Paper presentation 45.20, American Educational Research Association
New Orleans, April 1994
Objective
Basing her comments on the writings of Michael Lerner in Tikkun magazine, “Hillary Rodham Clinton speaks appealingly of a political morality based on the Golden Rule,” says Chicago Tribune columnist Clarence Page. Lerner and Clinton are correct in asserting that we need to rediscover and re-invigorate our spiritual values, though there is nothing new in this assertion, and Page is correct in his opinion that conservative columnists who say religion is spirituality, and that there is therefore nothing in need of re-invigoration, are wrong. Research on the spiritual dimension of disability, for instance, shows that the quality of spiritual experience has little, if anything, to do with religious church attendance, bible reading, prayer, or the taking of sacraments (Fisher & Pugliese, 1989).
The purpose of this paper is to propose a research program that would begin to prepare the ground in which a political morality based on the Golden Rule might be cultivated.
Theoretical Framework
Implementing a “political morality based on the Golden Rule” requires some way of knowing that what I do unto others is the same as what I would have done unto me. To know this, I need a measuring system that keeps things in proportion by showing what counts as the same thing for different people. A political morality based on the Golden Rule has got to have some way of identifying when a service or action done unto others is the same as the one done unto me. In short, application of the Golden Rule requires an empirical basis of comparison, a measuring system that sets up analogies between people’s values and what is valued. We must be able to say that my values are to one aspect of a situation what yours are to that or another aspect, and that proportions of this kind hold constant no matter which particular persons are addressed and no matter which aspects of the situation are involved.
Technique
Is it possible to measure what people value—politically, socially, economically, spiritually, and culturally—in a way that embodies the Golden Rule? If so, could such a measure be used for realizing the political morality Hillary Rodham Clinton has advocated? L. L. Thurstone presented methods for successfully revealing the necessary proportions in the 1920s; these were improved upon by the Danish mathematician Georg Rasch in the 1950s. Thurstone’s and Rasch’s ideas are researched and applied today by Benjamin D. Wright and J. Michael Linacre. These and other thinkers hold that measurement takes place only when application of the Golden Rule is possible. That is, measurement is achieved only if someone’s measure does not depend on who is in the group she is measured with, on the particular questions answered or not answered, on who made the measure, on the brand name of the instrument, or on where the measure took place.
Measurement of this high quality is called scale-free because its quantities do not vary according to the particular questions asked (as long as they pertain to the construct of interest); neither do they vary according to the structure or combination of the particular rating scheme(s) employed (rating scale, partial credit, correct/incorrect, true/false, present/absent, involvement of judges, paired comparisons, etc.), or the brand name of the instrument measuring. All of these requirements must hold if I am to treat a person as I would like to be treated, because if they do not hold, I do not know enough about her values or mine to say whether she’s receiving the treatment I’d prefer in the same circumstance.
In order to make the Golden Rule the basis of a political morality, we need to improve the quality of measurement in every sphere of our lives; after all, politics is more than just what politicians do, it is a basic part of community life. Even though the technology and methods for high quality measurement in education, sociology, and psychology have existed for decades, researchers have been indifferent to their use.
That indifference may be near an end. If people get serious about applying the Golden Rule, they are going to come up against a need for rigorous quantitative measurement. We need to let them know that the tools for the job are available.
Data sources
Miller’s Scale Battery of International Patterns and Norms (SBIPN) (Miller, 1968, 1970, 1973), described in Miller (1983, pp. 462-468), is an instrument that presents possibilities for investigating quantitative relations among value systems. The instrument is composed of 20 six-point rating scale items involving such cultural norms and patterns as social acceptance, family solidarity, trustfulness, moral code, honesty, reciprocity, class structure, etc. Each pair of rating scale points (1-2, 3-4, 5-6) is associated with a 15-30 word description; raters judge national values by assigning ratings, where 1 indicates the most acceptance, solidarity, trust, morality, etc., and 6 the least. Miller (1983, p. 462) reports test-retest correlations of .74 to .97 for the original 15 items on the survey as testing in the United States and Peru. Validity claims are based on the scale’s ability to distinguish between values of citizens of the United States and Peru, with supporting research comparing values in Argentina, Spain, England, and the United States.
The SBIPN could probably be improved in several ways. First, individual countries contain so many diverse ethnic groups and subcultures whose value systems are often in conflict that ratings should probably be made of them and not of the entire population. The geographical location of the ethnic group or subculture rated should also be tracked in order to study regional variations. Second, Miller contends that raters must have a college degree to be qualified as a SBIPN judge; the complexity of his rating procedure justifies this claim. In order to simplify the survey and broaden the base of qualified judges, the three groups of short phrases structuring each six-point rating scale should be used as individual items rated on a frequency continuum.
For instance, the following phrases appear in association with ratings of 1 and 2 under social acceptance:
high social acceptance. Social contacts open and nonrestrictive. Introductions not needed for social contacts. Short acquaintance provides entry into the home and social organizations.
Similar descriptions are associated with the 3-4 (medium social acceptance) and 5-6 (low social acceptance) rating pairs; only one rating from the series of six is assigned, so that a rating of 1 or 2 is assigned only if the judgment is of high social acceptance. Instead of asking the rater to assign one of two ratings to all six of these statements (breaking apart the two conjunctive phrases), and ignoring the 10-20 phrases associated with the other four rating scale points, each phrase presented on the six-point continuum should be rated separately for the frequency of the indicated pattern or norm. A four-point rating scale (Almost Always, Frequently, Sometimes, Rarely) should suffice.
Linacre’s (1993, p. 284) graphical presentation of Rasch-based Generalizability Theory indicates that reliability and separation statistics of .92 and 3.4, respectively, can be expected for a 20-item, six-point rating scale survey (Miller’s original format), assuming a measurement standard deviation of one logit. 360 items will be produced if each of the original 20 six-point items can be transformed into 18 four-point items (following the above example’s derivation of six items from one of the three blocks of one item’s descriptive phrases). If only 250 of these items work to support the measurement effort, Linacre’s graph shows that a reliability of .99 and separation of 10 might be obtained, again assuming a measurement standard deviation of one logit. Since not all of the survey’s items would probably be administered at once, these estimates are probably high. The increased number of items, however, would be advantageous for use as an item bank in a computer adapted administration of the survey.
Expected results
Miller’s applications of the SBIPN provide specific indications of what might be expected from the revised form of the survey. Family solidarity tends to be low, labor assimilated into the prevailing economic system, class consciousness devalued, and moral conduct secularly defined in the United States, in opposition to Colombia and Peru, where family solidarity is high, labor is antagonistic to the prevailing economic system, class structure is rigidly defined, and moral conduct is religiously defined. At the other extreme, civic participation, work and achievement, societal consensus, children’s independence, and democracy are highly valued in the United States, but considerably less so in Colombia and Peru.
Miller’s presentation of the survey results will be improved on in several ways. First, construct validity will be examined in terms of the data’s internal consistency (fit analysis) and the conceptual structure delineated by the items. Second, the definition of interval measurement continua for each ethnic group or subculture measured will facilitate quantitative and qualitative comparisons of each group’s self-image with its public image. Differences in group perception can be used for critical self-evaluation as well as information crucial for rectifying unjust projections of prejudice.
Scientific importance
One of the most important benefits of this survey could be the opportunity to show that, although different value systems vary in their standards of what counts as acceptable behaviors and attitudes, the procedures by which values are calibrated and people’s personal values are measured do not vary. That this should turn out to be the case will make it more difficult to justify and maintain hostile prejudices against others whose value systems differ from one’s own. If people who do not share my values cannot immediately be categorized as godless, heathens, infidels, pagans, unwashed, etc., ie, in the category of the non-classifiable, then I should be less prone to disregard, hate, or fear them, and more able to build a cohesive, healthy, and integrated community with them.
The cultural prejudice structuring this proposal is that increased understanding of others’ values is good; that this prejudice needs to be made explicit and evaluated for its effect on those who do not share it is of great importance. The possibility of pursuing a quantitative study of value systems may strike some as an area of research that could only be used to dominate and oppress those who do not have the power to defend themselves. This observation implies that one reason why more rigorous scientific measurement procedures have failed to take hold in the social studies may be because we have unspoken, but nonetheless justifiable, reservations concerning our capacity to employ high quality information responsibly. Knowledge is inherently dangerous, but a political morality based on the Golden Rule will require nothing less than taking another bite of the apple from the Tree of Knowledge.
References
Fisher, William P. & Karen Pugliese. 1989. Measuring the importance of pastoral care in rehabilitation. Archives of Physical Medicine and Rehabilitation, 70, A-22 [Abstract].
Linacre, J. Michael. 1993. Rasch-based generalizability theory. Rasch Measurement, 7: 283-284.
Miller, Delbert C. 1968. The measurement of international patterns and norms: A tool for comparative research. Southwestern Social Science Quarterly, 48: 531-547.
Miller, Delbert C. 1970. International Community Power Structures: Comparative Studies of Four World Cities. Bloomington: Indiana University Press.
Miller, Delbert C. 1972. Measuring cross national norms: Methodological problems in identifying patterns in Latin America and Anglo-Saxon Cultures. International Journal of Comparative Sociology, 13(3-4): 201-216.
Miller, Delbert C. 1983. Handbook of Research Design and Social Measurement. 4th ed. New York: Longman.
Livingcapitalmetrics's Blog 