Idea: A study is called “The Language that Gets People to Donate: Phrases that Predict Financial Donations to Children in Need.”
My predictive model will test whether certain words used by in-person brand ambassadors predicts whether or not a random passerby will commit to sponsoring a child for 12 months.
To control variables, I can provide a very strict script to my brand ambassadors, control for the types of ambassadors I send out, and control for the audiences they will be approaching. I will always use the profile of the same child to tell the story.
I will quantify my results by measuring the number of people who ignore the brand ambassador, the number of people who walk away within 2 minutes, and the number of people who end up committing to a sponsorship.
My variables will include the sales giftedness of my brand ambassadors as that can vary widely, but I will attempt to control for that before it becomes too large of a variable that we can’t wrangle.