Ventana software support forum

Posted by Jonah Fogel <jfogel@utk.edu>
Since there seems to be some interest in analyzing causal
loop diagrams and cognitive mapping in general I thought
we should start a new thread...

Martin has mentioned a few papers that are very useful for
understanding similarity between individuals' cognitions. I
have been studying under Dr. Joan Rentsch (University of
Tennessee- I/O Psych Department). The main thrust of her
research has been generating a theory of shared cognition
(of which Martin's suggested readings are a part). Dr.
Rentsch articulates a theory known as Team Member Schema
Similarity (TMSS). To learn more about TMSS I'd suggest
reading:

J.R. Rentsch and R.J. Hall, Members of great teams think
alike: a model of team effectiveness and schema similarity
among team members. In: M.M. Beyerlein and D.A. Johnson,
Editors, /Advances in Interdisciplinary Studies of Work
Teams: 1. Series on Self-Managed Work Teams/, JAI Press,
Greenwich, CT (1994), pp. 223–262.

Rentsch,-Joan-R; Klimoski,-Richard-J,Why do 'great minds'
think alike?: Antecedents of team member schema agreement.
Journal-of-Organizational-Behavior. Vol 22(SpecIssue) Mar
2001, 107-120.

In regard to methodologies for the measurement of schema
(i.e. cognitive maps) I'd like to suggest: Mohammed,-Susan;
Klimoski,-Richard; Rentsch,-Joan-R, The measurement of team
mental models: We have no shared schema, Organizational- Research-Methods. Vol 3(2) Apr 2000, 123-165.

Team research literature (in general) and shared cognition
literature (specifically) are useful avenues for future research
because they are focused on linking cognition to team
effectiveness through improving communication and interactions
between individuals (i.e. team processes --trust, conflict,
information sharing, etc.). Systems Dynamics (both hard and
soft systems) require that individuals make explicit their
cognitive structures. If we can link these two fields via TMSS
(or some other shared cognition theory) we should be able to
develop a robust process for evaluating and predicting the
performance of SD projects. ...at least that's one notion
I'm putting forward in my work.

And lastly, in regard to the whole notion of shared cognition
I'd like to suggest two other articles that might be useful:

Cannon-Bowers,-Janis-A; Salas,-Eduardo; Converse,-Sharolyn,
Shared mental models in expert team decision making., In
Individual and group decision making: Current issues. (pp.
221-246). Castellan, N. John Jr. (Ed). (1993). Hillsdale,
NJ, England: Lawrence Erlbaum Associates, Inc. x, 315 pp.

Cannon-Bowers JA, Salas E, Converse SA. Cognitive psychology
and team training: shared mental models in complex systems.
Human Factors Bulletin 1990;33:1–4.

be well,

Jonah Fogel
PhD Candidate

University of Tennessee
Forestry, Wildlife, and Fisheries
274 Ellington Hall
Knoxville, TN, 37996
Posted by Jonah Fogel <jfogel@utk.edu>
posting date Sat, 15 Apr 2006 11:51:20 -0400

Posted by <rgd6@cornell.edu>
Jonah:

I am not familiar with the literature you mention, but as I read your post a
questions/points came to mind.

One component of successful team members is the openness of their mental models to adaptation and change in the face of new, or even contradictory, information. That is, it is not necessarily that their schema are similar to start, but that a team exercise with such individuals can build a shared
schema, or at least an understanding other people's schema.

Presumably this is why the various tools for explicitly describing mental models are important -- to help others understand, and correct(?), thought processes.

Posted by <rgd6@cornell.edu>
posting date Mon, 17 Apr 2006 08:51:24 +0700

Posted by ""Alan McLucas"" <a.mclucas@adfa.edu.au>
This is an interesting discussion and a challenging area of research. The following notes of caution are needed if a method of comparing cognitive maps is being sought:

1) Cognition belongs to an individual and how individuals view the same problem can be quite different. See Eden, C. and Ackermann, F., 1998, ' Analysing and Comparing Idiographic Causal Maps' in Managerial and Organizational Cognition: Theory, Methods and Research' Eden and Spender (eds), Sage 193 - 208.

2) Meaning and emotion are linked. See Flood, R., 1999, 'Rethinking the Fifth Discipline: Learning Within the Unknowable', Routledge, London: 110.

3) We humans are not totally rational - bounded rationality is a reality in our thinking. Simon, H.A., 1972, 'Theories of bounded rationality', in 'Models of Bounded Rationality', M.I.T. Press, Cambridge.

4) We use mental shortcuts (heuristics) which have become ingrained in our way of thinking. Gigerenzer, G., Todd, P. and ABC Research Group, 1999, 'Simple Heuristics that Make Us Smart', Oxford University Press. The application of heuristics to thinking about a particular problem may not be revealed through cognitive mapping.

5) We do not necessarily wish to have the ways we really think investigated (nor are we likely to be completely honest when asked), invoking defensive routines or offering our 'espoused theories' rather than 'theories-in-use'. See Argyris, C., 1994, 'On Organizational Learning', Blackwell, Cambridge, MA.

5) Knowledge-Power influences our thinking, particularly within an organizational context. See Flood (1999:116-122).

Comparing (individual) cognitive maps (held by individuals in the context of generalised problem situations any broader than that which Jonah describes) and (group) concept maps, is problematic. Here I differentiate between a cognitive map developed through some elicitation process, workshop or interviews with individuals, and maps created by groups of individuals working together on a specific problem. This differentiation acknowledges that cognition belongs to an individual and in a group setting a concept map is an artefact created after dialogue, re-framing and compromise amongst a group of individuals. See McLucas, A.C., 2001, 'Decision Making: Risk Management, Systems Thinking and Situation Awareness', Argos Press, Canberra, Australia.

Where the intent is to produce a consolidated map which brings together the ideas of several individuals, then this can only be done with the intimate involvement of those specific individuals (perhaps through a nominal group process).

I trust this is helpful to the general discussion.
Regards,
Alan

Dr Alan McLucas
School of Information Technology and Electrical Engineering, UNSW@ADFA, Australian Defence Force Academy, Northcott Drive, CAMBPELL ACT 2600 AUSTRALIA Posted by ""Alan McLucas"" <a.mclucas@adfa.edu.au> posting date Tue, 18 Apr 2006 09:34:55 +1000

Posted by Jonah Fogel <jfogel@utk.edu>
Richard,

You've hit on one of my favorite lessons from this literature.
Schema are the heuristic devices we use to interpret, store,
and infer meaning to information. For a group to 'learn by
doing' it must have a group norm of openness and flexibility.
We know that's the case from experience but here's the
psychometric to help detect it! In team literature this norm
is captured in part by the term 'psychological safety'. The
norm (a similar schema that members hold) itself can be
established, maintained, and updated through three methods: 1)
socialization, 2) communication, 3) training. Socialization
what facilitators might call 'modeling behavior' or exhibiting
the behavior of openness and flexibility. In an organizational
setting you could think of this as a cultural norm, ""that's just something we do here"". Communication is the actual act of
behaving open and flexible. And training is, as you say, could
be a (effectively facilitated) team building exercise.

One of the strengths of Team Schema Similarity (TMSS) is its
acknowledgment that for a team to effectively communicate and
coordinate each member must not only share similar content of schema
but also that they need to have accurate understanding of each others
schema. These two attributes are known as schema congruence and
schema accuracy. The thought goes something like this...if two
individuals hold similar schema content but are not aware that they
do (high congruence/ low accuracy) then they can't coordinate
effectively because they are unaware that they are in agreement. And
if two individuals have dissimilar content and they know it then they
are in a state of conflict (low congruence/ high accuracy). Its only
when congruence and accuracy are high that team members can interpret
each others intentions correctly, coordinate effectively, and
anticipate (adapt) to changing conditions easily.

Schema can also be broken down into task (task work schema) and team
(teamwork schema) related ones. Task schema deal with an individual's
understanding of how to use equipment or step through the parts of a
task. Team schema deal with the relationships in a team. They relate
to understanding one another and the 'rules' or norms for the group.
Using this distinction might shed light on why some SD projects are
more successful than others. If a team has not established its work
processes and role expectations before they try to tackle a task its
going to run into coordination issues. Teamwork schema could be
monitored for appropriate levels of similarity (or dissimilarity)
before embarking on a task (e.g. strategizing, consensus building,
negotiation, etc.) or their development as a result of model building.
Likewise task work schema are likely negotiated during model building.
The communication that is needed to describe a model is by definition
task related if it is designed to find policy levers for improving the
system in question. 'Hard' SD is goal oriented (i.e. focused on the
improvement of a process or task). I think one of the biggest strengths
that SD has to offer is its stated goals of integrating the need of the
team members to learn and the need to improve a process or move towards a
goal. Parsing these two themes out as they relate to team cognition may
add some clarity to the group model building literature. ...but I digress
from my speculation...



Thoughts?
be well,
Jonah
Posted by Jonah Fogel <jfogel@utk.edu>
posting date Mon, 17 Apr 2006 09:21:36 -0400

Posted by Jonah Fogel <jfogel@utk.edu>
Along the same lines that Alan is thinking (if I understand his post correctly).
Doyle, Trees, and Radzicki (1998) list a number of factors that need to be
considered during elicitation and representation of schema (SD authors commonly
refer to schema as 'mental maps').

If I understand Alan's post correctly he's referring to ""the problem of aggregation"".
Collecting many individual's maps and linking them together to form some
aggregate map is not the same as a group map generated through dialog.
Also see...
Mohammed, Su
san, Klimoski, Richard, Rentsch, Joan R. The Measurement of Team Mental Models:
We Have No Shared Schema Organizational Research Methods 2000 3: 123-165

For those interested in defining methods for eliciting and representing cognitive
maps I'd also refer you to:

Measuring Change in Mental Models of Dynamic Systems: An Exploratory Study
J. Doyle, S. Trees, M. Radzicki, June 1998.
available at http://www.wpi.edu/Academics/Depts/SSPS ... ch/Papers/

This page has a number of papers that I've found interesting.

be well,
Jonah
Posted by Jonah Fogel <jfogel@utk.edu>
posting date Tue, 18 Apr 2006 11:32:41 -0400

Posted by ""Alan McLucas"" <a.mclucas@adfa.edu.au>
Further to Jonah's reply,

I am also saying that combining (aggregating) maps without the direct involvement of the individuals whose cognition was mapped in the first instance (without engaging these individuals in a group process), is problematic.

By treating the separate maps as having concrete or material existence (that is to 'reify') we are assuming that the individuals concerned will not change their minds even in the face of additional information, new evidence or being witness to the actual processes of aggregating.

If cognitive maps are combined by applying a working assumption that it is reasonable to treat the separate maps as having concrete or material existence, the resulting map will be little more than an artefact of the thinking of the person who constructed the aggregated map.

Given the option to review the process through which maps were being combined or aggregated, the original individuals are very likely to challenge the process. At some stage they are likely to re-frame their thinking and, as a consequence, question the aggregated map's validity.

Regards,
Alan

Dr Alan McLucas
School of Information Technology and Electrical Engineering, UNSW@ADFA, Australian Defence Force Academy, Northcott Drive, CAMBPELL ACT 2600 AUSTRALIA Posted by ""Alan McLucas"" <a.mclucas@adfa.edu.au> posting date Thu, 20 Apr 2006 12:01:53 +1000

Posted by Jonah Fogel <jfogel@utk.edu>
Alan,

Thanks for that clarification. My personal preference is to not combine
or aggregate maps at all. But for some purposes it may be allowable... especially, if the researcher is not interested in cognitive mapping for
educational purposes (e.g. there is no need for members to learn from one
another). For example, a researcher might be interested in mapping during
a process modeling workshop. Here the objective is to aggregate the
collective knowledge of the group without respect to group function...its
more of a procedural or institutional level improvement process. Maybe? or
Maybe not?

As I said my preference is to not aggregate maps for the reason you mentioned.
We miss opportunities for individual level learning and (in the context of a
project) we miss out on opportunities to build social capital.

be well,
Jonah
Posted by Jonah Fogel <jfogel@utk.edu>
posting date Sat, 22 Apr 2006 16:22:11 -0400