Reference enquiries and physical visits to the library are standard performance measures that all Queensland public libraries should collect. These performance measures are incorporated into the annual Statistical Bulletin and the national public library statistical report.
The collection method for these statistics can be either a continuous year-round count or by way of four annual samples [each lasting two weeks]. These samples are then used to determine an annual estimate, where the totals of the sample periods are multiplied by 6.5 to provide estimated annual statistics.
A reference enquiry is defined as a transaction where a staff member finds or recommends relevant resources within the library, assists in the use of these sources or answers the enquirer’s question either through the use of library tools or by reference to other sources of information, e.g. the Internet. Bibliographical enquiries should be included. Directional and administrative enquiries should be omitted. Each seperate transaction by the same library user should be counted as a new enquiry.
Physical visits sampling
An important distinction with this measure is that it is reporting visits, not visitors. A physical visit to the library is defined as a visit to the library by a member of the public, where the same visitor returning numerous times is to be counted on each separate entry to the library. It is recommended that the same sample periods are used for both reference enquiries and physical visits.
Sample periods in 2010
The sample periods are set for 14 calendar days, beginning on a Sunday and concluding on a Saturday. This ensures the full spread of library opening hours are included, whatever combination of hours or days your library is open to the public.
- 14th February – 27th February
- 09th May – 22nd May
- 19th September – 02nd October (School Holiday Sample)
- 14th November – 27th November
The basic approach to data collection – including both high and low seasons (School Holiday) – is methodologically sound. Collecting data for eight full weeks involves quite a bit of work, I believe. A smaller, but more dispersed, sample might do the job. I suggest doing two full weekly cycles, taking one new weekday each month:
- a Monday in January
- a Tuesday in February
- and so one
The counting days should be selected in advance.
If a few libraries try out both, the two methods can be compared – which may give the conclusion that the less time-consuming approach is good enough.