4.7
436 个评分
145 条评论

## 课程概述

We have all heard the phrase “correlation does not equal causation.” What, then, does equal causation? This course aims to answer that question and more! Over a period of 5 weeks, you will learn how causal effects are defined, what assumptions about your data and models are necessary, and how to implement and interpret some popular statistical methods. Learners will have the opportunity to apply these methods to example data in R (free statistical software environment). At the end of the course, learners should be able to: 1. Define causal effects using potential outcomes 2. Describe the difference between association and causation 3. Express assumptions with causal graphs 4. Implement several types of causal inference methods (e.g. matching, instrumental variables, inverse probability of treatment weighting) 5. Identify which causal assumptions are necessary for each type of statistical method So join us.... and discover for yourself why modern statistical methods for estimating causal effects are indispensable in so many fields of study!...

## 热门审阅

WJ

Sep 11, 2021

Great introduction on the causal analysis.The instructor did a great job on explaining the topic in a logical and rigorous way. R codes are very relevant and helpful to digest the material as well.

MF

Dec 27, 2017

I really enjoyed this course, the pace could be more even in parts. Sometimes the pace could be more even and some more books/reference material for further study would be nice.

## 101 - A Crash Course in Causality: Inferring Causal Effects from Observational Data 的 125 个评论（共 147 个）

May 5, 2019

Very interesting studies.

Aug 29, 2020

Very well presented.

Sep 11, 2017

enjoyed it very much

Feb 22, 2020

Enlightening.

Jun 4, 2021

w​onderful!

Nov 22, 2019

Overall a great course. Better than other courses on causal inference on coursera. However, some of the topics (e.g. within the IPTW and IV methodologies ) were presented in a sort of general manner (intuitive). Which is obviously not a fault of the instructor and is due to the strong research nature of these topics. Personally, I can't think of presenting, for instance, 2SLS or insights on IPTW in more detail within a crash course. Perhaps, increasing the number of weeks to 6 or 7 in order to include more detail on, e.g. 2SLS would be a good idea. What definitely helped to make up for those missed details is the practical examples parts with R. Keep up the good job!

Oct 12, 2019

Clear course most of the time and a very interesting subject. The teacher covers the concepts from many angles: conceptual understanding, math, examples and R code. I like how there is little "fluff", you learn a lot for the time given and I don't feel any of the concepts covered are unnecessary or esoteric. The only negative is that the course could've benefited from more practical assignments. There are 2 R code assignments: could've been more. I was thinking about giving it a 5 or 4 stars and decided on 4 in case a non-perfect score actually makes the instructor improve the course.

Apr 30, 2022

It was very fluid and well-detailed. The sructure of each video was clear with a lot of nice examples.

However I found the content too much specific (usually on Biological questions), which makes most of the tools used here questionable for others fields. For example, some of my great questions are :

1- How do I estimate causal effect if the treatment is continuous ?

2- What if I have a set of treatments and want to analyse the causal effect of subsets within them ?

It would be nice to take the content of this course on a more general view :)

Aug 24, 2017

Very approachable as someone with a Masters in Statistics, probably tough if you are not comfortable with notation and concepts of intermediate prob/stats. Extremely clear and concise presentation. Coverage of methodology is a little weak, there is not enough discussion of the dangers of doing causal inference on observational data, nor of the dangers of the proposed methods. For instance, propensity score matching is ineffective or even harmful in the face of hidden confounders, which in the real world you almost always have.

Sep 23, 2020

It is a great course for those who want to better understand how causality works, statistically speaking.

Until the 3rd week the classes are very well exemplified and detailed, great to follow.

Then, it is difficult to follow the explanations, impacts of the models, etc. - a pity.

The interpretation of analysis results, variations and other subtleties is not the focus of the course. If you expect to see analysis and interpretation of results right away, this course is not for you.

May 6, 2018

I have an economics background and during my undergraduate studies I took several statistics and econometric courses. The contents delivered in this course complemented my knowledge very well from another point of view. I would definitely enjoy a more advanced course dealing with other methods. The only aspect I would improve is providing the slides for further study. Other courses in Coursera do this and, honestly, I often consult the slides.

Oct 19, 2021

Great professor and teaching. This course was a great introduction to causal inference. I remain a little unsatisfied however on a few concepts which I found insufficiently explained. In particular, the link between DAGs & d-separation and the 2nd part of the course is not very well explained. I would recommend to first follow the EdX course "Causal Diagrams: Draw Your Assumptions Before Your Conclusions".

May 2, 2020

The contents of this course are extremely concise and useful. The course prioritizes some of the important techniques used for causal inference. The practice tests , quizzes and data analysis tests were helpful to learn better. The lectures weren't inspiring or exciting and self-motivation is necessary to be able to stick with it. However, I would recommend this course to anyone interested.

Jan 26, 2022

The contents covered in the lecture are excellent. I've gained a much better understanding of Causality thanks to this course. The only complaint I have is that the dataset required for the coding assignments has not been updated, and therefore does not have the exact same features as mentioned in the instructions.

Dec 9, 2018

Content was useful for understanding causal inference in a variety of situations. Presentation was sometimes slow even on double-speed. Lectures were generally structured from abstract to concrete, which was much harder to follow than if it were presented in english first and then made abstract (Mayer, 2009).

Sep 29, 2020

The material is useful and well-presented by Prof. Roy. Although recipes are provided for solving relevant problems in R, more familiarity with R will be required for applying them. Students should be prepared to develop that familiarity on their own.

Jun 11, 2020

The course is well structured and the slides are well prepared. Professor clearly explains the formulas and makes you easily understand everything that is written on the slides. However, I would love to see some more examples from the social sciences.

Aug 31, 2020

Course is great for a general overview! That said, the discussion forums are poorly monitored and one of the exercise datasets needs to be updated. In any case, don't expect more from a Coursera course!

Mar 16, 2019

Very easy to follow examples and great coverage for such an important topic! The delivery sometimes get repetitive and I wish we talked more about how the uncertainties are derived.

Nov 21, 2020

A high quality course that delivers what it says in the title. Well-paced introduction to the potential outcomes framework, with a nice balance of theoretical and practical aspects.

Dec 15, 2021

It will be better to give reviews of related applications in specific AI areas (e.g, computer vision, NLP, etc.) at the end of each of the sections of the lesson.

Dec 15, 2018

very good content. Story line is highly concise. However, Lecturer could be more stream-lined the the way of explaining. He sure is a skilled guy, however.

Jul 15, 2018

Excellent course. Could use a small restructuring, as I had to go through the material more than once, but otherwise, very good material and presentation.

Nov 15, 2021

A​ consise course on causality; watched on 2x speed because the instructor speaks rather slowly; really bad formatting of quiz questions.

Feb 10, 2019

I thought this was a good overview and I'm glad I took the course, but I would have preferred more hands on programming assignments.