Housekeeping

Who am I?

• My name is Andrew Pua.

  • Assistant professor in quantitative economics at Xiamen University
  • Focuses on methodological research and the analysis of data which track subjects over a period of time

• I usually teach difficult courses to heterogeneous audiences, so it is hard to perfect the material.

• I expect students to push themselves out of their comfort zone.

Course materials

• We have the Teams channel ECONOMETRICS.
• I have set up a separate website for the course materials, resources, and links of interest.
• There may be pre-recorded videos for the course, but I will minimize the use of such videos. I understand that there are other things which compete for your attention.

• If you want to have a video recording of the lecture, please make a request in advance.
• I can make a video recording provided that there are legitimate reasons for you missing a meeting.
• If I write things on a virtual blackboard, I will share these with all of you.
• The best way to “replay” the lecture, is to “replay” what you thought you heard and processed. Make notes, ask questions, and invest in reconstructing your understanding of the material.

• Cosma Shalizi makes the following arguments against video recordings:

  • “… the value of class meetings lies precisely in your chance to ask questions, discuss, and generally interact. (Otherwise, you could just read a book.)”

  • “They tempt you to skip class and/or to zone out and/or try to multi-task during it. (Nobody, not even you, is really any good at multi-tasking.) Even if you do watch the recording later, you will not learn as much from it as if you had attended in the first place.”

• Slides are usually released a week or so ahead of time.

  • They are usually in draft mode, meaning they could be changed depending on our circumstances.
  • I can assign readings but I have not been successful in encouraging students to prepare in advance. So, the website, the slides, and the weekly preparations should give you signals of where we are going.

Reaching out

• Email me at andrewypua@outlook.com.

  • Please put me in your email whitelist.
  • I will try to respond within two working days.
  • If I do not, please send a reminder.
  • I check my spam folder about twice a week.

• Starting from September 29, Microsoft Teams office hours are every Thursday and Friday from 1400 to 1430 Bari local time. Bring lunch/dinner/snacks if you want.

  • Just join the meeting if you want to discuss course-related material.
  • Do not use the private messaging feature in Teams.
  • Attachments do not work very well in Teams.

Reaching out: an example

• One of your classmates reached out to me about R installation and an exercise in fasteR.

• What they did was to reach out by email, attach clear screenshots of the problem and provide background for the problem.

• Even more important is the sincere attempt to address their problem locally first, Googling stuff and tracking similar issues. They also reported their experience and their own processing of these aspects.

• The problems were:

  • Warnings from an R installation about the language encoding
  • Typos in Lesson 4 about the years when the Nile’s flow level exceeded 1200.
  • An exercise in Lesson 10 which was not formulated well

• I am sharing these with you in case you have encountered similar problems.
  

• As a result, communication was smoother and my response was faster (no pun intended)!

How should you ask questions?

• Questions related to the course are definitely encouraged, but …

  • Show that you have thought about the question so that our conversations will be more fruitful.
  • Try to find your own answer first to your question. Sometimes the answer is already available in the materials.

• Asking questions in our Teams channel ECONOMETRICS (rather than privately) is for the benefit of everyone.

Private vs public?

• Asking questions publicly is meant to improve communication skills in a virtual environment and encourage cooperation amongst yourselves.

• Screenshots can help, but take time to think about how to formulate and write your question.

  • Describe the issue.
  • Provide an example.
  • Discuss how you have approached the issue.
  • What errors did you encounter?

Asking questions in class

• Feel free to interrupt with a question during the lecture. Your classmates will more or less thank you for the distraction. There is no need to raise hand anymore.
• If you are shy, that is ok. At the beginning of the class, I will supply a Witeboard link. This will be our burner whiteboard.
• There is no need to sign up/create accounts for this burner whiteboard.

What is this course all about?

• They call this course applied econometrics.

• From Mastering ’Metrics book flap: “Applied econometrics, known to aficionados, as ’metrics, is the original data science. ’Metrics encompasses the statistical methods economists use to untangle cause and effect in human affairs.”

• From the introduction: “Economists’ use of data to answer cause-and-effect questions constitutes the field of applied econometrics, … The tools of the ’metrics trade are disciplined data analysis, paired with the machinery of statistical inference.”

• From Maddala (2013): Econometrics is “the application of statistical and mathematical methods to the analysis of economic data, with the purpose of giving empirical content to economic theories and verifying them or refuting them.”

  • Econometrics is already an applied thing!!

What would be covered?

• Based on coordination with the instructors of Applied Econometrics 2, I was broadly tasked to cover linear regressions and instrumental variables.

• Since I am handling a heterogeneous audience, I need to cover the most important parts of probability theory and statistical inference, which are usually prerequisites of an econometrics course.

• But the focus and sequencing are different from the standard approaches.

• I emphasize:

  • the data on hand and the data you wished you had
  • what could be learned from applying a method or an algorithm, whether blindly or with the help of additional assumptions
  • computational aspects which you may encounter in your own work
  • recent developments and lines of thinking in linear regression and instrumental variables

What should you be able to do during and after the course?

• Understand conceptually and intuitively core econometric methods with a focus on whether they can be applied to a particular empirical context

• Be comfortable enough with the common mathematical and statistical thinking driving these methods in order to apply, build, and extend them

• Be able to move from a theoretical question or a theoretical model to a relevant econometric model and the econometric method to be used

• Be able to interpret the results and not just be able to push buttons using econometric software

• Be prepared enough for Applied Econometrics 2 and your future research projects

What books to use?

• This is hard to answer because it depends on your tastes too.
• Furthermore, you have to be discriminating about the books you read.
• I draw from multiple sources and it is not easy to find a treatment that balances aspects of applied econometrics.
• In the next slide, I give some suggstions for books.

• Everything in probability and statistics until regression:

  • Dekking, Kraaikamp, Louphaä, and Meester, A Modern Introduction to Probability and Statistics. Springer-Verlag, 2005.

  • Wasserman, All of Statistics: A Concise Course in Statistical Inference. Springer-Verlag, 2004.

  • Aronow and Miller, Foundations of Agnostic Statistics. Cambridge University Press, 2019.

• More or less standard textbooks:

  • Wooldridge, Introductory Econometrics: A Modern Approach, 6th ed. Cengage, 2016.
  • Stock and Watson, Introduction to Econometrics, 3rd ed. Pearson, 2013.
  • Goldberger, Introductory Econometrics. Harvard University Press, 1998.
  • Verbeek, A Guide to Modern Econometrics, 5th ed. John Wiley & Sons, 2017.

• Sticking to a particular flavor:

  • Huntington-Klein, The Effect: An Introduction to Research Design and Causality. CRC Press, 2021.

  • Cunningham, Causal Inference: The Mixtape. Yale University Press, 2021.

  • Angrist and Pischke, Mastering ’Metrics. Princeton University Press, 2014.

  • Angrist and Pischke, Mostly Harmless Econometrics. Princeton University Press, 2009.

• Odd men out:

  • Berndt, The Practice of Econometrics: Classic and Contemporary. Addison-Wesley, 1991.
  • Freedman, Statistical Models: Theory and Practice, 2nd Edition. Cambridge University Press, 2012.

Isn’t math harder than application?

• On the surface, maybe.

  • Part of your graduate education is to look beyond the surface. Being discriminating but not dismissive is the key.
  • It is easier and more tempting to cut corners in applied econometrics, but doing mathematics cuts corners by making assumptions.
  • Laying out assumptions on paper make assessment relatively easier. Inductive reasoning and our own prejudices shape the way we apply things.

• In my experience, applied econometrics may be harder than econometric theory in three senses:

  • (From Edward Leamer) “… economic theory is fiction, data analysis is journalism.”
  • The burden and accountability of applied econometrics seems higher, as you do not know how other laypeople would use your findings.
  • It is hard to think carefully about the world we actually live in.

• In the words of Angrist and Pischke (2017), “Econometrics at its best is distinguished from other data sciences by clear causal thinking.”

• A counterpoint is from Diebold’s blog entry: There is another “half of econometrics” which emphasizes careful modeling.

• Bottom line: Applied econometrics has both elements!

Course grading

• The final course grade is based on

  • a final exam (60% weight)
  • individual paper (30% weight)
  • weekly assessments (10% weight).

• The passing mark for the course is 60%. I use a 0-100 scale.

  • For EGEI students, I will be converting to the Italian scale. I will update you soon about this.
  • For IGP students, the 0-100 scale is the default scale.

Weekly assessments

• One part of weekly assessments involve practice quizzes.
• Another part of the weekly assessments involve exercise sets.
• As long as you sincerely attempted a solution (the so-called good old college try), no matter how wrong or embarrassing, you get automatic credit if asked to submit before the deadline or asked to present solutions and findings.

• You may be asked to share your screen (might be better for you to use the Teams app) when presenting or discussing your solutions.
• If you copy or buy solutions from anywhere, then that is an automatic failure for the course.
• You are allowed to discuss with each other, but you have to give credit for ideas, approaches, and resources you did not independently discover.

Why are weekly assessments designed this way?

• To make you aware of your mental mindset (see Farrington (2013)):

  • “I belong to this community.”
  • “I can succeed at this.”
  • “My ability and competence grow with my effort.”
  • “This work has value for me.”

• To give you a chance to look inward

  • Why were you asked these questions in the survey?
  • How did you go through fasteR?
  • How does your attention span work?
  • Why was the assignment designed this way? What were you supposed to do?
  • How much time do you spend?

• To expose yourself to others
• To help identify misconceptions in a just-in-time fashion
• To get chances to apply what you have learned in a low-stakes, possibly uncertain, environments

The individual paper

• Submit during the final exam date in a reproducible format.

• Details about this format will be made available as we progress through the semester, as you will be using R and R Markdown (or Quarto) along the way.

• The grading of the individual paper does not depend only on the final output, but it also depends on the milestones which are to be set as we progress through the term.

  • The weekly assessments and the assigned tasks before our weekly meetings all provide useful inputs into the progress over the term.

• Milestones include but are not limited to:

  • Choosing a baseline article
  • Evaluating your understanding of the baseline article
  • Gathering the data by following instructions
  • Conducting the analysis
  • Writing the report

• Milestones are meant to spread accountability over the entire term, to keep you on your toes, and for you to have something concrete and real at the back of your mind as we progress through the course.

Final exam setup

• The final exam is set up as follows:

  • graded per individual
  • in written form and closed book format
  • includes R

• It is tentatively scheduled on January 2023.

• You are expected to know how to read R code, program in R, and understand R output based on the topics discussed.

• You are also expected to know details about your individual paper.

Why is the final exam set up this way?

• It is extremely difficult to assign credit to collaborative tasks, no matter how well designed.
• Typically, statistical software will only be used when a student finds it necessary.
• Having self-study and regular practice over the term and making that practice part of assessment provides incentives for follow through.
• Including details of your individual paper is one way to assign credit for the process.

Attendance policy

• I do not check attendance every meeting.
• But, if you miss four classes without advance notice and sufficient reason and I detect it, then you fail the course automatically.
• You can choose not to attend classes, but there are consequences whether or not they bind for you or not.
• I encourage sleeping in class, with your camera turned on.

Academic and research integrity

• There are rules and regulations which set boundaries on what members of the academic and research community could do.

• EGEI and IGP have their own rules and you should spend time being aware of them.

• Why should you care about academic and research integrity?

  • Education is not a theatrical affair, although humans have a fondness for theatre.

• I would argue beside these rules and regulations as follows:

  • Your degree is only as good as what you, your peers, and all other members of the community have put into it.
  • Sometimes even the suggestion of wrongdoing, however defined, can taint one’s reputation and to some extent, other’s reputations as well.
  • Actively changing one’s habits and workflow go a long way.
  • When you find yourself asking “Is this kosher?”, stop and reach out to us.