制作回归表格的10个戒律
中文翻译简版：
如果你是我的学生或者我是审稿人，以下是必须的：
报告观察值的样本量、R2、残差的方差（Root MSE in Stata）
报告因变量和模型估计方法
因变量的标签能够直接体现变量的名称和含义（不要用字母标签）
为变量选择合理的单位
演讲时用的表格要能够看清
在同一列中报告系数的标准误，而不是在系数边另起一列
打印版本的论文中在正文中插入表格，而不是把所有表格都放到文末。
尽量报告标准误，而非T值或P值
用上标a,b,c代替 * 号来表示显著性水平，以结省表格空间
只报告双尾检验的显著度
（编者注：以上10条除了第9条持保留意见，其他的我表示都赞同，特别是第1、3、4，5。回归表格应该能够让读者在不用再去阅读上下文就能看懂。）
详细英文版如下：
The 10 Commandments for Regression Tables
By Keith Head
These commands are organized according to how controversial I think they might be. If you are my student or I am your referee these commands are not optional.
1.Report the number of observations, the r-squared, and the root mean squared error for each regression.
2.Report the dependent variable and the estimation method. in the table’s captionif it is common to all specifications or as a column heading if it variesacross specifications.
3.Use self-explanatory labels for your explanatory variables. Crypticabbreviations or symbols from the model section force the reader to page backand forth to understand your results. With five or six columns of regressionresults there should be enough room to use words to describe each regressor.Put the symbol used in the model in parentheses below this.
4.Choose sensible units for variables. The coefficients should not be very small(e.g. 0.000032) or very large (e.g. 75432.8). As a rule of thumb, coefficientsshould only use the first two or three places to the left or right of thedecimal point. One exception is the case where variables are unit-free becauseyou are estimating a log-log model. In that case coefficient size is inherentlymeaningful.
5.The presentation version of the table should be in large type. Don’t show atable full of tiny numbers and say “I know you can’t read this but…” Ifnecessary, place some of you control variables in an auxiliary table so you canfocus attention on the variables of interest.
6.Put standard errors in the same column as the coefficients. Regression packagesput standard errors along side coefficients as separate columns but you shouldput each regression as a single column in your results table. Columns should beused for 4-8 alternative specifications and samples. Thus the standard errorshould appear below the related coefficient in parentheses. Use the Stataestout package . You have to install this by typing the following code on the Statacommand line:
ssc install estout, replace
7.Insert key tables inside the body of the paper. Journals insist upon tables atthe end for the final submitted version of the paper. This does not mean youshould do it for working papers or first submissions. There is a reason why theprinted version of your article puts the tables back into the text: it iseasier to read a paper that way without having to constantly flip to the end tofind results and then flip back to the text for interpretation. By putting thetables in the text you will also be more aware of whether your paper has theright mix of text and tabular information.
8.Display standard errors, not t-statistics or p-values Unless the test that thecoefficient is not equal to zero is the only conceivable test of interest,display standard errors. These give readers a direct view of the precision withwhich you are estimating the coefficient. They are useful information for avariety of possible tests and are still valuable even if the reader prefers notto engage in classical hypothesis testing at all.If my arguments have notpersuaded you, let me appeal to a higher authority:
We’renot only (or even primarily) interested in formal hypothesis testing: we liketo see the standard errors in parentheses under our regression coefficients.These provide a summary measure of precision that can be used to constructconfidence intervals, compare estimators, and test any hypothesis that strikesus, now or later.
Angristand Pischke, Mostly Harmless Econometrics, p. 302
9.Use “a” (1%), “b” (5%), and “c” (10%) superscripts to show statisticalsignificance, if you show it at all. Using multiple asterisks (***) to displaystatistical significance wastes scarce horizontal space in a table. In myperception, a table stuffed with asterisks looks like you are showing off. Ifyou really like asterisks, and there is something to be said for followingcommon practice, then just pick a level of significance (five percent orpossibly one percent) that seems appropriate for your study and then use asingle asterisk for that level. You could use a squiggle for coefficients thatare only marginally significant. Another approach that is gaining favour is toput the significant coefficients in bold font.
10.Report significance for two-tailed tests only. You may think it is OK to useone-tailed tests if your theory tells you the sign of the coefficient. However,this potential justification is overwhelmed by the common practice of usingtwo-tailed test criteria. Many readers, will view the use of one-tailed testsas a cynical ploy to exaggerate the significance of your results. Thus, withinfinite degrees of freedom, variables are significant at the 5% level fort-stats over 1.96, NOT 1.645.
——END
英文内容转自：
http://blogs.ubc.ca/khead/research/research-advice/regression-tables
相关阅读：