5.4 Research Approach The next element is the reasoning of the research (Sutrisna 2009) which refers to the logic of
6.5 Social Feedback ‐ Data Capture and Processing Strategy
6.5.2 Data Structures Data Dictionary
The SOFI system is a generic model which needs to be configured into a given scenario as required. This approach is viable as most businesses have similar processes but all have specifics which make them unique. The first activity in setting up a SOFI analytic is to ensure the Spheres are names with representative labels for the scenario being modelled. For example; in the generic model the first sphere is named "customer". In a school this would be identified as a student and in a hospital, a patient.
Definitions for the Generic “Spheres of Influence” Objects
The SOFI system contains data structures that represent human input on key determinants of organisational or enterprise change, represented by data objects called influence objects or "Spheres of Influence". A first step in devising an appropriate organisational model is to develop a set of Spheres definitions that are appropriate to the enterprise, community or organisation. The labels and definitions for each Sphere are thus adapted to the use in the organisation from a lexicon of the end user. From the literature one specific set of spheres and lexicons that have been found to be useful in modelling the interaction among functions or departments in many organisations include eleven such determinants such as those shown in Table 6.5.
Table 6.5 ‐ Organisational Characteristics Influence Observation
Culture Standards, quality of service, quality standards, values
Leadership Direction, management
Marketing Marketing, communication, information, presentations, competition,
market capability, market share
Strategy Any activity planning. This includes references to specific strategies that
have been documented or implemented as well as strategies that have
been discussed but not yet written down
Finance Strategy, planning, advising, milestones, feasibility studies, review finance
indicators, measures, performance levels, reports, costs, penalties,
evaluation, risks, analysis, validation
Operations Operations, performance, methodologies, processes, implementation,
technology
Development Development, R&D, training, design,
Sales Negotiating, buy‐in, needs, expectations, account teams, pre‐sales support,
post‐sales support, RFPs, offerings
Structure Agreements, contracts, commitments, services, policies, procedures,
accountability, service requirements, infrastructure, data‐warehouse
Staffing consultancy, service delivery, human resources
Staff Material relevant to issues such as recruitment, hiring, removal, promotion,
workload, motivation and individual employee’s needs
Figure 6.11 ‐ Generic implementation of the Influence Map that is used for perceptive data visualisation.
SOFI System © Phillip Cousins, Diane Downs 1985‐2016 SOFI Executive Systems LLC. Applies to all maps,
matrices, diagrams, rules, functions, Logic, Infographs, Analytics and other elements related to the SOFI system.
Used with permission.
A graphical representation of one example of such a Spheres model is shown in Figure 6.11. This arrangement of eleven spheres on the Influence Map was derived from the analysis of frequency of examples that illustrate the Spheres that are most likely to have influence on each other appear as neighbours. This means that those parts of the community that are more likely to have a direct relationship are positioned closer to each other than those which typically do not. In a typical spheres model representation, a traffic signal system is used to
indicate survey responses. Thus, the surveys are preferably configured so that responses may be graphically coded as a sphere with a green, yellow or red visual indicator. The green indicator is typically used to indicate a positive response, yellow to indicate that the respondents feels this is an area that requires caution or further investigation and red to indicate a problem or urgent concern.
The spatial distribution of Spheres in this manner can also be used to reveal relationships among the eleven Spheres, classified according to Extended and Remote. The neighbourhoods represent relative distance between the Spheres. These spatial neighbourhood relationships are further defined using a Scoring Matrix for the overall Influence Map for example, a chart or "Scoring Matrix" can be developed that shows the neighbourhood definitions for each sphere in the illustrated in one such scoring matrix or palate shown in figure 6.12, the immediate neighbours are coded in green, extended neighbours are coded in yellow and remote neighbours are coded red.
Figure 6.12 – SOFI scoring matrix showing how the relationships between the influence objects are defined.
Virtually all survey responses in a large, complex organisation appear as Immediate and Extended Neighbours. In the display (as shown in Figure 6.12) only the first two Neighbourhood links are typically populated with data. The Remote Neighbours will not typically be able to be populated without a special enabling function; this avoids unnecessary complication of the data object models. The restriction to populating
Immediate and Extended Neighbourhoods also prevents visualisation problems using the Influence Maps and populates the model with data instances that are relevant to the interaction of the closest neighbours.
The neighbourhood property of the sphere object relates to various functions in the map analyser. Neighbourhood analysis provides feedback on the structure of a questionnaire and on the complexity of an individual's reasoning and other attributes. The scoring matrix analysis can also be reported upon. The colour of the cells in the scoring matrix which are used to represent the neighbourhood to which the relationship has been assigned in the examples shown in figure 6.12. For example, the Development/Staffing cell is coded "green" to indicate that these are immediate neighbours. However, the "Sales/Structure "cell is coded yellow, to indicate an extended neighbour relationship. The "Sales/Operations" cell is coded red, to indicate a remote neighbour relationship. Also, note that the absence of a number in the cell indicates that the designer of the survey has provided no associated survey question for this cell. The result generates up to the possible measurement points. However, in an eleven Sphere model, it is typically not the case that all possible combinations of extended neighbours are represented or needed in the scoring matrix. Further properties of the spatial array are revealed in a neighbourhood matrix, shown in Figure 6.13.
Figure 6.13 ‐ Extract of a neighbourhood matrix illustrating the relationships between influence objects and the
arrangement of the influence map array.
The matrix is used to examine clusters of influence objects along the axis from top to bottom of the influence map in Figure 6.13 “Strategic" activities tend to cluster toward the top of the map, operational in the middle and tactical toward the bottom. Using this approach to modelling survey responses, robust measurement procedures can be used to get reliable data to analyse soft issues like customer satisfaction and awareness. Tools and procedures can also be specified to collect hard data such as financial metrics. The data collection procedures include design specifications for survey instrumentation, data collection and configuration of questions relating to each influence object.
In Figure 6.12 and Figure 6.13 the relationship between the adjacencies and the questionnaire is discussed. The questionnaire needs to be similarly analysed and prepared to be relevant and meaningful to the community who are going to be asked to use it. The use of "yes", "no", "don’t‐know" type answers does make this much easier but can position the researcher with a little context as to why an answer was given and what could be done to improve the situation. This is improved when the context is further enhanced through
the use of free text notes from the respondent. This is a particularly important issue if we are going to make sensible use of the tool as part of the overall Test Bench methodology. Consideration also should be given to the number of questions. The 11*11 nature of the SOFI structure does drive the user to a potentially large number of questions, so they need to appear relevant and concise. This is of particular importance with the reciprocal questions which are very easy to slip into the mistake of asking a question that is too similar to the direct one. This not only leads to a failure in establishing the required information, but also gives the user a feeling that he may be wasting his time, which may influence the quality of their work. The identification of SOFI Worlds is an important design consideration. The Worlds allow the data to be analysed from the point of view of a single community group. Clearly as a tool to understand perception and its implications it is important to see the data from each of these points of view.