" Only 8.2 per cent of the respondents identified as women, which the authors say is "unfortunately a common problem with many developer surveys.""
Sorry, if 8.2 per cent of the respondents were actually women, that means that your statement is bullshit. Perhaps everyone who answers your next survey should check the box for female. Then what would you do, admit that the rest of your survey results are likely garbage?