Chatbots For Social Change/Prototypes/McGail/Beliefs and Why

= Summary =

The idea is to keep things incredibly simple, by querying for beliefs, understandings, and identities of conversants, and asking them to clarify what that means to them. This sets up a constrained environment to solidify some aspects of the storage of belief systems, working with possibly conflicting belief systems, and bouncing some off others.

The original research design consisted of the following questions:
 * 1) Which of the following do you identify as? Choose the option which most impacts how you view the world.
 * 2) What does it mean for you to be a [response from Q1]? That is, what do you believe which makes you a [response from Q1] ? Please give three distinct answers.
 * 3) You just mentioned [Statement 1]. Could you say a bit more about what this means?
 * 4) You just mentioned [Statement 2]. Could you say a bit more about what this means?
 * 5) You just mentioned [Statement 3]. Could you say a bit more about what this means?
 * 6) Are there any specific life experiences that have influenced your identity? If so, when and how?
 * 7) Now we would like you to tell us two things that come to mind when you think about these other groups that you don’t identify with. There is no right or wrong answer; you may think of particular words or phrases, or perhaps nothing at all. [List of ideological categories that the respondent didn’t choose]
 * 8) Keyword 1: _____________; Keyword 2: _____________ [for each category listed]

There were other versions of this, but the basic premise is to elicit iterated clarifications, to try to "get to the bottom" of their understandings of their belief systems. The study puts reported understandings, discursive explanation, as primary data to be explained. I think something like this could prove a nice model for a chatbot-based data collection scheme, but need to reflect more on what to do and whether it should be worth it.