| A general cost effective model was developed to test patterns of staffing for reference service in university libraries. Specifically, the model provided evaluative data on alternative staffing strategies for the reference desk.;The procedure for implementation of the model, to be carried out by library management, is based on statistical decision theory wherein, first, the states of nature are listed. These are the distribution of the patron arrival, the distribution of the types of patrons, the distribution of the types of questions asked by the patron, and the required time to answer these questions. Second, the strategies available to the person in charge of making staff assignments are identified; they are the number and skill level of staff members required. Finally, the consequences, or the result of interaction between the states of nature and the strategies, are demonstrated. In order to bring about the consequences, a simulation model is developed at this stage. The states of nature which make up the framework of this model together with the strategies' attributes establish the model inputs. The outputs are the results of simulating these data. The outcomes of the simulation model plus other effectiveness measures mentioned in this study, such as cost of staffing, the adequacy of answer, and consideration of the cost of staff idle time in the case of multi-server strategies, form evaluative data for the management of a library. These data could then be weighed according to the policies of a specific library. Ultimately, it would be possible to select the most cost effective strategy suitable to a certain case or situation by considering the trade-offs between the effectiveness measures.;The measurement tool developed in this study supplies any given library administrator with the information needed to predict future results for the strategies selected. This is advantageous since it may be expensive or inconvenient to implement various staffing strategies on a trial basis.;Other advantages of the model are the following. First, it provides a guideline for collecting meaningful data on reference desk activities and the types of patron being served. Second, the outcomes of the simulation model allow a reference head, or the person in charge of making staff assignments, to base staffing requests on sound documentation. Third, the implementation of the model enables library management to develop standards of performance for use in setting levels of activity at the reference desk and in planning the usage of idle time for each member of the reference desk.;The method of development of the model was based on statistical decision theory and queueing theory. The computer simulation technique was used to produce measures such as staff service time, staff idle time, cost of patron delay time, and number of questions lost by staff members who were unable to provide answers to specific types of questions. This was based on the estimated probabilities of patron arrivals, types of patrons, types of questions, and the time needed to answer questions by staff members who were determined to have various levels of skills. The operation of the simulation model was further verified with the representative data for five single-server strategies; namely, the student assistant, the Library Technical Assistant, the Library Associate, the Librarian, and the Senior Librarian.;The model developed in this study is applicable to a wide range of situations. It could be used by all reference departments regardless of the type of library, the type of staff classification, or the type of questions or service provided. |
