A defendant charged with burglary, larceny or assault in the D.C. Superior Court here who negotiates a guilty plea with the prosecutor usually ends up no better off than if he had gone to trial and been convicted, according to a new study of the city's court system.

That finding, received with some skepticism by prosecutors meeting in Indianapolis, appears to contradict the conventional view that defendants who plea bargain get shorter sentences than those who opt for a trial and are found guilty by a judge or a jury.

The study also found, however, that in robbery cases, significant concessions in terms of sentences resulted when a defendant waived his or her right to a trial and pleaded guilty.

The analysis of plea bargaining is one of 17 studies on the criminal justice system in the District of Columbus that were conducted by the Washington-based Institute for Law and Social Research with a $1.5 million grant from the federal Law Enforcement Assistance Administration.

The findings were drawn from statistics compiled by the U.S. attorney's office in Washington through the Prosecutors Management Information System (PROMIS), a computer-based data collection system. The results of eight of the Washington studies were discussed yesterday in Indianapolis at a meeting of prosecutors and judges from across the country whose offices use the information system.

Based on an analysis of 1974 felony cases in D.C. Superior Court, the study found defendants who pleaded guilty in robbery cases were sentenced to probation in 43 percent of the cases, while those found guilty at trial received probation in only 24 percent of the cases.

Fourteen percent of robbery pleas resulted in prison sentences of three years or more, compared to 32 percent of robbery convictions at trial, according to the study.

In burglary, larceny and assault cases, however, the sentences were substantially the same, whether the defendant entered a guilty plea or went to trial and was convicted, the study found.

In one of its studies, the Institute for Law and Social Research analyzed decisions in D.C. Superior Court to try to determine what factors influence a judge when he determines whether a defendant should be held on a money bond.

Under District bail laws, money bonds are to be set only if there is a likelihood that the defendant will fail to appear in court when ordered to do so. Otherwise, release on personal recognizance or other forms of conditional release should be granted.

The institute said if found, however, that a judge's decision to set a money bond had almost no correlation to standards bearing on the likelihood that a defendant would fail to appear, or to the possibility that the defendant would be rearrested prior to trial.

For example, the institute said it found that drug users are not held on bond any more often than are other defendants, although suspected druglaw violators have the worst record of failure to appear and of rearrest.

The institute asserted that the number of persons in jail prior to trial could be reduced by 40 percent if judges made their bail decisions with more weight given to which defendants are most likely to appear again and not be arrested in the meantime.

In a study of sentences handed down in Superior Court in 1974, the institute said it was unable to explain the reasons behind 40 percent of the judges' sentencing decisions.

The institute attributed that "at least in part . . . to differences in judicial philosophy among . . . the judges." The institutes findings also indicated that a defendant's fate at time of sentencing can simply depend on the identity of the presiding judge.

U.S. Attorney Earl J. Silbert, the chief prosecutor in Washington, said that the study findings - particularly those in the area of bail decisions - are significant but need to be further refined before any conclusions can be made about the operation of the criminal justice system.

William A. Hamilton, president of the Institute for Law and Social Research, said that although most of the data in the Washington studies is drawn from 1974 statistics, researchers feel that the overall findings are reflective of current operations in the Superior Court.

Hamilton said that the complete results of the Washington study are expected to be released by LEAA later this year.