Re: IETF Diversity Question on Berlin Registration?

At 10:28 AM 4/18/2013, Pete Resnick wrote:
> I noticed this post from a few days ago, but I think instructive to 
> talk about. And this is not picking on James; I think it's likely 
> that there are many folk who have similar perceptions, and I think 
> it's useful to think about.
> On 4/12/13 3:37 PM, James Polk wrote:
>
> > Eyeballing the IETF (and I've missed 2 meetings since IETF45, been 
> > a WG chair for 8 years, and written or revised over 300 submitted 
> > IDs) there is consistently about a 70-to-1 ratio of men to women.
> Your "eyeballing" had you put the ratio at about 70:1. I wouldn't be 
> surprised if this was a common view. However, when the whole 
> diversity conversation started, a few people quickly scanned through 
> attendance lists just to do a guesstimate of the actual ratios over 
> the past 10 years. They were seeing rates somewhere between 10:1 and 
> 18:1 (with so much variability due to guessing on the basis of 
> names), and it's seemed pretty consistent over the last 10 years. 
> Over the past 5 years, the ratio of Nomcom members (randomly 
> selected from the community) is about 10:1, which is consistent with 
> those numbers.
I believe I did myself a disservice in assigning such a high ratio 
without saying it "feels like 70:1", which it does. But I'd truly be 
surprised if it's only 10:1 - and you can't make effective and 
accurate estimates based on guessing the gender of the names of an 
international organization like the IETF is.
Now, and this is purely from a defending myself pov, the feel of a 
70:1 crowd that's right in front of you from a 18:1 crowd isn't that 
much, especially if you're part of that larger number. There is a 
matter of the number of the group you happen to be part of is so 
overwhelming you tend to guess on the high side.

> That's a factor of between 4 and 7 difference between an "eyeball" 
> guess and a rough calculation. I think that's likely an 
> unintentional sampling bias of your (and many other folks) eyeballs. 
> And I think it's because we have a tendency to subconsciously 
> discount the numbers of people who do not appear in leadership, or 
> even simply don't behave "the way the rest of us do".
I have to disagree with you here. In my mind, this "feels like a 70:1 
ratio" has nothing to do with the ratios of leadership to the 
community within the IETF, it's from the gathering places and 
hallways where there are just wave after wave (after wave) of men. 
It's also from sheer lack of numbers of women within the WGs I've 
attended over the years (SIP related, MMUSIC, GEOPRIV, ECRIT, the 
whole TSV area, etc).

> This isn't to say that we should spend all of our time on this 
> question by collecting statistics; that's just navel gazing. But we 
> do want to understand the nature of the problem and not let our 
> guesses and subconscious biases get in the way.
whether it's 70:1 or 18:1 - that's significantly more of one group 
that the other.
Thanks Pete for making this point, and causing me to clarify what I 
originally wrote.
James


> pr
>
> --
> Pete Resnick<http://www.qualcomm.com/~presnick/>
> Qualcomm Technologies, Inc. - +1 (858)651-4478