4% Antelope Valley Crime Hike Tied to Methodology : Statistics: Had the traditional year-to-year comparison been used instead of looking at the prior quarter, the figure would have been 19%.

When Los Angeles County sheriff’s deputies in the Antelope Valley last week released their tally of the region’s latest statistics on reported crime, homicides were said to be down a whopping 45%.

The news ought to have been heartening, except for one thing: The Antelope Valley has had a record number of 32 slayings thus far this year, more than in all of 1991 and double the totals for each of 1989 and 1990.

So why the discrepancy?

The answer lies in an unusual crime reporting method instituted by the region’s new commanding officer at a time when crime has been rapidly rising in the Antelope Valley. It is a tallying method that, according to experts, tends to dramatically reduce reported increases in crime.

According to a commonly accepted way of compiling crime statistics, reported major crime was up 19% in the Antelope Valley for the first nine months of 1992. But according to the new method, crime was up only 4%. That was the figure released by deputies and widely reported in local media last week.

Deputies acknowledge that crime in the Antelope Valley has been on a serious upswing all year, due partly to demographic changes in the population and the effects of the recession.

But sheriff’s officials in the Antelope Valley say they are not trying to play down the crime numbers to make themselves look better or to appease the public. In fact, they say, their method presents a clearer picture of crime in the region and allows them to respond to crime trends more quickly.

Other major police agencies surveyed by The Times, however, do not rely on the same method as the Antelope Valley station.

*

Most police agencies compare crime statistics with comparable time periods from the year before--the first quarter of 1991 with the first quarter of 1992, for example. But Antelope Valley deputies compare each quarter with the previous quarter--fall with summer, summer with spring and so on.

“The more the crime rate is rising and looking bad for you, the more you want to compare to the prior quarter. The increase is always going to be a smaller number,” said James Beniger, a USC sociology professor who is an authority on how statistics can be manipulated.

Beniger said he agreed that the method most often used by police agencies--year-to-year comparison--gives the most realistic picture because similar, and longer, periods are being considered. Comparing one quarter with the next, because of the shorter time frames involved, tends to show a smaller rate of change.

How much of a difference do such comparisons make?

Using the traditional method, burglary was up 20% in the Antelope Valley for the first nine months of 1992, when compared with the same period in 1991. But using the Antelope Valley deputies’ summer-to-fall 1992 comparison, burglary was down 8%.

Likewise with the homicides. Antelope Valley deputies correctly, albeit narrowly, reported that the number had fallen from 11 in the second quarter of this year to six in the third quarter, a 45% decline.

But through 10 1/2 months of 1992 ending Friday, the Antelope Valley had 32 homicides, double the 15 from 1989, the 16 from 1990 and exceeding the tally of 31 for all of 1991. In comparison, the valley’s population increased only 30% to 286,000 between 1989 and 1992, deputies said.

Sgt. Bob Denham, a spokesman for the Antelope Valley sheriff’s station, said the area’s new commander, Capt. Tony Welch, ordered the new reporting method after he arrived in February. Previously, Antelope Valley deputies had used the conventional year-to-year comparisons.

Welch was unavailable for comment but relayed his responses through Denham.

Denham said Welch believes that comparing one quarter with the next provides “the most valid picture of what is transpiring” and that year-to-year comparisons are outdated. Looking at one quarter to the next gives Welch a more timely way to assess changes in the crime picture, Denham said.

While admitting that anomalies like the homicide numbers can occur, Denham said Antelope Valley deputies are not trying to mislead the public on crime. “There’s no attempt to paint a different picture on that. But on the other hand, there’s no effort to make the crime picture worse than it is,” Denham said.

However, other police departments agree with Beniger and use the year-to-year approach.

“We do it this way because it gives people in our community a better picture of what is going on. To compare two quarters together might not show an accurate picture,” Long Beach Police Lt. Linda Fierro said.

Police in Los Angeles, San Francisco, Long Beach and Santa Monica all compile and release monthly or quarterly crime statistics that correspond to comparable periods from the year before. Some of these departments also do year-to-date comparisons.

The Los Angeles County Sheriff’s Department and the Federal Bureau of Investigation only release annual crime reports comparing an entire year with the prior one. In addition, some sheriff’s stations, such as Santa Clarita and West Hollywood, prepare monthly crime reports using the traditional year-to-year method for comparison.

None of those agencies said they rely on the method employed in the Antelope Valley. “That would never fly in this town at all,” San Francisco Police Department statistician Steve Lutes said. “The main problem is there is seasonality with some crimes.”

That means that some crimes, such as homicide, generally tend to increase in the summer months when there is more activity than in the colder winter months, said Beniger, the USC professor. Thus, comparing one month or quarter to the next can paint a misleading picture, he said.

“You’ve got to take it and put it in some kind of context,” Lutes said of crime trends.

Denham argued that the Antelope Valley, unlike many regions, does not have a crime pattern that varies significantly with the seasons.

But the valley’s numbers do fluctuate widely using the station’s new system.

For instance, reported major crime increased 10% when comparing the last quarter of 1991 to the first of 1992. But under the next quarterly comparison, crime dropped 10%.

Denham said the station plans to continue using quarterly comparisons but will provide year-to-year figures upon request.

Times correspondent Blaine Halley contributed to this story.

Crime Statistics

The Antelope Valley sheriff’s station in recent months has switched its reporting of crime statistics to a method that tends to minimize large crime increases occurring in the valley.

Most law enforcement agencies use year-to-year comparisons, such as the first nine months of 1992 compared to the same period in 1991. Each of four law enforcement agencies randomly contacted by The Times use similar methods.

The Antelope Valley sheriff’s station also used that method until recent months. Then, amid sharp increases in reported crime and under the leadership of a new commander, deputies switched to comparing the latest three-month period with the prior three months. Experts say that can be misleading.

The Antelope Valley Picture

Using Common Antelope Valley Police Method Station Method (Jan.-Sept., 1992 vs. (July-Sept., 1992 vs. Jan.-Sept., 1991) April-June, 1992) CRIME CHANGE CHANGE Homicide Up 33% Down 45% Rape Down 6% Up 19% Robbery Up 47% Up 16% Assault Up 21% Up 8% Burglary Up 20% Down 8% Larceny Up 13% Up 3% Auto theft Up 25% Up 29% Arson Up 1% Down 50% TOTAL Up 19% Up 4% Population Up 5% Up 5%

Source: L.A. County Sheriff’s Department.