Rules for Parsing Unsigned Arrow Diagrams

For OLS to be unbiased, you need Cov(X,epsilon) = 0 (exclusion restriction for OLS)

For IV to be unbiased, you need

  • Cov(Z,epsilon) = 0  (exclusion restriction for IV)

AND 

  • Cov(Z,X) is not zero    (relevance of the instrument)

How do you tell if a covariance is zero or not zero?

It works the same way for all 3 cases. Let me call the two things you want to know if the covariance between is zero or not A and B. (A and B come from the set {X,epsilon,Z). 

The covariance is NOT zero if EITHER

  1. There is a path following the one-way signs from A to B

  2. There is a path following the one-way signs from B to A

  3. There is a path following the one-way signs from something else to A, and a path following the one-way signs from that same thing to B. 

To show a covariance is zero, you have to check a lot of things. You need:

  1. There is no path following the one-way signs from A to B

  2. There is no path following the one-way signs from B to A

  3. There is no other thing from which there is a path following the one-way signs to A, and from which there is a path following the one-way signs from that same thing to B. 

The Analysis Task

The Analysis Task is now posted on Canvas.

Understanding the data:

This link takes you to the public Dropbox folder with the data files. Start by looking at the README file. Our Well-Being Measurement Initiative Research Assistant Jeffrey Ohl can answer your questions: johl@umich.edu Make sure to include Jeffrey's email address on any question about the data. He'll do most of the answering about the data himself. You can do almost anything for the analysis task; it just needs to be interesting.

  1. This is a link to take the Baseline survey so you can understand what data is available and what questions the data are based on: https://wiagl.gitlab.io/survey-baseline/?workerId=[enter your name, or your number plus same random numbers]

  2. This is a link to the Life & Psyche survey so you can understand what data is available and what questions the data are based on: https://ucla.qualtrics.com/jfe/form/SV_8kK2HMh6YrGSEF8. This is the survey that has most of the psychological indexes. (Baseline only has a few.) It has some other miscellaneous questions, too. Only some of the people who did Baseline went on to do this survey.

  3. This is a link to take the Bottomless HIT survey so you can understand what data is available and what questions the data are based on: https://wiagl.gitlab.io/survey-bottomlessa/. Only some of the people who did Baseline went on to do this survey (an overlapping, but different subset than those who went on to do the Life & Psyche survey.) You don't have to do all of this—just keep going until you have an idea for what analysis you want to do. The very first Block is a repeat of what is on Baseline, but it gets different after that.

Relevant Powerpoint File:

The analysis task is due by 11 PM Saturday, March 18. It needs to report the analysis with tables or figures and also have text that clearly explains the analysis. The idea is that this is like one section of a paper.

If you have an idea of what to do for the analysis task, just send me and Jeffrey an email and I'll give a reaction of how interesting I think it is, and maybe a suggestion for a tweak.

Seeing the analysis and its explanation as one section of your term paper. (Your term paper is due later, at 11 PM on Wednesday, May 3—the evening after the last class.) The idea is to make this analysis part of a larger discussion.

Including figures and tables, the analysis task should be at least 5 pages. I'll take a risk and not put an upper limit on the length of the analysis task. (The term paper beyond the analysis task should be between 5 and 10 pages, with closer to 5 being preferred.)

How to structure your writeup of the Analysis Task:

You can design a different structure, but a typical writeup could look like this:

  1. Here is an interesting question or questions. The answers matter (people care or should care) because: …

  2. Here is a statistical analysis that seems to have some bearing on this question or questions:

  3. On the surface the statistical results seem to say: …

  4. However, the following confounding factors could be giving rise to an illusion, making it seem like something is there that isn’t or that something is bigger or different than it really is.

Don’t forget to talk about the confounding factors! (4.)

Here is a Q&A about the analysis task:

Q:

What is the level of analysis you are expecting for this assignment? I’ve taken some stats classes, so I’m familiar with hypothesis testing and regression, but since this class doesn’t have a stats prerequisite I’m not sure how in depth I should go for this assignment.

Since most aspects of wellbeing are correlated with each other, it seems to difficult to use regression to analyze relationships between these aspects without running into reverse causality, cousin causality, or both. My knowledge of stats isn’t sufficient to avoid these problems in cases where instrumental variable regression isn’t a viable alternative. I’m wondering what you would suggest that I do to avoid this issue.

A:

At the low end, it could be simply some scatter plots or bar charts or other interesting graphs.

I don't expect you to have consistent estimates of anything, rather to be able to discuss any biases there might be in the estimates you do get, relative to something interesting. Please make the attempt to figure out the sign (+ or -) of any bias you discuss, and say what that would mean for the truth of the interesting thing one might care about. If there are multiple biases, try to figure out the sign of each one, even if all the biases put together can't be signed because some biases are likely to be + and others are likely to be -. Also, discuss whether you think a bias is likely to be large or small.

Advice for the Analysis Task:

  1. Use lots of graphs. I love scatterplots, but other types of graphs and figures can be good, too.

  2. It’s fine to do some statistics on individual variables, but make sure you do something that relates pairs of variables to each other.

  3. Do some formal statistical tests.

  4. When you test more than one hypothesis, set it up so you can do the multiple hypothesis test correction using the False Discovery Rate procedure!

  5. Make a distinction between being significant at the 5% level and being significant at the 1/2 % level.

  6. If something isn’t statistically significant, you say “I can’t reject the null hypothesis that …” NOT “I reject the alternative hypothesis.” If you want to reject a hypothesis, you have to set it up as a null hypothesis.

  7. Recognize reverse causality and cousin causality, including the consumer-theory-esque model I gave in class of how resources broadly construed help all good things, leading to the general principle (with only a few exceptions) that “All good things are positively correlated.” (This is a statement about the cross section.

  8. Define variables in full. You need to act like your reader doesn’t know what the abbreviations mean. So write out the full text of the aspects, and describe fully all other variables. (You will see that we do this in our papers.)

  9. Don’t order response categories alphabetically! They need to be ordered logically. For example, political leanings should be ordered from Left to Right and levels of education should be ordered from less to more.

  10. When you have interesting results for several variables that are along the same lines, think of creating a simple index to get more statistical power. That is, take simple averages of similar variables and treat that simple average as an index.

  11. Think about how nearly statistically exogenous your right-hand-side variables are. Other things equal, regressions with more nearly statistically exogenous right-hand-side variables are more interesting. That doesn’t mean you can’t do other things. Just think about this dimension.

  12. Think seriously about scale use. Any statistical analysis you do with aspect-of-well-being data you can probably do both with the raw aspect ratings and with (aspect rating - average of calibration questions). Doing both of those analyses will be much more interesting than just the one analysis.

Paper on "12 Rules for Life: An Antidote to Chaos"

This assignment is now posted on Canvas. If you are way ahead of the game, you could even submit it now!

3-5 pages

Due Tuesday, February 28, 2023 at 11 PM

Choose one of the 12 rules that you think, if more fully implemented, would make a big positive difference in either your life or in American society (or in your home country’s society). Lay out how it could help you get more of what you want (or reduce your suffering) or help people in our society more generally get what they want (or reduce the suffering of people in our society). Go into depth. Also, answer: “Are there ways you would modify the rule to make it even better?”

If you disagree with all 12 rules, choose one and write about how it is bad.

Meditation App Assignment

Get a meditation app for your phone and use it. Late in the semester, I’ll ask you to write half to one page about your experience doing that.

Some Meditation Apps (some free, some that charge for a subscription)

I personally use Waking Up and 10% Happier. One of my Economics colleagues likes Headspace. I don’t know anything about Calm and Balance, they just came up when I put “meditation” in the app search box on my phone. Also, take a look at this article:

Paper on "Happiness: A New Science," by Richard Layard

3-5 pages

Please put your name inside the document! I print them out to read them, and it is then hard to match up the pieces of paper with names. I will subtract points if you don’t put your name in there.

Documents can be Word documents or pdf.

Make sure to provide evidence in your paper that you have read the whole book. (For example, you might choose appropriate quotations from the book sprinkled from beginning to end, or simply address issues in the book that are raised near the end as well as issues that are raised near the beginning.

Due 11 PM Monday, February 13, 2023.

Themes/Questions to Answer:

What is the “conventional wisdom” about happiness that Layard presents?

Where is that conventional wisdom right and where is it wrong?

Where the conventional wisdom is right, what are the implications for your own life?

Note: Make sure that you don’t sound too much like ChatGPT. A hint about that is to be personal about “what are the implications for your own life.”

Snow Day

Happy Snow Day! I look forward to seeing you all on Friday. 

Congratulations! You made it to the course website. (I’ll use Canvas only for assignments, announcements and a bare minimum of other things.) I strongly recommend that you bookmark this website, which is softwarely an offshoot of my blog:

https://blog.supplysideliberal.com/ethics-happiness-and-choice

Ethics, Happiness and Choice - Confessions of a Supply-Side Liberal

If you wind up on the home page of my blog, you can get back to the course websites by clicking “Resources” at the top and then clicking on the name of our course.

The three big goals I have for this course are:

  1. Help you get knowledge and tools to make you happier.

  2. Provide perspective that can help you better understand and be more tolerant toward those who have different beliefs, including those who have different political beliefs in part because they didn't go to college. 

  3. Give you tools to help use statistics to find out the truth, rather than let other people deceive you with "Lies, damn lies and statistics." Lies, damned lies, and statistics - Wikipedia

Along the way, you'll also get some practice in writing and critical reading, and get your hands on happiness data. 

That means this course emphasizes ethics and happiness, and has less of an emphasis on choice, except insofar as choices interfere with happiness. The title of Economics 460 covers so much ground, it would be hard to give a fully thorough treatment to everything in the title of the course! Yangwei Song focuses more on choice in her Economics 460. "Happiness" has many meanings; I take "happiness" in the title of the course to have the broadest possible meaning: everything you find personally fulfilling, everything you want or desire, and what you get from doing your ethical duty, helping others, and making the world a better place. 

The posts dated 2023 are for this semester. Those are the ones to focus on. The posts dated 2022 are from last Spring. You can browse in those to see more of what is coming. 

Last time I taught the course, the big mistake I made was leaving too many of the assignments until late in the semester, so students were overwhelmed at the end. I won't make that mistake again: readings and assignments will come thick and fast at the very beginning, so that things are easier at the end of the semester. In particular, note that you will need to plan time to read two books by February 13: Happiness: A New Science by Richard Layard, and the one your group does a class presentation on. For you to get started reading the book your group is making a presentation on, I need to get a short assignment telling your preferences on which book by this Friday night (January 20) at 11 PM. I'll give you a preview of each of the 6 books that are options for your presentation on Friday, so you might want to plan to do that short assignment after class but before 11 PM on Friday. 

--Miles

Ideas for Using ChatGPT

The key rule is that you need to treat text (words) from ChatGPT the same as any other text (words) that you find online. You need to clearly label those words as due to ChatGPT. And, of course, you shouldn’t have a ridiculously high percentage of words in your paper be someone or something else’s words. The same rule applies to any other AI. The way to cite text from ChatGPT is to clearly set out the prompt you gave ChatGPT before the text you got from Chat GPT.

That said, as long as you clearly label text from ChatGPT as coming from there, there are many interesting ways to use it in a paper. Here is an article with some good ideas.

In addition to being an automatic F if you are determined to have used ChatGPT without clearly labeling the words from ChatGPT as coming from ChatGPT, one of your tasks in each paper is to make your own words sound different from ChatGPT. You will get a lower score if you sound too much like ChatGPT.

Extra Credit

Extra Book: One of the two extra credit opportunities for this class is, for half the credit you would get for the other book papers, to read and write about one of these five books as you wrote about the other books, focusing on what you learned:

(You can only get extra credit for one. You need to send in the paper on the extra book before the last class to get credit.)

The Daily Stoic: Subscribe to the daily email from The Daily Stoic at https://dailystoic.com/ and get extra credit for up to three 200-word reactions about a daily message that was especially meaningful to you, explaining why it is was meaningful to you.

Book Choices for the Group Oral Presentation; First Assignment

Timing: The presentations will be during the 6 classes from Monday, January 30, through Friday, February 10.

Six Books for Six Teams:

  1. Factfulness: Ten Reasons We’re Wrong about the World—and Why Things Are Better Than You Think by Hans Rosling

  2. GDP: A Brief but Affectionate History by Diane Coyle

  3. The Happiness Hypothesis: Finding Modern Truth in Ancient Wisdom by Jonathan Haidt

  4. Beyond Order: 12 More Rules for Life by Jordan Peterson

  5. The Coddling of the American Mind by Jonathan Haidt

  6. The Tyranny of Merit by Michael Sandel

First Assignment, Due 11 PM Friday, January 19 (I pushed this back a day because of our snow day on Wednesday; I’ll give you a preview of each of the books on Friday. But I need it at 11 PM Friday so I can assign you to book groups over the weekend and you can get started reading.)

1. Write a few sentences about what drew you to this class and what you hope to learn and get out of it. 

2. List in order your 1st, 2d and 3d choices for the additional book you want to read and do a group presentation on. 

3. Write a few sentences on why you are particularly interested in reading your 1st choice. (Optional: you can write about your 2d and 3d choices, too if you want to.)

Note:

I tried to post this assignment on Canvas. You can submit your answers there. (If you can’t figure out how to submit it on Canvas, you can send it to me in the body of an email.)

Not this time:

The Age of Em: Work, Love and Life When Robots Rule the Earth by Robin Hanson (Related blog posts laid out here.)

Quiz

****The Quiz will be in class on Wednesday, April 20th****

Shorter Readings to be Tested on the Quiz (you do not need to read the appendices to any of the papers)

Note: If you have any trouble downloading these papers from the Dropbox link, just get them from the Norlin Library website.

Statistics Principles to be Tested on the Quiz

  • Statistical algebra

  • Whether or not OLS and IV are unbiased given an arrow diagram

  • Whether the direction of bias is 0, +, - or ? given an arrow diagram with +’s and -’s on the arrows

  • Using the false discovery rate to deal with multiple hypothesis testing. Note that “Who Leaves Mormonism?” and “A Well-Being Snapshot in a Changing World” use the false discovery rate approach. (These are optional readings that I won’t directly test.) On using the false discovery rate approach to a multiple-hypothesis-testing correction, remember that the 1st hypothesis has to be significant by the FDR standard before you can check for whether a second hypothesis is significant by that standard, and then the second hypothesis needs to be significant by the FDR standard before you can check whether a third hypothesis is significant by that standard, etc. That is, you have to go in order and be able to jump over each hurdle, or you have to stop.

Arithmetic with Diminishing Marginal Utility

Posts on Statistics

  1. Evidence that High Insulin Levels Lead to Weight Gain

  2. Why a Low-Insulin-Index Diet Isn't Exactly a 'Lowcarb' Diet

  3. Cousin Causality

  4. Can Religion Reduce Suicide?

  5. Less is More in Mormon Church Meetings

  6. Who Leaves Mormonism?

  7. Mental Retirement: Use It or Lose It—Susann Rohwedder and Robert Willis

  8. Data on Asian Genes that Discourage Alcohol Consumption Explode the Myth that a Little Alcohol is Good for your Health

  9. Less Than 6 or More than 9 Hours of Sleep Signals a Higher Risk of Heart Attacks

  10. The Surprising Genetic Correlation Between Protein-Heavy Diets and Obesity

  11. Where is Social Science Genetics Headed?

  12. Exorcising the Devil in the Milk

  13. Are Processed Food and Environmental Contaminants the Main Cause of the Rise of Obesity?

  14. Livestock Antibiotics, Lithium and PFAS as Leading Suspects for Environmental Causes of Obesity

  15. How Rising Anorexia Can Go Along with Rising Obesity: Both Can Be Caused By Environmental Contaminants

  16. How Lithium May Have Led to Serious Obesity for the Pima Beginning around 1937

  17. Henry George Eloquently Makes the Case that Correlation Is Not Causation

  18. How Dating Apps Are Making Marriages Stronger

  19. After Crunching Reinhart and Rogoff's Data, We Found No Evidence High Debt Slows Growth

  20. Examining the Entrails: Is There Any Evidence for an Effect of Debt on Growth in the Reinhart and Rogoff Data?

  21. Frightening New England Journal of Medicine Projections for the Rise of Obesity

  22. Hypotheses about Salt and Blood Pressure

  23. Beyond the Classroom: Using Title IX to Measure the Return to High School Sports (Betsey Stevenson)

  24. Justin Briggs and Alex Tabarrok: Fewer Guns, Fewer Suicides

  25. Let's Set Half a Percent as the Standard for Statistical Significance

  26. Adam McCloskey and Pascal Michaillat: Calculating Incentive Compatible Critical Values Points to a t-Statistic of 3 as the 5% Critical Value after Accounting for p-Hacking

  27. Adding a Variable Measured with Error to a Regression Only Partially Controls for that Variable

Also, I referred to “Title IX and the Evolution of High School Sports,” by Betsey Stevenson.

Book Choices for the Group Oral Presentation; First Assignment

Timing: The presentations will begin in the third full week of the course.

Six Books for Six Teams:

  1. Factfulness: Ten Reasons We’re Wrong about the World—and Why Things Are Better Than You Think by Hans Rosling

  2. The Age of Em: Work, Love and Life When Robots Rule the Earth by Robin Hanson (Related blog posts laid out here.)

  3. GDP: A Brief but Affectionate History by Diane Coyle

  4. The Happiness Hypothesis: Finding Modern Truth in Ancient Wisdom by Jonathan Haidt

  5. The Coddling of the American Mind by Jonathan Haidt

  6. The Tyranny of Merit by Michael Sandel

First Assignment, Due 1 PM Tuesday, January 18:

1. Write a few sentences about what drew you to this class and what you hope to learn and get out of it. 

2. List in order your 1st, 2d and 3d choices for the additional book you want to read and do a group presentation on. 

3. Write a few sentences on why you are particularly interested in reading your 1st choice. (Optional: you can write about your 2d and 3d choices, too if you want to.)

Note:

I tried to post this assignment on Canvas. You can submit your answers there. (If you can’t figure out how to submit it on Canvas, you can send it to me in the body of an email.)

Required Reading, Listening and Watching