NCJ Number
19417
Date Published
1973
Length
42 pages
Annotation
A DISCUSSION OF THE USE, LIMITATIONS, RANGE OF VARIANCE, AND OTHER ISSUES INVOLVED IN DECIDING WHETHER GROUPS OF CHRONOLOGICALLY ORGANIZED DATA (TIME SERIES) SHOULD BE USED FOR EVALUATING TEAM POLICING.
Abstract
AN EXAMPLE OF A TIME SERIES INCLUDED IN THIS PAPER IS THE NUMBER OF BURGLARIES REPORTED TO THE CHARLOTTE POLICE DEPARTMENT FOR EACH MONTH FROM 1963 TO 1973. SEVERAL QUESTIONS ARE DISCUSSED IN THIS PAPER. AMONG THESE ARE: 1) WHEN SHOULD TIME SERIES BE CONSIDERED AS A WAY OF EVALUATING TEAM POLICING, 2) HOW CAN SHORT-RANGE FORECASTS SERVE AS A STANDARD FOR EVALUATING TEAM POLICING, 3) WHAT PROBLEMS NEED TO BE CONSIDERED IN COMPARING ACTUAL DATA WITH SHORT-RANGE FORECASTS, 4) HOW CAN A LIKELY RANGE OF VARIANCE BE ESTABLISHED FOR A SHORT-RANGE FORECAST, AND 5) HOW CAN A DECISION BE MADE AS TO WHETHER THERE HAS BEEN REAL CHANGE IN THE TIME SERIES ATTRIBUTABLE TO TEAM POLICING. THE AUTHOR CONCLUDES THAT TIME SERIES ANALYSIS MAY HELP IN EVALUATING THE IMPACT OF TEAM POLICING, BUT THE RESULTS WILL NOT BE AS CLEAR-CUT AS THEY WOULD BE IF EXPERIMENTAL AND CONTROL GROUPS WERE USED. IT IS FURTHER STATED THAT WHEN USING A TIME SERIES ANALYSIS WITHOUT EXPERIMENTAL AND CONTROL GROUPS, ANY ATTRIBUTION OF CAUSE MUST BE BASED SOLELY UPON JUDGMENT. THE APPENDICES INCLUDE OFFENSE STATISTICS REPORTED TO THE CHARLOTTE-MECKLENBURG POLICE DEPARTMENT. (AUTHOR ABSTRACT MODIFIED).