Re: [libreoffice-design] Re: LO Writer UI Analysis

Hi Astron,

Am Mittwoch, 30. Januar 2013, 17:50:46 schrieben Sie:
> Hi Björn,
>
> > Sorry for stepping in here. The results are just what it says in 
the
> > post. Yes, we did an analysis of data that was not gathered for 
this
> > kind of analysis in the first place. As it cannot be used to 
proof any
> > hypothesis (no statistical study can anyhow - read Popper on this
> > topic) - the data cannot falsify the hypothesis that more detail 
is
> > worse than less detail - but it can falsify the hypothesis that 
there
> > is no difference between more and less detail icons in this 
particular
> > setting. This is a value.
>
> So, first of all, the hypothesis makes some sense to me, logically.
>
> However, the problem here is that the categorisation seems pretty
> random: "low-detail" icons have small text on them, have delicate
> lines, contain many elements, etc. You don't seem to have published
> how you categorised the icon, either (maybe I haven't looked hard
> enough). 

You haven't: We did an expert rating (that what I wrote - and to add 
some more flesh to it:) with two experts (Heiko and me) working 
independently and discussing were categorization differed.

> If you look at the comments below the post, I am clearly not
> the first to have noticed. This is, btw, not the only concern about
> the validity of the more-is-worse analysis.

Most people complaining did not read the article well. Unfortunately 
you find this behavior often in the net. But on the other hand, I do 
not want to censor even the less appropriate comments.

So yes, you are right - there are issues to the validity. Take it as 
I said before: it is an indicator for the mentioned hypothesis and 
perhaps someone gets encouraged to pick it up and validate the 
findings.

You know, it really makes me sad how initiatives like these are 
always bashed upon. During my studies I had a course where we 
examined the 10 most-cited studies in Psychology from a 
methodological viewpoint. None of them was even close to bulletproof. 
Nevertheless they are valid - mostly because people re-did what was 
done before using (slightly) different methodology.

Back to our study: all raw data is freely available. Take it, 
categorize the icons the way you think it is appropriate and start 
discussing the difference in results. If you are not satisfied, 
create a study on your own - the tool you can make this with is 
freely available - again discuss the results. Untill then: I do not 
only have an opinion - I have some data that indicates a direction. 
Please be a bit more moderate if you haven't. I am moderate too - 
e.g. I do not claim anyone to do anything based on the results 
(except perhaps trying to falsify the findings on a data base).

I really hope to get some support here - we need to go ahead based on 
research (as good as we can). You see in this and other threads, what 
happens when we all just talk about our own preferences...

All the best,
Björn


-- 
Dipl.-Psych. Björn Balazs
Business Management & Research
T +49 30 6098548-21 | M +49 179 4541949

User Prompt GmbH | Psychologic IT Expertise 
Grünberger Str. 49, 10245 Berlin | www.user-prompt.com 
HRB 142277 | AG Berlin Charlottenburg | Geschäftsführer Björn Balazs

-- 
Unsubscribe instructions: E-mail to design+help@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/design/
All messages sent to this list will be publicly archived and cannot be deleted