- Zoos: A Historical Perspective from Smithsonian Institution Libraries
The National Zoological Park is an accredited member of the Association of Zoos & Aquariums.
AZA
From Erika Bauer, the Zoo’s first Fujifilm Curatorial Resident in Animal Care
June 2008
Last month I told you about a session I participated in with a group of State Teachers of the Year, and promised to share some more detailed information about my presentation on how behavior research studies are conducted. This information will prove particularly useful to you if you are inclined to develop an animal behavior study, or devise animal behavior-related activities for your own children or students. But even if you’re not planning anything quite so formal or if you're simply interested in this work, read on to learn a little more about animal behavior research. You just might approach your own observations differently in the future. Either way, I hope this is an interesting diary entry for you!
Some Challenges for Animal Behavior Researchers
First, let’s discuss some of the challenges faced by animal behavior researchers. There are a few things scientists must try to avoid:
Anthropomorphism is a common criticism when it comes to interpretations of animal behavior. We need to be sure we are not attributing human characteristics, reasoning, motivation, or emotion to non-human animals. - Observer bias is another potential pitfall. Observers might have preconceived notions about animal behavior, and depending on how data collection observations are structured, those biases might be inadvertently reinforced.
For example, an observer may believe that otters are incredibly playful. And if this observer watches a group of otters, she might focus on the most playful individual of the bunch and record that otter’s behavior, while ignoring the otters that aren’t so playful. One way to prevent this is to decide which individual otters will be observed before of the observation begins. - Observer effects can also be an issue. This is particularly true for animals that aren’t accustomed to the presence of humans. "Observer effects" are the behaviors of the animals are changed by the fact that an observer is present.
I once observed a group of students collecting data on one of the bears at the Zoo. When the observation began, the bear was resting. But the students wanted to see more activity. So they called out to the bear “Hey bear! DO something! Over here! Yeah, move it buddy!” Sometimes I even see parents and teachers doing the same sort of thing. If you are trying to collect data about the behavior of an animal, this is a no-no. The bear got up and moved when the kids yelled but probably wouldn’t have done so if the kids weren’t there causing a disruption. Therefore, the data they collected were not a valid measure of the animal’s behavior under normal conditions.
To avoid observer effects, many field biologists conduct their observations of behavior from blinds, so they are hidden from the animals they are studying. Absolute concealment might not always be an option, but researchers should do whatever they can to minimize the impact of their presence on the animals they are observing.
Keeping all these things in mind, how do researchers collect behavioral data in observational studies of animals?
Using an Ethogram
One of the most common tools in behavioral studies is the ethogram. An ethogram is a list of the behaviors, with each behavior’s corresponding definition, that will be recorded during observations. The purpose of the ethogram is to provide objective descriptions of behaviors that result in high inter-observer reliability. In other words, the definitions must provide enough detail that multiple observers will obtain the same results if they observe the same animal(s) at the same time.
Here’s an interesting ethogram activity:
- Choose an animal for your students to observe and have them come up with a list of behaviors that the animal might exhibit, or simply provide a list to them.
- Using just the list, have each student observe the same animal at the same time and record the animal’s behavior each minute (set a timer so they all record data at the same moment).
- After the observation, have the students discuss what they recorded, and focus on the differences in their data. Why did each student choose the behavior he or she recorded? What aspects of the animal’s behavior was he or she taking into account? A discussion of the disagreements can serve to highlight the importance of having behavioral definitions that everyone understands and agrees on.
- Then, have the students develop definitions for each behavior on the list, and conduct another observation.
- Again, discuss the results, and refine the definitions as needed.
- Continue the process and see how the proportion of agreements versus disagreements changes as the ethogram is refined.
- Once they have a high proportion of agreement (higher than 90 percent is great), they’ve created their own very useful research tool!
Here’s an ethogram tip that I like to share with my students: Remember to include catch-all categories for behavior. In all of my ethograms, I include an "other behavior" category, and a "not visible" category. These "catch-alls" allow for the possibility that the animal you observe might exhibit a behavior that doesn’t fit into any of the behaviors you’ve thought to include in your ethogram, as well as the reality that it is sometimes impossible to confirm what the animal is doing, depending on your viewing angle.
More Complex Methodology
As you get into it, behavioral research methodology becomes even more complex. Researchers must decide which sampling and recording methods they will use to collect data. I’ll describe a few of the most common methods here, but this should not be interpreted as an exhaustive list of options.
In the above ethogram activity example, students used a "focal animal" sampling method. This means they focused their observations on one animal for the entire period of observation. The recording method they used was "instantaneous time sampling," which means they took note of behaviors at particular time points (in the above case, data were recorded once each minute, at the exact moment the timer beeped).
Time sampling recording methods provide estimates of behavior rather than exact measures, whereas continuous recording methods provide an exact measure of behavior for the observation period. If students wanted to collect data on an entire group of animals, they might switch their sampling method from "focal animal sampling" to "scan sampling."
In the above example, if the same group of animals was observed using scan sampling, each time the timer beeped, the students might have scanned the group of animals from right to left, noting the behavior of each individual animal (rather than just the one). Sampling and recording methods vary depending on the study.
The Data Sheet
Another important tool in behavioral research is the data sheet. If observers are collecting data using pen and paper, data sheets should be designed to facilitate appropriate documentation and minimize the amount of time observers must spend looking at the data sheet rather than the animals. Here are some tips for creating good data sheets:
- Create a table that’s easy to navigate.
- Provide a place for all the information you want to record, so you don’t forget anything (if the date, time of day, and/or weather is important to note, have a spot designated for that information on the data sheet).
- Use codes for behaviors to reduce the amount of time spent writing (for example, instead of writing "resting," write "RE"—just make sure all your codes are different!).
- Provide a little "cheat sheet" indicating the abbreviations of each behavioral code directly on the data sheet, so you can quickly confirm the code if you happen to forget the abbreviation.
- Leave a spot for comments and notes at each observation—sometimes strange and unexpected things happen that have an effect on the subject’s behavior and you might want to capture that information.
In conjunction with all of these tools (a well-defined ethogram, well-chosen sampling and recording methods, and a well-designed data sheet) behavioral researchers should arm themselves with patience! Observational studies of animal behavior take time. And animals often behave unpredictably or may be inactive for long periods of time. Impatient observers might find themselves very frustrated when they are unable to collect the data they need to answer their questions right away. But the unpredictability of animal behavior is also part of the fun. And the animals we observe should be appreciated for all of their behaviors, even the ones that don’t seem particularly exciting to us as observers. After all, some of the most insightful information we have about animal behavior was discovered unexpectedly or after huge investments of researcher time. You never know what new insights remain to be found.
P.S. Last month I mentioned the animal 
web cams that can be accessed through the Zoo’s website. You can watch many of our animals in real time from any computer with Internet access. If it’s difficult to take students to the local zoo, why not design an activity for the classroom that involves watching the animals at the National Zoo remotely?
You can observe several species at our Zoo—from flamingos to naked mole-rats to lions. Several of these cameras remain in static positions, so you might not always be able to see animals. But here’s a tip: FONZ volunteers follow the pandas with the panda cam, so that’s your best bet if you want to conduct a remote behavior watch that’s sure to catch an animal on camera.
