Much of the violence associated with robbery is "recreational" and unnecessary to completion of the crime, according to a study released yesterday by the Law Enforcement Assistance Administration.

Victims who resist robbery are more likely to be injured, the study said, although many victims who are injured had not resisted the robber in any way.

The chance of being injured is higher when the robbery is carried out by a "non-professional" and the likelihood of injury dramatically increases when three or more robbers are involved in the incident, according to the study, conducted for LEAA by the Washington-based institute for Law and Social Research.

Research based on cases from 1973 to 1976 in D.C. Superior Court shows that robbers who caused injury tended to be less professional and more violence-prone than other robbers, the study said.

"It seems the people who injured their victims were primarily focusing their criminal activity on hurting people... rather than trying to gain money or property," said William A. Hamilton, the president of INSLAW.

The study found, for example, that unarmed robbers were more likely to be rearrested for a crime involving assault than were robbers who used a gun.

The study suggests that such "gratuitous violence" might be reduced "if robbers who injured their victims were singled out for relatively harsh treatment in the court."

As a result, the researchers recommend that prosecutors and judges -- who already give special attention to crimes involving guns -- also put additional emphasis on robbery cases in which the victim is injured.

The study, titled "Does the Weapon Matter?" focuses in particular on the nature of robbery and the possible consequences for victims depending on the type of weapon used.

For example, the study determines that robbers who carry lethal weapons, such as guns, tend to exercise some caution in their use because they fear the harsh penalties that come with a robbery-murder charge.

A robbery victim is better off -- in terms of possible injury -- if his assailant is armed with a gun rather than some other kind of weapon or no weapon at all, the study said. The report was authored for INSLAW by Philip J. Cook and Daniel Nagin, both of the Institute of Policy Studies at Duke University in North Carolina.

The mere sight of a gun usually is enough to intimidate a victim and make assault unnecessary, whereas other robbers need a show of force to carry out a theft, the study said.

The study also found, however, that victims are five times more likely to be killed when a gunman decides to attack compared to robbers armed with such weapons as knives, clubs or tire irons, or unarmed.

Most of the study research data was drawn from cases in D.C. Superior Court from 1973 to 1976. The study also included analysis of robbery-murder cases in Atlanta in 1976 and 1977 and on data included in a U.S. Census Bureau crimes survey of 26 large cities in the United States.

The study found that victims in unarmed robberies were physically attacked in 71 per cent of the cases, compared to 20 per cent for robberies in which a gun was used, 36 percent for robberies with a knife and 56 percent of the cases where other weapons were used.

While there are severe sanctions for killing someone in the course of a robbery, "the data shows there are no extra sanctions for injuring someone," Hamilton said in a telephone interview.

"Statistics show that the fact that an injury occurred wasn't increasing the likelihood of conviction, incarceration or the length of incarceration," Hamilton said

"What happened is that we may have limited our attention on the question of weapons and not gone beyond to look at another dimension -- which is that someone has been hurt," he said.

The study found that prosecutors in the District put a high priority on weapons possession cases and that the judges in the local court go along with the prosecutors' policy when they make their sentencing decisions.

According to the study, armed robbery defendants are more likely to be incarcerated than are unarmed robbers, and they are given prison sentences that ar twice as long as those for unarmed defendants.

Sixty percent of the persons prosecuted in the District for weapons possession had no prior record and were not rearrested during the time period studied, the researchers found. The study suggests, based on rearrest rates, that youthful weapons offenders and persons with prior arrests for crimes of violence should be the focus of efforts by both prosecutors and judges in terms of weapons offenses.