Files written in English from Website of Keiji Saito 
I collect files written in English from Website of Keiji Saito.
Sitemap (by category) is a simple collection of links.
Sitemap (updated order and update info) 
shows update informaton.
COE Discussion Papers [COE-F-81]
The university of Tokyo, Economics
A fallacy of wage differentials: wage ratio in distribution
[study field: education, labor economics]
(PDF, Version1.0, approx. 5000words, 730KB: fallacy_wage_differential.pdf)
- This paper demonstrates a fallacy of wage differential between educational
statuses.
- I upload Simulation of Ratio between Averages in the archive. (Excel:en_average_ratio.xls)
COE Discussion Papers [COE-F-82]
The university of Tokyo, Economics
Do schools form human capital? Distributional divide and cohort-based analysis in Japan
[study field: education, labor economics]
(PDF, Version1.0, approx. 6600 words, 700KB: human_capital_cohort_japan.pdf)
- This paper presents empirical wage analyses with Basic Survey on Wage
Structures of Japan.
summary in PDF
Learn fallacy of wage differentials with cohort data in 10 minutes
[study field: education, labor economics]
(PDF, 140KB: wage_human_capital_10en.pdf)
- This paper is 6-page summary of the above two papers.
- Even if you are unfamiliar with economics and mathematics, you can see
the key points in 10 minutes.
summaries in HTML
Making Histogram in Excel - Just Paste (en_paste_histogram.xls)
- Just paste your data in the Excel.
- You can choose maximum, minimum, and interval count in the histogram.
- The output graphs are [Original samples], [Frequency Histogram], [Share Histogram], and [Cumulative Distribution]
- The maximum value of sample size is 65518. The maximum value of interval count is 100.
Simulation of Ratio between Averages (en_average_ratio.xls)
Visual Statistics and Econometrics by Monte Carlo
- Random numbers are regenerated with holding down [F9].
Visual Statistics and Econometrics by Monte Carlo
- I write important comments in both Japanese and English.
- We can assign arbitrary values in Input Part.
- These files show the degree of bias and accuracy in each setting.
Versatile do-file
- I write important comments in both Japanese and English.
- These do-files are self-running by making sample.
Calculation in Each Group (calc_each_group.do)
- Calculation of arithmetic mean, geometric mean, and cumulative value in
each group.
Two-stage Summary Table (two_summarize.do)
- Summary table by two-dimensional group.
- The output statistics are [non-missing, mean, s.d., min, percentile1-percentile99,
max],
Ranking (ranking.do)
- If three IDs get the highest points, the ranking of the second highest
points is 4.
F test by Coefficient of Determination (Ftest_R2.do)
- If we use many fixed effects, this might be useful.
Histogram and Export (histogram_out.do)
- Though Stata has "histogram" command, this might be useful to export the histogram.
Correlation Matrix by ID and Round (corrlationM.do)
- Correlation matrix and meeting count by each i and j.
Loading CSV: The first row contains the vector names (load_ncsv_en.m)
- This script loads CSV using the first row as the variable name. (e.g. sample_co.txt)
- If your data is Excel, insert the variable names in the first row and save
as CSV in advance.
- This script can load space-delimited (e.g. sample_sp.txt). In that case, delete the line head % in [% Data= dlmread(char(Datafile),' ',1,0);].
- You'll be able to load mat-format with one line after you save once.
![[Sample Output]](arc/stata/c_m_atte.png){kind=link}
![[Sample Output-1]](http://keijisaito.info/arc/stata/ols_iv.png){kind=link}
![[Sample output-2]](http://keijisaito.info/arc/stata/ive_1st.png){kind=link}
![[Sample Output]](arc/stata/sele_heckit.png){kind=link}