The industrial lot and its sampling implications

It is shown that wrong inferences are drawn from samples because of failure to understand what constitutes a lot and because of failure to verify the existence of a lot in the practically useful sense. Time-tried methods of making valid inferences from samples without increased cost are pointed out.     

Impact of DOCLINE on interlibrary loan service at the National Library of Medicine

In March 1985, the National Library of Medicine (NLM) began implementation of DOCLINE, its automated interlibrary loan request routing and referral system. By the end of fiscal year (FY) 1987, over 1,400 biomedical libraries in all seven regions of the Regional Medical Library network were using the system. This report summarizes the findings of an analysis of the interlibrary loan (ILL) requests received in FY 1987, comparing the results with a similar analysis done in FY 1984 to describe any changes in the requests or service which might be attributable to DOCLINE implementation. DOCLINE has had a substantial impact upon ILL loan service at NLM. An increase in the number of ILL requests (35% over FY 1984) can be attributed to the speed and ease with which requests may be routed to NLM through DOCLINE. Requests which cannot be filled by local or regional libraries are automatically routed by the system to NLM as the library of last resort. NLM's fill rate for serial requests has declined, however, from 78% filled in FY 1984 to 67% filled in FY 1987. Some of the decline results from the 11,000 requests that NLM did not fill because the borrowing libraries were not willing to pay the NLM charge for filling the loans.