Congrats :D

I think in that case you can just say you got the PhD, although I'd imagine there's no harm either in having Expected Completion written if you felt particularly paranoid ;)

Have you passed your viva? Because if you haven't then IMO I don't think you can say you have it yet since that can still prevent you from it (not saying it will! Just caution :) ).

More concrete advice, what I've been making do with is having "Expected Completion: October 2013" under the qualification.

I disagree, a taught MSc and a research one are different degrees, and it does make a difference - that's almost like saying there isn't a difference between the exam and research side of a PhD!

To elaborate - if you are given a choice between two students:
- one has a taught MSc with some relevant courses and a minor project that isn't particularly relevant,
- the other has a research MSc whose thesis was on a topic very close to the proposed PhD topic,
it's pretty clear that the research MSc would have an advantage.

However, don't see this as a discouragement! I don't know much about the exact admittance policies in USA/Canada, but what I've found to be an important admission criteria in general is that the student demonstrates a clear understanding of and interest in the proposed PhD topic.

This can come in the form of good grades in relevant courses and/or research experience. I would imagine that if you got fairly good marks in relevant courses and your research project is relevant to the program that you're applying for, then you'd be in with a good chance. Just be aware if you do mediocre in your exams and work on a project that has little to do with any of the profs' expertise in the program then you are lowering your probability of admittance.

*bump*

If you've ever used the the internet for researching things, your input in this study will be much appreciated.

Please spare a bit of time this weekend to help out with this, thanks!

Please take part in my user study, pretty please: http://goo.gl/XyTKBz

The purpose of this study is to comparatively evaluate methods for results clustering in the context of exploratory search. It won't take more than 30 minutes of your time and it is an important part of the evaluation for my thesis so I would really appreciate your help.

Thank you!

(P.s. You'll also be in a chance to win one of 5 x €10 amazon vouchers if you complete it. Your odds are better than the lottery!)

Default interface is command line so not very user friendly (unless you already have an interest in programming). RStudio is definitely the nicest interface I've seen and is what I use by default.

I wouldn't call it a program but it is a fairly high-level (as in, you don't need to know yucky stuff like allocating memory or pre-defining variables) statistical programming language (it is definitely made for the statistically rather than programming-minded!). If you've done any sort of programming before then it'd be easy enough to pick up, otherwise there may be a bit of a learning curve.

That said, I personally think it's really well worth the effort. The trade-off between amount of time spent learning on it and how much time you'll save after if you want to do any regular statistical analyses and plotting graphs (they are super pretty) is well worth it.

For example, say you want to to generate 10 random numbers from a uniform distribution between 0 and 1 and you define that list of 10 numbers as x:
"x <- runif(10,0,1)" (runif being an abbreviation for Random UNIForm).
Then to calculate the mean, all you do is just type "mean(x)"
If you want to plot that as a histogram, it's just "hist(x)"

Plus it's open software - so it's free which is always a bonus :)

If the test you've been given is to assess your ability with R then isn't it kind of unethical to be doing this? You are not going to have a personal R tutor when you are on the job, why would you have one when you are being assessed on it.

However, if the reason of the assessment is not to assess your ability to code in R, but say, rather on your statistical analysis or reasoning, then I may be able to help since syntax questions can be solved quicker this way (PM would be better).

Either way, I've found that this page has some good resources for R:
http://www.computerworld.com/s/article/9239799/60_R_resources_to_improve_your_data_skills

Good to hear your reply iqbalu - you sound like you are in a reasonably good position :)

I guess the one thing you need to make sure (and it sounds like you are already aware anyway) is how much interest and therefore how much time your potential supervisor would likely spend thinking about your research problem. But since he's offering you a position then he perhaps he is already prepared to do so. Again, check with current PhD students to get a feel for this, I'm just speculating ;)

More interesting is that you mentioned that other PhD students were requesting a postdoc to join the team (makes total sense to me!) - do you know if that's likely to happen? I would think that if there is a lot of uncertainty around the supervisor's availability then you would be wise to make sure that there is a 'backup' of sorts. I know that in some more structured (and arguably more successful, meaning less dropouts/failures) PhD programmes actually require that there be a secondary supervisor or a mentor which serves this purpose.

Other than that, it sounds like you have a decent offer here, just double-check with the current students, especially those near finishing. Of course a PhD at the end is what you make of it, and that includes putting effort into evaluating your options :)

Good luck!

Hi iqbalu,

If your personal preference is very different from what this supervisor can offer then I would likely say no if I was in your position. Like you, I prefer discussing ideas on a regular basis, but I went ahead with a busy supervisor. I am now looking to finish with a research masters instead, after 3 years as I lost all motivation after 2.

However, it certainly isn't an impossible situation. So if I were to start over, here's my (more optimistic) advice / some questions I think would help with your decision:

- do you have a concrete project in mind, and how much do you already know about that subject area? If you don't, then working with a someone who can only respond to you once a month can seriously hinder progress, especially in the beginning. Even more if your preferred working style is having someone to discuss ideas with. This can be overcome if there are other senior researchers within the group who you can meet on a more regular basis. Are there reliable people like that in the group?

- how was your working relationship with this supervisor during your MSc? Better yet, if you know any of his current/past PhD students, ask them what it's like to work with him on that level. Don't ask about whether he is nice or not, more important is feedback and working relationship (does he respond to email, offer opportunities for teaching/attend conferences). This will be someone who will effectively be your mentor for the next 3-5 years. How often do people have the same manager/colleague for that long, by choice? This is the kind of choice you're making.

- what is the research group like? Both in terms of research interests (potential for collaboration) and as friends (for support!). They can work with or against you. Most groups are great and you'll make lifelong friends from them, but it's not always the case (there are some horror stories from this forum!).

Hope this helps.

I don't have access to this but I'd suggest emailing the author(s) for a copy (you'd be surprised how enthusiastic most authors are about people wanting to read their work ;) )

My first impression is that it looks good for someone who doesn't already use LaTeX (like myself) as the learning curve may be less steep. It's also quite neat that you can export as HTML and LaTeX, not only PDF :)

The biggest problem that I can see is that the documentation is sparse and very 'techie'.

Installation of Word is a no-brainer (MS Office Suite has to be easy to install and use, given its target market is like everyone under the sun), and although LaTeX is a little more complex, there is an abundance of online support. And from my not-so-deep peruse of how to get the environment for rST up and running on my machine, I'm finding the documentation very unwieldy.

The fact that you have to have python up and running would probably be confusing too. I can't really imagine anyone who isn't of a technical background and who doesn't already know how to script (in which case, would probably already be using LaTeX) be persuaded to use it otherwise.

Ok, well without looking at what you used to run the experiments exactly, I don't think I can help you :/ Sorry!

It sounds like you need to understand the code you've been given, or at least its structure so you can modify it to export what you need. I had assumed that with each experiment the output was exported to some textfile for analysis.

Wow, people actually use Cogent outside of coursework? :P

Can you export the data as a .csv file? (Or even just .txt) That would be my choice of format. I'd turn to Google for anything more complicated...

However, I'm a bit confused now since you say you use SPSS for analysis, yet you were trying to run a Matlab script to compute stuff like means? Does SPSS not support that? (I haven't used SPSS before, but hear it's the default software in Psychology/Social Sciences.)

I would advise that you use R (here's a good tutorial: http://personality-project.org/r/r.short.html) for the analysis you describe. The most important part IMO is getting the data frame, i.e. data format correctly. E.g. have columns as male, female, etc., and rows as the subjects.

It is possible to do them on Matlab too, but not realistically possible if you don't understand Matlab itself given that you would have to define a lot of your own functions to calculate standard stats analyses measures (well technically there are some, but again, if you are unfamiliar...).

To repeat, R has loads of libraries, and its original focus was for the statistics community so you're more likely to find what you need there. Plus it's free :)

(Aside: I used to use Matlab but now found R much more useful, esp. for statistical analyses. Matlab is good if you are dealing with big matrices, and it does produce some pretty plots. But, I would now say R has caught up in its plotting mechanisms and it's easier for small-medium sized datasets.)