Home > Preview
The flashcards below were created by user
famay
on FreezingBlue Flashcards.

Effect Size
it estimates the deg to which the null hypothesis is false compared to the alternative null. related to Power, in that if ES is High then power is as well.

Test Statistic
is a special value computed from a sample that is used in deciding btwn the Ho and H1. A zscore.

A P value
is the smallest alpha that would allow rejection of the null. (p <.05)

Parametric Test
The null hypothesis is about a specific population parameter.

Nonparametric test
When the hypothesis specify shape of the population

Null Hypothesis
 1. no change of difference
 2. proposes a specific value for the parameter
 3. helps to construct the sampling distribution

Rejection Region
Values specified by the decision rule. often set by alpha, alpha.05 directional CV1.64 the rejection is anything after this value.

Significance level
is represented by alpha.

Critical Value
is also called a Z score and is the value of the test statistic that begins the rejection region.

Type I error = alpha
Occurs when the Null is rejected when it really is correct. always equals the alpha, so if alpha is .05, you have a 5% chance that your making a Type I error.

Type II error
Is made when the null is not rejected when it really is incorrect. ( beta ) related to power!

1 Step in Hypothesis Testing
 1. Assum
 pop norm dist/samp sz
 samp random w/i
 data  interval /ratio

2 Step in Hypothesis Test
 state the nul  H0 there is no change or diff in the pop an is specific
 state the alternative and direction .

3 Step in Hypothesis Testing.
Test statistic  Z score and st error of the mean.

4 Step in Hypothesis Test
 Decision Rule used to decide which hypothesis provides a better description.
 Choose a Sig level alpha.

5/6 Step in hypothesis
 5 sample and compute
 6 Decide and conclude.

Power
 1(beta) is the probability of making the correct decision of rejecting and incorrect Ho.
 "it's the chance of finding an effect that truly exists"
 Power is incred, by decre (beta)

