A
two day course with Merrelyn Emery
Most statistics in social science belong within the reductionist, non-systematic
approach to empirical data and analysis. This draws its inspiration and
approach from the reductionist models used in the so-called hard sciences and
sticks with that despite accumulating evidence and approaches within these
sciences themselves that the phenomena under investigation are systemic. The
non systemic approach takes its base from Fisher’s statistics of regression etc
which were developed specifically for strictly controlled experiments. These
statistics have never been able to develop forms of systemic, or ‘cluster’
analysis with more than a handful of variables.
For those who recognize that the social sciences face even greater
challenges with the systemic, and open systemic,
nature of their subject matter, people in
environment, than those dealing with inanimate matter (and there has long
been such a stream within the social sciences) the inadequacy of the above
approach has been a matter of concern. Many have responded to aspects of the
challenge and there are indeed now a raft of proven approaches which taken
together provide a comprehensive base for reliable systemic data collection and
analysis. Not only is this approach comprehensive but it is also relatively
simple for anyone who understands the basics of scientific method, statistics
and number theory.
This course provides a comprehensive and hands on introduction to the
systemic approach. Every segment has examples attached which are worked through
during the course. “If you can’t do it, you don’t understand it.”
***This course assumes basic
statistical knowledge***
This is a definite
prerequisite. If you do not have a working knowledge of descriptive and
correlation based statistics, please do a basic course in statistics for the
social sciences before you enrol.
***Please bring a pencil, eraser,
calculator and
plenty of large square graph
paper (not < 0.5 cm2)***
Key reference materials and
notes are provided
Course Content
1. Questionnaire design -
contextualization and technical factors
The fundamentals of adequate Questionnaire design are simple and
constantly ignored. The result is frequently data which is either misleading or
uninterpretable in any rigorous sense. When the demand is for data applicable
to systemic multivariate analysis, the demands on data collection are radically
increased. Depending on the purposes of the study, the dimensions of the
Q’naire design from the point of view of content can undergo a further quantum
leap. For genuinely scientific studies based on OST, there is of course a
further requirement of understanding the conceptual framework itself.
2. Multivariate analysis -
causal path (Emery - McQuitty)
Building on McQuitty’s work with linkage analysis, Fred Emery developed
this rigorous form of hierarchical causal path analysis which allows the
researcher to see the patterns and total interrelationships within a body of
data. This method breaks totally with the reductionist approach and allows the
dynamics of any open system to be explored. The number of variables in any
given study is unlimited. It has been widely used in studies from market
research to trends in the extended social field and can be used in conjunction
with any variety of more conventional statistical tools.
After the initial practical learning using small examples from ‘job
satisfaction’ and ‘phone usage and marketing’ studies, participants will gain
further learning and experience by using this method in conjunction with the
following statistics.
3. From Qualitative to
Quantitative
(a) The Geisser Index is a variety of correlation which allows the
transformation of qualitative into quantitative data. Any data from qualitative
sources such as semi structured group interviews (otherwise known as ‘focus
groups’), Searches etc is appropriate.
Participants practice with data re ‘improving quality of work life’ and
then find the systemic interrelationships in the data using causal path
analysis.
(b) The Tau Correlation is the
most reliable and accurate form of rank correlation. It allows any data to be
ranked and then transformed into rank correlations. This means that data in
such forms as percentages or means can also be transformed and similarly
entered into a correlation matrix amenable to causal path analysis.
Participants practice with data from a study of ‘intercultural
perceptions’ and again subject the transformed data to causal path analysis. In
each example, different dimensions of this method will be learnt.
4. Building Scales and the
Master Matrix
For studies with a number of variables larger than say 50, there is a
need to break the analysis into two levels. This applies particularly when
there is reason, for example, to believe that the causal path for females will
be markedly different from that of males. This will involve the researcher in 3
separate analyses, of the total sample and the 2 genders. Other groupings of
the data may be required.
The first level analysis involves compiling scales from the systemic
data which become a master matrix. This master then becomes the base for the
second levels of analysis. Knowledge of the conceptual framework and a good
working general knowledge of classic social science, and open systems
theory, is highly desirable if not
essential in being able to generate a master matrix, the analysis of which
makes a contribution to the accumulation of social science knowledge. The
answers from this systemic analysis frequently cast a totally different light
on the subject than analyses based on separating out one or two variables at a
time.
What you will Learn
How to make meaning of data with a much higher probability of it being
close to reality.
Who Should Attend
Anybody who does empirical and/or action research who wants to know
what the overall data is telling them.
A
two day course with Merrelyn Emery
This two day module explores additional concepts to the introductory
course and takes others to much greater detail and levels of understanding. As
with the introductory course, all work is dealt with both theoretically and
practically. Participants will be theoretically briefed before working in
groups to answer questions, solve problems and plan pieces of work. Some
examples and questions are set to allow participants to think their way through
the concepts, others will involve participant’s own examples. Please be
prepared to discuss projects and programs in which you are currently working or
will have to work on in the near future.
*** Prerequisite: Previous
attendance at the introductory course***
If you have not attended the
introductory course, you will have to prove that you have gained the required
knowledge from other sources.
Key reference materials and
notes are provided
Participants will be
required to study the material beforehand
and be prepared to think conceptually about it.
Course Content
1. The Type IV Extended
Social Field of Directive Correlations
It’s origin, current nature and future.
2. Directive Correlations
Detailed exploration and their use in planning and problem solving. The
rest of the course builds on directive correlations as a major tool in open
systems work.
3. Consciousness and
knowings
Their nature and role in the open systems conceptualization of human
beings and their behaviour.
4. Maladaptions
Their theoretical origins from the open system and parameters of
decision making, their current nature and future.
5. The ABX Model
Explored in terms of directive correlation and its usefulness in
planning large scale projects
What You Will Learn
·
Greater detail of a wider range of open systems concepts
·
How to use conceptual tools for more precise planning and problem
solving
Who Should Attend
·
Those who have a theoretical interest in open systems
·
Those who want to be able to function as practitioners with a more
comprehensive and reliable range of easily applicable but quite precise tools
Last Updated: February 4, 2000