We are establishing the collection of user ethnicity data in the eHealth Forum. This is to enable the subanalysis of data in the Forum and its polls along ethnic lines when needed, and to ensure that our data can be reported on a Maori, iwi, and hapu level when required. A bonus is that we will gain visibility of our ethnic diversity (or lack thereof).
To this end, I have imported the ethnicity model that we developed for healthforum.nz, based upon the 2018 NZ Census. I’ll present it here so that you can take a look and make suggestions if we have missed something.
We ask two questions, the first structured and the second unstructured. Please give us your feedback!
Ethnicity Category
Which broad ethnic category do you most closely align with? Not displayed on your profile
Kia ora Nathan
Appreciate your work. I understand there may be some limitations in the ability to collect more than one ethnicity in the structured question but I do think it is problematic to ask what ethnic category someone most closely aligns with. The HISO document HISO 10001:2017 ethnicity data protocols (health.govt.nz) provides some guidelines for collecting this information and also how the question should be asked.
I note stats New Zealand recommend 6 ethnic groups with a minimum requirement of 3, maybe if we are unable to multi select then the question could be re-worded 'which ethnic category do you belong to? - if you belong to more than one ethnic category please list in the freetext box provided.
Ngā mihi nui
Erin
I wish that I had found that when starting this journey!! It is really helpful, and in the abscence of any better thoughts we should probably comply to it as much as we can.
I’ve had a crack at the technical side of things - please give the new joining wizard (a bit rough around the edges) a go:
Previously we had to handle this with groups, but now we can do multi-select user fields. I’ve moved over to the new way, and will migrate our data into this sometime this week. You should notice that the ethnicity groups disappear shortly, and your ethnic groups will appear in your profile.
This might sound like a political response, (its not) but I feel that it important that “Pakeha” also be a formal response. Whilst my ancestors came from Ireland, I have no links back to Europe or Ireland. I am a New Zealander. I feel much more aligned to ‘Pakeha’ than I do to ‘New Zealand European’. My ancestors arrived here around 140 years ago – I was born a Kiwi as was my father, grandfather and great grandfather.
I feel like I have little connection to anything European.
I agree with you, Chris, that ‘NZ European’ isn’t a great fit for many, including me. However, that is a point to take up with those who set NZ’s ethnicity data collection standards!!!
I’ve now fully implemented the ethnicity data collection mechanism, and it seems to both work great and be fully compliant with the HISO standard. I’ve also merged all the data and tidied it up so it is all looking sweet now. Many of our older members don’t have an entry for their ethnicity as it has only been recently implemented; you can add yours in your profile if you wish.
Regarding the ethnicity groups, I’ve only retained the @maori one due to the importance of being able to easily identify our Māori colleagues while CiLN and the eHealth Forum are working towards ensuring that we are as equitable as we can be. Is this okay do you think?
I’ve just implemented gender recording in the eHealth Forum. In order to ensure that we collect this adequately, folk will be directed to their profiles when they next login (still working out how to do that).
Thanks Nathan, gender is always a sensitive field.
StatsNZ have been going through a ‘what to ask’ for a year or two now, and have consulted with the community along the way:
They’ve gone with Male / Female / Another (please specify), which allows the respondent to self-describe… good.
It also allows the respondent to reply ‘prefer not to say’ which always needs to be an option in what is for some a very invasive question.
I’d tend to omit the suggested ‘Unspecified or unknown’ as I think that’s for others who are filling out the field for someone else (and can’t guess their gender). So a self-filler-outer knows what to put and can put ‘unknown’ in the ‘Another’ category if that’s how it is for them. I’d be surprised if anyone did though… we’ve all got an idea of a ‘something’ that we are?
I’ve just thought of a way that this could be implemented on the technical front (without cluttering up the form). And as nobody actually ever selected ‘Unspecified or unknown’, I felt I could safely delete that as an option.
And looking at this in more detail, I see an official update occurred just 2 weeks after Mitch’s post:
Stats NZ’s detailed coding rules arrived after our May 2022 update to the HISO standard, but we’re still aligned in our use of the three categories plus write-in option (plus the ‘unknown’ code for absent information, which I agree is not needed if you’re dealing with the person directly). Besides staying in touch with Stats NZ, we’re involved with ISO work to develop an international standard for sex and gender recording standard - @micah_davison is leading this
The coding specification has more of a statistical focus than the Consumer Health Identity Standard, and goes into detail to support production of statistics in instances where a collection allows for a single response for gender and only outputs a single a response, and where a collection allows for multiples responses to gender and outputs multiple responses.
In the context of patient information, it is possible responses would not be coded back in the same way. At this stage I don’t think the NHI text data under the ‘another gender’ response category is coded, but left as it is entered. @joel.brown could you please clarify that is correct?
Will this not cause some confusion, if we have two different data sets being generated from the same input? One is being used for planning and the other for service delivery. Or am I missing something really obvious?
