Construction and validation of an expert system prototype to determine the accounting treatment for business combinations
Abstract
The purpose of this research was to describe the construction and validation of an expert system prototype, PURPOOL, that determines the accounting treatment for business combinations. APB Opinion No. 16 defines generally accepted accounting principles for business combinations and specifies that a business combination must be recorded as a purchase or as a pooling of interests. However, the two methods are not alternatives for recording the same business combination.
An expert system consists of four components: the knowledge base, the inference engine, the user interface, and the explanation facility. The source of much of the knowledge programmed into PURPOOL's knowledge base was the twelve criteria contained in Opinion No. 16. All twelve critieria must be satisfied in order for a business combination to be treated as a pooling of interests.
Therefore, the rules constructed and included in the prototype's knowledge base tested each of the twelve criteria contained in Opinion No. 16. In this way a business combination is evaluated, and the expert system determines if the combination is a purchase or a pooling of interests.
The validation of PURPOOL used a framework consisting of nine elements. The nine elements were the following: content validity, criterion validity, construct validity, objectivity, economics, reliability, systematic variance, and error variance. Each of the framework elements was described and related to PURPOOL during the validation process. In addition, twenty-nine cases were developed and used as a part of the validation of the prototype.
The construction and validation of PURPOOL demonstrated that complex accounting issues, such as business combinations, can be evaluated using an expert system. Other accounting areas that would be suitable for similar expert system development were determined to be leases, earnings per share, and foreign currency translations.