top of page

Multidimensional Scaling


240 Scaling Project of Problems Facing t

Page from questionnaire 'Scaling Project of Problems Facing the Nation' .

In the spring of 1971, in cooperation with psychologists, Myron Wish and Douglas Carroll, multidimensional scaling experts at Bell Laboratories, E.A.T. began work on multidimensional scaling projects using the INDSCAL program (Individual Differences Scaling). The scaling program is a way to use simple similarity judgment questions – that is judge how similar two words or entities are  in pairings from a long list in order to generate a map where these judgements fall and cluster in groups that can reveal people's basic , but often unacknowledged, opinions or assumptions, so that the researcher one can, among other things, identify the major or general nature of those opinions. E.A.T.'s first scaling project working with Carroll was "A Scaling Project of Problems Facing the Nation." Twenty-two national problems were formulated and incorporated into a questionnaire. Included were: consumer exploitation, deterioration of public education, inflation, misuses of technology, pollution of the environment, public and private greed, urban decay, and war. The questionnaire contained long lists of pairs of these problems – e.g. inflation – pollutions of the environment, or greed  -- urban decay.  


Those taking the survey were to determine the similarity between the two problems on a scale 1 to 10 It was first taken by twenty-five subjects, mainly E.A.T. staff and friends. Carroll processed the data from these similarity-judgement questionnaires and found that the results were such that the axes in a three-dimensional space could be interpreted as “local political - national political”; “technological - non-technological”; and “moral individual - large scale organizational.” The aim of this initial study was to correlate national problems with technical and scientific resources applicable to these areas in order to develop guidelines for allocating scientific and technical resources toward solving national problems. For this purpose, plans were made to correlate the results of the scaling on national problems with those of "Scaling: Science Questions," a study Bell Labs had already started.


Two other scaling projects were initiated by E.A.T. during Telex: Q&A in August 1971, with participants at the New York, Stockholm, and Tokyo telex stations. One was to compare their cultural attitudes toward ten goals of life using a questionnaire developed by Carroll, "Dissimilarities of Goals of Life." The goals included goods, power, money, services, information, and freedom. The studies revealed interesting cultural differences.


The other scaling project that was organized by Bob Whitman was to compare cultural attitudes in the different cultures toward common object words. Artists in the US, Japan and Sweden chose common object words and each assembled a list of seven common object words (nouns such as vase, mushroom, man, and water). Three questionnaires were formulated in the necessary form from the respective lists and distributed at the three stations. Subjects in each country filled out the questionnaires judging the relatedness of the pairs of words within each of the three sets of words. The INDSCAL model at Bell Labs was applied to the data. resulting in nine sets of responses. While the results showed no significant differences in perception of the relatedness of the object words in the three cultures, there were some small differences in the dimension along an axis interpreted as man – nature, as this dichotomy appeared to mean less to the Swedes than to the more city oriented participants in Japan and  the US, whereas another dimension, the male –female axis, seemed to have greater importance to the Swedes.  


E.A.T. later proposed to extend the project by scaling relatedness judgments on all twenty-one words combined in one questionnaire to be taken by persons in the three countries and then, as part of a cultural research program, to have participants in fourteen countries take the questionnaire.

In 1977 E.A.T. collaborated with Carroll and Tapas K. Sen who had developed a new methodology “impact scaling” to generate the aggregate judgements of experts for forecasting aspect of an organization relevant to planning, which had been applied successfully within AT&T, and they were interested to generate a case study outside this organization. Billy Klüver initiated a series of discussions with Dean George Schillinger at Polytechnic Institute of New York about applying these methods to a study in futures planning for engineering education, and organize  a meeting at which Carroll and Sen introduced impact scaling studies and their results at AT&T. consensus developed that the study should be focused on administrative and academic planning for the Polytechnic Institute of New York, with the idea that it could be applied to other academic institutions. E.A.T. developed a proposal  for administering such an impact scaling study by  Carroll and Sen to be submitted to the Institute.

bottom of page