When you start statistics with practical examples, people tend to presume certain things (e.g. independence is given like in munchausen example) so I sort of understand this desire to keep every definition abstract/pure for solid foundations but damn, this course goes way too far. I took statistics at uni and this was a refresher on the specialisation track but the way Bayes' rule was covered made me doubt what I knew. Oh also, the instructor does some of my biggest pet peeves which are (1) using his preferred notations without actually reading it out loud first (2) unnecessary use of synonyms which just distract me from what he actually means (3) reading from slides without any context as to how these concepts are used.

Also, the concepts are elaborated on a seemingly random basis. Mean is left at "center of mass" like we just came out of physics 101 but the area under a curve is dragged out with random wood cutting analogies. I am just surprised at how all over the place this course is so far. Anyone starting from scratch, I highly recommend probability and statistics reading and some basic calculus elsewhere first. Otherwise you will get frustrated with this course.

This course is a fucking shitshow. Not only does Brian Caffo not explain anything, he has a tremendous gift to confuse people and make them forget / not understand anymore what they already knew. Great fucking course. Not. Hated it from the first minute.

The teach methods changed too drastically starting with this course. Much more prerequisite knowledge is required then is included int his concentration course set. No foundation or warning is given regarding this change and prerequisites. I had to seek out and spend countless hours on many other learning resources to get through this course and still don't understand what this course was trying to teach.

Poorly organized content and the lectures are presented in a confusing way. The lecturer obviously knows the material well, but is not able to present it well. He should use more sample problems annd examples. In addition, I am having trouble getting my submission graded, although I have already graded 6 fellow students.

I've taken several statistics, data science, and R courses. This is one of the worst. I took others' advice, and I also strongly suggest looking to other sources to learn Statistical Inference before taking this course. Khan Academy, DataCamp, Udacity, Duke (Coursera), and Columbia (edX) all have great courses. Though they vary in depth, each leaves you with a good understanding of the concepts they teach.

I'm sure he is super nice and smart, but Caffo is an awful instructor. I kept waiting for him to actually get to the explanation of what he is talking about, but he never does. He explains the equation...with the equation. I really like statistics, but he was beginning to change my mind and somehow I was actually unlearning everything I had learned from other courses in the past. I would highly suggest taking this somewhere else, I like the one from Duke on Coursera, she actually explained things and I could understand!

I couldn't make it through this course because I can't stand looking at the instructor's face on the screen. It is very distracting. He is also not very clear in his whiteboard explanations, too much scribbling. I prefer courses taught by Roger Peng.

The only reason for enrolling is to complete the data science specialization, though it may make you reconsider continuing with it. The instructor and provided materials fail to adequately explain the concepts this course is supposed to cover, and do not prepare students for the quizzes or assignments. If you don't know statistics you won't learn it here. If you know statistics, you don't need this course.

very unclear and monotonic lectures

Very horrible course. This course is a good example of how to design a bad online course and teach a complex material so it's more difficult to understand. I'd give other courses a shot instead of this one, unless you want the specialization.

First, the course should have listed R programming as a prerequisite. Second, instead of only reading the definition of terms, he should have explained what they are and give example. I spent more than 4x of the video lecture length, trying to think and follow, but I still didn't understand. No example, no explanation, just *reading* the definition right off the text book using mathematical term. Math is complex, but it can be explained so everyone can understand. Third, I simply have a hard time understanding why the professor has to talk so fast and edit his video and/or make slides so text pops up in order to speed up the lecture instead of writing along as he speaks and explains (but there was no explanation in the lecture anyway so pausing the video to slow down wouldn't help). It's difficult to catch up or take notes when he edits video so text pop up very quickly. I don't think there's a restriction on the video. Overall, horrible experience.

There are better (Free!) courses out there to learn basic inferential statistics. Khan Academy is a great place to start, and Udacity has a great class that gives a good intuitive understanding. The main reason to take this class is if you are trying to finish the entire Data Science program. Otherwise, look elsewhere for an intro stats class. The instructor clearly knows the material, the class just does not do a good job of transfering that knowledge.

Do not take this course to learn statistics for the first time!! you would feel so helpless and you may hate the entire subject.. this course is great as a review or revision for someone who wants to recap his knowledge, but if you are learning this for the first time I really really recommend that you go through all the videos of khan academy (statistics playlist on Youtube ) and study it thoroughly, then come here to recap your knowledge.

Course is compressed with lots of statistical concepts. Which is very good as most must know concepts are imparted. Lots of extra reading is required to gain all insights. Very good motivating start .

The material is obviously invaluable but I thought the lectures themselves were lacking.

I'm in the middle of the course and I'm thinking seriously to abandon it... The instructor is simply very bad (he might be very knowledgeable, but he cannot teach – at least in an online manner). I rarely leave negative reviews, but this time I couldn’t resist…

Terrible professor!. Too much theory, too little coding. However, the book is great. I recommend do not watch the videos just go to the book!

It is one of the most important courses in Data Science. It covers most of the mathematical portion and it is hard as well for a non-mathematical student.

For a minimum, every sentence will have any of the four words like distribution, sample, probability, variance, mean, median, standard deviation, etcetera. We have to spend enough time and to be very careful in understanding each and every sentence.

But this course was nicely categorized by Brian Caffo & others, This presentation was the simplest one on probability & statistics, I ever saw and it covers majority of the basic concepts.

Thanks to Brian Caffo, Roger D. Peng & Jeff Leek.

Initially it seemed all Greek and Latin and difficult to go through. With patience and slowly going through the course material again and again things started becoming easy and interesting. Now I do understand the importance of statistical modeling and how to predict the population behavior. I have learnt a lot to apply the linear regression, confidence intervals, t-testing, poison and binomial testing, and the p-values. All in all a very good experience and the Coursera team has been a great help.

At this point in the specialization, I was really worn out by the effort that I needed to put into this course (I solved the homework questions too). While I have no problem with the math, some topics like power should not even have been discussed or should have just been discussed in passing. Caffo spent a whole week on that. After taking the Applied Data Science in Python Specialization, I have a feeling like this course and Regression Models can just be merged, while logistics regression could just be transferred to the machine learning course.

Fortunately, the final assignment was very easy compared to the previous courses and one could finish it reasonably in a day (Reproducible Research final assignment itself took me almost a week) .

Previous courses in the specialization did a very good job in relating reality vs. theory. This particular course provided great amounts of mathematical theory to support learning, however, in my opinion fails to guide the student to relate real life cases with the theory, making it harder to understand and thus not as useful as it should have been.

The instructor's way of explaining things was not that good. Could not understand most of the concepts.

The information is so important and useful, but I found the presentation of the material to be fast and not very interesting, and therefore it was hard for me to retain the material. I learned a lot, but I would need to invest a lot more time to realllyyy grasp everything in the course. It wasn't presented in a way that made it easy to learn, so I need to spend more time going back over things to really get it.

There're lots of practice on manually construct statistics. I'm not sure if it's necessary to do that since we could just use R code to do it. I think how to interpret it and use these statistics in examples would be more important. There're some examples, but could be more and interpret more in depth if there were less focus on the calculation.

Very difficult course even for someone who had learnt Mathematics and Statistic at the University level. Many concepts were very tersely explained with very few examples. The course book definitely helped. I would say two semesters of Statistics were squeezed into this course. Homework work exercises were very interesting and interactive.

I get this course since I really want to learn on boosting. However, I think the course pace is too fast and should be include more 'non-greece symbol'. I have a background in biostatistic and epidemiology but I'm still having a hard time understanding lecture. I would say that this course is the most difficult among this specialization (practical machine learning is much easier than this). I bet many students will have a problem in this course since it's a 'beginner' level specialization.

My suggestion is to 'slow down' , expand the lecture with more drawing/picture explanation, more r coding and get rid of those greece-symbol as much as possible.