Calculating the median of a scale: statistically correct?

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

MaLin
Posts: 2
Joined: Sun Mar 02, 2014 8:31 am

Calculating the median of a scale: statistically correct?

Postby MaLin » Sun Mar 02, 2014 8:40 am

Hello everyone,

I want to calculate the median of a scale (consisting of multiple items) via SPSS in the following case:

I have a questionnaire in which the participants are supposed to rate 4 Items (v16-19) regarding their use of media according to their frequency (never – often/ SPSS: 0, 1, 2, 3). In order to do so, I first calculated the median for every participant (e.g. v20=Med.3(v16,v17,v18,v19)). After this, I calculated the median, Q1 and Q3 out of those 60 different medians.

Is this correct? Or do I calculate the median of a scale differently? E. g. by (somehow...) putting all the data I gathered from the 4 items into one variable and then calculate the median out of this data?

Unfortunately, I need the median and not the mean.

Help would be greatly appreciated.

Greetings
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Calculating the median of a scale: statistically correct

Postby GerineL » Mon Mar 03, 2014 8:48 am

It might depend on the scale you used / whats common with this scale.
But it would make sense to me to first compute a scale score (e.g., mean of the items * amount of items), and then take the median from that scalescore.
MaLin
Posts: 2
Joined: Sun Mar 02, 2014 8:31 am

Re: Calculating the median of a scale: statistically correct

Postby MaLin » Wed Mar 05, 2014 12:28 pm

Thanks, I'll keep that in mind for other analyses!

I can't use this procedure now though, because I want the median to be somewhere between 0-3 in order to compare it to another variable. If I calculate a scale score, I end up with something like 8.5.

I can't believe I'm having trouble with descriptive statistics…. meh.
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Calculating the median of a scale: statistically correct

Postby GerineL » Wed Mar 05, 2014 1:24 pm

can't you just standardize your scales?

Who is online

Users browsing this forum: No registered users and 1 guest

cron