Covariates & coefficients

What environmental or survey variables should you record?

Including covariates in your analysis allows you to estimate detectability and density more accurately

View slides in full screen

When designing your field surveys and analysing your data, you need to consider which covariates to include.  This is a decision that you should make before going to the field, because the complexity of your models influences the number of sites you need to visit, and potentially the number of replicate surveys as well

We discuss this further in the Survey design course

🤔  Reflection: Which covariates should you measure?

1. If you’re studying with our Chinese water deer dataset, consider what might affect Chinese water deer in this coastal wetland
2. If you’re working with your own data, answer these questions for your own work

• What might affect the density of your species of interest?
  • Which of these density covariates could you measure during your line transect surveys?
  • What might you be able to extract from maps or remotely-sensed imagery using GIS?
  • Are these factors categorical or continuous?  Which measurement scale do they use?
• What factors might affect your ability to detect the species?
  • Would these detectability covariates vary by transect or at different times?
  • How might you quantify them during your surveys, so you can test your hypotheses on what influences detectability?
  • Are these factors categorical or continuous? Which measurement scale do they use?
• What hypotheses can you generate about density and detectability involving the covariates you’ve identified?

💬 Discussion: Covariates

Share you answers in our discussion area