Origin of a question

This is story of a question. In the voter survey we (Avidit Acharya, from Princeton University and Morsel team) were trying to get questions like why people vote and how they select party. First we made a question in which villagers were asked to rate parties in 1-10 as per their liking. But we were working in villages where literacy rate is about 50 percent. As it was clear, this question did not work.
Then second day of pilot survey, we tried to get answer according party caste (as BSP and SP are low caste party), it worked sometimes but we were not satisfied.
Next day we draw two questions; one for caste and one for class (we are trying to check a theory of Yogendra Yadav, according to which in Uttar Pradesh, people cast their votes on the basis of class not caste, poor Brahmins and Dalits did vote to poor party i.e. BSP in last assembly election). So, we sketched a ladder on cardboard and tried again but results were not very satisfactory.
In evening we were discussing the questionnaire and at once Avidit said, “Oh yes! Why not use money to get a sense what people think about parties that how they are redistributed public services and money.” Someone said “yes! It can work because even illiterate villagers understand the language of money. And they will do it for us.” “Let’s do that”, Avidit said. And then we invented two questions; one for the different castes and other for the classes. We tried to make this question very simple and conversational so that people could answer in one thought. We arranged three plates, one golden color (for upper class), other stainless steel (for middle class) and third clay plate (for lower class). Along this we put three caricatures (drawn by our team) on plates and then we bought some fake notes (used by kids in a game). That day of pilot campaign was very successful and we said to our team, “GO AHEAD.”