Federal Aviation Administration inspectors at Delta Air Lines found significantly more problems in the recurrent training program for crews flying the Lockheed L1011 than with the other types of aircraft Delta flies, FAA records examined here show.

The L1011 jumbo jet is the type of plane that crashed in a thunderstorm at Dallas-Fort Worth International Airport Aug. 2, killing 133 people. Microburst wind shear, a deadly phenomenon of rapidly shifting wind speed and direction at low altitude, is the prime suspect as the cause of the accident, although the National Transportation Safety Board has not made its final determination.

Hollis L. Harris, senior vice president for operations at Delta, said, "There is no correlation in my mind between [the Dallas accident] and our training department."

As a result of the special FAA inspection last November, Delta has abandoned one type of pilot proficiency check. Under that program, called "training in lieu of," three of the four senior Delta captains checked in a flight simulator with an FAA inspector present were kept from returning to the cockpit until they completed additional testing.

A pilot cannot technically fail "training in lieu of," but is trained until proficient in the skill required. More formal proficiency checks, which can result in failure, require pilots to perform successfully a number of maneuvers that would occur in a routine flight, plus emergency procedures, such as what to do in case of an engine fire. The FAA normally requires proficiency checks twice a year but permits some exceptions. Many checks are conducted in a flight simulator; others on an airplane.

It is not uncommon for a senior airline pilot to fail a proficiency check, then be retrained and retested before he can return to duty. Nonetheless, FAA officials were concerned that "maybe a little bit of complacency . . . comfortableness" had crept into the Delta program, according to Jonathan Howe, chief of the FAA's southern region based here.

Howe said, "The relaxed atmosphere was such that some of the training benefit simply was lost . . . . My view is that the company's senior management was simply not aware of this comfortableness that had tended to take place . . . . We addressed it; the company addressed it, and they have taken some fairly significant corrective action."

The FAA inspection was a special, short-term effort by Howe's regional office specifically to look at Delta pilot training. Inspectors who normally check pilot proficiency at other airlines were used, a practice the FAA has tried in recent months to ensure its inspectors do not become too comfortable with an airline because they deal with it every day.

A more thorough inspection of Delta's total operations and maintenance program is scheduled as part of an FAA program for all airlines.

Concern has been growing that overworked FAA inspectors, forced to spend too many hours reviewing credentials of new airlines spawned by deregulation, have been unable to maintain the level of surveillance they would like on the established carriers.

That concern is part of the reason Transportation Secretary Elizabeth Hanford Dole announced plans to hire 500 additional inspectors over a three-year period despite federal budget constraints.

For the most part, the FAA inspectors gave high marks to Delta, which has long been regarded within the airline industry as top drawer in every respect, from training to maintenance to the age of its airplanes. For example, FAA inspectors witnessed proficiency checks for seven Boeing 727 captains, all of whom passed. The Dallas crash was the first one for the airline in more than a decade.

Delta officials said they do not know why most of the FAA's negative comments were directed at the L1011. "I would have to believe they were looking with a little closer scrutiny at the L1011 program. That would stand to reason," said Capt. H.C. Alger, Delta's assistant vice president for flight operations. "We did have one weakness in one of our instructors that we have since corrected, and that probably accounted for quite a few" of the negative comments.

Among the FAA comments:

*A captain following a radio navigational aid was cleared for approach to an airport by an air traffic controller. The pilot immediately descended to the lowest altitude in the approach plan. Such an error was one cause of the Trans World Airlines crash west of Dulles International Airport in December 1974, when the captain, having been cleared for approach, descended into the side of Mount Weather, killing all 92 people on board. Instead of immediately descending after receiving an approach clearance, pilots are required to stair-step down to the runway at specific altitudes for specific segments of the approach.

*Another captain was rated unsatisfactory because of an improper aborted landing in low visibility conditions. The captain failed to notify air traffic control that he was executing a "missed approach," as the procedure is called. He pulled the nose of the aircraft too high to climb, and airspeed fell well below the minimum established for the procedure before he recovered.

*One inspector who observed a formal proficiency check that a L1011 captain failed suggested that "wind shear avoidance criteria" be discussed rather than stating that "wind shear should be avoided."

Federal regulations do not require wind shear avoidance practice or training at proficiency checks, but many airlines include them as part of the check. Delta officials estimated that they put pilots through a simulated wind shear event in "probably three out of five" simulator flights.

"We don't think our training process per se was then or is now relaxed . . . ," Harris said. "You may get some people who are not operating as efficiently or aggressively as you want them to, and I think that was the area in which [the FAA] had some questions for us . . . . "

Alger said, "I will concede that there were some cases where maybe too much teaching was being done as opposed to checking, but it didn't sacrifice or compromise safe- ty."

Harris and Alger said that Delta began an internal training review after the Dallas crash. The accident, they said, pointed out the need for improved wind shear detection equipment at major airports, a program the FAA is seeking to get under way.

"I tell our people that I'm convinced that 99 percent of the aviation community would have flown in the same weather conditions and 99.9 percent would have met the same fate," Alger said.