Chapter 14

Question 1

independent sample means

Answer 1

come from two randomly selected samples who are not related, matched on any variables, or representative the same group at 2 different points in time

Question 2

yoked

Answer 2

randomly assigned together

Question 3

properties of sampling distribution of differences between 2 means

Answer 3

• if both population means are the same and H0 is true (the difference is 0), half the means are positive and the other half are negative
• Describes the differences that would occur by chance between X bar and Y bar when H0 is true

Question 4

pairings and potential scores

Answer 4

If we had 3 scores in X and 3 scores in Y, there are 9 possible samples and means that could occur, thus there are 81 possible pairing of means of X with means of Y that could occur – each of them equally likely

Question 5

if sampling distribs for X and Y are normal…

Answer 5

• the means selected at random will also be approximately normal
• Even if not normal distribs, sampling central limit theorem also applies here

Question 6

assumption of homogeneity of variance

Answer 6

• Assumption that the variances in the two populations from which the samples are drawn are the same
• Because of this, it’s better to pool both samples (bigger sample size) and estimate the population variance

Question 7

what happens if sample sizes are quite unequal?

Answer 7

if one sample is much bigger than the other, the standard error of difference between indep means is inflated (and the smaller it is, the easier it is to reject H0)

Question 8

random assignment

Answer 8

• Although random sampling is ideal, random assignment can be used instead, and in most situations will result in the same statistical conclusion, however, it does not allow us to come up with a research conclusion
• Both random sampling and random assignment act as experimental controls

Question 9

assumptions of t-test

Answer 9

• random sampling w/replacement used (if violated, it’s okay - use random assignment)
• raw scores normally distributed (if violated, it’s okay - central limit theorem)
• random sample from X is independent from Y (if violated, not okay, use different equation)
• variance in X and Y are homogenous (depends on whether n is equal… if not, use Levine’s test)