UK perspective: it may depend on why the data is being collected. UK government guidelines [1] for general forms to do with benefits/services &c are fairly thi

UK perspective: it may depend on why the data is being collected. UK government guidelines [1] for general forms to do with benefits/services &c are fairly thin Organisations that have to maintain equality metrics in some form [2] have a more complex approach. This second approach seems like the nearest to the Github situation. The example I have cited (third or fourth in goog search results) maps directly to 'protected characteristics' in diversity legislation in UK. [1] https://www.gov.uk/service-manual/design/gender-or-sex https://www.gov.uk/service-manual/design/gender-or-sex [2] http://www.ecu.ac.uk/guidance-resources/using-data-and-evidence/monitoring-questions/ http://www.ecu.ac.uk/guidance-resources/using-data-and-evide... Meta comment: Survey questions are tricky to formulate. Kicking ideas round in a group on a whiteboard with reference to external advice/exemplars seems to help get wording sorted and clarify the taxonomy. Genuine question (I don't work in the software world) is it normal for a lone web developer person to be left just making up the questions as they are coding them?