April 18 - 1 2 It clusters around a mean of 0, as predicted...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
April 18, 2008 IV. Steps for significance test o Write H 0 and H 1 o Select an alpha level, and find z critical or t critical o Calculate t or z; compare to the critical value of z or t o Decide to reject the null hypothesis or not, and interpret Significance testing for two population means (independent sample means) We use this test when we have: o A continuous dependent variable (either interval or ratio level) o We compare the mean of the dependent variable across two independent groups o Examples: Salaries for men vs. women Average re-arrest rate for youth in juvenile court and in adult court Begin with the null hypothesis o Assume that the two groups’ population means are equal – this is H0 H 0 : μ 1 – μ 2 = 0 Same as saying μ 1 = μ 2 μ = population average The sampling distribution again o We think of a sampling distribution of μ
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1 2 It clusters around a mean of 0, as predicted by H We calculate the standard error of the difference between means Calculating the standard error o There are 2 different formulas to use, depending on whether 1 2 (population variance for group 1) = 2 2 (population variance for group 2) o We will use only one assume that 1 2 2 2 Unequal standard deviations o For unequal population variances: 1 - 2 = ( (s 1 2 / n 1-1) + (s 2 2 / n 2-1)) o The formula for t is: t obt =( 1- 2 )/ ( (s 1 2 / n 1-1) + (s 2 2 / n 2-1)) Degrees of freedom o For unequal standard deviations, it can be more complicated o DF=n 1 +n 2-2 is still a good estimate (contrary to what your book says)...
View Full Document

Ask a homework question - tutors are online