new TESU degree - data science (statistics)

[SolarKat]

Have you confirmed that they are fine with you completing all the grad cert courses before your Bachelor's conferral?

Yes - strangely enough, all 4 of the grad classes are specifically approved electives in the BS Data Science & Analytics degree plan. It's just a happy overlap that I found while prowling around the TESU website. And I'm hoping they don't change it before December...

[davewill]

Any finesse to doing this as a second bachelor's? I'm interested in it for my son who thinks this might be a good direction to take his career. He has a BSBA in Marketing and Accounting. Going straight to a master's in such a challenging subject would be very difficult (but I'm looking at Harvard Extension School for that as well), but this (the BS+grad cert) could be a nice step that could pay career dividends all by itself.

[SolarKat]

Any finesse to doing this as a second bachelor's? I'm interested in it for my son who thinks this might be a good direction to take his career. He has a BSBA in Marketing and Accounting. Going straight to a master's in such a challenging subject would be very difficult (but I'm looking at Harvard Extension School for that as well), but this (the BS+grad cert) could be a nice step that could pay career dividends all by itself.

Does he want to deep dive into pure data, or just take a more data-intense view of his business foundation?

First - the TESU 2nd bachelors option: Statistics.com will be about $6K. Add the TESU residency ($2.5K) and graduation fee. That $9K might be better directed toward a masters outright. If he likes adding data to his business skills, many masters will cover it without requiring a deep data background as a prereq: ops research, project management, finance, etc.

Next, the TESU grad cert option. Honestly, I don't know how they could demand that you pay residency when you don't have the option to take *any* of the 4 classes via TESU directly. That's a question to ask TESU up front. Otherwise, I'd probably suggest doing the 4-course Statistics.com "Optimization Specialization Certificate" first (Optimiz, Integer/Nonlin, Risk Sim, Fin Risk - Opt & RiskSim both only require spreadsheet skills, so no worries on a deep foundation requirement...and the other 2 are built on the first 2). That'll be $1600 if you pay up front for all 4 using the specific coupon code for optimiz, plus a $65 records fee. And it's a total of 4 months of class, total...the trick is the timing, as I posted earlier. Three of these classes are ACE-accredited grad level for a total of 9 credits, so you might find a masters to roll them into. And optimization is a useful resume builder that immediately has meaning to hiring managers.

Then, he could take the Statistics.com Regression Analysis class, which would be the 4th class for the TESU grad certificate, which is more vaguely titled "Analytics". That gives him another college to list on his resume, obvious graduate coursework, in addition to the applied optimization resume item from Statistics.com.

But the bigger question is, what's the end game. If he really fancies a deep dive into pure analytics, the TESU 2nd bachelors will give him a taste of a lot of things, but he'll repeat them in the masters (predictive analytics, data visualization, modeling, etc). Rather than (or in addition to) the 4 electives that comprise the TESU grad cert, he'd be well served to take the Python, R, R/SQL electives...and then add on, beyond the mandatory degree/cert path: linear algebra, the calculus sequence, possibly databases, data structures, computer architecture...because pure analytics is about the collection/sorting/storage of data, the tools to analyze it, and the algorithms to tweak the efficiency/effectiveness. It starts to look a lot more like computer science than business. Which is why I suggest taking a look at job descriptions and masters programs for a variety of careers/emphases, and work backwards from there so he's not wasting effort on courses that won't move the needle for him.

[SolarKat]

If anyone is interested in the Statistics.com platform, they periodically offer free "preview" weeks in advance of the actual start of certain courses. It'll give you a taste of how courses are structured, without any obligation. I believe they give a coupon code for anyone wanting to continue to the full class...but they offer academic affiliate discounts too - I don't know which is the better value. Preview weeks are offered for 3 ACE-accredited options:

- Intro Stats Preview (the full course is 2 4-week modules, Stats 1 and Stats 2, with a proctored final)
- Predictive Analytics Preview (the full course is 4 weeks, Predictive Analytics 1 - Machine Learning)
- BioStatistics Preview (the full course is 4 weeks, Biostatistics 1)

Here's a link to the Statistics.com master calendar, in case you're looking at multiple options. And here's the master class list - anything with a checkmark beside it is ACE-accredited. (Though Stats 1 & Stats 2 *both* must be taken, to receive the 3cr lower-level ACE credit, transcribed at ACE as "Intro to Statistics 1 & 2".)

[SolarKat]

Just a tiny bit of inside baseball - there's a certain amount of grumbling from a number of students about how quickly the Statistics.com classes move, how there's not a lot of guidance (in advance) on problem sets, and how the hand-in quizzes/assignments may not reflect actual knowledge/ability. To some extent, they're correct - the classes are 3 college credits in fairly rigorous subjects (machine learning, network/graph theory, etc) in only 4 weeks, plus they introduce new software tools at breakneck speed.

The professors/TAs are highly responsive even at weird hours and over the weekend, which mitigates this to some extent. I think part of the problem is people waiting until the last minute to start the work for the week...which may work in a more slowly paced class...but it totally doesn't work in these. By the time they realize where they need help, it's the end of the week and time to move on. Also, the professors expect a certain...maturity...in approach, that learners are going to research topics, work practice problems, post detailed questions about things they don't understand, etc. These classes are anything but spoon-fed.

Further, for some classes, the assessed quizzes are literally only a handful of points, so even one miss does affect your grade. However, the quizzes are not really "quizzes" in the standard sense - they open at the beginning of the week, and you can revisit them as many times as you like prior to submitting your final answers. You're also allowed to ask questions in the forum about possible approaches, or software techniques to complete the analysis, etc...so it's not like a blind, timed exercise. I think people unfamiliar with the class are thinking it's a blind/timed thing and not opening it until last minute...which is a problem if they haven't been practicing the techniques all week. Since it's only a 4 week class, there's not really time to develop any sort of "rhythm" to your study...it's basically a mad dash (2 or more chapters in 1 week, often a new software tool or technique on top of that, plus fairly detailed software analyses/projects) for 4 weeks.

I guess my personal perspective is this: be prepared to drink from the firehose. I know that for many 12-16 week college classes, the amount of time required weekly runs less than the "three hours per credit hour" estimate I've seen around the 'net. Be aware that these 4-week classes can come very close to the suggested time investment. Statistics.com suggests 10-15hrs/wk. I agree with this number, particularly if you're determined to become proficient in the techniques vs just checking a box. And be prepared to dig for answers and to ask questions early & often. The materials are good (lectures, readings, software help features), but they expect self-directed, invested, proactive learners.

Finally, the classes I've taken to date seem mostly targeted toward people working in business/industry who're coming back short-term (one or a few classes) to learn specific skills relevant to their jobs...so there's already a body of knowledge that a non-career/young student may not have, and a specific interest in the course at hand. I don't think this precludes study in the program from a "degree" viewpoint, but it will probably require extra work by people not very proficient with Excel, or who aren't sure *why* things are done/of interest in data. Statistics.com has been a provider for career professionals - the foray into academia is definitely changing their student demographics. Not sure if this will foster changes in their system...they've been doing it well "their way" for better than a decade. For now, I think the college students (particularly those without business/career/reports experience) will continue to be in for culture shock when they start the data AOS for this degree.

YMMV.

[bjcheung77]

Thanks SolarKat for such info, very valuable. For me, I won't be taking any courses from Statistics.com for my degrees.
As my goal is the TESU BSBA and then the BSIT-Security at WGU, it would be an interesting mix of degrees for me already.