The study assesses the effects of these problems on the time-series analysis of UCR data and the advantage of alternative procedures for improving these effects. It also describes the methodology for assessing measurement problems in the UCR data. The study prepares a Suspect Data Key for all agencies with population coverage of 100,000 or more in the database that will flag suspect data points, indicate the sources of likely error, and provide adjusted replacement values. The data analysis proceeded in three stages. The first stage constructed work files for alternative subsets of reporting agencies according to specified criteria. The second stage performed a set of bivariate analyses in which measurement models of each construct and its indicators were estimated. The third stage involved multivariate estimation of the model, revision of the model, and investigation of the effects of alternative strategies to ameliorate the error upon the model estimates. 4 figures, 10 tables, 38 references.