Firstly, I would like to make it clear that my knowledge regarding Understand Society's data as well as STATA is limited, so I apologise in advance for asking what might be perceived as an obvious question.
I am currently conducting a project into the long-term care that stroke patients/survivors utilise. As both waves 7 and 9 collected observations that are relevant to this topic e.g. as they have variables that distinguish stroke patients as well as the formal, informal and social care that they use, they are consequently very much suited to investigating this issue.
Currently I am determining my wave 7 sample size from those understanding society wave 7 participants who have 'mentioned' having a stroke in the past year (since the last questionnaire) using variable hcond7. In total there are 109 patients. I however noticed that 2831 are proxies with values of -7. Is it safe to assume that these 2831 individuals have 'mentioned' having a stroke in a previous wave's questionnaire. Would there be any other reasons as to why they have proxy answers? I only ask as I was trying to find a way of incorporating previous stroke patients from other waves into waves 7 and 9, and a larger sample size would greatly help with my project.
Any feedback or opinions would be greatly appreciated.