RESOURCES
As I started to do research, I noticed there were quite a few things that I needed to know to “successfully” run an experiment, but no one taught me these things. I asked others and did Google searches until I thought I figured it out. Even then, it was difficult to confirm that some things were going to work properly, so it was a bit of crossing my fingers and hoping everything went smoothly. I’ve compiled some of these issues that I ran into and created directions for the steps I took in the hopes that it would be useful for some other graduate student out there. Of course, I am not claiming these are the best approaches—just that this is what I have done and it has been “successful.”
How to Connect Qualtrics to MTurk with CloudResearch
I have put together a document describing how I attach my Qualtrics survey to MTurk and how I setup payment for MTurk workers so that each respondent gets a unique payment code. I do all of this through CloudResearch.
You can find the steps I take here.
How to do a Power Analysis
Before doing any experiment, it’s important to do a power analysis to make sure that your sample size is larger enough to detect a treatment effect if there is one. While power analyses are just estimates and can vary quite a bit depending on the parameters you estimate, the power calculator from EGAP designed by Alexander Coppock is really helpful: https://egap.shinyapps.io/Power_Calculator/
You can find the steps I take and assumptions I tend to make when doing power analyses here.
How To Track Respondents Clicking on URLs in Qualtrics
We often want to have behavioral measures of respondents’ actions in our surveys. One way we can do this is by providing URLs in survey questions and asking respondents to click on them. We can track this behavior in Qualtrics using embedded data, JavaScript, and HTML.
You can find instructions on how to track link clicking here.
