Skip to main content
. Author manuscript; available in PMC: 2013 Jul 16.
Published in final edited form as: Neuroimage. 2012 Mar 30;61(4):1277–1286. doi: 10.1016/j.neuroimage.2012.03.068

Table 2.

Accuracy of Psychophysiological Interaction Estimates: 3 Conditions, Fixation

Beta Weights for Simulated Data Models
PPI
A		 PPI
B		 PPI
C		 A B C see
d		 con
stan
t		 sPPI sPPI+t
asks		 sPPIpl
us		 gPPI2 gPPI3
a a-1 0.00 a a-1 0.00 0.25 100 0.441
(0.500) 		0.487
(0.500) 		0.500
(0.500)		 1.000
(1.000)		 1.000
(1.000)
a a-1 0.5 a a-1 0.5 0.25 100 0.259
(0.500) 		0.297
(0.500) 		0.313
(0.500) 		0.625
(1.000) 		1.000
(1.000)
a a-1 a-0.5 a a-1 a-0.5 0.25 100 0.663
(0.500) 		0.719
(0.500) 		0.729
(0.500) 		1.457
(1.000) 		1.000
(1.000)
0.5 −0.5 0 a a-1 0 0.25 100 0.448
(0.500) 		0.500
(0.500)		 0.500
(0.500)		 1.000
(1.000)		 1.000
(1.000)
0.5 −0.5 0.5 a a-1 0.5 0.25 100 0.267
(0.500) 		0.310
(0.500) 		0.313
(0.500) 		0.625
(1.000) 		1.000
(1.000)
0.5 −0.5 0 a a-1 a-0.5 0.25 100 0.455
(0.500) 		0.500
(0.500)		 0.500
(0.500)		 1.000
(1.000)		 1.000
(1.000)
a a-1 0 0.5 −0.5 0 0.25 100 0.493
(0.500) 		0.487
(0.500) 		0.500
(0.500)		 1.000
(1.000)		 1.000
(1.000)
a a-1 0.5 0.5 −0.5 0.5 0.25 100 0.311
(0.500) 		0.297
(0.500) 		0.313
(0.500) 		0.625
(1.000) 		1.000
(1.000)
a a-1 a-0.5 0.5 −0.5 0 0.25 100 0.708
(0.500) 		0.719
(0.500) 		0.729
(0.500) 		1.457
(1.000) 		1.000
(1.000)
a a 0 a a 0 0.25 100 −0.110
(-0.110)		 −0.110
(-0.110)		 −0.110
(-0.110)		 −0.220
(−0.220)		 −0.220
(−0.220)
a a 0.5 a a 0.5 0.25 100 −0.292
(−0.110) 		−0.300
(−0.110) 		−0.297
(−0.110) 		−0.595
(−0.220) 		−0.220
(−0.220)
0.5 −0.5 0 a a 0 0.25 100 0.500
(0.500)		 0.500
(0.500)		 0.500
(0.500)		 1.000
(1.000)		 1.000
(1.000)
0.5 −0.5 0.5 a a 0.5 0.25 100 0.318
(0.500) 		0.310
(0.500) 		0.313
(0.500) 		0.625
(1.000) 		1.000
(1.000)
a a 0 0.5 −0.5 0 0.25 100 −0.110
(−0.110)		 −0.110
(−0.110)		 −0.110
(−0.110)		 −0.22 (−
0.220)		 −0.22 (−
0.220)
a a 0.5 0.5 −0.5 0.5 0.25 100 −0.292
(−0.110) 		−0.300
(−0.110) 		−0.297
(−0.110) 		−0.595
(−0.220) 		−0.22 (−
0.220)
a a b 0.5 −0.5 0 0.25 100 −0.083
(−0.110) 		−0.080
(−0.110) 		−0.081
(−0.110) 		−0.162
(−0.220) 		−0.22 (−
0.220)
a a b 0.5 −0.5 0.5 0.25 100 −0.088
(−0.110) 		−0.080
(−0.110) 		−0.081
(−0.110) 		−0.162
(−0.220) 		−0.22 (−
0.220)

Beta weights were multipled by a design matrix with the following regressors: PPI for condition A, PPI for condition B, PPI for condition C, HRF task regressor for condition A, HRF task regressor for condition B, HRF task regressor for condition C, seed voxel timecourse, and a constant. All datasets used 100 different values for a. Values represent the mean estimated interaction weight of A-B (the mean expected value of the interaction of A-B). Bold values indicate models where the model did not capture the true interaction. a and b are random numbers generated from the randn function in MATLAB®, which generates random numbers with a mean of 0 and a variance of 1.