Melinda, > My own feeling is that if we were to find that the > numbers supported the notion that there's bias > present in the system we probably couldn't do anything > about it without tearing the organization apart, so, I am actually a little bit more optimistic about it, for a couple of reasons. First, it does not take genius to understand that we have biases. We all do. And I'm not trying to minimise the problem by presenting it as a common occurrence, but it is a fact that we have biases. People tend to prefer status quo, select similar people to themselves for various roles, carry society's stereotypes about various tasks, etc. Not surprising that the IETF has that too, but of course it does not make me happy. We can try to reduce the bias, however. What helps with this is education, information, and good examples. As for the information & education part, I think the discussion we've had on diversity has already made the issue better highlighted. When something is in people's mind they tend to take it into account. FWIW, personnel discussions that I've been involved as the chair have had a very serious discussion about whether there was something we could do here to improve some aspect of diversity. Again, this isn't the only factor to consider, and most IETF leadership discussions are pretty constrained, have limited pools of volunteers to choose from, and so on. But I think we'll see the effect of that in coming years. There won't be a revolution, but I think we'll see some effect. Secondly - and this is why I though some knowledge of our participant base would be useful - what I'd find useful is to get to a mode where we can look at progress. Is there a way to increase #countries #small companies #women etc? Be it about the participants or the leadership. Could we set a goal that we'll increase some aspect every year, 2014 to be better than 2013? It would be wonderful if we could track that somehow, while of course respecting people's privacy, not retaining any data, checking with experts, etc. And we could, of course, conduct a far more comprehensive surveys, I think we'll get to that as well. Later, perhaps when the design team comes up with some recommendations. And by their nature, such comprehensive surveys will be separate efforts. But Ray's question was prompted by the need to have the IETF-87 registration form up and running very soon and we wondered if some very basic information could be collected. (Registration questions are not the only way to do this of course, separate surveys are a possibility. And I've experimented with using heuristics to make a guess of population's gender percentages. If you look at my document statistics page, you'll see the beginnings of some of the analysis. With truly massive error rates, of course. This could be repeated for meetings as well, I suppose. Not sure this is a viable approach, though.) Jari