\

Macroevaluation - like the figures that were shown on NYPL and NLB.  Doesn’t necessarily point to how these figures are achieved, or whether they can be improved, or aid ways in identifying failure points Microevaluation – finds factors responsible for the levels of performance on a transaction level

How do we measure whether one library is more effective than another?

=

And others McGrath (68)

Keep in mind what is the parallel in digital libraries

ROC – Receiver Operator Characteristics – used by radarmen in WW II by servicemen to differentiate between radar blips as planes and false alerts.  Was only applied to signal processing / IR in the 1970’s.