Marie Müller

Going through my #thesis for my #viva and thought I'd share some quotes on the way:

'Future studies should investigate the roles of, and interactions between, multiple dimensions of exposure, including spatial dimensions (where/characteristics of space), temporal dimensions (when/how often/how long), activity dimensions (type of activity/with whom), and subjective experience of exposure (perceptions).'

#greenspace #mentalhealth #wellbeing #childdevelopment #adolescence

Jan 9, 2023 at 12:47pmWeb
Show threadMarie MüllerJan 9, 2023

As a prerequisite for the above (especially when you work with #longitudinal #secondarydata):

'As physical environment factors have not been a priority in large-scale, longitudinal studies in the past, more attention should be paid in the future to include data on the physical environment in these studies, or to make it more straightforward for the data users to link their own physical environment variables.'

#greenspace #mentalhealth #wellbeing #childdevelopment #adolescence

Show threadMycotropicJan 9, 2023
@marie_ahoi I work with a large study of 10k smokers and we took years to make the case that we needed at least zip code level environmental variables. We were required by IRB to hold the geospatial identifiers in a single laptop and we got one pass at the census data. We're going back but, correctly, IRB is very strict on how these data are used. I agree with you of course but it's not trivial to accomplish.
Show threadMarie MüllerJan 9, 2023
@mycotropic I agree, not at all trivial! I've worked with data from Understanding Society, a large, longitudinal study in the UK. It took me over 1.5 years to get access to sensitive data (postcodes), and I could only access these data from a PC in my office at UCL. I even had to change the IP address to a static one. I understand we need to value and secure such sensitive data, but it was frustrating, restrictive, and very time-consuming (especially in the context of a 3-year PhD).
Show threadMarie MüllerJan 9, 2023
@mycotropic I wish there was an easier way, but, of course, data protection cannot be compromised. That's why I am hoping some of these studies will consider physical environment from the start (e.g., they may want to have a GIS team on board). That way there may be interesting physical environment factors, e.g., air pollution, already linked.
Show threadMycotropicJan 9, 2023
@marie_ahoi They should also have a plan in place for revisiting the spatial data when people move or the environment changes substantially. In our case we have 21 clinical centers all over the US and participants can drive in many, many miles for a study visit. If we'd planned ahead for this then we'd have spatial AND temporal exposure data since we're at 15 years of follow-up now, at least one Census recalculation of ADI and rurality right? #COPD #COPDGene
Show threadMycotropicJan 9, 2023
@marie_ahoi I did my dissertation using spatial data but it involved mortality and in the US that's outside of IRB (for this analysis anyway). For our large study of #COPD ( #COPDGene ) though the participants are present and protected by #HIPPA so - higher standard. I think that going forward we'll pull all of the shape files, do all of our #kriging and other work before we even think about merging the spatial identifiers. Now what does THAT grant look like??!