Thinking of leaving math PhD - do what?

[RollingRust]

Hi everyone,

I'm looking for a bit of advice on how to approach my situation. Any input will be much appreciated, no matter how blunt.

My situation: I completed a BSc in Math (Honors - sort of a masters equivalent for those not familiar) in the middle of 2010. After finishing I kind of drifted around for about a 6 months, working on some small research projects but not really knowing what to do.

In 2011 I worked at a graduate role in an engineering company's computational modeling division. The program was terrible, and I was really depressed and desperate to get out, so I left after a year (was a 2 year program) and started a PhD in math (computational) at an Australian university early 2012.

So here I am now, in my PhD.. but I've realized i just don't love the subject. Not only that, I'm not really all that good at it, and to top it off I'm a bit under prepared. I'm thinking of quitting my program and going out into the work force, but I really have no idea what I could look at doing? At this stage I really just want a career where I feel useful which I can commit to.

I feel like I've made a right mess of things, and I'm just hoping people can give me a few ideas of what my options might be. Yes I could stay in the PhD, but would this really be a good idea? I'm about 4 months in now, and already feeling pretty doubtful about it.

Thanks

 

[anjelin]

If you do not love your subject then change it. In early times of postgraduate years (PhD) actually is difficult, especially if you do not have a good scientific advisor. But then when you enter into the research work it will be interesting and not as terrible as was in the beginning. If you do not need much money continue your work doing on PhD.

 

[RollingRust]

Thanks for your thoughts.

The problem is not my supervisor, or my particular research topic. Just being around other people in my department, I realize I don't have a love for the subject that they do. I don't want to study it 'for the love of it'. I find myself constantly feeling like I'm wasting time learning esoteric things that will never be useful to me.

Don't get me wrong - if you love the subject, who cares if it is esoteric and abstract. But that isn't me

 

[TMFKAN64]

I would say, consider yourself *very* lucky. If you realize that you don't love the subject after only 4 months, you've saved yourself a *lot* of time and effort.

If you don't see yourself doing what you are doing now for the next 40 years, get out while you still have a chance.

I can't advise you on what you *can* do now, but I can say with certainty that staying the course is not a viable option.

 

[ParticleGrl]

If you don't see yourself doing what you are doing now for the next 40 years, get out while you still have a chance.

Well, that's not true. Most math phds won't get to keep doing research math for very long after their phd. Its more likely that if you don't enjoy grad school there is no point to doing it because other than time-spent-in-grad-school there is little reward to the program.

To the original post, you are probably right to get out. There is no reason to get a phd in math other than a love of the subject- its not going to be a good economic investment.

 

[RollingRust]

Well its nice to have people confirm my thoughts. Cheers!

But any idea what options I might have? I do have pretty strong programming skills, and experience working in a variety of applied topics. But I'd rather not be a code monkey.

Honestly, not a lot gets me very excited. So I just want to find something 'worthwhile'. I don't mean saving the world, just a good reason to get up in the morning.

 

[TMFKAN64]

Well, that's not true. Most math phds won't get to keep doing research math for very long after their phd.

I don't think that matters. It's not that you *will* be doing what you do during your Ph.D. for the next 40 years... but if you don't *want* to do it for that time, you are better off doing something else.

 

[Skrew]

If you stop now going back to a PHD program may be more difficult.

 

[epenguin]

I'd agree with the strong majority - decide now and trust feelings.
If he sticks with it for a year of two he will have too much invested to stop and the undeniable, and at that point right, argument will be to complete. So he is now deciding the next 4 yrs and further really.

Skrew is saying he is already there

If you stop now going back to a PHD program may be more difficult.

I think this has be be resisted if this more difficult thing he will not be able to get back into is not what he wants.

I am not clear whether it is just math, which he has done 3 years in, or computing that turns him off. Most applied aspects I'd suppose involve some computing, and most employment would be like that, but he has already tried 2 projects at least that were not for him.

Trouble is he hasn't anything else lined up, any second interest that we know.

There are lots of areas, finance, ecology, health, admin, education, ... where the focus, the end, is one of those things and math or computing skills are valued means towards. Or he may be better to make a total break, he has to 'know himself' to decide.

 

[StatGuy2000]

Well, that's not true. Most math phds won't get to keep doing research math for very long after their phd. Its more likely that if you don't enjoy grad school there is no point to doing it because other than time-spent-in-grad-school there is little reward to the program.

To the original post, you are probably right to get out. There is no reason to get a phd in math other than a love of the subject- its not going to be a good economic investment.

ParticleGrl, I agree with your view that the OP should probably get out (after all, if he/she does not have a love of the subject, then it is pointless for him/her to continue with their PhD).

However, I disagree with your assessment of the worth of a math PhD (as opposed to a theoretical physics PhD, which you possesses and have a better grasp of its value given your experiences), if you include statistics as a field of math (which some argue is the case). Many employers (in both the US and Canada -- this may also apply elsewhere in the world) who hire statisticians often require their employees to have a PhD.

Furthermore, at least anecdotally, there seems to be a healthy market for tenure-track or research positions available in statistics, so those wishing to continue research in this area have more opportunities to do so (I am not immediately aware of any official data on this question).

 

[StatGuy2000]

Well its nice to have people confirm my thoughts. Cheers!

But any idea what options I might have? I do have pretty strong programming skills, and experience working in a variety of applied topics. But I'd rather not be a code monkey.

Honestly, not a lot gets me very excited. So I just want to find something 'worthwhile'. I don't mean saving the world, just a good reason to get up in the morning.

I think you should clarify what you mean about doing something "worthwhile". OK, you admit that not a lot gets you excited -- is there anything that interests you? An applied math background is a starting point for careers in areas as diverse as software development, data mining/analytics/statistics, finance, health admin, ecology, education, etc.

So there are plenty of options career-wise; what you need to ask yourself is if any of these fields are places you could see yourself working, and being satisfied in working there.

 

[RollingRust]

I'm aware that if I stop now it will be difficult to go back - it was already a little hard going back after having worked for a year. However, I don't know if that's a good reason to dedicate three or four years to it (programs here are slightly shorter than the states).

First of all, I do like computing and I do like math. And I am willing to have a job centered on those. I just don't 'love' math, in the sense that when I start reading more abstract material I can only think to myself 'this is cool, but what the hell is the point?'.

I think you should clarify what you mean about doing something "worthwhile".

Yes okay fair enough. I suppose vaguely what I mean, is something which has two qualities:
1) The work to have clearly defined goals which result in tangible benefits for my community (this can be nationally, locally or whatever).
2) Doesn't involve any kind of marketing (sorry if this offends anyone) or exploitation.

Okay, this is still a bit undefined, but basically I want to be able to go to work and know that what I'm about to spend my time on is going be useful to other people. Research could certainly fit this description, but the problem is that I don't think I'm good enough at math to make particularly meaningful contributions. Further, I think a good deal of math is purely an intellectual pursuit. DON'T GET ME WRONG - I do not look down on this, it just doesn't motivate me personally, I wish it did.

Interests. Statistics, ecology, uncertainty quantification, risk assessment, law, electrical engineering?, mountain biking :P.

Everyone has always said to me that a degree in math would open up opportunities in these kinds of fields, but I really have no idea how to find these opportunities. Presumably I'll need to retrain to some extent for most of them, and I'm not against the idea of doing postgraduate study, but I just have no idea where to start looking.

I think ideally, I would find an entry level job somewhere that they would be willing to train me up in something, but I'm open to any suggestions.

Finally, thanks everyone, I really appreciate your input, I am finding this quite difficult.

 

[ParticleGrl]

I do work for an insurance company in data mining, some of my fellow employees have masters and phds in statistics, computer science, physics, math,etc. However, most of the people I work with just have bachelors degrees (generally in business/finance,accounting, actuarial studies). I would think if you could find a company doing data work, you could get in the door without a phd. While this sort of work might not be worthwhile for humanity/society at large, its pretty useful for whatever company you work for.

 

[RollingRust]

Like I said, not hoping to save the world, just don't want to be totally irrelevant.

Particle girl, any tips on how I might get my foot in the door at a company like that? I'm not entirely sure its something I want to do, but certainly something I'd look into.

 

[StatGuy2000]

RollingRust, you had stated that one of your interests is statistics. If so, as ParticleGrl has suggested, data mining or business analytics is one avenue that is open to you.

Another option to consider would be to work as a statistician or biostatistician for the research/teaching hospitals, the pharmaceutical/biotech industry, or health-care consulting firms or organizations (the type of work that I am involved with, by the way).
Statisticians who work in the health-care sector are involved in such areas as the design and analysis of clinical trials, analysis of health economic data, or the analysis of genomic data.

Since the results of the analysis have an impact (directly or indirectly) on patient health, this could satisfy your requirement that the work be "relevant".

 

[RollingRust]

Thanks StatsGuy, both of those avenues do sound like options I would be very interested to look into.

Since you work in the statistics/health-care field, do you have any suggestions on what would be my best course of action to move into this field if I wanted to? Would it be best to try and change my PhD topic to something in that area? Or maybe do a Masters? Or could I jump right into work?

Something I'm actually looking at doing right now is implementing some of my current work into the package 'R'. I was hoping that if people were at all interested, then maybe It would be a starting point to network with people.

 

[StatGuy2000]

Thanks StatsGuy, both of those avenues do sound like options I would be very interested to look into.

Since you work in the statistics/health-care field, do you have any suggestions on what would be my best course of action to move into this field if I wanted to? Would it be best to try and change my PhD topic to something in that area? Or maybe do a Masters? Or could I jump right into work?

Something I'm actually looking at doing right now is implementing some of my current work into the package 'R'. I was hoping that if people were at all interested, then maybe It would be a starting point to network with people.

Whether you can jump right into work in the statistics/health-care field would depend greatly on what your background is in statistics thus far, and what experience you have in conducting statistical analysis or modelling or applied statistics

I would personally suggest you pursue at least a Masters in statistics (with an emphasis on applied statistics) or biostatistics. A PhD in statistics would also be valuable, in the sense that for many employers, earning a PhD is counted as having earned 2-3 years of work experience, compared to just having a Masters. Therefore, you will have an edge in applying for a variety of jobs in pharma, biotech, or health-care organizations; ditto for analytics or finance type roles.

Implementing your current work into R is an excellent idea if you intend to pursue further work in statistics as well, as R is widely used by statisticians, especially in academia or in research settings. It is also increasingly used by a wide variety of businesses as well. Knowledge of R would certainly also help in networking; there are a number of groups you can join on LinkedIn to share tips and such.

One further advice I would add is for you to acquire programming skills in SAS, since much of the data analysis conducted in health care or pharma (and also in business analytics) is conducted in SAS.

 

[nucl34rgg]

You could become an actuary. With your skills, you would probably be a good fit.

 

[seaofghosts]

My boyfriend had to quit his PhD in math because he was going through a divorce, and had to find a job, so he went into programming. He found that with a little self-teaching, a lot of programming companies were impressed with a master's in math and were lining up to hire him, even in the rough economy. He also finds it very rewarding and gets excited about learning new languages and moving onto better opportunities and more money.

 

[mSSM]

My boyfriend had to quit his PhD in math because he was going through a divorce, and had to find a job, so he went into programming. He found that with a little self-teaching, a lot of programming companies were impressed with a master's in math and were lining up to hire him, even in the rough economy. He also finds it very rewarding and gets excited about learning new languages and moving onto better opportunities and more money.

What country are you guys located in?

 

[RollingRust]

StatGuy2000: Thanks again. It sounds like something along the lines of either converting my PhD to Stats, or finishing a Math masters and then looking for a Stats PhD might be my best option. I work in a 'probability related field', but I don't have much experience with applied statistics.

If you don't mind, I would be very interested to hear a little bit more about what you do, how you found the work you're involved in etc. Maybe a PM is more appropriate? Sorry if I am asking too much here.

nucl34rgg: Something I've tried looking into. I am a little confused about the process, and to complicate matters it seems the vary a lot country to country. It is however something I would consider, do you have a specific idea of what I would have to do to go in that direction?

seaofghosts: I'm sorry to hear about your boyfriends trouble - but happy to hear he enjoys his work! I guess honestly, I always hoped this would be a backup career for me. Its nice to know that he found himself in hot demand.

I'm sorry if I'm sounding a little difficult in my replies - I really don't intend to me. I'm just a bit down about my situation and quite confused about concrete steps I can take.

 

[nucl34rgg]

I can refer you to a site with good information. I am looking into this myself currently. I do not yet know what it fully entails. I do know that there are a lot of exams to pass, and that you must pass some of them to even start work as an actuary. I think you would need to self study some financial mathematics and economics, and review statistics.

http://www.beanactuary.org/

 

[RollingRust]

nucl34rgg thanks for that link, looks like a lot of good info there.

It looks like its a pretty long and specialised path to follow, I had better do quite a bit of research before I consider it as a career.

Easy to worry that the job market might be different by the time I'm done though!

 

[StatGuy2000]

StatGuy2000: Thanks again. It sounds like something along the lines of either converting my PhD to Stats, or finishing a Math masters and then looking for a Stats PhD might be my best option. I work in a 'probability related field', but I don't have much experience with applied statistics.

If you don't mind, I would be very interested to hear a little bit more about what you do, how you found the work you're involved in etc. Maybe a PM is more appropriate? Sorry if I am asking too much here.

Hi RollingRust, I sent you a PM yesterday. Feel free to PM me back for further questions.

 

[seaofghosts]

What country are you guys located in?

We're in the US, Las Vegas.

 

[Chronos]

Programming skills are a huge selling point with most companies now. They don't need to hire an IT guy if you can double duty.

 

[hadsed]

I do have pretty strong programming skills, and experience working in a variety of applied topics. But I'd rather not be a code monkey.

...

Honestly, not a lot gets me very excited. So I just want to find something 'worthwhile'.

I'm not meaning to be accusing or critical, but is it possible that maybe you only have some kind of vague image of what these other jobs mean? For example, you say you don't want to be a code monkey (which betrays your perspective on programming work). Certainly there are people who are practically 'code monkeys', but these are people who remain stagnant at entry or intermediate levels, don't do useful work, aren't curious and interested in applying their brains to difficult problems, etc. There is a huge amount that goes into writing high quality, reliable, and efficient software. It requires a lot of practical and logical thinking to figure out how to use proper data structures, how to architect your object hierarchy, how to make it efficient for future revisions and additions to the codebase, logistical issues like making it easiest for lots of people to contribute, etc.

I'm projecting a bit here, of course--I come from a programming background, but having studied physics for two years I'm moving back towards computer science, partly because of the dismal outlook for careers in theoretical physics (and the same can be said for pure math). Computer science, however, is practically applied math (though in a different sense, and with different math). If you enjoy topics like graph theory and general logic for things like building and optimizing networks, working on computer architecture, coming up with more efficient algorithms, applying statistics, probability, and linear algebra in conjunction with artificial intelligence (read: machine learning) to data mining/computer vision/robotics/etc., you'll be very happy with CS.

It took me a long time to get interested in things other than what I had plans for (which was theoretical physics). But the more you read about these other topics, the more you think about them, and the better image you have about the work, all of which of course contributes to building up a sort of passion for doing it. For me, I've realized that I like to solve difficult problems, and some of the most important problems today rely on making computers faster, more reliable, or more efficient.

Also, perhaps you can make the jump a little differently. Say, do a Msc. in mathematics concentrating on... say, numerical linear algebra. Then go into CS and work in high performance algorithm implementation of your Msc. topics. "Going into" could mean do a Ph.D, or straight into industry. Or, say, you want to work for an aerospace company. So tailor your topic to work on numerical solutions of PDEs that are important in computational fluids research (it gets more specific, for example you can come up with a way to eliminate need for more artificial parameters so that simulations are more accurate and natural). I can't say this would be how you get hired by an aerospace company, but I know the basics of similar paths people have taken.

I'm saying all of this because of your comment earlier talking about when you're reading about cool mathematics, but wondering what the point is. Well it can be applied to lots of really awesome stuff, you just have to know where to look. As a person with lots of interests, I've spent a lot of time trying to figure out how to make the various jumps, from a physics student with heavy computational experience to actual hardcore computer science/engineering. It can be done, but most people don't do it this way which means you'll have to put your brains and social skills to the problem and forge a path on your own (and don't underestimate the social skills, some of the most valuable information I've gotten is from professors in CS departments who I've explained my situation to).