variable, has somewhat limited predictive value. This fact, coupled 
with a theoretical position recognizing existence of plural structure- 
function interdependencies within the urban social system, suggests 
the application of multi-variate analytical techniques to obtain more 
complete definitions of socio-physical relationships. Such procedures, 
including index construction and scale analysis, might be expected to 
yield more precise formulations regarding the predictive value of the 
photographic data. 
Exploratory investigations of this nature attempted applications 
of the Guttman scalogram model? to the six regional cities. This tech- 
nique offers a convenient procedure for combining several qualitative 
variables which, if scalable, form a single continuum defining the exact 
position or rank of each object (or subarea) in relation to every other 
object in the sample. It was hypothesized that the photo data categories, 
considéred as physical structural attributes of residential neighborhoods, 
represent a scalable content area. Since, for American cities, the photo 
items, in general, reflect status location and ability to pay for pre- 
ferred residential sites, the hypothetical continuum could be called a 
scale of ‘tresidential desirability." This scale, comprising the several 
physical structural categories might then be expected to correlate 
significantly with a variety of social data categories. Because of the 
mathematical properties of the Guttman scale variable, this technique 
simplifies the problem of gaining maximum predictive power from the 
series of photo data categories, and thus provides for more precise 
definitions of linkages with the urban social structure. Specifically, 
the outstanding advantage of the scalogram method is that the scale 
variable, used as a predictor, has a value equal to the multiple cor- 
relation of all the combined photo data attributes with the particular 
external variable of interest. 
Analyses of the six cities revealed preliminary supporting evidence 
of the scalability of the physical structural categories. In all cases 
it was possible to construct fully acceptable or closely approximated 
scales of residential desirability. The scalogram model for Chattanooga, 
Tennessee reflected the pattern of racial segregation in that city, In 
the case of Spokane, Washington, the scale types helped to define the 
ecological pattern of the social areas of the city by portraying rental 
value distributions, In Austin, Bridgeport and Spokane, the residential 
desirability scale scores correlated on the order of .85 with median 
annual income, Generally speaking, the exploratory tests indicated the 
residential desirability scale variable to be a considerably efficient 
predictor of the socio-economic status ranks of urban subareas. 
  
9, For a complete treatment of the Guttman scale analysis method 
see Samuel A, Stouffer, et al, Measurement and Prediction, Princeton, 
N, Jo: Princeton University Press, 1950. 
11