Welcome

We are faculty who study ethology, or animal behavior. Our goal is to understand how animals perceive their world, and we typically work with captive animals. Our research projects range in size from those involving only a few people to teams of 20+ graduate and undergraduate students. We are committed to quality data collection. The goal of this site is to provide resources and insights into how we approach reliability in our research projects. ​These principles apply broadly to animal science, veterinary science, and animal behavior research, and are not limited to direct observations of what animals do.

What is reliability?

Reliability is a way of evaluating similarity of measures or techniques, and can apply to many aspects of the research process. We use reliability as a way to confirm that our methodology can be understood and uniformly applied by all members of our team to reduce potential bias.​

While we use the term “reliability” to describe this concept throughout this site, you may have heard of this idea as “consistency”, “agreement”, or “repeatability”. These terms describe related ideas, of evaluating similarity, but have different connotations. “Consistency” and “agreement” both have statistical meanings, and we limit our use to those contexts. “Repeatability” accurately conveys our goal of scientific reproducibility; however, this term is not common in some circles, and may create confusion. For this reason, we use “reliability” throughout this site, with the caveat that it is not used as a referendum on personal traits and attributes (e.g. a researcher that is not reliable in this context does not mean that the researcher is not trustworthy).

Why does reliability matter?

Reliability allows us to ensure uniformity of the approaches our teams use throughout data collection: before starting, during, and afterwards. This process allows for 3 key improvements to our science:

Better definitions

We aim to refine and improve our initial definitions or criteria. We find that repeated testing and modification at the start of the research project results in more robust outcomes. ​

​Better processes

We find that the training of observers or experimenters helps us identify weaknesses in our experimental procedures and allows us to evaluate if these have been addressed.

Better reporting

Formally evaluating reliability demonstrates consistency and rigor in our scientific papers, in a way that can be replicated by others. ​

We are committed to quality and rigor in our science. We also believe that this process should be transparent, and work to implement practices associated with Open Science in our laboratories.

Next Steps

• 1. Foundations
• 2. Diving Deeper
• 3. Checklist
• 4. Resources

<

>

We recommend beginning here for the basics about why and how we look at reliability.

Let's go ->

Already familiar with the basics? Start here for a deeper look at how reliability can differ by data type, along with insights into troubleshooting and reporting your reliability process. ​

Let's go ->

Ready to incorporate these tactics into your own research process? Look here for a checklist of key considerations when preparing, running, finishing, and publishing your research project.

Let's Go ->

We provide additional information about reliability, including textbooks, papers, and example tests in this section.

Let's go ->

​Please note that this site is best viewed on a computer screen; we recommend Chrome for the best viewing experience.