I'm currently writing up my undergrad dissertation so feel free to ignore this as it's not strictly PhD related. But was wondering if there is a minimum/maximum for number of references you should have. At the moment I have 35 and about 4,500 words, my word count at the end can't be more than 6,000. Is there a point when I should stop or does it not really matter. (It's a science diss btw)

Asking how many references to use is like asking how long a piece of string is. The number of references depend on several things-
1. The popularity of your area (if there are tons of papers in this area, then you will inevitably find more references)
2. How in depth you are going (as its an undergrad dissertation with a 6000 word limit, you will not be going as in depth as if you were writting a PhD literature review, or a critical review paper)
3. Its not the reference list that earns you marks, its the quality of your work, so don't just add references to make the list get longer.

I also agree with Stu. But on the other hand, I can understand your concern. I recommend reading a few dissertations from past years to see how they used references. You should be able to have access to the selected ones at your Institution's Library. It is also important to refer to journals related to your subject; not only books. Good luck with your work.

My undergrad dissertation of 20,000 words contained 76 references. Most of these were very relevant but others were simply acknowledgements that several authors had come to same conclusion on some issue.

I tend to find review articles a good benchmark as to the amount of knowledge/literature there is in an area (obviously depending on how closely the review article matches your topic). Hence, if they find only a few references you can safely assume that you're not going to do much better than them.

The best advice I can give is that relevance is they key here rather than the trophy-hunting!

I always tried to put my OWN ideas in. Tended to get a good reception as opposed to the students that went for endless pages of references.

I personally found there tended to be very little development of students own ideas in written work, with rather a complete reliance on regurgitating someone else's ideas. In my experience students can be quite scared of putting their own ideas down if someone else hasn't said it first. I am obviously proposing this as an alternative academic lifestyle as opposed to the usual student methods [e.g. More is better re references]. Thought that was obvious.

That helps a lot! Thanks. Most of my references are from my intro where I'm setting up the idea etc, and a few from the disscussion where I'm criticising other papers that did the sameish expereiment but came to different conclusions. Another quick question, is it bad to have references that are quite old? I have the obvious historic ones from 1960s but also a lot are from 1990's. I do have recent ones as well, just not the majority

Not necessarily bad to have old references, but you must check that the research has not been superseded in the meantime. This is where review papers give some steer.

Alternatively, why not stick the said article's title and author into Google Scholar and see who has more recently cited that article and what they were talking about.

Katq, in my field, I need to refer back to the first works in the area, and so some references are old (1950's) but this is the 'story telling' of research, using principles and theory that have been around for ages, whereas 95% of the references are from the last few years. You shouldn't base your research entirely on old references (i.e. your reference list shouldn't be predominantly pre 2000).

Golfpro, you are still missing the point. If you are going to put your own ideas about, you need to substantiate them with references (which argue your point) and if possible actual experimental data.

Katq, your introduction should really be an intro to the work, and a review of the current literature, so nearly all the references will be here. So you don't need to worry about that.

No I got the point Stu. I am just very against disserations-by-numbers. Agree with you re experimental data btw.

Well, I'd pretty much never reference a review, only ever original papers. I know what you mean about undergrads taking all their ideas from references because if you find a well written review it's very difficult not to take the ideas on as your own if you don't know much else(which is kind of the purpose of a good review surely?). But if you are only referencing original papers you need to have drawn your own conclusions/ideas from them anyway.

I'm not saying you quote the review (unless that review makes conclusions itself). Of more interest are the papers that they read and making sure you read them for yourself. The review is merely an opening into the topic and the state of current knowledge.

I focus my effort on those reviewed papers which have direct relevance to my dissertation.

I'm not boasting about the 76 papers I read over 15 months, but what did appreciate was that reviews frequently quote articles wrongly, or miss the point of the paper, or selectively use those parts which support their own point of view and disregard those whioch don't.

It's also common for a review to be less than comprehensive. I was amazed that a recent review in my field completely ignored a very emminent piece of research that was very relevant. Why? Because that article was written by a competing research group (who themselves were ignoring the first research group's research).

Who said academics were thorough?

onwards&upwards yes your wrong. MSc with dist [2005].

We all shouting about our qualifications now? I can swim 10m unaided

No sorry if I came across as a little abrassive there O&U. You make a very valid important point. Notwithstanding that it's nice weather for swimming isn't it. Wouldn't it be super in this weather to go to one of those huge open air pools they have in London?