Guidelines for Complexity Factors when estimating

[anopres]

Has anybody found any reference guidelines for figuring out how to assign values to the TCF and ECF figures?

I've been able to find assistance on most of the TCF values by looking at the corresponding factors used in calculating Function Points.  I haven't been able to find any assistance on the ECF values.

ECF values are pretty subjective at face value.  It would be great if there were a set of objective guidelines for them.  For example, application experience could be the average number of years per developer working with the language.

I can't imagine how you would objectively determine Motivation levels.

[TrtnJohn]

I think these numbers start out bad and get refined with each new project you do.  My advice:  Take a guess.

[anopres]

I was afraid of that.  It would be nice if we had some definitions of what Karner was actually thinking about when he came up with his categories.  Without some objective guidelines, it's probably next to impossible to compare estimates between organizations using this approach.

I mean, how do you measure a team's Motivation?  I've been using a level of overtime that team members are willing to put in and then adjusting it with a whole host of other factors.  It's difficult to maintain consistancy beween estimates.  Oh well, maybe somebody will write a white paper about it.

[KP]

If you Google on Karner+ECF the first page gives half a dozen promising links you might follow, including this one: Project Estimation using Use Case Metrics

[sargasso]

Here's my bible on the topic:

http://heim.ifi.uio.no/~kribu/oppgave.pdf

bruce

p.s. you might also like to review this thread
http://www.sparxsystems.com.au/cgi-bin/yabb/YaBB.pl?board=UMLPRO;action=display;num=1098471280;start=10#10

p.p.s.  I am not aware of anybody ever having collated and published an updated survey of the values across the whole industry.  Any one else heard of one?

[anopres]

Thanks for all the help guys.  I eventually got fed up looking for other people's help on google.  There seems to be a lot of people doing studies about the process or case studies, but I can't find anybody that has tried to put some objective measures on the factors.

I've been using the EA tool for about 2 or three years now, so I just posted up the approach I use on my blog hoping to get some comments about where I've gone wrong.

Take a peek at http://mrmx.blogspot.com/2005/06/cf-software-estimation-part-3-working.html
if you feel like commenting or sharing your experience.