Content Analysis: the collective and the filmmakers
A consideration of the methodology of content analysis with a focus of the process of unitisation is undertaken before the actual research is reported. An analysis of attendance at meetings gives an indication of who was in the Exploding at any time and how active they were. The intensities in which filmmakers showed at Exploding are measured and could indicate a democratic approach to a pre-canonic formation. The proportion of women to men allows us to measure one aspect of accessibility and compare it to mainstream film production.
Content analysis can be defined as a research technique that makes inferences from the identification of objective characteristics within a text. These characteristics can most simply be measured in terms of quantities. Later researchers, such as Krippendorff (1980) put more emphasis on the contexts to which inferences are drawn.
Since the Sixties content analysis has not become as ubiquitous a research method as the questionnaire or focus group, which may be due to a lack of agreement on standard methodologies which has meant that results from different studies are not comparable. The main reasons which make content analysis useful are:
* The act of measurement is unobtrusive and is unlikely to act as a force that will influence the subject of study.
* As the method acts on documents from the past, which are unlikely to change, the cultural indicators derived in such a way can be checked and are likely to constitute reliable empirical data.
* Data collected in this systematic way is persuasive and difficult to refute as it is easily replicable.
On the other hand the necessary data reduction can destroy nuance and complexity. The imposition of standard measures and the production of averages flattens difference and obscures the value of the singular. The idiosyncratic may be argued to be particularly important in creativity and innovation. And although the process of measurement may be unobtrusive, the presentation and interpretation of results can still be made to appear favourable to a preconceived thesis.
Both these criticisms are valid only for poor content analysis that is not self-critical of the limits of its own unitisation and not supplemented with more qualitative methods. The key question is whether the units counted give a valid measure of the concept that the researcher associates with them.
Equal counting is a practical simplification that content analysts believe works well in most circumstances, but reality is probably much more complicated. (Webber, 1991, p72)
Although such standardisation is necessary it is important to note the implications of variations in value. In my analysis of the films shown, each film is valued as one unit in spite of a wide variation in the types of films and the impact that they have. Whilst the reduction of a film work to a unit value of one is obviously a crude mechanism it seems to produce results which are useful if the limitations of unitisation are kept in view.
Inferences derived from such totaling must not only be internally valid but must also be tested within the wider context in which 'the text' exists. It is often these wider inferences that are more significant but also more vulnerable to normative reductions. This sort of analysis must entail a qualitative vigilance to ensure that the objectivity of the method does not obliterate the value or diversity of its subject.
If the above limitations are borne in mind content analysis could be particularly useful in the study of art collectives in that it is suited to the analysis of a large amount of data that is inevitably thrown up by a large group or 'movement'.
Three aspects of Exploding Cinema have been investigated using this methodology:
1. Attendance at meetings (1992 - 1997) has been tracked to give a dynamic picture of the involvement of various people within the collective. It will be argued that this gives the best objective indication of the general influence of particular individuals in the collective, creating a model that can be used as the basis of discussion in an area that is often emotionally charged. This is based on recording the attendance at meetings as recorded in the minute books.
2. All the works shown at Exploding Cinema from 1992 to 1999 have already been recorded in a catalogue of film-makers and the films they showed, which has been extracted from the surviving show programmes. From this list data will be extracted on those filmmakers who have shown most works and whose work has been given repeat screenings. This shows those who are most energetic in contributing to the main programme. I also argue that this correlates with the process of canon formation. This tentative canon formation is then tested and expanded to include selected makers with one or two films on the list by reference to three further sources (The Vacuum compilation video and two European tour selections from 1994 and 1995). The implications of such a construction are then briefly discussed.
3. The proportion of men to women both showing and taking part in the collective is also examined and discussed in relation to the open access policy of the group. It would be unrealistic to expect an open access policy to compensate for the inequalities within society but at least it provides an objective measure against which these issues can be examined, discussed and challenged.
In each case I will first enter into a more detailed discussion of the units of measurement, or unitisation and the most likely sources of error.
The following quantitative assessments give us some insight into the influence of any individual, canon formation and gender inequality and access. All of these are key issues for any collective of cultural producers. Conclusions have been checked against the oral memories of the present collective and others.
A draft of this content analysis of collective membership was circulated to the Exploding Cinema collective. The feedback produced a number of minor corrections, which I will discuss later, but no major challenges to the main personnel.