Dear Allstaters,
Could anyone help me with the following problem ?
I have k variables (let say X1, X2, ...,Xk). Each of them have 3 possible
outcomes, i.e. 1 (favorable), -1 (unfavorable), and 0 (for no answer).
I calculate a total (composite) net preference as an average of (X1, X2,
...,Xk) divided by total samples (N), i.e. p = sum(Xij)/(N.k)
i = 1,2,...k and j = 1,2,3,...,N
How do I test a hypothesis that p = 0 against p not equal to zero ?
I don't have any prior information whether X1, X2, ...Xk are statistically
independent.
Thanks in advance.
Satrio Wiseno
PT. Grup Riset Potensial, Indonesia
|