Online Survey: Uni students & Stress! (2 Viewers)

[Cyph]

Originally posted by z_morris
Yes, good thought. This is not my only avenue of data collection. So im hoping that by using different groups or by going through subject co-ordinators too ill get a range of students. Can you think of any other ways??

I think all results from this survey are invalid. What about the people who see this and don't do it, like myself (reason being, I feel the results are inaccurate.. not because I don't want to help). Could say there's a fair chance that the people on this board are more enthusiastic about Uni then most people and the ones that do the survey are even more enthusiastic. You have no way for accounting of those who don't do this survey, it's why surveys in magazines where you have to send results in are totally inaccurate - those who do send it in must feeling strongly enough about the subject, good will won't work on all people.. maybe in an idealistic world.

The sample has not truly been randomised. Unless you attain a true random sample, your results are going to be way inaccurate.

I think you need to devise a method, where you select the people who undertake the survey, perhaps at the Uni and you need to ensure that you cover the whole Uni at random times and select random people as they come past to avoid discrimination based on looks, etc.

You may want to cover the library, the eating hall, eastern entrance, western entrance, northern entrance, southern entrance. People coming in from the west may be mainly arts/commerce students, while the east is all computer/engineering students. Even then, you need random samples of those students.

You will want to cover the areas at different days and different times and select random people as they walk past. Draw times out of a hat, roll a die, etc.

Maybe you could go and see a stat professor/lecturer to get them to fully explain it, as I've only done a level 1000 stat course, but feel pretty confident in saying your methods won't yeild accurate results seeing as you are only sampling a small percentage of the population, which is not as accurate as samplig the whole population (called a consensus, if I recall correctly).. so your sampling of a small percentage needs to be very accurate! The more data you have, providing it's randomised, the more accurate it is. Depending on how you analyse the data, you may need quite a bit of data.

 

[hipsta_jess]

argh, i did the same questions the other day for someones research at my uni. grr.

 

[Cyph]

Originally posted by iambored
yeah i always wonder about this with any research

I am always wary of research and experiments until I see the method(s) used. It is essential to include in any research/experiments done, so people can reproduce it if they wish ad see if they yield similar results. Of course, with assignments for Uni it doesn't matter

 

[iambored]

Originally posted by Cyph
Of course, with assignments for Uni it doesn't matter

and that's why you posted the longest post in this thread? which i completely agree with btw. but still, someone has to explain to me how ANY research is not biased in some way. maybe if the sample is chosen randomly off an electoral roll and they get a 75% response rate? besides that, i would say all research is biased.

and z_morris, you're doing a good job, there is nothing else you can really do

 

[z_morris]

Hmm....

thanks for your input guys... I understand that my results may not be generalisable to the complete population of university students. But Im hoping to say, of this sample, who responded to my survey, the stress, coping and health experiences show that..... Although this site is for people who are 'into' the whole uni experience, it is also for those who may be having difficulty, and who are seeking support. Alot of other students are procrastinating, and really its just another chat room as such but we all have studying in common. Hmmm definately stuff to think about. Plus I am looking at the stress experiences of uni students not how pro-uni life people are.... all students, in fact all people experience stress in many areas of their life.... so the fact that a number of my participants are visiting this site does not mean that they have the same amount/type of stress and they deal with it in the same way.

I guess we are used to dealing with 'fake' data in our stats classes and we just expect that everything should be perfect, when in the real world most things correlate at about .2 not .8 like we would like them to. but we get excited about .2 cause its what you expect from real data. hmmm. am i babbling??

Thanks again for everyone's input!! zoe

 

[Ragerunner]

Originally posted by iambored
and that's why you posted the longest post in this thread? which i completely agree with btw. but still, someone has to explain to me how ANY research is not biased in some way. maybe if the sample is chosen randomly off an electoral roll and they get a 75% response rate? besides that, i would say all research is biased.


and z_morris, you're doing a good job, there is nothing else you can really do

You'd find that yes it may be a bit biased, and the results may not be that accurate, but is accurate enough to make a lot of judgements. If you studied Psychology you'd find in unusual situations, psychologists may be able to predict your next action based on this research. e.g. the door in the face technique.

 

[z_morris]

...hmmm (x2)

I know, even if my sample is biased, which apparently every sample is. Well there has been previous research looking at the scales i am using...soooo if i compare it to that and find similar stuff then that is good.... cause it means that my biased sample is like someone else's biased sample...yay.

 

[Cyph]

Originally posted by iambored
and that's why you posted the longest post in this thread?

I didn't think a thesis is quite equivalent to an assignment from first or second year. I'd personally want to do a proper job on it.

Originally posted by Ragerunner
You'd find that yes it may be a bit biased, and the results may not be that accurate, but is accurate enough to make a lot of judgements. If you studied Psychology you'd find in unusual situations, psychologists may be able to predict your next action based on this research. e.g. the door in the face technique.

lol... inaccurate data leads to inaccurate predictions, no matter how accurate the method(s) used are.

 

[Ragerunner]

Originally posted by Cyph
lol... inaccurate data leads to inaccurate predictions, no matter how accurate the method(s) used are.

That's assuming EVERYTHING is inaccurate. There is of course some accuracy in these survey's / questionaires.

Because Psychology is based on cognitive processes and not everyone possess these exact processes there of course will be inaccuracies. When survey'd in large, you'd find it's actually pretty accurate.