RESEARCH STATISTICS
Analysing
Formative Model
DR CHUA YAN PIAW (UM)
Prof Dr Chua Yan Piaw
University of Malaya
Formative model
The arrows of indicators pointing to the latent
variable.
The formative indicators have no internal
consistency, they do not have conceptual unity.
The indicators are independent one another, but
all of them measuring the same concept of a
variable.
1
Example 9:
A study is conducted to examine the
effect of CMC competence of students on
their attitude towards Facebook.
CMC competence has three formative
indicators – Knowledge, motivation, skill.
Attitude has three formative indicators –
Perceived usefulness, perceived ease of
use, usage rate.
Formative model analysis: PLS-SEM and
Bootstrapping
Using the Data Example 9, analyse the
validity and reliability of the indicators of the
two formative models, and the effects of
CMC competence on attitude towards FB.
1. Convergent validity:
• Redundancy analysis
2. Multicollinearity:
• VIF
2
1. Convergent validity
To ensure that the indicators are able to
measure the same concept of the variable.
- Redundancy analysis
Formative model:
Indicator: β > 0.10
Path coefficient: > 0.70
2. Multicollinearity
To ascertain that the indicators are independent one
another, by showing that they are not correlated
strongly.
VIF < 5.0
3
Result
The effect of CMC competence on Attitude to FB is
β = 0.777 (p < 0.05).
Report also the effect size and predictive power
(Q2predict) of the model.
7
RESEARCH STATISTICS
DR CHUA YAN PIAW (UM)