Over the last half century, worldwide, data has been increasingly collected on attitudes, preferences, choice intentions and behaviour. And certainly during the past 30 years or so of my experience as a policy analyst and researcher in the fields of marketing, public policy and corporate development, there has been an increasing recognition that efficient success requires compatibility between personal preferences and corporate or community objectives.
There is continuing strong interest in collecting data which reveals public or target group desires and preferences. Yet data such as that on share of choice, despite its arithmetic simplicity, emerges with substantial mystery from the complex interplay of the option variations and personal evaluations. For instance:
Despite a host of sophisticated techniques of varying practical or explanatory efficacy, much scope remains for:
Options Analysis supports such efforts. It focuses on share of choice, appeal status, and share stability and changeability of choices generated through the differential appeal of options to individuals.
Key to Options Analysis methods is the separation of:
This separation enables the analysis of share of choice by way of the generic parameter components of each option's appeal distribution within any selection set of options. These reveal the character of option appeals contributing to preference and the types and amounts of changes required to change shares.
A corollary to this separation is the ability to focus independently on the appeal of each option and its modification with a clear understanding of how it needs to be changed to alter its standing relative to all other options in a selection set.
In terms of economy of application and ease of implementation, Options Analysis is just a step beyond a well executed political poll. But, it enables a quantum leap of insight in the analysis of complex data critical to a fuller understanding of the relations between competing options.
Of particular interest is the ability of MaceTech Options Analysis to provide insights and even enable prediction of shifts in share results which may be described as "dramatic", "catastrophic" or "chaotic" or otherwise exhibiting the non-linear behaviour typical of social systems.
Some of the larger and apparently inexplicable jumps we see from time to time in public opinion polling and attitude surveys will be capable of explanation with Options Analysis. Contrary to popular interpretation, such jumps are likely to reflect modest but critically poised shifts in appeal across a significant proportion of the population. Such measurements of appeal may play an important informational role in an environment dominated by a mindset which requires dramatic effects to have dramatic causes. The lessons of complexity theory with its non-linear effects are relevant here. This is just one significant area where Options Analysis is capable of making a practical contribution.
The creation of Options Analysis was stimulated by a combination of long-standing personal interest in the measurement and application of attitude and preference data and an intuition of how a new approach to the analysis of choice data for groups or populations might be developed.
The seed for Options Analysis development lay in one elderly text and an even older paper. J. P. Guilford's “Psychometric methods” (1954), which is still referred to in some contemporary psychometric courses, makes reference (pp. 256-59) to the earlier work of L. L. Thurstone, “Prediction of choices” in Psychometrica, 1945, 4, pp 237-253. Thurstone provided examples of the counter-intuitive implications of differing standard deviations of appeal distributions that were otherwise identical. He found major differences in share could be the result. Guilford noted the potential usefulness of exploring the implications of a full set of appeal parameters on share of choice.
Both Thurstone and Guilford referred to the potential significance of such research:
“… many problems of merchandising and of politics, in particular, may be illuminated by such fundamental principles of choices…” (Guilford, p259)
The absence of computer capabilities at the time, the nature of the subject matter, and the then current preoccupations of psychometricians with test development and factor analysis, all counted against early action on the suggestion.
It is perhaps a tribute to the counter-intuitive nature of the seed sown by Thurstone and fostered by Guilford, that there has been little sign of subsequent development. Yet the basic concepts are not obscure and are sufficiently simple and obvious to be generally compelling. Indeed, the key appeal/choice relationship may be validated from study to study or accepted as a working assumption.
It has taken over two years of methodological, technical and programming development for Options Analysis to be brought to the stage of availability for others. While appreciation of various patterns of aggregate appeal-choice relationships provides an understanding of some puzzling share results, Options Analysis, on the whole, remains a tool for analysis rather than an answer in its own right.
There are many avenues for future development and already there is a very long list of possible improvements and extensions to the MaceTech Options Analysis program. But future development needs also to be informed by the diverse experience, interests and preferences of a wide range of research users.
So, if you wish to add your experiences or suggestions to make MaceTech Options Analysis more user friendly or more penetrating in its analysis, please contribute.
Winston Gardner
Founder, Corporate R & D