Definition of marriage -relationship characterized by…
Person
Social roles (scripts)
Systems or organizational level
4. Note that the type of construct differs with the level
IV. The story of iron filings
Let a scatter of iron filings
represent individual scores on some measure
(e.g., marital satisfaction, MS);
If we use magnets of different
strengths we will see that the iron filings cluster
more or less around the magnetic bar when it is thrust into the filings;
each time
The magnets represent variables
(predictors), i.e., some predictors work well
(= correlate highly with MS) whereas others do not (e.g., nose length and
MS not related)
Our task is to find the variables that account for MS (=explain the variations in MS scores)
Where we look for these variables is what makes psychology different from other fields.
Notes on Constructs and iron filings (see K&B
Article #2)
I.Constructs are abstractions:
1. They point to interesting
"things"
2. Require some form of
support
3. How do we know (=what
is the evidence for) constructs?
4. Theories differ in
terms of the constructs they employ
II. Variables are indices of constructs
1. Equation: Satisfaction
= a + b + c+ ……..
2. Variables (a, b, c…)
as predictors
3. How well does a variable
"explain" variability of a construct
4.
Effect Size:
The more a variable explains (i.e., accounts for) variability the better;
Cyanide is effective…but some snake poisons are 4 times better (more
effective in explaining who will die from them!)
III. Causality
1. A causes B
A --> B
2. But maybe it
goes A --> [C] --> B
"C" is a mediating variable; the effect of A on B depends upon the
level of C;
as we have more or less "C" we will observe changes in B for the same
amount of A.
3. Pathways are
used to describe directions among variables; either direct, or indirect
4. The K&B model
describes pathways of variables based on the literature relating
to marital quality and marital
stability: The model asks -- What variables,
in what
combinations, allow us to predict to MQ or MS? If you ask your grandmother
she will
say eating broccoli is good for marriage. If you ask a 90 old man he will
say smoking
cigars will lead to a good marriage. Can you do better than that in your
choice of variables?