Exam 2 Review Concepts.docx - 1 Review Concepts t Test for...

1
Review Concepts
t Test for Related Samples Concepts
Used when you want to compare to means measured on the same people
two different times (Time 1 and Time 2). The variable must be scale (i.e.,
interval or ratio).
Repeated-measures research is needed when the data violates that
assumption of independence of errors. This happens when:
 You study the same individuals over time. That is, you have measures
for the same individuals for more than one point in time. Longitudinal
data.
 Matched designed studies where the individuals in one sample are
matched to individuals in another sample. Thus, the selection of one
individual is contingent on the selection of another individual.
Assumptions
• Dependent variable must be scale (interval or ratio)
• Random sample
• No outliers
• Assumption of normality
• Assumption of homogeneity of variance
Effect size R 2 (r2 = 0.01 small effect, r2 = 0.09 medium effect, r2 = 0.25
large effect)
Confidence interval
μ
D
= M
D
±t s
M
D
APA Style
t (degrees of freedom) = obtained t value, p < .05 if reject null hypothesis or
p > .05 if fail to reject null hypothesis
t(7) = 3.61, p <.05
ANOVA Concepts
Analysis of variance (ANOVA) – analyze the relationship between a
dependent variable that is interval/scale and an independent variable (i.e.,
factor) that is ordinal or nominal with more than two categories.
Benefit of ANOVA over multiple t-tests is that it controls for experimental-
wise error (that is inflating the test-wise error when doing multiple tests).

2
For ANOVA the difference between means is the descriptive statistics
component and the F-test (f distribution) is the inferential component.
ANOVA using variance within groups to model natural variation (or variance
due to chance). If the between-group variation is substantially greater than
the within-group variation, then the difference between the means is unlikely
to happen by chance and can be attributed to the treatment (or difference
between the groups).