|
Versions
Environmetric Studies
|
Construction of the
SETA
Design Goals and Assumptions
Oddly enough, very few
guidelines have been available to anyone interested in creating this
type of environmental assessment instrument. Consequently, a "right
tool for the right job" approach to designing and testing the SETA has
been used. A few overriding design goals were established at the
beginning.
- The presentation format and layout of the
assessment should resemble the primary measurement of psychological
type, the Myers-Briggs Type Indicator™ or MBTI™ tool.
- A forced-choice format can reveal
responses that are consistent with the underlying Jungian constructs
(Hicks, 1984).
- A mixture of phrase questions and word
pairs respects the different approaches to responding to items that
Myers and Briggs observed in people (Myers et al, 1998).
The SETA also needs to
be "generic" enough to be used in all the different contexts and
settings as the MBTI instrument. Hence, general terms (environment,
people, activities) are used in items. With the move to an online
version, however, the SETA can be tailored to the people and setting.
Item Development
SETA items were
generated from two sources (Craik, 1981). Behaviorally derived SETA
items were adapted from the validity studies and research projects
reported in the previous MBTI Manual (Myers & McCaulley, 1985), as well as from the growing
number of research studies related to the functioning of the different
psychological types. Theory-derived items resulted from taking an
interactional view of the Jungian hypothesis and the developing
taxonomy of environmental types (see Theory).
Using recognized
environmental subdomains or context-specific areas as guides to improve
content validity, items were reexamined across four subdomains that
were drawn from work by Moos (1979): the human aggregate, physical
elements and their arrangements, the organizational structure, and the
social climate.
Item and Scale Analysis
Item/scale analysis has
been accomplished in three ways.
- The first approach
relied on suggestions by Meir and Gati (1981) for analysis of
assessments that produce multidimensional profiles, including review of
non-response rate, scrutiny of the response distribution, and
examination of the correlations with both intended and non-intended
scales.
- The second procedure
has been exploratory factor analysis of SETA responses, using principle
components extraction, which helped reduce the number of items to 60
for Form B and to 76 on Form C. Form B results can be found in the SETA
Manual (Salter, 2000) and Salter (1995; 2002).
- Finally, a
subsequent confirmatory factor analysis was conducted on the overall
measurement model (Salter & Vandiver, 2002). That analysis
supported the existence of four interrelated latent constructs, as
measured by the 15 items on each of the SETA Form B's four scales.
|