When gathering personal data about respondents, you must comply with legal privacy obligations. Also be aware of ethical considerations, such as the importance of allowing people to self-identify. This can sometimes clash with the need to gather measurable data for evaluation.
Data from tick-box categories (eg ‘Male’, ‘Female’, ‘Other’) can be automatically collated but is offensive and othering. However open-ended questions, such as ‘How would you describe your gender?’ may generate different responses (eg woman, female, F, etc), which must be manually assessed. Balance the need to collect measurable data with the importance of allowing people to self-identify.
A good approach is to offer multiple-choice responses with the option to self-describe. For example, in a question about ethnicity, you might wish to offer a checkbox for ‘Arab or Arab Australian’. However allow people to express the nuance of their identity by adding a question such as ‘Would you like to be more specific?’ with space to write a response. All questions should be optional, with the ability to select more than one.