Directors Defend SPH Study, Criticize Fellow

The leaders of a federally sponsored air pollution study at the School of Public Health yesterday said a former research fellow exaggerated the importance of minor errors in their project.

The reason Armando Garsd found statistical errors in the project is simple, said his supervisor, Professor of Environmental Health John D. Spengler: Garsd was hired to check its accuracy.

"His job, in part, was to find errors," Spengler said. He added that all the faults Garsd cited were corrected and none of them were used in any published work.

Earlier this week, Garsd said Spengler dismissed him from a position at the Kennedy School's Energy and Environmental Policy Center (EEPC) partially because he questioned the data in the "Six City Study," a federally funded investigation of the effects of air pollution. Spengler is one of the project's principal investigators.

The mistakes Garsd found included a graph of sulfuric acid levels that reported some years' data at one-tenth of their correct magnitudes.


Garsd is suing the University for more than $100,000, charging that it fired him unjustly. But the complaint he filed in Massachusetts Superior Court in June 1987 does not mention the EPA study.

Instead, it charges that Spengler removed Garsd from his position as manager of an EEPC study of Spanish electric companies because he complained about an improper diversion of that project's funds.

Spengler, who spoke to The Crimson with the guidance of EEPC Executive Director Henry Lee, said that a University grievance committee had reviewed Garsd's case and upheld his dismissal.

In a letter dated September 5, 1986, Spengler cited "unauthorized entry into my office and reading of private, confidential correspondence" as his reasons for dismissing Garsd. Last week, Garsd acknowledged entering the office, but said he was looking for a paper Spengler had asked for, and that he had often used the room before.

But Spengler said yesterday that Garsd only used the office on a few occasions to make long-distance telephone calls.

"In no case does that justify going through someone's personal files, rooting around to find something, reading it, and acting on it," he said.

Spengler would not comment further on the lawsuit because it is still pending.

Concerning the EPA study, Spengler said Garsd's criticism of the whole project's integrity should not carry any weight.

"He's a statistician. He doesn't know one pollutant from another pollutant," Spengler said. "He was of a point of view that only a statistician wasqualified to analyze this data."

Other project officials also defended thequality of the study.

"I can assure you that if we had any doubtsabout each other's competence, it would have cometo the light before someone on the level ofArmando Garsd would have noticed it," saidProfessor of Environmental Science Frank E.Speizer, another of the project's principalinvestigators.

Speizer described the quality of Garsd's workon the study as "acceptable, but not particularlyinnovative or dramatic."

Another project researcher, Professor ofEnvironmental Science and Physiology Douglas W.Dockery, said Garsd had performed some"sophisticated statistical analysis," but that allthe errors he had found were "part of the normalpart of events.