It's surprising how different people interpret post-1980 decline of women in computer science differently. As the article suggests, introduction of personal computer is one of the best candidate. My interpretation is that this data supports "interest, not ability" thesis of Damore, and personal computer amplified pre-existing interest difference.