I imagine proving validity and that sort of thing would be by showing that the figures couldn't be easily achieved by a random answer to each question (or something).

So what, if I can demonstrate a pattern in the data? That sort of thing?

Do you have a choice between dissertation and a final project??

If so then re-consider!

Most people I speak to about final projects are enthusiastic and eager to explain what they are doing and how. When you ask people what they are doing their dissertation on the typical response is "OMG you don't want to know - it sucks and I'm sick of it"

My dissertations were in colouring in and typefaces, but the academic necessities were there:
• quote sources
• research based on individual effort
• being able to keep in the lines when needed

I don’t quite know what you are “dissertating” about, but if you have some kind of statistics to crunch, I would expect that you have the raw data to back it up: questionnaire responses, tapes, original data that kind of thing, as well as a repeatable and solid methodology. I would also expect that you could pass your data through some form of analysis and get result A, and pass some meaningless random stuff through the same system and get result B. What do you expect results A and B to be? What have other people undertaking the same kind of research done? Can you learn from their methods - as long as you quote them as a source, of course.

Is there some guidance from your tutors to help you with this? It’s their job to give you all you need to help you to pass.

I proof-read a dissertation for German friend, who had to do it in English.

The document seemed to be 75% quotes from external sources, 24% biliography and 1% drawing conclusions from said quotes. There was nothing original in the document, but he told me that is how it had to be.

How does that prove you have learnt anything or can come up with some original research?

OK, for anyone who has done a dissertation, I have some questions...

Firstly, is it really necessary to use a program such as SPSS over Excel to produce charts?

Secondly, how on God's green Earth do I "prove the validity" of the data and the "trustworthiness of the results"?

Well my methodology (As recommended) was basically approach every 3rd person on the street and invite them to participate (fill in a questionnaire).

As for analysis, I was basically going to describe the results, produce a few charts based on the data, and then contrast against what I found in the literature review. Does that sound about right, or am I going horribly wrong?

That sounds right, I'm doing mine now.

Oh and I'm just using excel for all my data/charts but the most advance thing I'm doing is T-tests.

If I may be so bold, why a T-test?