5 ms·
I know of two interesting data points here, actually. In the state of Maine, there were 3 programs which allowed high school students to simulate the state gove
by ekidd 14y ago
I know of two interesting data points here, actually. In the state of Maine, there were 3 programs which allowed high school students to simulate the state government (write bills, hold committee hearings, etc).
Boys State / Girls State: Open to the two best students from each high school in the state. Had a reputation for being incredibly boring, largely because there were few differences of opinion.
Model State: Open to any student who wanted to come. This featured intense debates and was generally awesome.
Similarly, the best programming conference I ever attended was Lightweight Languages 1 at MIT. There were about 30 attendees who worked on major open source programming languages and on academic research languages. The Perl people and the Python people and the Scheme researchers got into some really intense and amazing debates, because they came from intellectually diverse backgrounds.
Given these experiences, I actually favor diversity of opinions and backgrounds over speaking ability. Of course, I'd rather have both.
- oh_sigh 14y agoWhat makes you think women and men will have different opinions about software development based solely on their gender?
- ekidd 14y agoI interpreted angersock's question to be about diversity in general, and not specifically gender diversity. For me, this is more of an observed relationship (with an overly-small sample size) than any kind of well-formed theory. Diversity of background seems to be correlated with diversity of ideas, which in turn seems to be correlated with fascinating discussions.