Class Notes
(810,913)

Canada
(494,374)

Carleton University
(17,379)

Psychology
(2,647)

PSYC 3000
(97)

Deanna C Whelan
(2)

Lecture

# Week 7 Practice Questions - power calculations and multiple comparisons (1).pdf

Unlock Document

Carleton University

Psychology

PSYC 3000

Deanna C Whelan

Summer

Description

Week 7 Practice Questions – Power calculations and multiple comparisons
For questions 1-5, indicate whether the statement is a) True or b) False.
1. Reducing the PC error rate in a family of tests makes each test less powerful.
2. For 2 or more pairwise comparisons between treatment means, the familywise error rate is
less than the per-comparison error rate.
3. If an independent variable has 4 treatment levels, there are 8 possible pairwise comparisons
between treatment means that can be tested.
4. If there is no correlation between the two groups of a matched-pairs design, the formula for
the required sample size for a given power is the same as the formula for an independent-
samples design.
5. The coefficients in any given contrast must always add to 1.
6. You have just completed a study with the following four treatment means: M =2, M = 3, M
1 2 3
= 5, and M 4 8. Calculate the value of the contrast specified by the coefficients {-1, -1, -1,
+3}.
a) = -2
b) = -14
c) = +2
d) = +14
7. Referring to the previous question, if a t-test is performed on the specified contrast, what is
the resulting t-value? (assume n = 3 and pooled= 3.5)
a) t = 1
b) t = 2
c) t = 3
d) t = 6
8. Using Bonferroni’s formula, what PC error rate should be used to control the FW error rate
to 5% in post hoc tests among 7 means?
a) .014
b) .005
c) .007
d) .350
9. You’ve just completed your thesis. Your results indicated that a treatment reduces panic in
students in statistics courses from a mean of 5 panic at

More
Less
Related notes for PSYC 3000