Creation and maintenance of links in large hypermedia documents are di
fficult. Motivated by an application to a federal clinical practice gu
ideline for cancer pain management, we have developed and evaluated a
repertory grid-based linking scheme we call repertory hypergrids. Harn
essing established knowledge acquisition techniques, the repertory hyp
ergrid assigns each ''knowledge chunk'' a location in ''context space'
'. A chunk links to another chunk if they are both close in context sp
ace. We have developed a program to convert the hypergrid and associat
ed knowledge chunks to HTML and have made the hypermedia clinical prac
tice guideline available on the World Wide Web. To evaluate the scheme
, we conducted two analyses. First, we conducted a protocol analysis u
sing the paper-based guidelines. Six users of the guideline addressing
typical cancer pain management tasks made 30 explicit links. The repe
rtory hypergrid using a neighborhood size of 16 captures of 24 of thes
e links. With optimization, the repertory hypergrid captures 27 of the
links with a neighborhood size of 14. Second, 18 users addressed the
same tasks, six using the paper-based guideline, six using the hyperme
dia document with repertory hypergrid-created links (''TALARIA''), and
six using the hypermedia document with randomly selected links (''Ran
dom TALARIA''). TALARIA users found the required information significa
ntly more quickly than either the users of the paper-based guideline o
r of Random TALARIA, with no loss in accuracy. (C) 1995 Academic Press
Limited