联系方式

  • QQ:99515681
  • 邮箱:99515681@qq.com
  • 工作时间:8:00-21:00
  • 微信:codinghelp

您当前位置:首页 >> Database作业Database作业

日期:2024-12-12 05:13

ECN 3620 Econometrics

Fall 2024

Course Wrap Up

Thank you for taking Econometrics with me this semester. I certainly enjoyed this class, and I hope you feel the same way.

R Basic

 

Import data

 

Generate new variables

 

Create graphs

 

Get sample statistics

Basic Statistics

 

Sample distribution and population distribution

 

Standard Normal distribution and t distribution

 

Jarque-Bera test and related concepts

 

Find corresponding probability and critical values from the Z table

 

Value at Risk

 

Central Limit Theorem and confidence interval

 

Estimate vs Estimator

Simple Linear Regression

 

Coefficient related diagnostic: t test and p value

 

Hypothesis test and confidence interval of coefficient

 

R2, adjusted R2, and its components

 

Standard Error of Estimate vs. Standard Error of Forecast

 

Within sample and (pseudo) out-of-sample forecast

 

MAE (Mean-Absolute-Error), RMSE (Root-Mean-Square-Error), MAPE (Mean-Absolute- Percentage-Error)

Multiple Linear Regression

 

Test linear combinations of parameters: e.g. H0  : -β1  = β2 or H0  : 2β1  = 3β2  +12

 

Joint significance test: F test

 

Variable selection

 

Dummy variables, interaction of dummy variables with other variables

 

Residual related diagnostic: Homoscedasticity vs Heteroskedasticity

 

Applications: hedonic pricing, seasonality and trend, Interrupted Time Series design (ITS)

Special Topics and Models in Multiple Regression

 

Omitted variable bias: the direction of omitted variable bias

 

Multicollinearity: symptoms and remedy

 

Models with low R2

 

Nonlinear models: LnY = a + b X; LnY = a + b LnX

 

Probability models: linear, logit, and probit

 

Probability models: odds and odds ratio (optional)

 

Probability models: marginal effects, partial effects

Causality Models

 

Causality problems

 

Interrupted Time Series (ITS): graphs, regressions, and interpretations

 

Difference-in-Differences: graphs, tables, regressions and interpretations

Time Series Models

 

Components of time series data

 

Lag function and difference function

 

Mean stationary, first and second difference

 

AR, MA, ARMA, and ARIMA models

The following is a checklist of Econometrics modeling when you start your project:

1. Do you have relevant data for the question you are after? Do you have enough observations (at least 30 or so)?

2. If you have data, is there error in the data? You can check mean, maximum, minimum. Graph the data and see whether there are outliers.

3. Are you using the right unit of measurement? This is especially important when you are doing medical and healthcare research.

4. What types of data do you have? Time series, cross-sectional, panel, other?

5. If the data is cross-sectional or panel, you are most likely to choose a structural model, in which case, you should check:

a. What independent variables should be included? Are you imposing a causality relationship? If so, is it valid?

b. What functional form. are you employing? Linear or nonlinear? Why?

c.   Are   the   estimated    coefficients   consistent   with   theory   or   your expectations? If not, what can explain the difference?

d. What is the model’s explanatory power? If it is low power, are the coefficients biased? Can you still use the parameters to forecast or make policy and business decisions?

e. Is multicollinearity a problem?

f. Does the error term satisfy homoscedasticity? Is there a serial correlation in the error term?

6. If the data is a time series, you are most likely to choose a time series model, in which case, you should check:

a. Graph the data. Is it at least mean-stationary? Are the first difference, second difference, seasonal difference, or log transformation needed?

b. After necessary conversion, what is the correlogram of the data? What does it tell you about low-order and high-order correlations?

c. Use AIC or SIC to find the appropriate model.

d. After comparing a series of test statistics and forecasting evaluations, fine-tune the model.

e. Is the residual white noise? Conduct forecasting.

7. In some cases, you may have forecasts from the structural model, time series model, and judgment forecasting from the experts at the same time. Then, your best forecast  will  most  likely  be  an  average  of the  three.  This  is  often  called  ensemble forecasting.

Where can I get more resources: data, books and websites?

One of the most asked questions is where I can get more resources such as data, books, or websites for more information on Econometrics.  Here is a list of resources you may find helpful and interesting.

Data Resources

IPUM:https://ipums.org/

Integrated Public Use Microdata Series. IPUMS provides census and survey data from  around  the  world   integrated   across  time  and  space.  IPUMS  integration  and documentation  make  it  easy  to  study  change,  conduct  comparative  research,  merge information  across data types,  and  analyze  individuals within  family  and  community context. Data and services are available free of charge.

ICPSR: (http://www.icpsr.umich.edu/icpsrweb/ICPSR/)

Inter-University Consortium for Political and Social Research is an international consortium  of  about  700  academic  institutions  and  research   organizations.  ICPSR maintains a data archive of more than 500,000 files of research in the social sciences. It hosts  16 specialized collections of data in education, aging, criminal justice,  substance abuse, terrorism, and other fields.

Current Population Survey:http://www.census.gov/cps/The Current Population Survey (CPS), sponsored jointly by the U.S. Census Bureau and the U.S. Bureau of Labor Statistics (BLS), is the primary source of labor force statistics for the population of the United  States.  The  CPS  is  the  source  of  numerous  high-profile  economic  statistics, including the national unemployment rate, and provides data on a wide range of issues relating to employment and earnings. The CPS also collects extensive demographic data that complement and enhance our understanding of labor market conditions in the nation overall, among many different population groups, in the states and in substate areas.

CRSP: (http://www.crsp.com/)  provides monthly, quarterly, or annual updates of end-of-day and month-end prices on all listed NYSE, AMEX, and NASDAQ common stocks with basic market indices. Available on all Cutler workstations.

WRDS:   (http://wrds.wharton.upenn.edu/)   Wharton    Research   Data    Services (WRDS) is a web-based business data research service from The Wharton School at the University of Pennsylvania. It is known for its holdings of historical financial data from CRSP and COMPUSTAT. This data covers over 30,000 companies and includes security prices and trading volume, income and balance sheet items. WRDS also contains stock market indices, interest rates, mutual fund and executive compensation data, and a wide array of macroeconomic time series.

Bureau     of     Labor      Statistics,      Bureau      of     Economic      Analysis: (http://www.bls.gov/,  http://www.bea.gov/)   generally   macroeconomic   data   such   as employment rate, wage rate by region, consumer price index, GDP by region, Import and

Export etc.

Economagic:      (https://fredaccount.stlouisfed.org/public/datalist/159?pageID=8) there are more than 200,000 time series for which data and custom charts can be retrieved. Though the greatest utility of this site is the vast number of economic time series, and the easily modified charts of that same data, an overlooked facility of great utility is the availability of Excel files for all series.  The majority of the data is USA data. The core data sets involve US macroeconomic data (that is, for the whole US), but the bulk of the data is employment data by local area -- state, county, MSA, and many cities and towns.

Economic  Data    FRED:   (http://research.stlouisfed.org/fred2/)   Welcome  to FRED® (Federal Reserve Economic Data), a database of 19,599 U.S. economic time series. With FRED® you can download data in Microsoft Excel and text formats and view charts of data series.

US  Census:  (http://www.census.gov/)  public  resources  from  the  US  Census Bureau including population, economic, industry, and geography studies. The information can be accurate at zip code level.

MEPS: (http://www.meps.ahrq.gov/mepsweb/) The Medical Expenditure Panel Survey (MEPS) is a set of large-scale surveys of families and individuals, their medical providers, and employers across the United States. MEPS is the most complete source of data on the cost and use of health care and health insurance coverage.

NHANES:   (http://www.cdc.gov/nchs/nhanes.htm)   The   National   Health   and Nutrition Examination Survey (NHANES) is a program of studies designed to assess the health and nutritional status of adults and children in the United States. The survey is unique in that it combines interviews and physical examinations.

Pew  Research  Center:  (http://people-press.org/dataarchive/)  A   collection  of survey data from Pew Research Center For The People & The Press. Survey data are released five months after the reports are issued and are posted on the web as quickly as possible.

Books

Business Forecasting (5th edition) J. Holton Wilson and Barry Keating

*Introductory Econometrics: a Modern Approach, by Jeffery Wooldridge (pre- bundled with the student version of Eviews).

*A Guide to Modern Econometrics by Marno Verbeek

Econometric Analysis (5th Edition) by William H. Greene

Introduction to Econometrics by James H. Stock and Mark W. Watson Analysis of Financial Time Series by Ruey Tsay

*Applied Econometric Times Series (3rd edition) by Walter Enders

Introductory Econometrics for Finance by Chris Brooks *Stands for my personal favorite.

Additional Resources on Using R

If you want to learn R programming, the following are recommended readings. They are all freely available on the internet.

•   Forecasting: Principles and Practice, Rob Hyndman and George Athanasopoulos

https://otexts.com/fpp3/

•    Using R for Introductory Econometrics, by Florian Heiss

https://www.urfie.net/

•   Applied Econometrics Time Series, Walter Enders

https://time-series.net/home

•   R for Data Science, Hadley Wickham and Garrett Grolemund

https://r4ds.had.co.nz/

•    UCLA R resources

https://stats.oarc.ucla.edu/r/

•   Econometrics Academy

https://sites.google.com/site/econometricsacademy/

Websites

UCLA Academic Technology Services:

http://stats.idre.ucla.edu/  A  website  by  the  Institute  for  Digital  Research  and Education at UCLA. It has lectures, examples and videos on R, SAS, SPSS, and STATA.

Econometrics Academy

https://sites.google.com/site/econometricsacademy/home?authuser=0

The Econometrics Academy is a free online educational platform and non-profit organization. Its mission is to offer free education on Econometrics to anyone in the world.

Using Python for Introductory Econometrics

http://www.upfie.net/

This book introduces the popular, powerful and free programming language and software package Python with a focus on the implementation of standard tools and methods used in econometrics.

Using R for Introductory Econometrics

http://www.urfie.net/

This book introduces the popular, powerful and free programming language and software package R with a focus on the implementation of standard tools and methods used in econometrics.

IBISWorld

https://ezproxy.babson.edu/login?url=https://my.ibisworld.com

Search by NAICS code or keyword to find thousands of U.S. industry research reports, includes Global Industry reports with some China coverage.

The Economist:https://libguides.babson.edu/economist

•   The app and economist.com—distinctively distilled analysis

•   Digital newsletters—curated topical opinion

•   Audio version & podcasts—immersive listening

•   The digital archive—all our content since 1997

•   Webinars and conferences—intelligent debate and informed analysis

•   Flagship franchises—The World in and 1843 magazine

WSJ Economic Forecasting:

http://online.wsj.com/public/page/economic-forecasting.html

A collection of forecasting on US macro-economy including GDP, unemployment rate, housing, inflation. Forecasts are from various resources.

Institute of Business Forecasting:

www.ibf.orgoffers a variety of programs for business professionals and quarterly Journal of Business Forecasting: Methods & Systems a jargon-free journal on forecasts.

Forecasting Principle:

www.forecastingprinciples.com/  The  Forecasting  Principles site summarizes useful knowledge about forecasting so that it can be used by researchers, practitioners, and educators. It has link for researchers, practitioners and educators, and databases.

Federal Forecasters Consortium:

http://www.va.gov/HEALTHPOLICYPLANNING/FFC_2014.asp

The Federal Forecasters Consortium is a collaborative effort of agencies in the United States Government, as well as other interested parties in the academic and not-for- profit communities, who share an interest in the practice, planning, and use of forecasting activities by and within the Federal Government.

Science Direct:

http://libguides.babson.edu/content.php?pid=17543&sid=1839426

Select Science Direct. You need log in using your Babson email and password. It is the world's largest electronic collection of science, technology, and medicine full-text. It has over 2,500 peer-reviewed journals and more than 11,000 books. There are currently more than 9.5 million articles/chapters, a content base that is growing at a rate of almost 0.5 million additions per year.


 

版权所有:编程辅导网 2021 All Rights Reserved 联系方式:QQ:99515681 微信:codinghelp 电子信箱:99515681@qq.com
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。 站长地图

python代写
微信客服:codinghelp