IN THE BACK corner of the basement of the Science Center Library, Rand Corporation reports fill 130 feet of shelf space. At one end of the shelves are technical reports from the early '50s, when the new corporation was essentially a research and development arm of the Air Force. At the other end are sociological studies done for the Department of Health, Education and Welfare in the spring of this year. Between 1950 and now, Rand researchers seem to have studied subjects as diverse as any in the world, as well as a few beyond this world at NASA's request.

Many of the studies seem trivial and absurd. After encountering a study like "Resource Allocation Models for the Arkansas State Police," one is hardly tempted to wade through the remaining hundreds of thousands of pages on the shelves. Neither does the prospect of learning the arcane contents of "An Annotated Bibliography of Dynamic Cloud Modeling" set the pulse racing. Yet there are many striking and subtly disconcerting papers tacked away in the stacks. For example, how did Rand researchers get the extensive bibliographical data they included in a profile done in the late '60s of an elite six-man Vietcong guerilla unit fighting in the Mekong Delta? The report doesn't say, but it does outline each man's background, psyche and personal beliefs. The extent of the researchers' knowledge is striking. Their ability to gather so much information on even presumably hostile people is disturbing.

Rand's experts have written on basic physics, engineering, statistics, economics, sociology and world cultures and religions. They know how peasants in northeastern Thailand make their living--and how the economy there can be developed to fit a more Western mold. They have examined the role of the Catholic church in Latin American society. Their reports include countless mathematical models for efficient allocation of resources. Rand physicists have even stepped back to a cosmic perspective and scanned the whole globe to find "The Exact Solution for the Propagation of Electromagnetic Pulses over a Highly Conducting Spherical Earth."

The only type of writing the corporation doesn't seem to produce is creative writing, but it nonetheless devises classification schemes for creative writing from other sources--one Rand whiz worked out a method for computer indexing of translated mainland Chinese texts.

SINCE THE '50s the corporation has shifted its emphasis away from the hard sciences toward economic and social topics, but always its researchers look for systems and equations to classify their data. A 1973 study of how students decide what college they will attend looks at the students as "investors" who try to estimate the yearly earnings they can expect after attending different colleges. The study is reminiscent of the approach to education of Stephen S.J. Hall, onetime hotel chain executive and former vice president for administration at Harvard. Hall touted Harvard as "one of the finest universities in the world as far as the product they turn out."

Military ventures get a similar cost-benefit analysis. A Rand researcher in 1973 suggested that efficiency could be determined by defining military output in terms of "the capability to destroy 1000 tanks in a 90-day Central European war scenario, or the capability to deliver a given amount of bomb tonnage in Southeast Asia." His formulas did not include significant variables representing such factors as a ravaged European countryside or a decimated village in Vietnam.

Even religion comes under economic analysis. The study of the Catholic church in Latin America includes charts showing "religious personnel per 100,000 population and per capita GNP" and "Catholic students per capita GNP." The author suggest policy makers apply "organizational analysis" to predict how the clergy will behave and what the church's status will be.

A June 1976 study of '72 high school graduates strikes very close to home. It examines how school curricula affect later career choice. And it suggests the results be considered an "agenda for policy research," implying that someone should start deciding for students what careers they should choose.

RAND REPORTS do influence Washington policymakers. In April of this year a report suggested that physician shortages in rural areas could be partly relieved by linking federal aid to medical students with the requirement that the students later serve in rural areas. Congress passed such a resolution this fall.

So the question arises: how much should be known about the structure of society and how much should policymakers be allowed to decide from that knowledge?

The Rand Corporation resembles an eager biology student at his first dissection. It can carefully lay bare the anatomy of its subject but cannot comprehend its motivation. It can describe the details, but misses out on the whole animal.