Choosing the appropriate statistical test- please help This seems really simple, but I need some help choosing a statistical test. I conducted an experiment where I was trying to figure out what food 20 crickets preferred, potatoes or sugar. I put the mice directly between the two types.  17 went to the sugar and 3 went to the potatoes. I'm not sure if I should do a binomial test, a t-test, or something else. What test would be the best to determine which diet the crickets preferred? If you would please state the null and alternative hypothesis, that would really help. If my data needs to be formatted differently, that would also be very helpful to mention.
I use RStudio for my statistical testing in case it is relevant.

 A: The null hypothesis will depend on what made you perform the experiment. Did you start out with the idea "mice are stupid animals who do not care what they eat" or with the idea "Sugar is sweet. All mice are going to eat the sugar, not the potato" or with the idea that "Lab animals are fed with potatops when they grow up. None of them will eat anything else then potatos".
It is a binomial decision the mice have to make so a binomial test is appropriate. Depending on the considerations above you will have to set a probability p that defines the null hypothesis. Let's say, the null hypothesis was, that animals ate both options equally likely, so probability p = .5
In R:
> binom.test(c(3,17), p = .5)

Exact binomial test

data:  c(3, 17)
number of successes = 3, number of trials = 20, p-value = 0.002577
alternative hypothesis: true probability of success is not equal to 0.5
95 percent confidence interval:
 0.03207094 0.37892683
sample estimates:
probability of success 
                  0.15 

