How do you choose the appropriate statistical method?

#1
Hey everyone,

I know this might be very basic but I can't get my head around the stats for my final thesis.

I have done a user study with 30 users. Each over the users completed 8 tasks using 4 different visualisation methods (meaning every user completed 24 tasks overall). The time that each user took to complete the task was recorded. My supervisor would like me to use advanced statistics to analyse my results but I genuinely have no idea how to start. How can I tell if I have to use parametric or non-parametric tests or how can I tell if my groups were paired ? I would like to compare how well the users performed each task comparing the different visualisation method and if possible I would also like to compare the overall results of the 4 visualisation methods.


Thank you so very much :)
 

Karabiner

TS Contributor
#2
You have 2 factors, "task" with 8 levels, and "method" with 4 levels.
These are repeated-measures factors, since each participant completed
all levels of each factor. Since "time" was measured on an interval scale,
a repeated-measures analysis of variance with 2 within-subject factors
seems appropriate. This will produce tests of whether "task type" has an
effect on Performance, whether "visualization" has an effect, and whether
there is an interaction (i.e. whether visualtization methods have
different effects in different tasks).

If completion time is extremly skewed, it might perhaps be useful to transform
completion time using logarithmic transformation,

With kind regards

Karabiner