Statistics Flashcards
Goodness of fit for Poisson distribution
H0: modelled by Poisson H1: not modelled by Poisson Find mean X | P(X=x) | n x P(X=x) | fe V = classes - parameters -1 X^2 > critical value Reject H0
Geometric distribution
Number of trials till success Same probability P(X=r) = q^r-1 x p Ex = 1/p Varx = 1-p / p^2
Residual equation
Y1 - (a + bx1)
Spearman’s rank
Rs = (1 - 6 sum di^2) / n(n^2 - 1)
Association
Ex and varx of function X
E(g[x]) = integral g[x]f(x)dx Var(g[x]) = integral (g[x])^2 f(x)dx - (E(g[x]))^2
X + Y Poisson
X + Y distributed Poisson( x + u)
Goodness of fit for binomial
H0: can be modelled H1: not modelled Find n and p Fe is less than 5 merge cells If X^2 > critical value Reject H0
What critical values for a Wilcoxen
For two tailed take smaller one
One tailed , test if population mean is greater than M take W- and W+ for population mean less than M
Accept H1 if W is less than or equal to critical value
When using a cdf how do you find area between two points
Take away Fx - Fx1
Median for cdf
Fm = 0.5
What is the uniform distribution
Same distribution over value
Proof for ex of uniform
N+1 /2
Proof of variance of uniform
N^2 -1 /12
Describe geometric distribution
Probability of a success and not a success