jamovi

Hi,

I am trying to calculate the familywise error rate of a group of data, and basically all I want is to be able to create a transform that allows me to use if abs($source) > x use 1, else use 0. However, I am unable to add an absolute value command here.

I have made a workaround where I've made a compute variable of each t-test that's literally just taking the absolute value of it (the function is just =ABS(t1_2) etc) so that I can use this new variable abs(t1_2) as the source for my transform variable sig1_2, but this seems inefficient as I should be able to just do that absolute value within the transform as you can on SPSS.

I really really hate SPSS, but my tutor is forcing us to use it this term, so I've been trying to use Jamovi and translate the instructions over myself. This is the only hitch I've hit, and also the only place so far where I've seen SPSS be more efficient than Jamovi, so I'm hoping it's just because I'm missing whatever button lets me put "if abs($source) > 2.179, use 1" as the recode condition like I've tried to illustrate in the attached screenshot.

Thanks so much for your help!

Caitlyn

PS: I realise this step is just for calculating the per-comparison error rate, I said familywise because I need this step to be able to then add them up for the FWER

Hey @Caitlyn,
why not so with Compute, take a look: Cheers,
Maurizio

Updated:
if you prefer to use TRANSFORM

you could also use

IF $source > 2.179 use 1
IF $source < 2.179 use 1
ELSE use 0

IF $source > 2.179 or $source < 2.179 use 1
ELSE use 0

cheers