Lessons learned

Unfortunately this week I have nothing to report on the data collection front. Last week at the UA event  I was able to remind some survey participants to send me their missing data but I haven’t gotten any emails or new surveys booked. I thought I would thus take a little time to reflect back on the summer.

How did the summer evolve:

I tried to set up everything “just-right” but in the end I had to adapt to how both my stakeholders and research assistants worked so that I could still get the as much of the information I needed. Contacting stakeholders was hard at the beginning because I was a bit shy. It was also challenging to really get into the “groove” about how to make respondents feel comfortable while also insisting on getting the quantitative numbers (they had the information but they didn’t understand they had it or thought it was way too hard to get it). Both aspects got better over the season but got harder again in the fall. I was (and still trying to) surveying actors who were less organized and enthusiastic (which is probably why they didn’t respond quickly in April) and trying to get missing pieces of data from actors throughout the whole summer.

Each type of UA had its particular challenges. For the business and collective gardens it was great that one person was usually in charge, but they were often busy and had irregular schedules and did not always keep track of production or management as the main goals were social. For community and private gardeners, the hardest was to get initial contact as there isn’t a public record of participants. But once in place, with proper time commitment, we could measure at least the inputs and garden size first-hand with them. It was still hard because even when I provided detailed instructions, I ended up having to probe in person and measure things myself. I must say that the larger farms were much more of a challenge then I thought they would be. The larger farms were perhaps the hardest because of the managers didn’t have much time and records were not organized for easy sharing.They keep records but not systematically good ones, and it was a lot more time consuming than I thought actually get the records (often only paper copies) and to make sense of the records to then extract the information I was looking for.

Here are my lessons learned:

1. You really need to call and be quite persistent (while being nice and respectful of course).
2. There doesn’t seem a magical formula for people to fully answer your questions.
3.  Data collection takes a lot longer than you hope when you are asking people to answer you.
4. Technology is your friend but it isn’t always the solution (aka the tablet “situation”)
5. Managing field assistants is really time consuming.

What would I do different?

For points 1,2, and 3: Honestly I don’t think there is any perfect time or way to ask people for information but I think for farms I would have done data collection in the winter (because they are so swamped during the summer), done data collection with organizations at the end of the summer (not winter because a lot of the people that take care of gardens don’t work year round), and I would have kept the community and private gardeners as is because it was necessary to “collect” the data with them (i.e, looking at fertilizer bags and measuring gardens, and for all info to be fresh in their minds (that was a big reason we did collection for 2013 even though my real reference year was 2012).  I can’t insist enough about actually going to see the fertilizers used and writing down the ratio. I still haven’t come up with any better ideas about how to get people to return my calls or actually answer all my questions though. Perhaps some kind of incentive structure…..

For point 4: I would have just gone with paper survey’s and digitize right after on a normal computer (which is what I ended up doing at the end of the season). Also, back-up really frequently.

For point 5: Check in with field assistants way more than I thought was necessary at the beginning instead of realizing later on things were not done as they should of and having to redo them. I guess I would extent this to collaborators as well if you are relying on them for data collection.