Situation: The U.S. Air Force is preparing to expand its bomber force and is trying to decide on the most efficient locations for overseas bases.

Question: Where should the overseas bases be located for the most effective strike capability at the least cost?

Answer: In the United States. (You asked the wrong question.)

Reasons: Locating the bases in the United States will save $1 billion and will strengthen America's retaliatory capability, which in turn will increase the ability of the bomber force to deter an attack from the Soviet Union. Bombers can be maintained better and more cheaply in the United States and refueled in mid-air during missions.

The bomber study is the quintessential example of how the Rand Corporation tackles a problem. Never mind, Rand tells its resident brains, the question the client has asked. What is the problem? Has the client asked the right question?

If there is a "Rand approach," the bomber study typifies it. Gustave H. Shubert, Rand's senior vice president, says, "If there's a system, it's really based on the proposition that you have to be certain you're asking the right question. And the question asked may not be the right question."

Rand is one of the oldest, best known and most prestigious of the think tanks that cropped up after World War II. It brings together interdisciplinary teams of experts-- economists, engineers, social scientists. And although the term "think tank" conjures up an image of witty, smart-as-a-whip geniuses closeted in a room brainstorming, the process often is grueling, even boring, with mountains of technical detail, laborious gathering of data and rigorous questioning of assumptions, reasoning and conclusions.

The people who work at Rand are bright, sometimes brilliant, not always geniuses. Rand's Washington office, in the estimation of Paul Hill, director of operations, has about the same aggregate IQ level as a top Washington law firm, although the top pay is less: about $60,000 a year.

Rand started as Project Rand (for research and development) as part of the Douglas Aircraft Corporation after World War II when Gen. Henry H. (Hap) Arnold set it up to maintain the scientists and engineers assembled during the war to give the Air Force advice. Rand has kept a close relationship with the Air Force since both began, but now about 50 percent of its work is nonmilitary.

Rand has grown since the late 1940s, but it is far from the largest of think tanks. A decade ago its managers decided to grow no more. "There's no advantage in getting business for its own sake," explains Hill. Rand has about 800 workers (about 80 in its Washington office and the rest in Santa Monica); half are professionals and the other half administrators and support staff. The professionals have either PhDs or the equivalent in experience.

Rand is in the business of systems analysis -- not looking at a problem in isolation but attempting to understand the way the whole mechanism works. "I'll tell you what systems analysis is," Hill says, grabbing a piece of paper and sketching a crude airplane with large sections of its fuselage and tail shot out. "They used to have these airplanes and they came back and these parts were shot away. Now, before systems analysis, people would say, 'Ah, strengthen these parts (the ones shot away). Reinforce them.' After systems analysis, people said, 'Those are the ones that came back. Maybe we ought to strengthen the remaining parts.'"

In practice, the Rand approach uses strict application of common sense. First is the matter of the question. The senior person on a project holds a seminar and explains the problem and the planned approach. Other Rand professionals are encouraged--are expected--to comment and criticize. Shubert says the system is remarkably free from personality clashes. Rand professionals show a "willingness to accept criticism and ideas that I have not found at any university," he says.

The point of the seminar is to encourage the team leader to place his hypotheses where skeptical colleagues can have a look. "Now the place is internally competitive enough and self- critical enough that you can't get away with just a brute force shift of the question," says Hill. "If you don't have the evidence to suggest that the shift of the question is a better way to look at it than the old way, you'll be chopped up."

Along the way progress reports may be issued, allowing staffers to kibbitz and criticize. Near the end of the tunnel, a final draft is circulated inside Rand and to outside experts for more criticism.

Like most think tanks it owes its existence to the idea that administrators in the trenches can't be objective. Consider the city fathers of Cleveland, for example, who went to Rand not long ago asking how to reorient the city's commerce away from heavy industry and into the growing service sector. After studying Cleveland's strengths and weaknesses, Rand told the city fathers to forget about cashing in on service industries. Cleveland's strength would remain, Rand said, in manufacturing. Cleveland was told to look for new things to make.

Rand's recommendations often result from not much more than straightforward fact-gathering and analysis of data not available to its client. The Army went to Rand in the late 1970s wondering why its reserve ranks were dwindling, dropping from 638,000 soldiers in 1973 to 527,000 in 1978. The decline threatened the all- volunteer army, which depended on a strong reserve force. Rand was asked to determine what effect a bonus would have on getting weekend warriors to sign up again.

Rand found, first of all, that reenlistments in the reserves had dropped off because soldiers who had enlisted to avoid the draft were completing their terms and getting out. Once those enlistees had left the army -- for the most part by 1979 -- the waning of reserves had stopped.

Rand also found that reserve service was seen by some recruits as a way to supplement income, but by others as a way to spend some time with the boys, away from nagging wives and clinging children.

The Rand approach, Hill says, is an exercise in "redefining problems to make them more solvable." Rand has made a name for itself by standing problems on their heads, a process is calls the "paradigm shift." Frequently the outcome is proposals that challenge the conventional wisdom.

The process seems easier than it is, Hill cautions. "You have to build capital," he says. "You can't get an original idea just by demanding an original idea every day."