Chapter 3
Measures of Central Tendency and Variability
Advantages and Disadvantages of the Major Measures of Central Tendency
Advantages of the Mode:
1. Easy to find.
2. Can be used with any scale of measurement.
3. The only measure that can be used with
answer can be obtained directly from the F ratio:
2
RM =
df RM F
3 (2.42)
3 (2.42)
7.26
=
=
=
= .377
3 (2.42) + 12
3 (2.42) + 12
19.26
df RM F + df int er
This is a very large proportion, but it can be considered misleadingly large in that the error term
Chapter 15
Repeated Measures ANOVA
The same RM ANOVA procedures that are applied when each subject is measured several times
are used to analyze the results of a randomized-blocks (RB) design, as well. Imagine that a consumer
organization wants to answer
Optional Material: Two-Way RM ANOVA
Suppose that each subject in a study of reading methods is tested on two types of reading materials,
and therefore provides two reading scores. You would then have two within-subject factors: the
method factor, which in
sphericity assumption. The first of these, known as the Greenhouse-Geisser (G-G) correction, is by far
the more conservative and more popular of the two; the Huynh-Feldt (H-F) correction is rarely used.
The fourth variation uses what can be referred to as
Chapter 15 Section D
Repeated-Measures ANOVA
There is a fundamental difference between the ways in which data are entered for a repeated-measures
design and for an independent-groups design. When the same subjects are measured at all the levels of
your IV
compared to .
Curvilinear correlation: Occurs when two variables are related in such a way that the scatterplot appears
as a curve instead of a straight line. There are coefficients of curvilinear correlation that are sensitive to
such relationships.
Biva
Chapter 9
Linear Correlation
In order to review the calculation of Pearson=s r, I will describe a hypothetical study to determine
whether people who come from large immediate families (i.e., have many siblings) tend to create large
immediate families (i.e
Chapter 8
Statistical Power and Effect Size
Power analysis can be separated into two categories: fixed and flexible sample sizes
1. When the sample sizes are fixed by circumstance (e.g., a certain number of patients available with a
particular condition),
Two-group Confidence Interval
Although you would probably not bother to calculate a confidence interval for the above experiment
because of the small sample sizes and lack of statistical significance, we will find the 95% confidence
interval just for revi
Chapter 7
The t Test for Two Independent Sample Means
To review the statistical analysis of a two-group experiment, I will describe a hypothetical study from
the field of neuropsychology. A researcher believes that a region in the middle of the right hemi
Chapter 6
Interval Estimation and the t Distribution
One-Group t test
To illustrate the calculation of a one-group t-test and the construction of a confidence interval, we ask
you to imagine the following situation.
An anthropologist has just discovered a
Step 6. Make the statistical decision.
If the z-score you calculated above is greater in magnitude than the critical z-score, then you can reject
the null hypothesis. As alpha had been set to .05, we would say that the results are significant at the .05
l
Chapter 5
Introduction to Hypothesis Testing
Hypothesis testing can be divided into a six-step process. The six steps are as follows:
1.
2.
3.
4.
5.
6.
State the hypotheses.
Select the statistical test and the significance level.
Select the sample and col
Adjacent values: The upper adjacent value is the highest score in the distribution that is not higher than the
upper inner fence, and the lower adjacent value is similarly defined in terms of the lower inner fence of a
boxplot. The upper whisker is drawn
The Range. The range is the highest score minus the lowest. The number of books read ranged from 1 to
9, so range = 9 - 1 = 8. If the scale is considered continuous (e.g., 9 books is really anywhere between 8 1/2
and 9 1/2 books), then range = upper real
fatigue, and can be averaged out by counterbalancing.
Latin-Square design: A system for counterbalancing in which the number of different sequences of
treatment levels equals the number of treatment levels. Each treatment level appears once in each ordina