Data:
FILE = 
longitudinal_dyadic wide2.csv;

VARIABLE:
NAMES = id dyadid person idp lengthymean index
deps audits cudits sats trusts commits csts cons
micros isms psss vics sods 
ismp microp pssp vicp sodp satp trustp commitp
cstp conp depp auditp cuditp
depsl auditsl cuditsl 
satsl trustsl commitsl cstsl consl
satpl trustpl commitpl 
cstpl conpl deppl auditpl cuditpl
noncis pover29;

Missing = ALL(-999);

UseVariables are 
csts cons    
sats trusts commits;

cluster= dyadid;

within are ;
between =  ;

ANALYSIS:
ESTIMATOR = MLR;
Type = twolevel ;
MODEL:
%WITHIN%
NRIac BY csts cons;
[NRIac@0];


RQac BY sats trusts commits;
[RQac@0];


RQac with NRIac ;

NRIac(wpv1);
RQac(wpv2);

%BETWEEN%
NRIacb BY csts cons;
[NRIacb@0];
NRIacb(bpv1);

RQacb BY sats trusts commits;
[RQacb@0];
RQacb(bpv2);

RQacb with NRIacb ;


MODEL CONSTRAINT:
new(NRIicc RQicc);
NRIicc = bpv1/(bpv1+wpv1); !ICC for NRI
RQicc = bpv2/(bpv2+wpv2); !ICC for RQ



Plot:
Type = plot3;
OUTPUT: svalues residual STAND SAMPstat CINT Modindices(2) tech1 tech2 tech3;