What is the difference between "phase wise defect removal efficiency"
and "review effectiveness".

-Suyash

Hi Suyash,

Here are my suggestions, I'd be interested to know what others think
....

Suyash Agnihotri wrote:

I'm not aware of a single measure of phase wise defect removal
efficiency. I'd imagine it means: Measure number of defects detected in
each phase, e.g.
- defects detected in requirements reviews
- defects detected in design reviews
- defects detected in code reviews
- defects detected in test
- defects detected in pilot trials
- defects detected by customers during field operations

Use this measure to drive defect detection back up to the ealiest phase
where it's easier to fix. Some companies note, in their bug tracking
systems, where a bug was detected and where it _should_ have been
detected, Then they improve the test and review processes in the phase
where it should have been detected, to detect such bugs in future.


This would be a measure of the number of defects detected per
reviewer-hour. Effective reviews detect lots of bugs in an hour,
ineffective reviews detect none

.... although that is an overstatement.

To make reviews effective, you concentrate them on the areas that are
most likely to be buggy. So, for example, review code written by
novices, or complex code. Sample review stuff that's simple, or written
by known experts. And if, while reviewing something, you're part way
through and finding it's well done and not buggy, you might decide to
stop the review right there because it's clearly ok: that would be
another way of improving review efficiency.


Hope this helps,
Patrick

Thanks patrick

-suyash