“Primary” field data collection:
From a purely SFA and natural science perspective I might say (even though I am aware this not entirely true) this is the gold standard. Collecting your own data specifically for your research questions is in many ways ideal if you have no time or money constraints. (Note that it doesn’t mean not using other methods because you probably still want to verity your results).
Keys points I take into consideration for SFA:
- Particularly useful to collect “place-specific” P content measurements. (P content records might exist for wastewater management. But for flows operating within the city (i.e., smaller than standardized government record keeping) on-site measurement could be great (e.g., compost, runoff from UA sites, and soils)
- Would require a substantial amount of man-power to measure flows of food, UA practices and production. Even if we only did a sub-sample, we do not have a good idea of within city variability (although it is probably high) so targeting sample size and representative sample groups is more difficult. Same problem with P content sampling.
These are not problems within themselves, they just make field data collection perhaps beyond the scope of a 6 city comparison.
Keys points I look into consideration for local context:
- Some of the same pros and cons as for SFA but for different variables (e.g., rain, temperature, ect).
- For some of the more social factors observation or participant-observation (of people and practices but I think also documents) might be useful but time consuming. Such “field” data-collection can tell you something different that asking why people do what they do.
Based on these criteria (“selecting methods” posts parts 2,3,4, and 5) I am opting to use literature review, surveys, and review of official government and organization reports (I put this separate from literature review as the objective of the publications are different). By using all three of these data sources I hope to get the most accurate data possible, and when possible triangulate bias by looking at the same P flow or context factor with more than one data source. Next I will write about my experience developing survey questions, which is KEY to quantifying P in UA and thus to my whole thesis chapter.