XpdWiki

FrontPage
RecentChanges
XtC
FindPage
PageIndex
XpApprentices

Set your name in
UserPreferences

Edit this page

Referenced by
PersonalPractices...




JSPWiki v2.0.52


RepertoryGrid


What's a RepertoryGrid?

It's a tool for eliciting ontologies. There are various schemes for using them (variously formal), but here's a cheap and cheerful approach that's worked well for me.

In best Blue Peter tradition, here's one I made earlier:

1=tools 1=testing policy 1=facilities 1=testing as design
5=techniques 5=testing as such 5=thinking 5=testing as QA
xUnit 2 4 1 2
test-first 5 2 5 1
regression testing 5 1 2 5
pre- and post-conditions 1 3 5 2
when to stop? 3 2 5 3
test-testing 3 3 2 4

This was done as a "warm-up" exercise for a testing workshop I lead last year.

How it works is: think of half-a dozen or so things you want to think about in some area. In this case, testing:

xUnit
test-first
regression testing
pre- and post-conditions
when to stop?
test-testing

Then pick any two, say test-first and pre- and post-conditions, then try to think up an axis of variation that lets you distinguish them, say tools vs. techniques, and use an index number to indicate this.

1=tools
5=techniques
xUnit
test-first 5
regression testing
pre- and post-conditions 1
when to stop?
test-testing

Then place the other items on the same scale:

1=tools
5=techniques
xUnit 2
test-first 5
regression testing 5
pre- and post-conditions 1
when to stop? 3
test-testing 3

pick another two items and repeat. Once you have a few of these axes in place, look for patterns.

1=tools 1=testing policy 1=facilities 1=testing as design
5=techniques 5=testing as such 5=thinking 5=testing as QA
xUnit 2 4 1 2
test-first 5 2 5 1
regression testing 5 1 2 5
pre- and post-conditions 1 3 5 2
when to stop? 3 2 5 3
test-testing 3 3 2 4

If we look at testing policy vs. testing as such , we see that apart from xUnit, all the items are on the testing as such side, so that's a prompt to look for new items that are more at the policy end of the scale. Or we can look at test-testing and see that it is a fairly middle of the road item, so that's a prompt to look for an axis of variation that will distinguish it more strongly from the other items. And so on. In the more formal approaches you can look for more subtle patterns.

Keep iterating this process and pretty soon you can expose a lot of information about how your thinking s structured in some area of interest. --KeithB


Edit this page   More info...   Attach file...
This page last changed on 30-Jul-2003 13:46:49 BST by unknown.