And it was followed by a choice experiment This was a choice experiment with price and taste randomization We're then able to achieve statistical identification through having both price and breed, breed being tastes randomized in this slight randomized characteristics which are in red And this choice experiment involved giving people a choice actually to the survey. They were given a choice between one of two gifts And this gifts was two different discount vouchers that they could use to buy chicken with the vendors that were in the project that was working with So the first choice, we call option one that was always a random discount for safe crossbred chicken By safe, I mean the chicken that was produced the project that I talked about And option two was always a random discount for regular crossbred or local chicken This was just the normal chicken that was sold on the market that didn't have anything to distinguish of So it's an example of how I did randomization First of all, I can randomize the value of the discounts