Skip to main content
. Author manuscript; available in PMC: 2014 Sep 1.
Published in final edited form as: Psychol Methods. 2013 Jul 8;18(3):285–300. doi: 10.1037/a0033266

Table 4.

Brief Derivations of the Variances and Covariances of Facet-Representative and Domain-Representative Parcels

Facet-representative Parcel Domain-representative Parcel
Indicators N1 = T + S + e1
N2 = T + S + e2
N3 = T + S + e3
N1 = T + N + e1
V1 = T + V + e4
S7 = T + S + e7
Parcel
(13)I1+I2+I3
(13)I1+I4+I7
Variance of the parcel
var(13[3T+3S+e1+e2+e3])
var(13[3T+S+V+N+e1+e4+e7])
Restated as a covariance
19cov(3T+3S+e1+e2+e3,3T+3S+e1+e2+e3)
19cov(3T+S+V+N+e1+e4+e7,3T+S+V+N+e1+e4+e7)
Expanded with zero terms removed
19[cov(3T,3T)+cov(3S,3S)+cov(e1,e1)+cov(e2,e2)+cov(e3,e3)]
19[cov(3T,3T)+cov(S,S)+cov(V,V)+cov(N,N)+cov(e1,e1)+cov(e4,e4)+cov(e7,e7)]
Restated as variances
19[9var(T)+9var(S)+var(e1)+var(e2)+var(e3)]var(T)+var(S)+19[var(e1)+var(e2)+var(e3)]
19[9var(T)+var(S)+var(V)+var(N)+var(e1)+var(e4)+var(e7)]var(T)+19[var(S)+var(V)+var(N)+var(e1)+var(e4)+var(e7)]

Covariance between two parcels
cov(13[3T+3S+e1+e2+e3],13[3T+3V+e4+e5+e6])
cov(13[3T+S+V+N+e1+e2+e3],13[3T+S+V+N+e4+e5+e6])
Expanded with zero terms removed
19cov(3T,3T)
19[cov(3T,3T)+cov(S,S)+cov(V,V)+cov(N,N)]
Restated as variances var(T)
var(T)+19[var(S)+var(V)+var(N)]

Note. Figure 3 depicts the item-level and parcel-level models that describes facet-representative parceling. Figure 4 depicts the item-level and parcel-level models that describe domain-representative parceling.