The IBM Data Science Professional Certificate is one of many stepping stones on my journey to become a data scientist. This certificate program is comprised of 9 courses, taking the student from the basics of data science through to creating their own unique capstone project. It is a well-established, well-crafted online education program that has been around for many years, with 234,952 enrollees in Course #1 at the time this article was written (interestingly only 67,533 enrollees in Course #9…quite a drop off, perhaps that speaks to the program difficulty).
This certification is a great starting point for a career shift into data science. I highly recommend it to either beginners or intermediate level folks. Beginners receive a broad overview of the fundamentals of data science. Intermediate students will appreciate the capstone project for the experience gained by coming up with your own project, sourcing your own data, and writing your own code. An added bonus is kickstarting your portfolio with a mandatory GitHub account, writing a research paper, and writing a blog article. Curious students wondering whether a data science career might be a good fit will know if they make it to the end, and especially if they enjoy the last two courses.
How Much Money Does the Certificate Cost?
The course is free if you can complete it in 7 days, else $42 per month through to completion. Unless you can focus 10+ hours per day, you probably won’t complete it within the 7-day window unless you cut a lot of corners and short-change yourself of the full learning opportunity.
How Much Time Does the Certificate Take?
It took me 106 hours across 17 days working evenings and weekends. The course documentation indicates that on average a person takes 195 hours and up to 10 months to complete the program (6 weeks for the capstone course alone).
My actual hours per course are listed in Figure 2. One caveat, I just went through an intense year completing a graduate degree in data science and consequently moved pretty fast thru the course material and labs; so adjust your expectations accordingly. Also note that for courses 1 thru 8, I went 2x to 4x faster than most students, but then dropped to 2x slower for capstone project #9. This was intentional as I planned on using the capstone to kick-start my project portfolio and wanted to really understand the material and polish off the artifacts (Jupyter Notebook code, PDF research paper, and PPT presentation).
Is the Certificate Worth the Time and Money?
Absolutely! You cover a lot of material and get to practice your data science skills throughout the program. In the Capstone course you get to heavily exercise what you’ve learned. The capstone also requires that you start a portfolio in GitHub…something I’d been putting off, but am very happy I had to complete (my GitHub). The capstone also requires that you write a blog or create a presentation summarizing your findings. Although I went with the PowerPoint / PDF presentation to complete the certificate faster, I understand the benefits of publishing articles and writing blogs…so here I am, one week later, freshly signed-up on Medium and writing my first article/blog.
If you need more proof, this certificate is frequently rated in the top 8 to 20 data science certificates: here, here, here, here, and here. Completing this certificate will give you an extra edge, setting you apart from the competition that does not put in extra time sharpening their skills. The capstone course in particular gives you direct experience with real world data and tools that you could turn right around tomorrow and use in your job as a data scientist on a similar project.
What Social Media Badges are Awarded?
IBM offers pretty badges (Figure 3) for each individual course through Acclaim (by Credly). These are digitally verifiable and publicly accessible. Click my Acclaim link to see and example of how they work.
Coursera provides the final certificate (Figure 4) representing completion of the whole program. Although not as pretty as the individual IBM course badges, this certificate is the one you should post on social media. (Astute readers will wonder about the pretty IBM Certificate badge displayed above in Figure 1— it came from Google search, not Coursera…I suspect it is from the $400 sister-EdX IBM certificate.)
Click my certificate below to see what the details look like, and notice that as this article ages, the certificate ages just fine along with it. I prefer not to do certificates that expire unless it is a job requirement.
What Should be Posted to LinkedIn?
I happen to have a strong opinion on how to properly post ongoing certifications to LinkedIn. I believe it is best to post a single overarching certificate representing the program, or the single final or single highest-ranking badge in the series (Figure 5). It is NOT a good idea to dilute your certification section by posting each individual course badge. I also like to emphasize the level of effort behind the certification by including the count of courses or exams and the actual number of hours it took me to complete.
This sets my certifications apart from the pack who either have no certifications, or alternatively who litter the section with 20+ little five hour courses. Of course, there is an exception…if you only have one or two individual 5 hour course badges completed then it would be a good idea to post those for now, and replace them later with the final certification when that is completed.
Click my LinkedIn page for an example of good Certification posts.
That all said, it is important to still post each of those individual courses so that resume crawlers, google, etc. is aware. I prefer posting them in the “Accomplishments” section, specifically in the “Test Scores” sub section because when a viewer expands it, they will see each credential’s URL. Nowhere near as nice as links in the credentials section where you can easily click them, but still the right decision to maintain message control by showing only the highest priority certifications in that section.
Are there Better Certifications Available?
Sure. Harvard and MIT both have statistics and data science certificates out on EdX, as does Johns Hopkins on Coursera. Several excellent Python/R/Statistics focused certifications exist at Duke University, the University of Michigan, the University of Washington, and other large institutions. I plan on taking many of them.
To me, the point is NOT which is certificate is better or worse. Instead, the objective is to hit the material from a diverse set of perspectives, each reinforcing the last. Example, I just completed a graduate degree on this subject and yet was still able to take away many new ideas, new tools, and new techniques from this certification…especially from the capstone. The learning never ends, you just see more approaches more clearly and can execute faster over time. So, I suggest viewing this certification not as better or worse than others, but more as a great starting point on a long journey for those who already have a solid foundation in math or statistics or data analysis or software development. If you have no related experience at all, then there are simpler certifications with fewer courses than this from which you might better start. Likewise, there are courses and certifications that are much more advanced and should follow this one.
How are the Courses Graded?
Quizzes are automatically and instantly graded. You typically have 5 to 10 multiple choice questions from the course material. There are generally some difficult questions mixed in where you really need to understand the specifics to get it right. You either pass and move on, or fail and must retry. If you fail, Coursera will show you the pass / fail status of each question, but not necessarily the answers. At this point, you should go re-review the labs and videos before attempting a re-test. If you fail the test a 3rd time then that’s it, you’re done and cannot re-attempt it for 8 hours (when you fail the 2nd time, a warning pops up underscoring the importance of slowing down, studying and getting it right). Each re-test will have different variations of the questions and sometimes even completely different questions are asked. Take the quizzes seriously and make sure you understand the material.
Projects are “peer-graded”, meaning that other students currently taking the same course at the same time as you are required to review and grade at least one other student’s material. A detailed rubric is provided walking each student-grader through what is expected. From Coursera: “Reviewing another student’s work is a valuable way to learn. Providing quality feedback is a useful skill to master.” They recommend using a polite and positive tone when providing critical feedback, and to remember that English is not everyone’s native language (look past grammar and typos at intent and understanding). The screenshot provides one of many rubric grading points that helps you walk thru the grading process.
How has the Material Stood Up Over Time?
The answer is a mix of good and bad. The core concepts are all building blocks of Data Science and have held up fine and will continue to do so for a long time to come. That said, as with most technology-specific how-to’s, they age out. Some of these courses are no different. I had to do some reading between the lines to work through incorrect documentation on projects. Although occasionally frustrating, think of it as a simulation of how the real-world works where you frequently get handed a broken widget or process or code and need to “go make it work.”
A former professor of mine was once asked, “why is the documentation so sparse and some of the labs outdated and broken? That makes things much more difficult and time consuming than it otherwise could be.” His response has stuck with me. It was something like “This is a graduate level course. Your job is not just to find the answers, but to also figure out the right questions to arrive at the best answers”. So, although frustrating, it is best to accept it and realize other students are working through the incorrect documentation and successfully completing the labs, projects, courses, and certificate. Also remember that in the real world, you’re paid to solve problems, not complain about them. I know, harsh…but it’s a good reality check on yourself from time to time.
…in the real world, you’re paid to solve problems, not complain about them.
In Conclusion, this certificate is well worth the time, effort, and minimal cost. I highly recommend it if you are on a path to becoming a data scientist.
I hope these observations and opinions were helpful. If so, hit me up on LinkedIn or subscribe to my fledgling YouTube channel at DataResearchLabs.
If you’d rather watch this on YouTube:
https://www.youtube.com/watch?v=wl0X8cCEz00
