1. ## what test?

how do you solve this question
A farm shop had two identical bins, one containing GM sweet corn and one containing
non-GM sweet corn. Over the season, it sold 8160 ears of GM corn and 5430 ears of non-GM corn.
If we assume that each individual customer bought a dozen ears at a time, can we infer that people
prefer GM corn?

i know that 680 people bought GM corn and 452 people bought non GM corn.

but what test can we perform with just two information?

2. Originally Posted by alexandrabel90
how do you solve this question
A farm shop had two identical bins, one containing GM sweet corn and one containing
non-GM sweet corn. Over the season, it sold 8160 ears of GM corn and 5430 ears of non-GM corn.
If we assume that each individual customer bought a dozen ears at a time, can we infer that people
prefer GM corn?

i know that 680 people bought GM corn and 452 people bought non GM corn.

but what test can we perform with just two information?
Under the null hypothesis that there is no preference the mean number who bought GM is 566, and the SD of the number that bought GM is ~=16.8 and the distribution is B(1132,0.5)

Now use a one sided z-test (the numbers are large, so it is OK to assume normality, but use a continuity correction anyway).

(in fact there is no need for a formal test, the z-score is ~6.8 so we reject the null hypothesis without any further thought)

CB