Mixing items in a formative construct

Research topics can be discussed in this area.
Post Reply
sdevaraj
PLS Junior User
Posts: 1
Joined: Mon Oct 20, 2008 4:02 pm
Real name and title:

Mixing items in a formative construct

Post by sdevaraj »

When using formative constructs, should all the items be measured on the same scale or is it OK to mix item types?
For example, say Performance was measured as:
1. I am very satisfied with performance (measured on a 7 point Likert scale)
2.. The time I spend productively per day is ___ hours
3. The percent of products delivered on time is __percent

In this example, each item is measured on a different scale. Is it technically OK to use PLS to combine these as formative indicators of performance? Or should we use items that are measured on the same scale?
Thank you for your insights.
Best regards,

Sarv Devaraj
______________________________________________________________
Viola D. Hank Associate Professor of Mgmt
Mendoza College of Business
University of Notre Dame
Notre Dame, IN 46556, USA
http://www.nd.edu/~sdevaraj
User avatar
Diogenes
PLS Super-Expert
Posts: 899
Joined: Sat Oct 15, 2005 5:13 pm
Real name and title:
Location: São Paulo - BRAZIL
Contact:

Post by Diogenes »

Hi Sarv Devaraj,
it is ok, but remember that if some transformation could improve the linearity it is a good thing to try (thinking in the correlation's computation).

Best regards,
Bido
Post Reply