Impatiently biting his nails and chewing on a pen, Sen. Edward M. Kennedy (D-Mass.) held his first news conference on the tragedy at Chappaquiddick yesterday in an effort to rebut articles raising fresh questions about his veracity.

He opened it with a short prepared statement, turned it over to his experts, then took it away from them again to insist that he has always told the truth about the accident. At one point, he excused himself from the room abruptly, apparently to recheck some of the testimony in the 10 1/2-year-old case.

Obviously stung by the publication of two investigative articles within a 24-hour period, one in the Reader's Digest and the other in The Washington Star, Kennedy repeatedly made clear that he felt he had been treated unfairly. He called the back-to-back publication of the articles "pretty shoddy."

Kennedy faces the first test of his presidential campaign next week in Iowa, where polls show him trailing President Carter. At the news conference yesterday, held in one of the offices of the Senate Judiciary Committee, which he chairs, Kennedy expressed exasperation at the revived interest in Chappaquiddick "after 10 1/2 years," but made no other allusion to its potential political impact.

Turning to one of the issues confronting him, Kennedy said, for the first time, that he never gave any outward signs that he was in trouble while he was swimming back to Edgartown after the 1969 accident in which his woman passenger, Mary Jo Kopechne, was killed. Both the Star and the Digest articles challenged Kennedy's account of the swim, and they flatly disputed his sworn story of being swept north toward the sea and nearly drowning before reaching the shore.

"I know exactly what happened when I entered the water," Kennedy told reporters. "That's the way it was."

Doubts about the swim have been fueled for years by the testimony at a 1970 inquest of Kennedy's companions on the night of the accident, his cousin, Joseph Gargan, and his friend, Paul Markham. They said they watched Kennedy jump into the water, but added that they watched him until he was halfway to three quarters of the way across and saw no signs of his being in trouble.

Kennedy maintained yesterday that this was because he gave no such signs, even though he felt he was about to drown.

"There was no thrashing or outward expressions for help," Kennedy said. At the same time, he said, "I had a sense of motion and movement . . . of being swept [north] toward the [Edgartown] lighthouse. All of the trauma of the accident came back to me in a vivid way."

By contrast, Gargan, when asked at the inquest if he had been at all concerned about Kennedy's ability to make it across, testified:

"No, not at all. The senator can swim that five or six times both ways."

Kennedy saved his sharpest complaints yesterday for The Washington Star article, which, he charged, was "inaccurate . . . irresponsible, shoddy and incomplete." He also contended that it "distorted and misrepresented" statements of individuals who were cited as authorities for the study.

Kennedy had more difficulty in addressing the Reader's Digest article, although he insisted that it, too, was "an inaccurate story."

The Digest article cited two scientific studies commissioned by the magazine. The first, based on an IBM computer analysis of the accident, concluded that Kennedy had been driving much faster than he reported to police when his car plunged off the bridge on Chappaquiddick sometime around midnight on July 18, 1969.

The Digest's accident study said Kennedy was traveling at about 34 mph as he approached the bridge, saw it and slammed on the breaks sooner than he stated, and went off the bridge at a speed of 22 to 28 mph.

By contrast, Kennedy told police, and reiterated again yesterday, that he had been driving at about 20 mph as he drove down the dirt road toward the bridge -- a far less reckless approach. The judge at the inquest thought even 20 mph was "at least . . . negligent and possibly reckless," but Kennedy was never charged with anything except leaving the scene of an accident. He did not report it to police until the next morning.

Kennedy's campaign manager, Stephen Smith, said they have asked the Digest for the basic data underlying its accident study. Kennedy observed that the investigator for the Massachusetts Registry of Motor Vehicles, the late George Kennedy, also calculated his speed at 20 to 22 mph. The Digest, however, said Inspector Kennedy based his findings solely on the skid marks found on the wooden bridge, whereas its study showed the braking started sooner.

The Digest said it repeatedly asked Kennedy for an interview but was turned down. Kennedy said he had already given a number of interviews on Chappaquiddick, and complained that the magazine would not supply his office with its studies. He said he did not see "how I could have even responded to a study by a computer, outside of what I have testified to."

The other Digest study, based on currents and tides in Edgartown Harbor, concluded that Kennedy would have been pushed southward -- if he had been pushed at all -- when he entered the water. Unlike the Star study, which was based on a different hypothesis, Kennedy's experts -- admiralty lawyers Laurence J. Hoch and Timothy R. McHugh and MIT ocean engineering professor Jerome H. Milgram -- said they had no quarrel with Digest's tide and current observations.

Instead, Kennedy maintained, "The Digest's error is to misstate the time at which I entered the water." The Digest put this at sometime between 1:35 a.m. and 1:45 a.m. on July 19, 1969.

Kennedy insisted that he dove in around 1:30 a.m., when his experts said he would still have encountered a tug northward toward the Edgartown Lighthouse. But the testimony of Gargan at the inquest suggested that he could well have entered the water five to 15 minutes later.

Asked how he fixed the time in his own mind, since he was not wearing a watch that night, Kennedy said only that 1:30 a.m. was what "I testified" to at the inquest. Asked how he knew what to testify to, he said, "I made that judgment at the time." He indicated that he did not recall the basis for it.