ABSTRACT

Until now, psychology has had its greatest successes with problems that lend themselves to experimentation, quantitative surveys, or computational modelling, where there are numerous equivalent cases, with precise and complete data, and ample time for analysis. But there are other kinds of problems which are common and very important, yet notoriously intractable by these means, problems in the real world, where orderly situations unravel over time and have to be repaired. These situations are too intricately structured for standard experiments; too poorly understood for survey methods; too rich for the stochastic mathematics of sequences; and so on. They are “chain reactions”, with no standard initial conditions, or clearly separable causes and effects. Such problems call for a radically new approach, which starts with the kind of information that is available in practice (often incomplete, idiosyncratic, fragmented, partly qualitative, and rapidly changing) and works out increasingly effective ways of dealing with it, on its own terms. The methods may sometimes have to work with single cases and in real time. Better tools are essential to extract the relevant patterns from data, to extrapolate them into the future, and to steer them away from bad outcomes. This has been the aim of developing the “SERIAL” (SEquential Real-time In-depth case AnaLysis) approach to research. If we could understand how the present generates the future, we should have the key to conflicts, accidents, the breakdown of relationships, and crises of various kinds.