# To Lois on Construct Synthesis

BillJanie@aol.com
Wed, 6 Mar 1996 20:07:23 -0500

--PART.BOUNDARY.0.8468.mail02.mail.aol.com.826160841
Content-ID: <0_8468_826160841@mail02.mail.aol.com.317695>
Content-type: text/plain

Attach File

--PART.BOUNDARY.0.8468.mail02.mail.aol.com.826160841
Content-ID: <0_8468_826160841@mail02.mail.aol.com.317696>
Content-type: text/plain;
name="TOLOIS"
Content-Transfer-Encoding: quoted-printable

Lois,
=0D
The use of intelligence in many senses makes =

your problem a little confusing. In a real grid
we might find such construct labels. For now,
let's refine the labels a little for clarity's sake, =

understanding that in an actual grid the labels =

would be irrelevant- as far as the math goes. If =

we were conducting a grid study we would not =

need to refine these labels but for illustative
purposes it will help.
=0D
Lets change "Intelligence vs Foolish" to =

"Wise vs Foolish".
=0D
change "Foolish vs Careful" to =

"Careless vs Careful".
=

change "Intelligent vs Incompetent" to =

"Competent vs Incompetent"
=

keep "Intelligent vs Stupid"
=0D
The causul model is"( leaving off the contrast poles =

for illustrative purposes)
=0D
Intelligent------------->
+ =3D Wise
Careful--------------->
=0D
Competent
=0D
Or =

Intelligent + Careful=3D Wise
=0D

with competent not entering the equation. =

=0D
Note that the configuration does not require that =

Competent be uncorrelated with Wise. It may or =

may not be. Both Wise and Competent may =

share a common source in a fourth variable; =

Educated. If they are so rooted in Education, =

Wise and Competent would be correlated =

dependent variables. They would be correlated =

but without any causal connection between the two. =

=0D
In grid form we might ask a person to rate or rank =

50 persons (elements) along the 5 constructs.
=0D
Grid Elements (N=3D50)
=0D
Mary Willie Joseph ........ Mike
Constructs
(k=3D5)
Intelligent 1 5 4 3
Competent 3 7 1 1
Careful 7 6 2 5
Wise 5 5 1 4
Educated 1 4 7 7
=0D
These numbers are just random and not meant at this =

time to simulate real formal causation. =

In the real world we would be interested in discovering =

if the client is implicity synthesizing any of the =

constructs from the other constructs. We could also =

concatenate many peoples grids, so long as they =

all rated using the same constructs. Their elements =

could differ. If we contrasted two groups of people, =

we might find the two groups synthesize the constructs =

in different ways. =

=0D
Of course we would not be limited to human elements.
We could use musical compositions, commercial =

products, works of art, feelings, or what ever.
=0D
Does this help set the stage?
=0D
Bill =

--PART.BOUNDARY.0.8468.mail02.mail.aol.com.826160841--

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%