As a professor of statistics at the University of Maryland, College Park, I was dismayed to read William F. Buckley Jr.'s column {op-ed, Feb. 5} in which he used the statistical concept of significance and tolerance level to justify U.S. involvement in World War II, Korea and Vietnam.

This concept was introduced by the founding fathers of the statistical sciences for testing scientific hypotheses and for carrying out business decisions and has no applicability to the problems of "cost efficiency" of human life vs. "preservation of freedom and sovereignty."

Using the 1990 figure of the U.S. population of 250 million, Buckley concluded that our losses in the above-mentioned three wars -- spanning the years 1941-1975 -- of 500,000 people, i.e., one-fifth of one percent of the population, are apparently within acceptance limits. The question is what is the threshold beyond which the losses in human life are deemed unacceptable? Is it a universal constant (say, one percent or 5 percent as in business-investment applications), or does it vary from country to country or generation to generation?

Shall we use Buckley's yardstick to assess the "worth" of British losses in the Falkland Islands war or that of the people of the Soviet Union during World War II? -- Samuel Kotz