PDA

View Full Version : Research for Psych Class Involving Linux



firefly2442
February 11th, 2007, 12:52 AM
Hello all. In my advanced research class for Psychology I need to design and implement an experiment. The experiment has a few requirements. First, it needs to be a factorial design so there must be two (or more) variables. Second, one of the variables has to be manipulated by me (so an independent variable). And third, it has to be at least slightly related to Psychology.

I was originally thinking of writing two different search engines and having people search on both of them and compare how well they did. However, today I was thinking about comparing operating system user interfaces and see how well people perform tasks and how tolerant they are at trying new things. For example, I could setup four laptops, each with a different OS on it. One could be Windows XP, then Vista, then Ubuntu Gnome, and lastly Ubuntu KDE. I could tell people to do something and see how long it takes them to do it in each system. I imagine the task under XP should be pretty easy but under Vista, Gnome, and KDE might take some exploration.

So my question is, what other variables or experiment setups involving different operating system user interfaces could I try? Thanks. :)

jeffc313
February 11th, 2007, 12:56 AM
Hello all. In my advanced research class for Psychology I need to design and implement an experiment. The experiment has a few requirements. First, it needs to be a factorial design so there must be two (or more) variables. Second, one of the variables has to be manipulated by me (so an independent variable). And third, it has to be at least slightly related to Psychology.

I was originally thinking of writing two different search engines and having people search on both of them and compare how well they did. However, today I was thinking about comparing operating system user interfaces and see how well people perform tasks and how tolerant they are at trying new things. For example, I could setup four laptops, each with a different OS on it. One could be Windows XP, then Vista, then Ubuntu Gnome, and lastly Ubuntu KDE. I could tell people to do something and see how long it takes them to do it in each system. I imagine the task under XP should be pretty easy but under Vista, Gnome, and KDE might take some exploration.

So my question is, what other variables or experiment setups involving different operating system user interfaces could I try? Thanks. :)
try system bar/tray stuff on top, on bottom, on left, or on right see how that affects. make the desktops cluttered for some and not for others.

Garyu
February 11th, 2007, 01:44 AM
We talk all the time about user friendliness...

Do the same setup on a couple of computers, only differ the wallpaper and colors used. One desktop with a hostile wallpaper, dark colors, and so on. One desktop with happy wallpaper and colors. One neutral desktop with B/W or grays and no particular motive on the wallpaper, just shades of gray everywhere. Of course, you can't let the same person try different systems so you need a larger sample of people to make it statistically safe, say 30 people or so for each desktop. Then don't ask them how easy the system was, ask them how they felt about using it. :)

Bragador
February 11th, 2007, 02:04 AM
Don't forget to set up a control group. You can't just compare all 4 OS for example. You need a control group to compare your results with.

I've just finished my bac in psychology in december and this science rocks !

Go psychology !

firefly2442
February 11th, 2007, 04:34 PM
Great suggestions, go Psychology! ;)

That's an interesting idea about the wallpaper colorings.

ButteBlues
February 11th, 2007, 05:42 PM
Since you need a control group...

I would compare the opinions of long-time Windows users to newer computer users regarding the use of the 4.

firefly2442
March 1st, 2007, 04:32 AM
Well, I think I decided on a concrete topic:

I'm going to compare Gnome Beryl and Gnome Metacity(default). The questions I will be looking at are whether or not Beryl is more effective from a usability standpoint and more effective in encouraging people to think about switching to Linux. I might also look at how their experience with Linux (either Gnome Beryl or just plain Gnome Metacity) compares to their (probably) primary OS being Windows XP or Mac OS X. Right now I am looking to gather resources and references to other research papers involving desktop operating system usability and appeal. Can anyone point me to papers or resources that might be beneficial? Thank you!

:)

DoctorMO
March 1st, 2007, 07:23 AM
I'd just set up a gui tool with two vertical sliders, one 'Happy' and the other 'Sad' and a button that says 'This is how I feel' since the two variables are conflicting, it should be interesting not where they put the sliders but what the difference between the visa versa of them is.

tjtansey
March 1st, 2007, 07:48 AM
How about two older, but functional computers with blank HDD:
One gets an XP SP1 install CD, the other an Ubuntu live CD and document the reaction of the user as they try to get each fully functional? Of course the XP install could have adverse side effects](*,)