blog article, I listed all the topics that I am going to write upon in this coming year, covering a broad range of topics from educational system in Asian countries to entrepreneurship in the ed field to human capital theories, etc. But, now the question for myself is: which topic (or the more detailed subfield questions)on my list, is an important problem with real-world counterparts and that matter in substantive terms and moreover that my approach to the problem is new? Which part that I am working on will make the readers, who are familiar with the literature, after reading my paper, will be able to see the world differently? (Almost like all the high emotions I felt when I just enrolled to the doctoral program and started to read some brilliantly done papers, especially the ones about marriage institutions!)?

So now, looking back to the fields that I have ever worked on in the past 4 years, which fields I gained the most intellectual excitement when doing my own research? Human capital and economic growth, program evaluations, teacher's pay, university internationalization, gender econ, labor econ, even, cognitive sciences/ test scores/ the production function of education? Which one did I have the most ideas pouring out of my mind when I was taking classes, participating discussions and writing models? Which one did I write some models that later on I found that some scholars had already done so and obtained high reputation of the work, which, shows myself where my talents outbursted?

Basically, there are two types of econ researches: theoretical and empirical.
For theoretical ones, you can do a entirely new theory, or you can improve the current existing theories, or you can disapprove the current ones, "break the old and establish the new". For empirics, you can do an analysis, under certain theoretical framework, with a new dataset that no one else ever looked at it (but given the scarcity of good quality datasets, this direction, I would say, although may come with good policy implications by providing more knowledge about something we ingored before, but most of the time lacks academic rigorousness.) Empirics can also be in the form of adding some variables to current existing estimation strategies and thus improves both statistical but also economic significance of the discussion. Another type of empirics can be about using a very innovative methodology to look at some old or new issues and reach to some compelling conclusion.

I don't think I am writing an entirely new theory about one institutional dimension of human life. So, theoretical wise, I need to immerse more into the entire literature of my interested field and come up with new ideas on HOW I CAN IMPROVE the current existing theoretical and analytical framework.

Empirically, I also have a strong reluctance in only using some "untouched" dataset (e.g. something collected from field studies in China) to run a routine regression and claims novelty. Given the amount of training I have in econometrics, (besides my international mathematical contest gold medal, what else did I spiraled up these years mathematically?), I do not think that I am going to develop or invent some new econometric strategies. So, the most feasible way for me to thrive empirically will be conducting analysis using some sophisticated, large scale dataset to address some results that will make my theoretical analysis more convincing. Many discussions in recent decades about structural model and other estimation strategies in empirical studies might also conclude that my strength is in combined theoretical modeling plus some smart empirical evidence--- of course, this is more of what I WANT to pursue, because it is undoubtedly desired by most economists.

So, now the direction is much clearer, my applied field is about education, using applied theories of economics, I want to develop some innovative models based upon existing, well-established ones, to address one important question that has a real-world counterparts, and I should also work at making sure of some robust empirical evidence.

I should go back to the literature now again: to figure out, on human capital theory the most updated model, is there anything I can do about polishing its implications about gender roles and parental choices in education?

Thursday, June 19, 2008

China makes everything. Who is making China?

It is almost the NO. 1 question I asked myself very often recently. And I believe the answer lays in education. Human capital accumulation and technology advancement all depends on education (even R&D units in industries are functioned by school-produced researchers).

In terms of import and export, China is manufacturing everything the world is consuming in daily life, from socks and shoes to computers and cars (well, I put it here as an aspiration for the next 10 years). But, from entrepreneurs who come up with ideas to incubate new market and engine the growth, to all types of workforce supporting the economy, what factors are producing them?

Labor economists have done intensive work, tried to capture the best method to quantify the measurement of ability, skills, the add-on effects of education, and value of experiences. We all would not deny that all human are born equal. In the scale of a nation, Chinese babies are not doing better than babies in any other countries in terms of genetic IQs. So, it is those add-ons that matters how the economy will develop. Now, education. But what is education? How much comes from schools, how much from parental influences (family education), how much the community (cultural factors?)?

I have done courses in development econ, labor econ, sociology and educational philosophy. Still no idea where to start the huge project to answer: Who is making China?

I will try to narrow down my thoughts and try to formulate at least an outline to exhibit the facts in a structured way.

Wednesday, June 11, 2008

More Thoughts to Come

Organizational failure or born to fail in nature? The question is always in my mind when I read about school system reforms, or, more general, that of education. Economists are not giving effective enough solutions, neither do professional practitioners. Then how about any form of R&D, which has been fueling sustainable growth of all other sectors facing same transitional challenges? Is educational R&D, integrating IT developments, the way leading to a better future? Does the key, how to conduct Ed R&D, lay in our head or some external factors?

Being lucky enough, being a Ph. D student in Economics and Education, (though, coincidentally, "failing" too), I can explore potential answers from within economic researches, which are imperial in social science realms nowadays; I also got many precious exposures to strategic smart heads who share the same concerns over education. No need to brag at all, it matters significantly about the future of mankind.