Millisecond Forums

IAT D Score Calculation: Calculating Inclusive SD

https://forums.millisecond.com/Topic9076.aspx

By Sab Goh - 11/5/2012

Just wanted to check the SPSS Syntax for calculating the inclusive SD. I have two groups of pairs, which has resulted in four calculations for each numerator component (e.g., M1 M2 M3 M4 SD1 SD2 SD3 SD4 N1 N2 N3 N4). Group 1 (GRP1) are calculations with 1 and 2, and group 2 (GRP2) are calculations with 3 and 4.


When using the syntax to calculate the denominator for D-beip for each group, would I just replace only the calculations that change based on groups of pairs?


For example:


D_asis_numGRP2 = SQRT( ( ((N3-1) * SD3**2 + (N4-1) * SD4**2) + ((N3+N4) * ((M3-M4)**2) / 4) ) /
(N3 + N4 - 1) )


Thank you :)
By Dave - 11/5/2012

When using the syntax to calculate the denominator for D-beip for each group, would I just replace only the calculations that change based on groups of pairs?


As far as I can see based on the info you provided: Yes.
By Sab Goh - 11/6/2012

Thanks Dave :)
By newbieiat - 6/20/2014

Hi

Where does this 4 come from? In your D-score formula? 
Anyone that can help?



Best
By Dave - 6/21/2014

It's explained in the original post:

Sab Goh (11/5/2012)

Just wanted to check the SPSS Syntax for calculating the inclusive SD. I have two groups of pairs, which has resulted in four calculations for each numerator component (e.g., M1 M2 M3 M4 SD1 SD2 SD3 SD4 N1 N2 N3 N4). Group 1 (GRP1) are calculations with 1 and 2, and group 2 (GRP2) are calculations with 3 and 4.



It's not something that applies to D-score calculations generally. For that please refer to Greenwald et al. (2003) and this overview pertaining to Inquisit: https://www.millisecond.com/forums/Topic3444.aspx