When I was a kid, one of my favorite summer activities was staring at hummingbirds. I would sit for hours, moving as little as possible, while I took notes about everything I saw. (Yes, I was a pretty weird eight year old.) I wanted to ask the hummingbirds so many questions, but I don’t speak hummingbird! Observing them was my only option for trying to understand their behavior.
While it is literally impossible to ask a hummingbird to take a survey, there are many times with humans when a survey won’t work to collect the data you need either. Observation can be a great data collection tool when you want to see how different people interact with each other, a space, or a passive program. Observation is also helpful when it would be difficult for someone to answer a question accurately, like when you ask them to remember what they did or, particularly with children, if you ask them to give critical feedback or written feedback, both of which are sometimes developmentally inappropriate.
In this post, I’m going to talk about why you might choose observation as a data collection method. Next time, I’ll talk about the logistics of observations and how you can use observational data. To better understand why you would collect data with observations, let’s use our example evaluation question from throughout this blog series: “Does attending storytime help caregivers use new literacy skills at home?”
When we first outlined number data and story data, we talked about when to use each. We also outlined how to break your research question down into smaller questions. You really need to do that work to get to this point, so let’s go back and review what we did. Here are some of the sub-questions we identified within our larger evaluation question:
- Were caregivers already using literacy skills at home prior to attending a storytime?
- Are caregivers learning new literacy skills during storytime?
- Do caregivers use new literacy skills from storytime at home?
Would a survey work to collect this data? We certainly could ask caregivers all of these questions. But we would immediately bump into some of the problems that come up when people self-report data: 1) we are not great at remembering things accurately and 2) we want to portray ourselves in the best possible light (social desirability bias). Let’s take a look at how those challenges would impact the data collection for our questions.
- Were caregivers already using literacy skills at home prior to attending a storytime?
- They may not know or accurately remember which skills they knew before attending storytime and which they learned at storytime.
- Are caregivers learning new literacy skills during storytime?
- They may report that they are learning new literacy skills at storytime because they don’t want to hurt anyone’s feelings—even if they aren’t actually learning those skills.
- Do caregivers use new literacy skills from storytime at home?
- They may report that they are using new literacy skills at home because that feels nice to say—even if they aren’t actually using those skills at home.
So we could collect that data using a survey, but it may not be very accurate. We could get more accurate data by observing caregivers at home with children before they ever attended a library storytime and then continuing to observe after they started attending storytime. Then we could see for ourselves what skills they already knew and used at home, and which ones they learned at storytime. We could tally up how often they were using those skills too. Great! Let’s go follow people and their children around their homes 24 hours a day taking notes for several months.
What? You don’t think that’s going to be a thrilling success? Unlike hummingbirds, who don’t seem to mind too much or alter their behavior a lot while I am watching, humans mind quite a bit and can change their behavior when they are being observed. Additionally, do you know any library staff who have the time to do this kind of intense observational study? Yeah, that’s what I thought. The time involved in observation and successfully navigating privacy concerns are two major elements that you always need to consider.
What can we do that’s a little more realistic? Collecting data in the real world is often about doing what you can with what you have. In this case, it is unlikely anyone would let us come follow them around their home. We can, however, more easily observe caregivers and their children in the library. This would allow us to observe for indicators of caregivers learning skills during storytime and to observe if families are using early literacy skills while they are spending unstructured time in the library. Intrigued as to how we would do that? Come back for our next post where we’ll get into the nuts and bolts of how you can collect data using observation and pull out important takeaways from that data.
If you are an aspiring birdnerd, the two hummingbirds pictured are both species we have in Colorado and you can learn more about them here.