Item Parceling Strategies in SEM: Investigating the Subtle Effects of Unmodeled Secondary Constructs

Abstract

For theoretical and empirical reasons, researchers may combine item-level responses into aggregate item parcels to use as indicators in a structural equation modeling context. Yet the effects of specific parceling strategies on parameter estimation and model fit are not known. In Study 1, different parceling combinations meaningfully affected parameter estimates and fit indicators in two organizational data sets. Based on the concept of external consistency, the authors proposed that combining items that shared an unmodeled secondary influence into the same parcel (shared uniqueness strategy) would enhance the accuracy of parameter estimates. This proposal was supported in Study 2, using simulated data generated from a known model. When the unmodeled secondary influence was related to indicators of only one latent construct, the shared uniqueness parceling strategy resulted in more accurate parameter estimates. When indicators of both target latent constructs were contaminated, bias was present but appropriately signaled by worsened fit statistics.

The relationship between smoker and nonsmoker prototypes and smoking status among 14-year-old Norwegians

“…More specifically, in Paper I, items of each factor were summarised into indices. In Papers II and III, items representing each of the constructs were combined into summarised parcels (Hall, Snell, & Foust, 1999;Little, Cunningham, Shahar, & Widaman, 2002). An advantage of indices is that random error is to some degree parcelled out, leading to a more reliable measurement (Little, Cunningham, Shahar, & Widaman, 2002 Increasing the number of items can be a way of increasing alpha to an acceptable level in an instrument (Nunnally & Bernstein, 1994).…”

What is materialism? Testing two dominant perspectives on materialism in the marketing literature

“…There are different parceling techniques. For the purpose of the present study, random assignment was used (see Hall et al, 1999). The aforementioned models were contrasted to identify the most appropriate measurement model for the personality materialism construct.…”

Toward a Mediation Model of Employability Enhancement: A Study of Employee‐Supervisor Pairs in the Building Sector

“…The only exception is tbe latent variable learning value, for which we created two parcels of items as recommended by Hall, Snell, and Foust (1999). Bagozzi and Edwards (1998) referred to this as a Partial Disaggregation Model.…”