Great explanation

very interesting

Great course!

very helpful

good lecture

excellent

Very Good

gracias

Great!

good

good

N/A

很好啊

V

v

Very mixed feelings about this course.

Generally speaking, the course lectures are informative and well organized. Mentors are reallly of great help, they are doing a great job, honestly: they are very active, they give good insights, they know the subject matter.

But in the course lectures, there are occasions where concepts are used which were not formally introduced before their actual use.

One example: in the lectures on probability, the first "slide" in the lecture talks about random processes, outcomes of random process,... On the next slide, the notion of probability of an event is introduced, but the very notion of "event" was never introduced. It is introduced in the accompanying book, but if it is the case that the book chapters should be read PRIOR to watching the course videos, that fact should be made clear.

Further in the course on probability, some words are used "interchangeably" without the context making it clear why they can be used interchangeably. For instance, on some occasions, the concept of independent events is used, but then, later on, the discussion talks of independent processes. Which is which??? Is there a difference? If so, what is it? When do I need to use independent events as opposed to independent processes?

The graded assignments are of varying quality. The most disturbing thing about them is that, on some occasions, concepts are used in the quiz questions (either directly in the questions and answer choices, or indirectly in the "correction" for the quiz after you have submitted it) that were never touched upon in the course.

I have had two occasions of concepts not introduced in the course but used in the graded assignments.

The first occurrence of a gap between course content and quiz questions was on a quiz question about inference. I failed the question, and understood why I failed based on the course content litterally minutes after failing the question (and one mentor actually rightly corrected me). But the question "correction" (the explanation text you receive after submitting, as justification for what the correct answer is) referred to the concept of "two-sided hypothesis test". Where did THAT come from?? I checked and rechecked the course videos, no mention at all of it. I checked the accompanying book, and the first mention of two-sided hypothesis test is way way way further in the book, in a chapter that is entirely focusing on inference.

The second occurrence was in week 4. The course lectures cover two distributions: normal and binomial. The recommended reading in the book also focus on these two distributions (the recommended reading actually skips the section on geometric distribution, if I remember well). But in one of the quiz question, there was one of the possible answers referring to the geometric distribution. If it is the case that we are supposed to know and understand about geometric distributions, then the course content should cover the subject. Or at the very least, the course lecture should mention clearly that learners are advised to read about it in the accompanying book.

The guidelines for the project assignment (week 5) are not all that clear as to what is expected from the learners. Sure, there are instructions on where to find the info, what structure should be followed,... There is also a very nice "example" project (designed by one of the mentors), which provides a lot of useful info (how to filter missing values from variables,...).

But there is no real hint as to the depth of analysis we are expected to complete. This is definitely a source of confusion, not only for me, but also for a few other learners, from what I gathered in the discussion forums. The result is that the projects you get to review are of very disparate levels. Some end up in calculating one figure per research question, without any attempt at deriving trends or patterns, others do not include any plots at all,... The thing is that the peer review criteria do not really provide a good basis to ensure that learners did indeed assimilate the course contents. Most of the questions in the peer review assignment have a lot more to do with following a canvas and not so much with the course substance itself.

For instance, some of the peer review criteria have to do with the narratives for computed statistics and plots. The criteria are: "Is each plot/R outout followed by a narrative", "Does the narrative correctly interpret the plots, or statistics", "Does the narrative address the research question". But when the research question is a question of the type "What it the IQR for income per state", for instance, the narrative can be very short: "IQR per state shows that the state with higher variability of income is...". So, the narrative meets the 3 evaluation criteria: there is a narrative, it does address the research question, and it does correctly interpret the statistics. But it is not particularly useful.

I do understand that Internet-based peer review is challenging, and that you have to settle for "neutral" criteria that are easy to assess by learners. But the peer review grading "grid" as it currently stands is not "that" helpful in assessing whether the course contents has been assimilated.

To conclude, when I took the course, my initial plan was to follow the entire specialization. But after having completed the first course of the specialization, I have radically changed my mind, and will look for alternatives "elsewhere" to get the knowledge/skillset that I am after.

This course to me had some very clear un-explicit limitations, pros and cons:

- The lectures are fantastic and have a good sequence for beginners

- The course is very holistic in its approach, meaning that it covers theory and application very broadly and gives you a good sense of how different aspects of the field of statistics relate to eachother

- The coverage of the R programming language is insufficient for the requirements for using it in the final assignment, I can't stress this enough for beginners. I highly suggest you take a foundational course in R, highlighting syntactical structure of the language, prior to taking this course

- The labs are great for learning the primary components of R, but they don't give you real practice coding. There is very little to no explanation of certain functions in R and there are no videos on it. I do not feel at the end of this course I have a very good understanding of the structure of the language of R, I do however feel I was assessed as if I should have.

- I felt the quizzes were appropriately rigorous for a beginner such as myself.

Most important bottom line is: If you are a true beginner like myself I urge you to first take a course more targeted to R before starting this specialization. Otherwise, like myself, I think you will feel very overwhelmed at the end.

The course is divided into 5 sections, each of which you have a week to complete (if you want a certificate). The first 4 sections/weeks are well designed and involved a mixture of lectures (most were good), reading assignments in a textbook (free online access), practice problems, and a weekly quiz. Along the way students learn how to use R through a handful of walk-through examples. In general this works. That said, the last two R assignments are a mess. For the 4th week, the instructors put together a demonstration for using R to ask and answer some basic research questions. The document they put together for this demonstration, however, is so full of typos and grammar mistakes, and worse, heaps of nearly incomprehensible sentences and phrasings, that it is almost worthless. It was really painful to get through it. The final R task is to work with a real-world data set, ask a few research questions, and use R to do some basic statistical analysis of the data. Working with a real-world data set is great. That said, I felt as if the instructors were asking students to do far more with R and statistics than we had learned in the class. I saw many similar opinions about this assignment online. And in grading my peers, I noticed that other students didn't know how to complete the project either.

The course is great in terms of building foundational concept for data analysis and lab assignments were ok. The I think often the time listed for any class is a little underestimation of time commitment. The course has offered me a lot of new concepts to learn and was good refresher for many other. But the final project was a huge disappointment for me. The students have been given a huge datasets that leaves students struggling to figure out where to start. In order to understand the data, students have to go to various links to see what the data is, how it is collected and definition of each variables. Then there are more than 300 variables and you need to pick few to do something you think is interested. Finally, the project needs a good level of expertise in "R" and course does not teach you that at all. I would suggest that for future courses reduce the number of variables depending on what most students have been using. Draft a quick summary and report about key information of data and share that on course page rather than links to web pages, and finally let students use the software that they may feel interesting. Not everyone is skilled to use R.

I'd say the name is somewhat misleading, in the sense that very little of R is taught, even for an introductory course. You only get shown several functions without much context and no big picture understanding. Comes down to rote learning in the end. Probability-wise, since I've already had some background skills, this is also a shallow level look into the basics, with barely any theoretical setup or even a bit of historical background to some concepts to really flesh it out. The fact that we haven't even noted down the function for a normal distribution and explained the difficulty in calculating it by hand (therefore, using R or a table) is kind of a disappointment. I believe this is down to my own expectations, thinking this would be something useful to people who already have some background in STEM, but it seems to be aimed at total beginners. Will probably continue with the specialization, but I can't say my certificate is of much worth without it.

The videos, readings and quizzes are excellent, they are well organized and follow a logical sequence. The level is also suitable for a beginner and pleasant enough to watch after a busy day at work !

However Rstudio/r instructions and lab assignments need improvement. The specialization description puts a lot of emphasis on R giving the impression that these skills are also going to be taught from scratch. However there is not enough instruction and feedback. Judging from some assignments I have read it definitely seems like most of the students already have a comfortable working knowledge of R. If you are like me, a complete beginner, you will have to learn a lot from additional sources and your assignment will look like a mess (but you will most likely pass !)

I didn't get on with this course, I'm afraid. I found that the R explanations were somewhat lacking for what I needed. Some sections I could complete absolutely fine but then it got suddenly much harder to do what was required. While that's part of learning and language, it wasn't helpful for the stage in my learning experience (during each week's classwork).

I'm not an idiot. I was able to complete the full course, including the last stage project but it just didn't teach all that well. I don't feel much more confident in R.

I'd also say that the course relies quite a lot on the open source textbook. The instructor did write it - fair play - but I'm paying Coursera for the privilege of following something that's free.

The course lectures were very good and informative. However, this course does need some work. First, the text revision references were confusing. The homework assignments were a confusing as to where they should be performed; on our own or within GitHub. I have used R before and was using this course as a refresher. The course series definitely needs a optional introductory course in use of R, R Studio, GitHub, and R Markdown language. Similar to the JHU Data Science specialization. Finally, the course project was a bit deep for introductory Probability and Data. Need to make the course project less demanding or drop the need for a final course project until later courses

The instruction in statistics is understandable and complete. The final homework problem/test is way outsized for this introductory level course. The meager instruction in R - tucked between the video instruction in statistics - is insufficient to tackle the huge data set that is provided for open analysis. The data set (brfss2013) is too large for R Studio Cloud to handle, which is very limiting. The statistical questions that could be asked from the data are infinite and beyond what was taught in the course. The Week 5 final project ended up being months of frustration without the solid support in R.

Problems with the course: Despite getting high grades, I felt like I had forgot much of the material by the end. There should be more quizzes, a mid-term, and more peer-reviewed projects. Didn't teach R basics which made it hard to learn the language and complete the final project w/o a lot of research. Content was generally great.

Problems with the platform: Video skips randomly, submitting ID was buggy (wouldn't save), and the chance that you might not get enough peer reviews and have to delay to the next session is nuts. Just require that people submit more reviews!