I would agree that responding to a question regarding the strengths and
weaknesses of a repertory grid is a bit non-specific, as there is no single
grid.

I would, however, disagree that a grid is totally free from interviewer
bias. Decisions about the number and type of elements, construct
elicitation process, selection of a ranking/rating method and analysis
(whatever form) all involve critical decisions which allow for individual
'bias', personal preference and subjectivity.

In addition to the useful citation John provided there are two good recent
books on grids - and some older reviews:

Fransella, Bell & Bannister: A Manual for Repertory Grid Technique
Janckowicz: The easy Guide to Repertory Grids

Some basic considerations (bearing in mind the caveat that there is no
single GRID):

1. although they contain a lot of information grids are small data sets and
hence any error by the interviewer or interviewee (e.g if a wrong value is
entered or a person makes a wild guess) can make a large difference.

2. representativeness of elements, an understanding of issues related to
grid size, number of construct/elements, elicitation methods and an
understanding of how the data can be analysed (these comments are
also applicable to other approaches) all need prior careful consideration

3. people may not readily understand what a grid is/or advise you & it may
be difficult to establish validity/reliability by established methods

4. The question has to be asked, is a grid the best way to answer the
question you have

Having said this grids have been used for a wide range of applications and
have particular utility when there is an interest in examining individual's
points of view/constructs/subjectivity.


regards

Bob