Research – Page 2 – Junjun's Home

Writing a Mathematical Equation on Your Blog

In fact it is quite rare for a blog writer to put some mathematical models on the blog unless it is an article related to some very professional courses. However, low propability can hardly mean never happen. Taking myself for example, I would like to post some of my notes on econometrics and my feeling of research proposal with the blog readers, though I can write them with a Latex system and then compile it to a PDF document, providing it as an attachment for download. While I don’t think most readers would prefer such behavior. Reading that directly can be a much more pleasant experience, especially when there is not many complicated equations.

Thanks to the flexibility of the WordPress system and its countless expanded plug-ins, the solution for this can be quite simple. The only thing you are willing to do is to be familiar with the LaTex syntax, which, under most occasions, would not be a big problem for most users who are willing to do so. Even though you are unlikely to type the raw code, Mathtype can help you transfer the equation. The only work thus left for you is copy and paste.

Taking a search at WordPress.org, with the keyword of ‘latex’, you are likely to find the following plug-ins.

Some Plug-ins for WordPress

Among the plug-ins that are available, I made an attempt for the following ones on my blog. Here is a summary about each of them. For details, you can refer to their websites.
1. WP Latex
  
  This is an official plug-ins for Latex from WordPress.com, with its default service based on wordpress.com. No cache is needed. You can freely change its style through the css file in settings. The only problem is that its original server ‘s.wordpress.com’ is forbidden in China, you have to change it to ‘wordpress.com’ to make it work. Besides, you can also set up your own service, but I wonder if it can violate the regulation from the host.
2. Easy Latex
  
  Comparing to the official plug-ins, this one offer more flexible features. Besides the colors, you can also freely change its size directly through its GUI. Cache option is provided so that the readers can stil read the equations on your blog even when the latex service is out of work.
3. Youngwhan’s Simple Latex
  
  This is the simplest plug-ins for use among the three. No other option is presented besides the address of the latex service. The service for this plugin, unlike WP Latex and Easy Latex, is provided by some open source project. Though it is convenient to set up, its flexibility is a question.
My Choice

I finally choose the official plug-ins with the consideration of the limitation of file numbers on my server. And there is no free CGI support. Which means, I can hardly construct my own system and can hardly bear too many cache files. For those whose server can support CGI, a self-constructed system should be preferred considering its stability.
An Example

As a final, I shall present an example. This is a measurement of likelihood that is common to see in econometrics.

[latex size=2]{{R}^{2}}=\frac{SSE}{SST}=\frac{{{\left( \sum\limits_{i=1}^{n}{\left( {{y}_{i}}-\bar{y} \right)\left( {{{\hat{y}}}_{i}}-\bar{\hat{y}} \right)} \right)}^{2}}}{\left( \sum\limits_{i=1}^{n}{{{\left( {{y}_{i}}-\bar{y} \right)}^{2}}} \right)\left( \sum\limits_{i=1}^{n}{{{\left( {{{\hat{y}}}_{i}}-\bar{\hat{y}} \right)}^{2}}} \right)}[/latex]

Where [latex]y[/latex] stands for the real value and [latex]\hat{y}[/latex] stands for the estimated value from a regression model like:

[latex]y=\alpha_0+\alpha_1*x_1+\alpha_2*x_2+\alpha_3*x_3+\cdots+\alpha_n*x_n, n\in {{Z}^{+}}[/latex]

The only problem left here is: how to adjust the alignments? I want them to be centered, but not left-aligned.

Some Tips for Rookie Empirical Researchers

最近大家都开始做毕业论文了。鉴于现在国内外的学术潮流，大家都开始做实证了。哎，说实话，我们80后这一代可真是累，穿衣服、吃饭、看书什么都要赶时髦，甚至连现在这“人人得而诛之”的毕业论文都开始给我们找麻烦。您说洋洋洒洒数万字的规范研究有什么不好，既可以天马行空，也可以写的很潇洒。若有哪位牛人从小熟读唐诗宋词、楚辞汉赋，此时论文写作便是该君最为得意的时候。可惜，古人云，天有不测风云，今人曰，研究有风险，如今实证研究大行其道，不做点实证研究，毕业论文不放几个表格，不放几张统计图，不去写上一些数学模型，如何能一讨恩师的“芳心”？

正所谓闻道有先后，术业有专攻。由于本人恩师在实证研究上的痴迷，我也就比较早的感受到了实证研究的痛苦与残酷，而自然在如何避免“踩地雷”方面有了自己的一点心得，这里就大家最常见的一些问题来给出一些个人的见解，以求抛砖引玉。

我们一般做研究的数据可以从哪里取到？数据一般使用什么格式？

一般我们的数据都不需要手工搜集。我们学校（中南财经政法大学）已经购买的数据库可以搞定大部分的财务报告数据和公司治理等方面的数据。而CSMAR则可以找到股票回报率方面的数据（我们学校没有高频数据）。如果觉得数据还是不够用，则有两种处理方法：
1. 手工搜集：我们可以通过巨潮网取得所有上市公司公开发布的报告，包括定期公告和临时公告。
2. 请求外援：我们可以通过其他学校的朋友来帮忙。比如Wind的数据质量很高，但是我们学校没有购买。于是我们可以通过自己的人脉去那些购买此数据库的学校和机构来求助。
数据格式可以有各种各样，比如常用的Excel格式，数据库DBF格式，以及各种统计软件的格式等。具体使用什么格式，可以根据自己的情况来定。Junjun早期比较喜欢Excel格式，因为其处理灵活，但是现在更倾向于DBF，因为它能够具有更好的后期处理接口，同时格式定义更规范。
我的数据应该如何来合并？异常值和空值如何处理？

一般来说，我们做研究的数据来源非常广泛，有手工搜集，有来自于不同的数据库，甚至还有部分数据来自于网站。那么，如何把这些数据进行合并是个非常要命的问题。这个问题也曾让Junjun郁闷了很久，想必也是大多数初级选手的绊脚石。大多数情况下，我们可以有三种方式来进行数据文件的合并：
1. 利用Excel进行合并
  
  这是最简单，但是也是最容易出错的一种合并方式，尤其是当你对Excel的各种函数并不是很熟悉的时候。常用的函数有Vlookup，Hlookup等。其实Excel不仅仅可以用来合并，同时它也是数据处理不错的工具。Junjun早期的数据处理也是基于Excel做的。只是Excel的运算效率有点低。Junjun曾有一项最高纪录：一句复杂条件匹配函数在笔记本上计算了20个小时。
2. 利用Access类进行合并
  
  这是一种比较常用的方式。简言之，就是使用Access等自带的SQL（结构化查询语言）来进行处理。SQL具有远比Excel更优秀的算法，效率非常高。但是Access也有缺点，那就是数据库对格式要求远没有Excel灵活，这就导致了一些情况下数据类型不匹配问题的发生。
3. 利用专业软件包进行合并
  
  这也是一种高效率方法，但是Junjun在这里并不推荐给新手，除非你准备未来投身于学术界。通过专业软件包的一些关键语句可以很好的合并，效率绝对不逊于Access，但是需要Hand Coding。对于没有Programming的童鞋而言，这可能就是一种杯具了。
对于空值和异常值的处理，最简单的方法就是剔除。当然，也有一些备选方法，比如对于异常值采用Winsorize方法，对于空值通过EM算法来填充等等。只是这些方法需要更好的技巧和Programming的技术，此处就不赘述了。
我应该使用什么统计模型来分析呢？

不得不说，这个问题非常高深。高深到Junjun现在还是深受其害。按照统计学的说法，我们应该根据不同的数据样本来选择不同的估计方法。什么？听着像废话。那就说点实用的，但是却是没有科学根据的。一般情况下，如果你跟Junjun一样，怕瞎弄闹笑话，那就假定自己的数据服从正态分布（Normal Distribution），采用OLS。而如果你的被解释变量是两分类的虚拟变量，请考虑使用Logistic回归，如果是多分类，可以使用Logit。而如果数据有存在很大的偏倚，则请考虑Tobit。具体的方法，可参考相关计量经济学和统计学教材。
我应该使用什么统计软件？或者说，什么统计软件好用？

这也是一个新手容易问到的问题。这个问题看起来幼稚，事实上却并不简单。其实在Package的选择上，完全遵循着咸菜萝卜各有所好的道理。以工作平台为例，Junjun非常喜欢在Linux下运行计算软件、利用Latex进行论文写作等。而更多的同学可能更觉得Windows+Word平台比较合适等等。

Junjun也曾试用过很多统计软件，如Eviews，SPSS，Stata，SAS和R。就我个人感觉而言，SPSS无疑是最容易上手的。Junjun最早的一篇论文就是使用SPSS完成了所有的工作。但是，SPSS的低效率也实在让Junjun难以忍受。但是不可否认，SPSS最容易找到中文教程，毕竟这是最早进入中国市场的统计软件包。目前最新的版本是17.0（不知道最近更新了没～）。

Stata则是统计软件中的新秀。该软件的特点是短小精悍。定制很灵活，可扩展性强，适合中高级用户。由于这个软件的高扩展性，某些任务在此软件下会非常容易。但是，需要说明的是，大多数关于此软件的教程和说明皆为E文。国内尽管也出版了几本教材，但是在人大论坛上却是恶评如潮。愿意使用此软件的童鞋，请作好阅读大量英文的准备。

SAS则是统计软件中的大哥大，价格昂贵。但是速度很快，支持并行处理。但是语法严格，必须要求正则表达。

R是一款开源软件，通俗来讲，是不要钱的。风格介于SAS和Stata之间。该软件在Linux下效率很高。
我的结果不显著怎么办？是不是可以改数据（结果）？

这是Junjun今天打算回答的一个最重要问题。其实这个问题也是让Junjun最烦的。其实实证结果不显著很好解释：那就是您所假设的关系并不存在。而解决这个问题，也不是那么简单了。除了技术层面的问题，这个问题的解决也涉及到了研究思想方面。在这个问题的理解上，Junjun要感谢他的老板，香港中大的TJ Wong和Joseph Fan教授，夏威夷大学的Shirley J. Daniel教授，台湾大学的李书行教授以及财大的诸多教授。

就Junjun个人的观点，结果不显著可以遵循以下几个方面来进行解决：
1. 我的研究思路有问题否？
  
  这是最基础的问题，但是也是最容易被忽视的。长久以来，我们总是有一种错误的观点。那就是我们先把实证结果做出来，然后再来考虑如何进行解释。殊不知，这种做法就如同当年大跃进一般的盲目。Junjun也曾是这种说法的坚定拥护者，但是在去年的培训中，被范博宏教授给狠K了。其实大多数情况下，我们的结果不显著，或者没法解释，都是因为研究思路的问题。
  
  作为一篇实证论文，最为核心的就是研究假设。所有的步骤，包括引言、文献回顾、研究设计都是基于此。一个合乎逻辑的假设应该有着完备的理论背景和坚实的现实基础。如果说，君的论文假设是拍脑袋想出来的（垃圾文章上的假设也归于此类），那么，君就得好好想下，您论文的假设是否符合逻辑。最简单的方法，您可以使用常见的基础理论，比如产权理论、交易成本理论来尝试从理论基础来推导，看假设基础是否坚实？同时，作为中国市场上的实证研究，我们同时要关注，这个命题是否具有现实性？相关的法规是怎么说的？制度背景是非常重要的。
  
  如果上述仍然不足以解决问题，您可以试着去阅读一些相关文献。尽可能去阅读国际顶级期刊文献，如TAR，JAR，JAE，RAS，JF，CAR等等。相比国内的文献，这些文献做的更严格。同时，关注下TJ Wong和Joseph Fan等人的研究。他们对中国的研究对我们具有很好的启发作用。
2. 我的研究样本是否有问题？
  
  这个问题其实与上述的现实性基础具有异曲同工之妙。作为研究样本，最容易被人攻击，而事实上也最容易出问题的就是样本具有Bias。而Bias的种类和产生都千差万别，最常见的就是Selection Bias。发现此类问题的最好方法就是进行描述性统计分析。同时，我们需要注意我们的研究样本与制度背景的匹配。如：在全流通环境下去研究非流通股股东显然是不合时宜的。
3. 我的数据处理是否出错？
  
  Junjun最喜欢说一句话，统计学是一门数学，而数学是严谨的。当数据处理出现错误的时候，最后数据分析的结果也会忠实的告诉您：这种关系不存在！或许您会认为这个问题非常幼稚，而事实上，Junjun曾也就此犯过错。这里有个真实的例子：Junjun曾在2008年做过一篇论文，由于当时技术尚不熟练，出了很多未知错误，导致结果非常牵强。在从美国开会回来对此文章进行大量重新修正的时候，Junjun仔细严格的重做了所有的数据，但是没有变更统计学模型和估计方法。而修改后的文章的结果显著性和预期吻合非常好。这就说明，在思路没有错误的时候，技术也是很重要的。
  
  常见的错误有区间配比错误，数据匹配错误等。还记得我前面提到的数据合并问题否？合并的错误可是很容易导致严重后果的哦～～
4. 跟导师沟通
  
  这是最后一招，也是Junjun用的比较多的一招。导师总是站的比我们高，看得比我们远的。Junjun曾因为论文数次骚扰了我敬爱的老板，甚至让老板亲自帮忙审阅了好几次英文论文稿。大家应该知道阅读英文论文的感受～～，更何况还是Junjun的Chinglish。
我是不是可以改数据（结果）？

这是今天的最后一个问题。其实这个问题不用Junjun回答大家都知道答案。大家都应该知道什么叫道德，什么叫做学术道德。如果一位科研工作者不能恪守学术道德的话，那他恐怕就没有资格再做这项工作了。目前中国的学术造假已经不少了，都已经驰名中外了，我们就不要再为此添砖加瓦了吧，哈哈，毕竟这可不是什么好事儿是吧？

写在最后：上述这几个问题是最近同学们问Junjun回答的最多的。但是Junjun在准备博士入学考试，所以没有大量的时间来回复。但是Junjun非常乐意就自己知道的来帮助大家。然而，Junjun跟大家一样，也只是一个在读的硕士研究生，眼界、知识难免有局限。这里所作的回答也只是基于Junjun个人的理解和感悟。不对的地方，还请大家一定要指出。

大家如果有新的问题，可以在下面写评论。第一次在我博客评论的同学可能会遇到评论审核。请大家不用着急，我会在第一时间来审核并回答的。谢谢大家的支持！

THFD: Free Lunch for Researchers

It is widely acknowledge for the researchers on finance and accounting as well as related fields that the access to the data for their researches are undoubtedly essential. Unlike the traditional research framework that our predecessors did in 1950s and earlier, when narrative researches are mainstreams, the research nowadays, defined as contemporary researches in most literatures, are taking great preference on analytical methods and quantitative analysis. For researchers in China, especially for the fields of accounting and finance, empirical research is very popular these days. While, the data for these researchers who are conducting empirical researches means a lot, because they can directly decide the availability of a proposed research.

For most researchers that are employed in top institutions, the access to these data would not be a problem since they are mainly purchased by the library, clients within the certain IP range would be permitted for their visits. But for the graduate students and some researchers who are not in the top institution, the access would be a problem since the library of their institution would not will to, or without sufficient financial ability to afford, to purchase such service. Thus for them, doing contemporary research and publish their papers in top journals would be a nightmare. In fact, due to the lack on access to COMPUSTAT and CRSP, most researchers in China can hardly keep pace with the scholars in US and Europe for top researches, though that does not matter much since we can still do the research within China. Frankly speaking, the market of China itself can be regarded as a good example of emerging markets.

However, for those researchers and graduate students who even do not have adequate access to the data in China, it would be a nightmare. Without sufficient data and strong support for the conclusions (anticipated) in a manner that is based on strict statistical analysis, the research would be regarded of little value and credibility. In the old days, including me, I would use a VPN connection to some other top universities in China, acquiring the data from their databases, especially when I am taking my vacation off campus. Since the VPN service is only available for the faulty in my school.

But now, thanks to the carefulness of Junmei, it is her hard working and strong interest in research that eventually lead to the finding of the THFD, a free solution for the data access for most researchers in China, which is hard to believe in but is indeed the fact. According to the instruction on its site, this database is free to all the researchers who are interested in the China related questions. To start use this service, you shall visit its official site, and you will see the following site. Clicking the registration button would lead you to the site for a free registration. In fact, the interface for this database is quite similar to the Julin, a paid database provider in China. PS: A later test find that the Julin is also the provider of the data in THFD.

After registration, log in with your own username and password, you will be redirected to the main page, where you can start your data collection intermediately.

In an integrated site, you can conveniently collect various types of data in a single sheet and export it as Excel spreadsheet. But the structure for this spreadsheet would be two dimensional, you would need to transfer it into a long structure, which is more beneficial for most analysis. Such operation is quite simple in professional packages like Stata and R.

Considering that the potential readers for this post would be current and potential researchers who are at least, of some familiarity to the usage of such data, and this post is not intended to be an user guide, the details for the usage is not provided. For details and information, please refer to the help center provided in this database.

For users who are using Windows Vista / 7 and Internet Explorer 7/8, please turn your Compatibility View on. And, Internet Explorer is preferred. I have tried to visit this site with Chrome, but some unexpected errors occurred.

Please intuitively follow the copyright law in both China and the country of your residence when using this service. There is no English interface, which may bring you some trouble if you are not good at Chinese.

Accounting Theory: Positive or Narrative?

Since the development of contemporary research of accounting in 1992, when numerous HK scholars started to conduct the empirical researches under the Chinese stock market, a discussion between the traditional (We call Narrative Accounting Theory) and the contemporary (We call it as Positive Accounting Theory, PAT) theory has been started for decades until recent years. Now the empirical research has been a popular trend in China, which can be easily observed from the conference proceedings, working papers and published papers, especially those from young scholars.

But apparently the research nowadays in China is losing its way. Grounded on the accounting research framework that is introduced by the William R. Scott (2006) and Watts and Zimmerman (1986), two popular textbooks for contemporary accounting research, the directions has been lost. You can easily find the similarity of the papers from the scholar of accounting and those from other fields like banking, finance and economics. Taking the recent hot topics like political connection, lend contracts for example, which are both based on the corporate finance, has little connection to the fundamental understandings of the field of accounting, for both the accounting professors and accountants who are working as role of practitioners. For a brief introduction on books of accounting theory, refer to a post published previously.

The reason for this bias may come from the development process itself. For most researchers, they are in fact quite familiar with the traditional theories and thus can write good papers with such stuff. But the transition of methodology changed their way. A preliminary assumption for empirical research is its good competence on computer programming and sufficient understanding on econometrics and statistics. Frankly, it is a disaster for most scholars who seldom receive strict training on quantitative analysis when they were graduate students. Such problem even exists for the new generation students like us. The lack of training on fundamental instruments can cause great problems on research design, that can partially explain why so many errors on modeling of accounting papers. And, besides the fundamental theory, some new theories in finance and economics are introduced, like asset pricing, market efficiency, etc.

The transition on both supporting theory and instruments cost too much time and energy for both scholars and researchers in this field. Given the publication as a key standard of promotion, an easy way for publication but without deep thinking on its fundamentals are needed. Admittedly, this is a necessary step that can hardly be avoided. In fact, I did such foolish things in the past two years. Comparing to the papers discussing accounting policy and methods, such as recognition on loss and revenue, the topics on corporate finance can be much easier to understand for both researchers and readers. That’s why we can see many corporate finance papers in the journal of accounting these years, but few papers on financial accounting can be found. What’s more, quite a lot of scholars are not from the major of accounting as their undergraduate major, and they may have missed essential courses to help them understand the core accounting questions.

In fact, by doing accounting research on accounting itself, rather than the corporate finance, though it may be more attractive to most scholars, a scholar may face the coherence on the PAT and narrative accounting theory. The narrative accounting theory help us understand how an optimized accounting plan would be, and the PAT showed us the incentives from multi-parties that can affect this optimization. In fact, William R. Scott is just trying his best to build up this link, providing us the understanding of the fundamental theory that is not presented in a published paper, whose work is of significance to the graduate students like us.

Many discussion on the conflicts between these two types of theory, presented on both blogs and academic journals, even in quite a lot of books, are quite superficial. Only when linking both these theories can the high quality research be produced. Though, I am also still thinking about how to. Just like a previous post that I wrote in this blog, we need to think real questions and do real research. The link on these theories, in fact, provides its preliminary step.

NOTE; THIS ARTICLE ONLY REPRESENTS MY OWN OPINION AND CAN BEAR NUMEROUS FAULTS. FOR ANY DISAGREEMENT OR FURTHER DISCUSSION, YOU ARE WELCOMED!

Do Real Work, Think Over Real Questions

The application to the PhD Program may be regarded as the greatest failure on my individual plan for my future in the past years. While frankly, such efforts, though bring me no benefit on its proposed directions, show my a good lesson. That is, do what you can do, but never make attempts on those you are not familiar with. Doing something particular in a unsuitable environment, where you can hardly find reference, would directly bring you to the failure. Positioning is too essential for a reasonable decision-maker, without reasonable positioning, things can go terrible.

I was supposing the prize from the conference held at my birthday can bring me more courage, recovering myself back to the old days when I am fully prepared and encouraged to both the challenges and academic research. However, I found it didn’t work, I still remained what I was, no significant change at all. My confidence on previous research, the pursuing of a PhD program overseas has been totally abandoned. Well, I am so lucky that I registered for the PhD admission exams of Sun Yat-sen University, where can bring me acceptable reward, if I am successfully admitted.

A first reason may be from the research itself, as I hereby conclude, after a deep consideration. Comparing to the majority of postgraduate students in China, I am a lucky dog. I have a good supervisor, helping me building up solid foundation on my research. However, for the empirical research, a contemporary research methodology adapted in China since 1992, we are both rookies. We select our topics from existing literature. That is right and reasonable, why? We are professor and student respectively, neither owns sufficient practical experience. To tell it frankly, we don’t have the first touch of the real world. What’s the consequence then? We don’t have our own story. How will it be then? We get topics from papers, not from the real world. My emotion on research may thus be greatly mitigated when such problem is aware of, but without any good solution.

The other reason may directly come from the selection of graduate school. Undoubtedly, it is a hard decision. Given a high certainty on going overseas, I shall not have to fight for the top schools in China. It is a time-consuming and tiring job, without any pleasure at all. However, considering the limited reputation of our school outside China, and the limit on number of candidates going overseas for PhD, there is no reference. That is, I can hardly evaluate my own probability of admission to international graduate school. Under such situation, giving up top schools in China can be too risky. That’s why I finally made such decision. While,such decision can hardly mean peace, it reflects much more the selection itself.

I am writing this post for two aims, one is to express my depression, which has been done in previous paragraph. The other one is to identify my understanding on ethics of research. Do real work and think them over should be a key determinant to the future research. The interest of research can only come when you are doing a real question. In fact, I was firstly supposed to write a proposal on privatization of state-owned firms in China. After an attempt for 15 days, I gave that up. This abandon directly wastes my time of 15 days, but I get this lesson. And I am now starting to aware of why many scholars overseas are so interested in research. They are doing real research grounded on real world!

This is an article sharing my own feeling and understand on research. Your comments are welcome.

A Special Birthday: Academia and Entertainment

估计傻瓜都知道工作是工作，娱乐是娱乐，两者不可混为一谈。否则会一事无成。但是呢，在今年的圣诞节，也就是本人的生日之日，这两者却似乎找到了一个不错的交点。在这美妙的一天里，我同时完成了学术任务和娱乐任务。可谓工作与生活兼顾。不过说实话，我也累得够呛。这个故事还得从三天前的一个电话说起。那天晚上，哥们正在忙着拜读新买的《四书集注》，正在为《大学》开篇中的“齐家治国天下平”而激动不已，心情澎湃，却忽然接到研究生部的通知说要参加今天的第二届会计学院研究生学术研究研讨会，并作大会报告。一开始也没有觉得什么，毕竟咱也是久经沙场的人嘛。但是师弟的下半句却让我心慌了很久。因为他告诉我，我的报告被夹在了两位学术牛人之间，更何况，其中有一位还是在下的恩师。这下好了，关公面前耍大刀了。但是呢，我们是社会主义一块砖啊，哪里需要放哪里。于是，在下只有恭敬不如从命。

所谓喝水都能噎死人，阴沟里照样翻船理论，同时援引本人的一贯遭遇，此次演讲注定不会那么顺利。事实再一次验证了统计学的重要性。先不说我只有两天时间来准备PPT和讲稿，光演讲时候的PPT自动往前跑就让我非常郁闷，以至于我今天毫无演讲的性质和激情。脑子里的那些理论，假设都被抛开了。更多的，我一边在讲，一般在小心防备着我的PPT自动往前走（刚测试发现，这可能是03和07的不兼容性的一个Bug，出现于某些特定情况）。同时，这个报告也说明了一个问题，永远不要用中文去报告自己用英文写的论文。那过程绝对是自虐。今天我不得不一边看着PPT，一边把脑子里的英文换成中文再说出来。于是开始了舌头打结，语无伦次。好在，这不是太高级的会议。不过我那敬爱的老板的表情还是让我着实来了一身冷汗。哎，太年轻啊，定力不行。

下午跟着同门好友跑到西苑唱歌去了。你别说，KTV还真是个发泄的好东西。尽管这种事情多少有点损人利己。在给被人带来了极其恐怖的噪音之后，自身的感觉会变得非常轻松。同时，我也跑回去领了个奖。在唐院长最后宣布我名字的时候，那感觉还是非常爽的。上台领奖时候还是很兴奋的。难怪那些影星影帝会在颁奖典礼上语无伦次的感谢什么CCTV，MTV之类的。哎，看来如此感谢，如此言语也都是非条件反射啊。

晚上在南苑吃的，建霞弄来了一个漂亮的蛋糕。如果说蛋糕本身没什么，关键在于情意之类的话，那恐怕有点假。毕竟我还是个好吃的主儿。不过跟同门的师兄师妹们一起过生日的感觉真是不错的，温馨而充实。加上我的4+1（此为张院长原创，4为四位学术牛人，1为在下），我真是在生日这一天找到了学术和娱乐的完美结合点。但是，我绝对不认为这是个帕贝托改进，因为，我实在是太累了。

其实坦言，学术研究的确是需要激励的。我很羡慕学术班的人有如此好的条件。但是呢，也为大多数人的身在福中不知福而感到可惜。当然，人各有志，学术也不是唯一出路。正如我以前所言，在这里，我们选择的学术道路可能是曲折的，甚至是郁闷的，在这样的环境下能够坚持下来是要点道行的。所幸，本人遇到了很好的老板，也得到了比较多的激励。在这些激励下，我现在的成果其实还是比较悲惨的。之所以要感谢老板，原因很简单。有句老话“师父领进门，修行在个人”。但是问题是，如果没有师父领进门，您老难道还打算在门外修行不成？当然，释迦摩尼在菩提树下得道。可是，我们皆凡人。

原本以为今年一年都不会有学术会议上的记录，没想到今天却填上了这一空白。尽管这会级别不高，收益却不低啊。

Graduate Students in China: Depression Only

摘要：如果说人生是一场戏，那么在天朝上研究生将会是您人生重要的一课。本文以经济学基础理论为理论依据，通过非现实，调侃等非正常逻辑思维和手法展现了我天朝研究生生活之特色。本文发现，同发达国家，以及大多数新兴市场的研究生生活相比，我们的生活充满了更多的激情与刺激，同时也需要更多的博弈过程。限于篇幅，本文仅抛砖，并未提供严格数理分析以及福利分析。此等可作后续研究之方向。

或许我就是一个后知后觉的人，以至于到了研究生快毕业的时候才对研究生生活，尤其是中国研究生的悲哀，有了一次初步的了解。话说，做学术的人需要淡泊名利，更需要有那种执着到犯傻的地步。现在我懂了。因为在你能够真正的看破红尘之前，想做很好的研究真是难上加难。高水平研究本身已经对人的智商和体能是一个严峻的考验，但是要做到完全抗干扰，却是万万难以办到的。在这里，我们终于知道了一般均衡严格假设在现实条件下的苍白无力，也知道了Nash教授几十年前那薄薄的几张纸的毕业论文所包含的能量。

在转型经济和新兴市场中，监管往往是非常薄弱的，而薄弱的监管往往会带来一大堆的问题，比如人性的丑恶可以极大的得到发挥。因为在这个环境里，具有强制性的法律对过度的行为并没有约束力，而基于人性最本质判断的道德标准则可能由于人文素质等原因而趋于消逝。这其实是一种很危险的状态，因为秩序已经不再存在，任何人都可以做本性的决策。而亚洲教授，无论是在北美，欧洲，大洋洲还是亚洲本身，口碑都不是很好，或许就是基于这样一种制度背景吧。难怪某BBS上有位哥们说，在亚洲出生不是我们的错，但是在亚洲上学那就真是我们自己的不对了。但是在预算约束下，这种行为无疑成为了“被次优化”的一个结果。

作为转型经济的一个特例，我们伟大的天朝，则是更是发扬了这种曲折不挠的精神。基于人民是XX主人的原则，研究生们也就理所当然的成了学校和自己的主人。在亚洲其他地区的那种高额奖学金加补助来招纳Push对象的资本主义策略是我们所不耻的。更多的，我们是自觉的新新人类。我们自己支付自己的学费，然后感激涕淋的领到了10%的回扣，我们享受着1.3-1.5倍的平均物价和房租，享受着农民工般的待遇。更有甚者，著名的投入产出模型在此也失去了诺贝尔奖的光辉，因为这个模型实在无法解释在产出几乎为0的情况下，却有着高额货币资本和劳动力资本的双重投入。此种状况，想必凯恩斯教授从坟墓里爬出来都没法解释啊。算了，还是去求脚下马克思教授和恩格斯专家吧，或许资本论的第四卷题目可以是：Higher Education in Socialism: Asymmetric Role on Input and Output

这里还是个博弈论的试验场。博弈论的精髓“根据对方的策略来决定自己的策略”得到了完美的诠释，贝叶斯均衡在这里有着完美的用武之地，因为你发现先念概率永远是笑话。想考博士么？来吧，我们绝对“以人为本”。怎么说？你有丰富的科研经历？你有很好的外语基础？哈哈，对不起，亲爱的，以”人”为本说的可不是你，是你的推荐人。没有推荐人？你以为你是商鞅？没人推荐，恐怕您的一身武艺只能自己拿个DV拍下来看着玩了。再说了，就算是商鞅，他老人家还不是照样被完美分解了。同样，这里也需要转型经济理论来说事儿。在转型经济下，（省略若干字），Relationship可是最重要的，跟美国教授拿着自己的声誉去推荐人是两码事儿。要知道，我们是礼仪之邦，我们讲究礼尚往来，同时要求迎来送往。

不过既然是转型经济，那就不是个均衡状态，就会有发生转变的时候。事实也如此，在某些经济和人文素质已经发展到了一定水平的某些区域，此类单纯的犯傻行为已经得到了逐步的纠正。Output已经不再是0，货币资本终于开始由方程的一方挪到了另一方。严重倾斜的天平慢慢开始平躺。但是新问题也来了。基于这些地区特殊的治理结构，这些地区中的学校开始了蜜月期。什么叫做蜜月期？非诚勿扰！

谨以此文献给同在一条战壕里的兄弟姐妹们！

注：对阅读此文深感困惑的朋友们而言，尝试一下研究生生活将有助于促进您对本文的理解。

声明：此文系头脑发热胡思乱想基于非理性思维所作，仅供饭后消遣，持不同观点者请绕道。

To Be A Teacher: Why Only For PhD?

Quite occasionally I get the ticket for admission of a employment conference this morning, accompanying with my friends I participated the employment conference as my first time of doing so after the last participation of such activity two years ago in Nanjing. But this conference, frankly speaking, is not a good one, no matter whether it is evaluated under the organization or its size. While, it does offer quite numerous positions for those who are willing to be a teacher.

For those who are holding a master’s degree and with strong willing of being a teacher in a college, the experience today seems to be a dilemma. If you are willing to be a tracked for a teaching position in an acceptable university, a PhD degree seems to be a must. But if you can accept the comparative low salary for the similar position in a technical school, then you have the opportunities of being accepted. While once accept such position, new problems will occur. The low probability of being a professor and tutor for doctoral students, and the enforcement of taking the Ph.D study can all be coming problems that are waiting to be solved.

There has been quite long for me to consider, but never provide a substantive answer that, whether it is necessary for a teacher holding a Ph.D degree, especially for those technical school. It is of no doubt that teachers with good academic training, and holding an advanced degree can obviously help students broaden their horizon, rather than keeping a narrow vision on the skills and knowledge points printed on the books. However, with the concern that the main task for the undergraduate students is to build a solid foundation, I don’t know if the research experience of a Ph.D students or a Master’s student would help. What I argue more practically, is to hire some teachers with working experience in an accounting or consulting firm, who can deliver some practical skills.

However, for the training on academics for graduate students, things are totally different. These students require more training on the research issues, such as the theory explanation, research methodology and some introductory lectures on contemporary research. For undergraduate students on accounting, they are not required to know Williamson, Coarse, Watts, Beaver, etc, but for graduate students, good familiarity on such classical works is to be a must. So with full experience of research can help the lecturer deliver a better and deeper understanding on these issues. For these purposes, Ph.D graduates is a good source to get the candidates of the lecturers of the issues on graduate levels.

This is only quite an initial consideration on the issue of employment for colleges, for any comments, you are welcome to follow and discuss.

Nobel For Chinese Scholars: Is It Still Far?

今天早些时候，瑞典皇家科学院宣布了今年（2009年度）的诺贝尔物理学奖得主。其中我们再一次看到了华人的身影。黄锟，作为一名在光纤通讯领域做出了卓越贡献的物理学家，他继早年的杨振宁，李政道，以及近年的朱棣文之后，又一位华裔物理学奖得主。让我大感意外的是，今年物理学奖的两个项目均是对现当代数码生活起到了极大促进作用的成果。光纤通讯技术极大提高了数据传送的效率，使得国际通讯的成本变得低廉；而CCD的发明，则使得我们今天数码相机的使用成为了可能。与杨振宁、李政道和朱棣文他们的研究项目偏重于理论相比，今年的成果无疑具有更大的实际意义。

按照惯例，在每年的这时候，各类媒体都会大张旗鼓的来讨论诺贝尔奖，其中已经讨论了多年的一个话题就是中国何时能够产生诺奖得主。其中有褒有贬。而解释无非有中国的教育体制路后，实验室设备落后，科研能力低下，没有创新精神等。但是很奇怪的是，这类文章往往出自于媒体之手。本人并不反对媒体的讨论，但是事实上，很多的问题往往非业外人士所能了解，正如陶渊明所言“不足为外人道也”。于是本人决定从自身经历出发，继续老调重弹一番。

不可否认的是，我们的教育制度的确是有些问题的。但是也不尽然。以初等教育为例，我国以及东亚地区的学生（包括日本，韩国和我国台湾）往往都具有很好的基础。这源于这些地区的严格的初等教育。笔者数年前曾在Times亚洲版看到过一篇介绍韩国（South Korea）的孩子们如何痛苦的文章，其中所涉及的韩国，日本等教育与中国大同小异。正是如此严格的基础教育，使得他们在高中毕业之时已经有了相当不错的英语，数学，物理，化学，地理，历史等方面的基础。或许有人说美国的学生在大学里可以非常努力，然后赶上我们的水平，使得我们大学生的毕业水平并不高于他们。

其实这个解释就有问题。我们可以做如下假设：假定人的努力程度有限。而大学期间水平的提高就是关于努力程度的一个函数。为了方便分析，我们忽略掉方法所导致的效率差异，即我们认为学生努力无差异。基于此，如果两者学习努力程度接近，那么水平提高程度应该类似。而事实上，两者的提高水平差异巨大。这就不能不利用努力程度来解释了。其实作为80后的大学生，我们可以扪心自问，我们在大学期间，本科也好，研究生也罢，我们都做了什么。我们学到了什么，我们的大学老师都在讲些什么？作为一个市场经济环境下的教学机构，适当的面向市场并没有错。但是，当一所大学或者学院，甚至研究生院，开始套用职业技术学院的价值观的时候，我们不得不为这个制度感到悲哀了。我实在不清楚这类教育中的短视行为是否是转型经济的特点？望相关专业读者赐教。

正如Joseph和TJ所言，很多问题都可以归集到制度层面去解决。而这个问题也不例外。以今年的两位物理学家的项目为例，他们的研究成果距今已经有了几十年，充分经历了社会的考验，确被证明为对人类做出了巨大贡献。这说明，一个好的研究，需要花时间去做，同时也需要花时间去验证。倘若没有一个开放，自由和包容的学术环境，这是万万不可能的。大家或许不会忘记五十年代的“大跃进”，而事实上，现今的中国学术界正在经历着“大跃进”。从高校老师的职称到研究生的奖学金评定，一律都按照论文来说事。我并不反对这种科研评价方法，只是，他们已经承受了他们不能承受之重。对理科生而言，这种压力更大。于是就产生了从定题到投稿SCI期刊只需短短两三个月的奇迹。尽管说研究是站在巨人的肩膀上，但这种只在肩膀上跳舞，却不会更上一层楼的研究本身就让研究的意义打了折扣。

从浙江大学的贺海波，到武汉理工大学的周祖德，我们都发现了上述“大跃进”的影子。而在这样的环境下，学术道德直接被无视。而作为一个学者，正直和良心是他们的核心竞争力。如果做不到这两点，就不配做一个真正的学者。可是很不幸，我们看到，在出现问题时，教授把所有的责任都推给了学生。试问，这样的教授可能培养出诺奖的得主么？答案诸位心中自有分量。

现如今的中国，已经不比从前。科研的条件和实验器材已经好于过去百倍，甚至于有实验室都已经达到了国际水准。在硬件问题可以得到保障的情况下，我们是否该考虑下人性本身的问题呢？只有这些问题都做好了，我们才有可鞥培养出自己的诺奖得主。否则只能是任重而道远。

Accounting Research in China: Graduate Students

It has been approximately two years since the very first touch on the contemporary research of accounting through a talk to my present tutor, Prof. Wang, a young scholar. Until then I found the existence of a new methodology called “Empirical Research”, a kind of methodology that adapts the data from the capital market for some test on the conclusions from other kind of papers, usually analytical ones in US and narrative ones in China. But the popularity of such methodology apparent brings the students of accounting, like us, great challenge with following reasons:

Most graduate students of accounting here in China, is mostly owning the background of arts, that is, they seldom getting full and sufficient training on science, such as computer programming skills and mathematical models. But unfortunately, the very fundamentals of empirical research requires skills on programming and mathematics, otherwise, you may have to confront great trouble when dealing with data and models.
Without sufficient knowledge on mathematics, they seldom have good opportunity to take some course on branches of mathematics, such as real analysis, game theory, differential equations, etc. These courses provide fundamentals for research tools in the subject of economics, which is widely used in analytical accounting research. That can partially explain why there is little analytical researches in China.
Unlike the students in US that are entitled of excellent computer competence, we feel great trouble on programming due to insufficient trainings in high school and college. Undoubtedly, most modules that we use are available through the internet, but we fail to get a chance for some modifications that promote our efficiency. And, improper program sometimes can even abandon a good works that should have been published sometimes. The only cause for this is the poor computer competence.

Though so many problems occur, they can hardly be obstacles for researchers in accounting, except for the students like us. The limitation on skills of data management and statistical model selection can be easily solved through the cooperation among scholars, which is getting popular in China. But such cooperation among students is still quite seldom to see. I personally interpret it as the following reasons:

Very few students can fully master the knowledge needed for a qualified research under a present style of higher education that enforce much on books but seldom focus on usage.
Cooperation can cause some conflicts, such as the name rank for publications, which are to be crucial criteria for researchers and students.

Besides the problems on technology and background knowledge, funding should also not be overlooked. As a graduate student, getting the chance for some top-level academic conference seems to be beneficial. But such opportunity is quite rare in China especially for the students who are qualified and willing to join some international conferences. For this, I often admire my friends in Hong Kong and Taiwan since their abundant funding from their department. An apparent example for this is the refusal of funding for my participation of APJAE symposium from a leader of my department last year, with the only reason that I am a student, but not a faculty!

For followers who are willing to pursue their masters’ and PhD degree on accounting, you are suggested to get a good professor, and prepare yourself sufficiently for fundamental knowledge. They are both easy to solve. But once you are willing to learn more from others from conferences, it can be difficult under most occasions unless you are following a good professor who are about to pay the travelling cost for you.