I want to be able to switch to part-time work at some point by taking on projects and working part time and/or on and off. Data processing, analysis, knowledge synthesis (data mining+), and creative ways to present data in order to generate or support hypotheses are my strong suits. Also, I find myself policing the statistics on the majority of papers I review, so I figure those are skills in short supply, and I am ahead of the crowd. My idea was to get an online MS in applied statistics, for credibility in recessions, etc. I learn best when left alone with a book, so the degree itself ($$) would really only be for show - very expensive and potentially useless unemployment insurance.
However, I have recently poked around various statistical packages and I found that the main packages cost big money and crunch big data. I figure the companies that can afford the software are likely to already have in-house statisticians, because statistics is often the business model, not means to an end. Now, I am questioning whether paying or even studying for the degree makes sense if it possibly does not offer required skills. Much more importantly, I am questioning the whole part-time plan. In my old vision, there were data processing gigs that I could get here and there. But there possibly could be few side projects for someone without prior big data experience.
1. True/untrue? Thoughts? Are small data projects still available somewhere?
The online program at Penn State (well ranked, ~$25-30K) offers the following elective courses:
Design of Experiments
Epidemiology Research Methods
Design and Analysis of Clinical Trials
Statistical Analysis of Genomics Data
(possibly useless as clinical projects have professional (PhD) statisticians in-house)
Applied Time Series Analysis (for e.g., finance)
(likely useless as market modeling has been using partial differential equations 20 years ago, now they have likely moved on to even more esoteric methods)
Applied Data Mining and Statistical Learning (Data Mining tools: exploring data with regression, PCA, discriminate analysis, cluster analysis, classification and regression trees (CART).)
(possibly useful on big or small data)
2. Is there a market for part-time projects with any of these skills? If not, is there something else that I should be focusing on in terms of statistical skills?
3. Would the last course give me enough skills to get a credible start on big data processing?
4. Is it possible to contribute meaningfully with just one course in time series analysis? Finance I figure not, but are there other uses?
5. Does anybody have experience with the freelance market for statisticians? How workable is the part-time plan at all, data big or small?
6. Given that I have a quantitative STEM PhD from a very highly ranked university, and will have 20+ years of experience in quantitative science research, can I put the degree on my CV/LinkedIn as 'self-study' and sound credible? Is there a way that this is a good idea besides saving money?
7. Could my data mining/presentation skills, together with some non-basic statistical skills (such as the last course) be useful somewhere (e.g., presenting business plans for small businesses, helping smaller research projects in non-STEM fileds)? Could I just skip the whole online statistics program and save myself time and money? if so, what other skills would be useful to acquire? Which statistical skills would be central?
I'd be thankful for any insight.