| A stochastic, discrete-event computer simulation model was constructed to examine the factors that affect a library's ability to satisfy the needs of its patrons. The model provides a controlled environment in which alternative library policies about the circulation of the collection and patterns of use by patrons can be evaluated and analyzed by categories of failure. The collection can include monographs, and bound and unbound periodicals of different subjects and loan periods. The patrons can be subdivided into categories depending on their borrowing privileges and propensity to use the library.; The model assumes that a patron only continues to use the library if the library's service is satisfactory. Therefore, the model not only records the outcomes, the time and effort spent on the requests from patrons who do use the library, but also potential requests lost by the library as an indication of failure. Unlike most other models, this model allows users to be flexible when the requested items are unavailable by either recalling the items, browsing for substitute items, returning at a later date to continue the search or abandoning their request.; The effects of six policy variables (loan-period, purchasing duplicate copies, weeding rates, browsing frequency, binding and reshelving intervals) and five descriptive variables about the characteristics of the collection and the patrons (rate of obsolescence and quality distribution of the collection, volume of demand, temperament and persistency of the patrons) were examined. Among the six library policies, the model found loan-period policy to be most effective and browsing and reshelving-interval the least in improving system performance.; The results show both limitations and promises for the model. On the one hand, the model takes considerable time and computer resources to run and the results can be generalized only with great care. On the other hand, the model provides quantitative predications and unique insights which are difficult to obtain from other methods. The model's flexibility in handling combinations of policies and assumptions and adaptability to local circumstances also make it a useful management tool for policy evaluations. |
