This course gave me a hard time. The lectures from week 3 and 4 seemed difficult, some more resources would be helpful.

Like in all courses of this specialization, the peer assignment was a real bottle-neck in the completion of the course.

some courses are not understandable, and there are many mistakes and bugs in tests needed to be fitted

This is certainly a good introduction. But it did not help me to be independent on bayesian statistics

The explaining for some bayesian methods are unclear, which make it harder for new learner to follow.

Slides poor compared to 3 earlier modules and instructor not as engaging. However, the labs are good.

Find it hard to follow the lectures. The labs and supplement material is good though.

A useful course, but very demanding. Many of the lectures are fast-paced.

I would recommend to include more weeks; slow down and go deeper

Some contents are just too difficult to understand fully.

Very heavy information very quickly otherwise - great

Too such for such a complicated topic.

Good course, but need more details.

not very interested

tough course.

huge jump

Unfortunately I found this course to be inferior to the Inferential Statistics course. Below are some comments which will hopefully help you to improve the course.

- The lectures more or less repeated the textbook verbatim. When learning, it's helpful to be exposed to different examples. The lectures did not help if I didn't understand something from the textbook.

- The textbook was filled with spelling and grammatical errors, which affected my understanding at times. Ditto for the supplementary materials (although I did appreciate their presence).

- The pop-up windows for the questions during the lectures were poorly formatted (at least using Chrome).

- I would recommend redesigning the course. Begin by clarifying the philosophical differences between Inferential and Bayesian statistics. E.g. discuss the likelihood principle, the conditionality principle, etc. I had to look this stuff up myself to understand the fundamental differences between the two approaches.

- An explicit guide of when to apply different priors would have been useful. E.g. I have this data, I have this belief about it, I want to get this type of answer, so which prior should I use?

- I think a deep-dive on a smaller amount of material would have been better. It would have been helpful to slowly and deliberately go through each step of Bayesian inference, rather than rushing through a larger number of examples.

- Similar to the point above, it would have been helpful if you did not wrap the R functions. Instead, show us the source code and explain what each line does.

- The amount of time spent on Bayesian regression was puzzling, as the results were noted to be numerically equivalent to frequentist regression. Why not teach us something new, and only highlight the difference in interpretation?

Thank you, and I hope you can improve this course in the future. I would not recommend it at this point, although I would recommend the frequentist course.

I wouldn't call this a bad course, but it's difficult for me to recommend it to someone. It started out very strong. Mine did a great job of teaching and explaining the concepts in the first couple of weeks, and I felt they provided a lot of value. Unfortunately, Weeks 3 and 4 felt very rushed and difficult to follow. David and Merlise had very robotic delivery as if they were reading off a teleprompter, and the concepts were not clearly explained at all. It made the 5-10 minute videos take upwards of a half hour to actually understand and take proper notes on. For a course emphasizing its focus on R, I felt fairly underwhelmed by the actual R content within it. Each week has a lab assignment, all of which are pretty well-done, but to be honest, I did most of my learning during the final Peer Reviewed Project where I did a lot of independent study filling in the gaps created by this course. Most of the actual Bayesian Statistics work within R was just using a functions contained within a library created by one of the course's authors. While they're certainly functional, it didn't quite align with my expectations given how this course is advertised. The peer-reviewed project is also problematic. Given the low volume of students at this point, it takes weeks, sometimes longer, to get your assignment graded. I even got locked out of the course while waiting for my grade, so it was a bit nerve-wracking. It's not the fault of the authors, but it's something to consider if you're operating on a strict deadline for this. If you're genuinely interested in the subject matter and don't mind putting in a lot of effort to self study and finding out answers for yourself, I think you'll do fine in this. However, in my opinion, if you're going to pay money for a course, it should provide enough value to make the price you pay worth it.

This course covers a good amount of bayesian statistics. However, the presentation/videos starting from week 2 really sucks. They change instructors for difference topics and obviously some instructors are not very good at explaining other than reading the material.

The videos skipped many medium steps that are actually very crucial for understanding the concepts. And no suggested reading materials at all either. Also the quiz are not very well designed either. For example, some quiz are much more simpler than the course material, which makes it not helpful at all to understand the course material itself. While some times it is the opposite.

The first three courses in this specialization are very good, but somehow this course are way below the quality of the previous ones.

The previous courses of the specialization were much better. This one is too fast paced and confusing. The math for this course is significantly harder than for the previous, but in my case it was not the math what was making it hard. The videos are hard to follow. I answered some of the quiz questions based on intuition and what looked reasonable rather than actually knowing how to solve them. Usually in the previous courses the project felt like the hardest part, but on this one the project felt like the easiest. What I did like about the course is that it has good breadth of topics in Bayesian statistics.

This course, compared to the others in the specialization, was a bit of a mess. The lectures were hard to follow with fewer exercises to check your learning than in previous courses. The "text" seemed to just be a bad transcript of the lectures with all sorts of errors. The labs were confusing and sometimes included incorrect or outdated instructions that caused me to waste a lot of extra time trying to figure out what was wrong. I enjoyed doing the final project, though, and learned a lot doing that.

The course presents interesting material but it is not easy to follow. It is a huge jump from the previous courses and requires far more hours to understand all the (math-heavy) material than the stated. The slides feel a bit chaotic and the language/sentences during the explanations could be much simpler. At times it feels that the instructors limit themselves to reading formulas one after another, making it hard to find a connection between them and how they are applied.

This course makes a valiant effort to provide as much coverage of Bayesian statistical methods as the prior three courses in the "Statistics in R" specialization do for Frequentist statistical methods, but the lack of supporting material (e.g. reading/text exercises directly paired with each lesson) really hampers this. The videos are quite informative, but if you don't catch on to the material based strictly on the videos, the weekly quizzes can be a bit frustrating.