The difference between formative and reflective constructs can be tricky. This document contains short notes to develop an understanding of the terms, the differences, and how to identify a formative/reflective construct.
The content is taken from a highly cited paper published in the MIS Quarterly.
1 of 5
Download to read offline
More Related Content
Reflective and formative constructs differences
1. Reflective, formative, and multidimensional constructs
A. Definitions
B. Diagrammatic representation
C. Mathematical representation
D. Decision rules to identify reflective/formative constructs
A
1. Reflective construct
When measures (also called indicators or items) are used to examine an underlying
construct that is unobservable (i.e., a latent variable), the measures can be referred to as
reflective indicators or effect indicators. The unobservable construct, which consists of
the reflective indicators and the error term for each indicator, is called a reflective
construct.
2. Formative construct
Indicators that determine a construct are called causal or formative indicators. Constructs
comprised of these causal indicators along with a disturbance term are called formative
constructs or composite variables.
3. Multidimensional constructs
Multidimensional constructs are constructs with more than one dimension, and each
dimension can be measured using either reflective or formative indicator
2. B
Structural models comprised of all reflective constructs are reflective models; however, if at least
one construct within the model is formative, the model is considered to be a formative model
5. Document created by Aun Ahmed (aun.iitrpr@gmail.com)
The content in this document is wholly taken from the following paper
Petter, S., Straub, D., & Rai, A. (2007). Specifying formative constructs in information systems
research. MIS quarterly, Vol. 31, No. 4, 623-656.
D