APPENDIX 2.3: SCORING METHOD USED TO GENERATE SAMPLE OF FILMS
FOR ANALYSIS
The method is as follows: each mention of a film title (see
Appendix 2.1) is recorded as a point on a table for that year
(see Appendix 2.4). The total number of points for each film is
the film’s 'score'. A similar process is undertaken in respect
of star names (see Appendix 2.4). The top scoring films are then
listed and the score for the star(s) appearing in the film, if
any, is added to the film's score (see below, scoring procedure).
In practice the addition of the star score to the film score made
little difference in the top two films in the year, but was
helpful in distinguishing between the rest of the films recording
multiple points on the table for their year of release (see
below, Different methods applied to 1954).
Clearly some films and stars achieved duplicated points by this
method by virtue of citations recorded twice, such as
Kinematograph Weekly's Most popular and consistent star for which
a point was recorded against the film cited and against the star
in question. The consequence of this duplication, however, was
rather a magnification than a distortion (see below, 1954) of an
already existing difference, and the effect was potentially
available to all titles. In fact this method allowed
clarification of rather small differences and, I think, allowed
account of the important factor of the star's box office drawing
power - their significance in bringing audiences into a
particular film. A more serious problem concerned the scoring for
successful films which were released late in the year, and which
therefore tended to appear in adjacent years in different
listings. In the year I tested with different scoring methods,
1954, the inclusion of adjacent year's listings made a difference
only when US box office assessments were included, but not when
I restricted my calculations to UK data only. The inclusion of
critics' awards, such as the Academy, the British Academy, the
Golden Globe and so on also made a difference. However since the
method selected excludes both critics' awards and US box office
data it seemed reasonable to ignore listings of adjacent years.
One further problem is the changing context of my sources. The
marked decline in cinema audiences in the latter part of the
period is evidenced in the demise of the fan magazine
Picturegoer, in 1959. The machiavellian relationships between
British and American sources of production finance in the
declining industry of the later fifties may well be a factor in
the non-appearance of MPH listings specifically related to the
British box office between 1957-1962. However my sample groups
of films for analysis are drawn from years when there were at
least two different sources of data available: thus for 1945-46
I used Picturegoer, Kinematograph Weekly, and Motion Picture
Herald (star listings only); for 1955-56 Picturegoer,
Kinematograph Weekly and Motion Picture Herald (stars and films);
and for 1963-65 I used Kinematograph Weekly and Motion Picture
Herald (stars and films).
The top scoring films in all years are therefore, by definition,
those which appear in listings from all possible sources (though
not necessarily in all possible categories employed by the
109