Research across multiple systems: Probabilistic Population Estimation (PPE)
Today, social service administrators are examining client service utilization using cross system analysis, because often a client's needs require accessing governmentfunded services from multiple organizations. One technical problem that arises is that organizations do not share common unique identifiers from which to link one individual’s information together (i.e., system #1 uses Social Security Number (SSN) and system #2 uses Personal Identification Number (PIN)). Different methods have been employed to deal with the issue of working with information across data sets when there is no common unique identifier. Probabilistic Population Estimation (PPE), Caseload Segregation/Integration Ratio (C/SIR), and Probabilistic Population Matching (PPM) are methods used in our shop. This paper discusses the use of SAS® to perform the PPE & C/SIR methods of cross system analysis. These methods accurately identify the number of individuals who cross multiple systems without using a unique ID, while keeping the identity of an individual confidential. PPE is a statistical procedure for deriving unduplicated counts of the number of people represented in data sets that do not include unique person identifiers that do not share common personal identifiers (Banks & Pandiani, 2001).
Scholar Commons Citation
Haynes, Diane; Larsen, Rebecca; and Mehra, Shabnam, "Research across multiple systems: Probabilistic Population Estimation (PPE)" (2001). Mental Health Law & Policy Faculty Publications. 487.