Rule Of Thumb

Question 1

SRS versus subsets

Answer 1

It is acceptable to use a subset rather than a SRS for a study if the size of the population is at least ten times the size of the sample.

Question 2

Most likely value

Answer 2

Of the sampling distribution,B(n,bo), for the statistic on samples of size n under the null hypothesis that b = b0 is nb0.

Question 3

Central limit theorem

Answer 3

States that, given certain conditions, the sum of a sufficiently large number of independent random variables each with a finite mean and finite variance, will be approximately normally distributed.

Question 4

Standard cut-of p-value

Answer 4

P = 0.05

Question 5

P-values close to the critical value

Answer 5

Marginally significant

Question 6

Very small p-values

Answer 6

Highly significant

Question 7

Normal approximation of the binomial distribution

Answer 7

For large values of n and values of b0 close to 0.5, the sampling distribution for sample proportion for samples of size n of a bernoulli trial is very close to the normal distribution. Nb0, b0(1-b0)/n, with mean mu = b0 and standard deviation theta = b0(1-b0)/n.

Question 8

ANOVA

Answer 8

We fit two models to our data - full model (allows the mean response to vary between the different treatment groups) and a reduced model (assumes that the treatment means are the same). If the full model fits the data much better than the reduced model, we conclude that the treatment means really are different.

Question 9

Chi-squared assumptions

Answer 9

Independence and no small expected values.