AFTERMATH OF AN EVALUATION
| Published date | 01 January 1971 |
| DOI | https://doi.org/10.1108/eb026505 |
| Pages | 1-10 |
| Date | 01 January 1971 |
| Author | F.W. LANCASTER |
| Subject Matter | Information & knowledge management,Library & information science |
THE
Journal of Documentation
VOLUME 27 NUMBER 1 MARCH 1971
AFTERMATH OF AN EVALUATION
F.
W. LANCASTER
Graduate School of Library Science, University of Illinois
As a result of the comprehensive evaluation of MEDLARS conducted in
1966-7,
several recommendations were made to the National Library of
Medicine as measures appropriate to the improvement of effectiveness and
efficiency of the system. This brief paper summarizes these recommenda-
tions and discusses steps since taken by NLM to implement them.
IN THE TWO-YEAR period 1966-7 the National Library of Medicine
undertook a comprehensive evaluation of the MEDLARS demand search
service. This study was fully reported in
19681
and
19692,3.
The purpose of
the
present communication
is
to outline steps taken by NLM to implement
some of the recommendations made by the author
as
a result of the evalua-
tion programme.
USER-SYSTEM INTERACTION
A critical aspect of all information retrieval operations is the interface be-
tween the user and
the
system.
In particular, the system should give the user
every possible assistance in the formulation of an accurate request statement
(i.e.
one that clearly reflects his actual information requirements). If a user
makes a 'poor' request (too broad, too specific, or simply vague), it is
obviously unlikely that the search
results
will be of maximum value to him,
except purely by
chance.
In other words,
a
poor request will tend
to
produce
poor search results despite the quality of the indexing, the search strategy,
and the system vocabulary.
The MEDLARS evaluation revealed that imperfect request statements
were largely responsible for approximately
25%
of all the recall failures and
1
To continue reading
Request your trial