@Silverfish, in relation to your p-hacking Q: didn't you have another question some time ago about a form of publication bias that occurs more on the reader's side - something like a "readers' bias"? I vaguely remember this question and I think it was well upvoted but remained unanswered... Not 100% sure if it was your question or somebody else asked it though. In any case I tried to find it but failed. Weird.

@amoeba I did indeed! :-) I'm afraid I deleted it for the time being to give me some time to think about it some more.

It seemed to contain the kernel of an interesting idea and even if not completely original was less widely known than I expected...

I have been considering taking a masters degree in statistics or similar (sadly my statistical knowledge is mostly self-taught - I took the graduate diploma of the Royal Statistical Society, which offered exams and certification but no lectures or classes) and figured it might make an interesting thesis.

Incidentally, if anybody here feels they "know" me a little from my posts, particularly what I might enjoy or find interesting, cares to suggest a course they think I might like, please feel free.

It is a bit off-putting when, as a self-taught person, the general assumption when one reads about postgrad study is that you're doing it right off the back of undergrad. (Goodness knows what I'd do for those places that want "letters of recommendation"!)

Also off-puttingly expensive - not so much the fees (so long as they are in "new car" rather than "new house" territory) but the opportunity cost of largely shutting my business down for a year or so would be rather painful.

(When folk in their early 20s embark on Masters/PhDs, I'm sure the concept of the opportunity costs of not going out to join the exciting world of work must cross most of their minds, but I doubt it actually hurts so much as if they'd been earning first!)

Hey Silverfish. Are you only considering to do MSc or also PhD?

@Silverfish: Have you thought about part-time &/or distance learning?

sheffield.ac.uk/postgraduate/taught/courses/purescience/…

@Scortchi I know there are some online part-time/distance learning courses out there - within UK universities I think Sheffield's MSc might be highest-rated online one (or at least: the one from the highest-reputed university, which is not the same thing, nor necessarily very meaningful)

Birkbeck do part-time courses too, but with in-person attendance required - which means either living in London, or near to it and travelling in regularly. Both rather disruptive and, this being London, costly.

There is a bit of a trade-off here: due to a chronic condition, my health is very inconsistent. If I have a bad patch, then I would likely have to drop out (unless the course was unusually flexible). Healthwise I only narrowly got through my 1st undergrad degree, and it did scupper an attempt I made at a 2nd degree through the Open University (part-time and distance-learning, but assignment deadlines and exams are rather inflexible).

On that basis stretching out a degree over additional years seems to rather dramatically increase the probability of failing to complete it. (Which I think tallies with data about part-time students having high non-completion rates in general.)

Doing a one-year course involves "biting the bullet" somewhat, in terms of disruption if you're already established doing what you're doing. So I'll have to think about that quite carefully.

AFAIK the Statistics MSc at Sheffield U. is the only one offered by a UK university to take distance-learning students since Sheffield Hallam stopped their Applied Statistics MSc. It is highly rated.

@Scortchi Thanks. I did notice they specifically accept the RSS grad dip which helps. Their course structure seemed to involve sticking to a lot of deadlines with quite regular assessment, which rang alarm bells for me as it would require (for me) a long run of unbroken good health to keep up.

(Which was basically the problem with the OU degree. I was very lucky my first degree almost entirely revolved around end-of-year exams, as if all the essays and assignments I simply had to skip had been compulsory, they'd have kicked me out years before!)

(In contrast the RSS diploma, now defunct sadly, was perfect - read the books in your own time, take the exams when you're ready. And if you're not ready this year, take it the next. But with no teaching, all you were paying for that was for the certification. I wouldn't want to pay "new car" type money for that kind of experience, just to get another shiny certificate.)

Most degree courses are like that now - you can't rely on swotting madly through the last few weeks of Easter term.

Yes, it's a shame about the RSS diploma - never really understood why they got rid of it.

Browsing a second-hand bookstore a while back I came across a textbook for the degree-equivalent diploma from (iirc) the Institute of Physics. I think they scrapped theirs when they saw that the polytechnics would be able to increase the supply of physics graduates.

The RSS claimed that university is a suitable alternative for people but it would be financially beyond the reach of many. I've been successful enough that the cost isn't something that would physically stop me going back to uni (though whether I'd find it financially worthwhile or desirable is a different question!) but there must be plenty of people who'd find the money a deal-breaker.

And in an age where there is more separation between teaching (online learning, MOOcs etc) and certification (which is a big problem for the self-taught) it struck me that there is something of a gap in the market there for someone like the RSS.

I fear you're right about the modern masters degree (and probably most of the undergrad ones as well) being an almighty slog all the way through. But thanks for the advice about how well-regarded the Sheffield course is.

@amoeba If I ever find something I'm absolutely fascinated by I'd probably want to do a PhD in it (though a lot of people I know have found them rather dispiriting!) As it stands I have a general intellectual curiosity about many things, but I think I lack the specific obsession that may be a prerequisite for driving through 3/4/5 years of PhD.

I tend to learn a lot anyway, but do think it would be better structured/I would be better disciplined if I took on a Masters. And while a PhD would almost certainly have a large negative financial return for me, a Masters would likely lie between a manageable negative and a small positive (depending on whether i went into industry with it)

There is something @Sycorax put nicely in chat, that a Masters degree lets you work on more interesting problems - I also suspect this is true and this kind of sentiment is probably my main driver for thinking about an MSc!

@Silverfish But if you just want to work on cool problems, consulting isn't a terrible career. I worked for such a company for a few years. Got great exposure to a wide range of problems and data types and analysis methods. It's really the reason I'm able to write useful answers here in the absence of any real credentials.

@whuber Perhaps I've been spoiled by my industry -- we don't have to gather physical items of any kind, instead trafficking in information which can be gathered automatically by various means.

The only wrinkle is when the automatic methods are susceptible to incurring bias, or you need information that is not automatically collected. But we have the resources to directly address those problems also.