I've already written a review but it seems to have been removed...

This is an awful course, there is very little purpose to it whatsoever, it is basically a module in markdown which will in all honesty not have much application for most learners.

In addition, the course is not at all balanced / laid out well, there is a peer assignment in week 1, which you need to have covered week 2's content for.

Lastly, the recording quality of some of the lectures is awful, it is clear that they have simply used some recordings of an actual classroom session of a related course instead of recording for Coursera.

In all honesty, this entire specialisation is of awful quality, it is not a data science course, it is a "here's a few useful things in R" course, and the instructors should be ashamed that their institution makes money from it.

Too expensive for such a simple course

Course material is not enough to understand course as whole. Instructor just reading slides and not enough examples demonstrated.

T

h

is is just a tiny tiny part of data science. It's actually part of any science. Why to waste 20+ hours by learning some very specific tools such as "knitr" when the thing I really want is to learn data science?

The teacher is focusing too strongly on one particular R package. There is about 1 hour of interesting general knowledge and 19 hours of working with "knitr".

Instructor not as good. A big fail when you need the 2nd lesson in order to complete the 1st programming assignment.

Whilst I can see why the idea of reproducible research is important there was't really enough material in this course for the full four weeks - and in fact a lot of the videos repeated the same information.

I don't think it requires a separate course for this topic. possibly combine it with other courses and introduce neural networks.

I felt this course could have been added as sections to other courses. One separate course for this topic is a waste of money.

1 for Knitr, otherwise it's waste of time.

Seems like the course lectures are out of order or duplicated. In order for us the complete the first project, it required us to view Week 2 videos as well. I wish this was noted/told to the students in the course.

Completely irrelevant material.

The useful part of this course is the material about markdowns. The rest is just an endless repetition of the same idea over and over again: people must be able to reproduce your analysis. The final project lacks clear directions. Is it too much to ask to provide short descriptions for the columns included in the analysed data base, or at least for the most confusing ones? I do not think that in a real life situation a data scientist is supposed to interpret variables like PROPDMGEXP or CROPDMGEXP based on their names alone. Even if he guesses the meaning of these variable there is no guarantee that this guess is correct.

I often feel like people completely ignore the "science" aspect of data science (read any data science career question on quora for example). This course does an excellent job of introducing key aspects of the scientific method that you might not have encountered if you've never done an experiment before. The final project is a lot of work (mostly data cleaning) but very fun and informative.

My favorite course, at least it gives me an argument why scripted statistics is awesome and can be applied to a number of data related activities. Recycling chunks of code has proven useful to me.

should be included much more in course 1. it would have been great to know up front how easy it is to mix text and code, not from a reproducibility standpoint, but just to take notes.

Without taking this course wouldn't have fully understood the importance of reproducible research in data science. Thank you so much. I recommend this course for all data scientists.

The two projects were interesting and built on skills learned in the previous four courses in this specialization that focused on using the R language. The video lectures were largely repetitions of the course text, which is fine, some people prefer videos, others prefer texts. (I read the text and was fine with skipping around the videos and/or playing them at 2x speed.) Perhaps the most useful skills learned in this course were during the projects where we did some data cleaning and analysis, then wrote up our results in an R markdown file (.Rmd) and published to Rpubs. Overall enjoyable experience. The most useful "hack" was learning how to preserve markdown files in the Rstudio settings so that when you push your .Rmd s to github, you get a nice readable markdown file.

I have learned a lot in this course especially the final project. Not only did I learn to make every part of the project reproducible, but also had the chance to apply all the I have learned in the previous courses. The final project can be either easy or very challenging and demanding depending on how you choose to do it. I chose to give it my best in the different parts(reproducibility, processing, analysis, reporting, visuals...) especially in the data processing part, it took me a lot of time but I have learned a lot through it. I felt like working on a real data science project. I also plan to work on it again and improve it/experiment with it after I learn text mining ... Thank you!

This course was not so convient for me as other Data Science Courses. Therefore it even seems easy than the rest of speciality I was not able to get to the end until first assignment has been moved to the 2nd week. But starting and leaving it for few times I got so needed experience with report preparation, reseach process description and converting all the steps of my study into a story, which could be understandable for someone else. Thanks everyone! I enjoy my new skills and sell them for good price!