PHOTOGRAMMETRIC ENGINEERING
tionships between physical and social char-
acteristics of the city. By this means, certain
physical-structural-spatial data categories
could be developed as objective criteria for
photographic interpretation of urban demo-
graphic and social data. The ratios and per-
centages of structural type combinations re-
ferred to above constituted the first explora-
tion of this concept. The Rochester findings
offered preliminary evidence of the practical
utility of aerial photography in this new role.
The technique involved classification of the
Rochester subareas on a nine-point socio-
economic status scale derived from U. S.
Census data. Photographic data on combina-
tions of building types were correlated with
the constructed scale criteria. Clear gradient
patterns of increasing and decreasing per-
centages of structure types were found to
match the gradations of socio-economic sta-
tus areas. Predominant at the lower end of
the scale were heavy proportions of all struc-
tural types except one- and two-family homes,
while at the upper end, the one outstanding
feature was the single-family structure. Fur-
ther investigations of this nature included ad-
ditional photo data categories and more pre-
cise techniques for specifying the complex
socio-physical inter-correlations.
These continuing studies produced detailed
ecological analyses of several cities in the
United States and also elaborated the theo-
retical framework and design for the ongoing
research. In addition, a method was devel-
oped for the mathematical expression of mul-
tiple inter-relationships characterizing urban
physical and social structure, the center of in-
terest in this second phase of the project.
One of the key studies in this series was a
statistical analysis. of ecological data for
Birmingham, Alabama. In this work, the
Guttman scalogram model was adapted to
the construction of a scale of residential de-
sirability based on housing types, density pat-
terns, land use characteristics and ecological
location. This physical structural scale was
then correlated with a similar Guttman scale
comprised only of social structural data cate-
gories. À high positive relationship was found
between the two scales, to the extent that the
physical-structural-spatial data comprising
the residential desirability scale, accounted
for seventy-eight per cent of the variation of
Birmingham's subareas on the socio-economic
status scale. In terms of methodology, this
study demonstrated that the Guttman scale
analysis technique, conventionally applied to
attitude data in social psychology, is also an
excellent model for defining complex rela-
tionships in the spatial patterning of intra-
urban social and physical characteristics.
In addition to this statistical analysis and
scaling of Birmingham data, there were simi-
lar investigations using census tract data for
six other U. S. cities. This work centered on
four items of physical-structural-spatial in-
formation which could be obtained by photo-
graphic interpretation: (1) the location of the
tract relative to three concentric circular
zones having a midpoint in the central busi-
ness district; (2) the description of the tract
in terms of land use characteristics; (3) the
prevalence of single-family homes in the
tract; and (4) the density of housing in aver-
age number of dwelling-units per block in the
tract. Statistical analyses of data for all these
cities revealed many consistent and signifi-
cant relationships between categories derived
from the four items and a variety of social
structural categories. These empirical findings
thus provided further evidence of the predic-
tive value of the photo interpretation infor-
mation.
The next step in the project was a contin-
uation of the above-mentioned scale analysis
methodology. In this case, the technique was
applied to data for the same sample of six
United States cities. Similar to the Birming-
ham results, the four physical structural
items, comprising a total of twelve data cate-
gories, were found to constitute a scale which
defined mathematically their joint relation-
ships with several aspects of the social topog-
raphy of these cities. The scalogram for Chat-
tanooga reflected the spatial pattern of racial
segregation in that city. In the case of Spo-
kane, Washington, the scale types helped to
portray intra-urban social areas by indicating
housing rental value distributions. In Austin,
Texas, and Bridgeport, Connecticut, the
physical structural scale correlated on the or-
der of .85 with the ecological patterning of
income class groupings. Generally speaking,
these results showed that the so-called resi-
dential desirability scale variable, combining
all the photo data categories, was a consid-
erably efficient predictor of the socio-eco-
nomic status ranks of urban subareas.
Among other related studies, special atten-
tion was given to technical problems regard-
ing the accuracy of the photographically col-
lected data. In this part of the work, it was
discovered that discrepancies in the photo-
graphic observations were distributed non-
randomly. This situation provides a basis for
constructing systematic correction factors
through knowledge of the nature, amount and
direction of the photo data errors.