From OpenWetWare

VPL        Home        Contact        Trial Data        Lab Members        Visualization        Reports       

Sign up on Amazon MTurk

  • Amazon MTurk
  • Search for "Volunteer Prediction" to find our HIT.

Amazon's MTurk allows us to give you a unique identifier and pay you for your research efforts.

Predicting Volunteering

Hello. My name is Alex Davis. I am a researcher at Carnegie Mellon University studying human behavior related to climage change as part of the Center for Climate and Energy Decision Making. My research involves getting people to change their behavior to reduce carbon dioxide emissions that result from consuming electricity in their homes. For example, in a study we want to conduct, we will give people a device that shows people their household electricity consumption in real time.

We will offer this program to a number of people, and they can freely choose to participate or not. However, if only the people who will benefit from the device participate, and those who do not benefit do not participate, then we will mistakenly think that the device is effective for people in general, when it is actually only effective for people in the study. Unfortunately, getting people to participate in a program like this is hard, and we don't know why some people agree to participate and others don't. If we can build a statistical model to predict who will agree and who won't, then we can control for the fact that some people refuse to participate.

We don't know how to do this. We don't know why some people agree to participate, and others don't. We think that once we know the right questions to ask, or the right measurements to take, then we can develop statistical models to best predict who will participate and why.

This is where you come in. We want you to help us to come up with the right questions, or right measurements, or whatever you can think of, to help us best predict who will participate and why. You can use any information or knowledge that you have to try to come up with things, as long as you tell us what sources you used.

You will be part of a team of collaborators trying to solve this problem. Every couple of weeks we will conduct a trial, where we take the best measurements you and your collaborators can come up with and give them to real people who will decide whether to participate in the study or not, then answer your questions. After every trial, we will give you access to the materials we used, a data summary, and the raw data at this website. This will include the correlations between the measures you suggested and their actual decisions.

If you come up with a good measure or method of predicting who will participate, then you will receive a bonus payment. This research may take several months or even years, depending on whether we can get money to conduct our research.

Recruitment Documents

This link contains all the recruitment documents used. To collect data we randomly sampled people in Pennsylvania and mailed them both the recruitment document and the survey including all of the questions.