Chapter 12

Question 1

hypothesis

Answer 1

always statements about population parameters, not samples

Question 2

null hypothesis

Answer 2

• hypothesis being tested
• Aka: statistical hypothesis
• Can be false, but can never be “true”; can only be “conceivably true” (because we don’t have enough evidence to reject it)
• Even when we conclude it’s false, we can never be 100% sure

Question 3

alternative hypothesis

Answer 3

contradicts H0, the one the researcher is trying to prove true

Question 4

rejecting vs. retaining H0

Answer 4

• Reject: obtained sample statistic has low probability of occurring by chance if H0 is true
• Retain: we don’t have sufficient evidence to reject H0 (it potentially could be true)

Question 5

non-directional/two-tailed tests

Answer 5

• HA states that population parameter is different than the value stated in H0
• Two-tailed statistical tests are more common than one-tailed tests and should be used whenever both mu >K and mu
• if it’s the first time testing something, use 2-tailed

Question 6

directional/one-tailed test

Answer 6

• HA states that the population parameter differs from the value stated in H0 in a particular direction
• Generally only appropriate when it’s unlikely that mu could be less than or greater than K -> more powerful than 2-tail tests in these situations
• if means must differ in one direction only, use 1-tailed (ex. IQ of gifted vs. normal students -> gifted will either have same or higher, not lower)

Question 7

type 1 vs. type 2 errors

Answer 7

• Type 1: data leads us to reject H0 when it’s actually true (alpha)
• Type 2: data leads us to accept H0 when it’s actually false

Question 8

assumptions of the z-test

Answer 8

• A random sample has been drawn from the population
• The sample has been drawn using the “with-replacement” sampling plan
• The sampling distribution follows the normal curve
• The standard deviation of the population of scores is known

Question 9

what varies in the z-test?

Answer 9

• X bar: normally distributed variable (varies)
• Stdv: constant
• Mean: constant

Question 10

t-test

Answer 10

• Used when standard deviation is unknown
• Value of t is not normally distributed – there is a family of distributions
• z and t differ negligibly for large n’s, but when n is small, z and t can differ considerably
• The critical value of t (the minimum t-value at which Ho will be rejected) vary with the degrees of freedom, df = n –1
• The smaller the degrees of freedom, the larger the tcrit

Question 11

p-value

Answer 11

probability when H0 is true of observing a sample mean as deviant or more deviant than the result obtained