How can we make sure we have experimental control in a study?

If something in a study threatens to change the results, we can

1. manipulate it: treating it as a factor and making it a variable.
2. control for it: we cannot treat is as a variable, but maybe we can control it so that we this confounding variables is the same for all.
3. Record/measure it: at least we can evaluate in some ways if it had an impact and in what measure.
4. Hidden effect: maybe we aren’t even aware of its presence.

References

• measurement-errors
• variable-types
• confounding-variables
• measuring-results
• running-an-ab-test