Summa Week 9 Part II Flashcards
SStotal total variance in the data
= SStreatment variance due to the treatment + SSerror errors in model
ANOVA is an Omnibus test that tests for overall…
difference between groups
ANOVA is an Omnibus test tells us that group means are…
different
ANOVA is an Omnibus test that doesn’t tell us exactly where…
the significant difference lies
What else do you use with ANOVA?
priori comparisons and/or post-hoc tests
What are the 5 posthoc tests?
Bonferroni Fisher's LSD Scheffe Tukey's HSD Dunnett's C
Tukey’s HSD stands for
Tukey’s honestly significant difference test
What is another name for Studentized Range Statistic?
Tukey’s Honestly Significant Test
What is another name for Tukey’s Honestly significant test
Studentized Range Statistic
Tukey’s HSD test results in a ____ number HSD
single
HSD determines the _____ difference between treatments that is necessary for ______ _______
minimum
statistical significance
In psychology, Tukey’s HSD test is probably the most ____ ______ and _____ commonly used post-hoc test
most well understood
most commonly used
Tukey’s HSD is conservative or liberal?
conservative, i.e. the familywise error was smaller than Bonferroni
HSD formula =
q x square root of (MSwithin/n)
HSD formula: has q that means
the table q is such that r refers to means that are k steps apart, where k is the number of treatments
MSwithin is found where?
from ANOVA in SPSS
n refers to for HSD formula…
number of scores within each treatment
We reject the null hypothesis if HSD < …
/x_1 = x_2
The Bonferroni adjustment formula is
divida a by the number of comparisons (c) and require that each test be significant at that level (.05/c) it is a bit conservative
What is strongly recommended for Bonferroni adjustment?
when you are making all possible pairwise comparisons among means. It is much too conservative in that case.
When is Bonferroni best used?
when you only want to make a few comparisons
BSD formula
t(a/c)/2 x square root of (MSwithin x (1/n1 + 1/n2))
If a = .05, c = 3, and there are 12 degrees of freedom, what is BSD?
t(.05/3)/2) = 2.681 MSwithin = 4.5 n1 = 5 n2 = 5 =2.681 x square root of (4.5 x (1/5 + 1/5)) = 3.60
Fisher’s LSD means?
Fisher’s least significant difference procedure
Fisher’s LSD is the most conservative or liberal?
liberal, and is the only test that requries a significant overall F-value before continuing
What posthoc test is the only test that requires a significant overall F-value before continuing?
Fisher’s LSD
LSD formula
t x a/2 x square root of (MSwithin x (1/n1 + 1/n2))
What is Dunnett’s C?
a pairwise comparisons test based on the studentized range
what pairwise comparisons test is based on the studentized range?
Dunnett’s c
Which posthoc test is appropriate when the variances are not equal?
Dunnett’s c
what is an advantage of a priori procedures?
limits the alpha
If the a priori comparisons are planned, then you test them ______ any correction
without
Each ___ test for the ________ comparison is treated like any other ___ test. You look up an ___-critical value in a table with df____ and df_____
F F F comp error
T-test for a priori uses MS____ and t_____ at df_____
error
critical
error
What a priori test do ou use with correction?
Bonferroni or Dunn’s test
What is a common requirement with Bonferroni and Dunn’s tests?
they both require equal sample sizes
What are orthogonal contrasts?
comparisons that are a priori and hypothesis driven, enabling the researcher to analyze similar groups as one against other (uses fractions to discern the degree to which each are represented in the comparison
What is the first rule when choosing contrasts?
constrasts must not interfere with each other (they must test unique hypotheses)/independent
What is the second rule when choosing contrasts?
each contrast should compare only 2 chunks of variation
What is the third rule when choosing contrasts?
you should always end up with one less contrast than the number of groups
If you are testing the effects of Viagra on libido using three groups, what are some possible contrasts?
1) only treatments with low or high doses
2) placebo vs treatments
How are contrasts mapped out according to what we’ve learned?
SSm variance explained by the model =
low + high dose variance explained by experimental groups (that are high and low - contrast 2), vs.
placebo variance explained by control group (contrast 1)
When creating orthogonal contrasts, groups coded with positive weight are compared to groups coded with…
negative weights
The sum of weights for a comparison should be…
zero
If a group is not involved in a comparison, assign it a weight of…
zero
For a given contrast, the weights assigned to the group(s) in one chunk f variation should be equal to…
the number of groups in the opposite chunk of variation
If a group is singled out in a comparison, then that group should…
not be used in any subsequent contrasts
What is a linear trend in trend analysis?
an upward line
What is a quadratic trend in trend analysis?
a hill-type line
What is a cubic trend in trend analysis?
a line that goes up and down and then up again
What is a quartic trend in trend analysis?
a line that goes down and then waaaaaaay up, and then down, and then up (shape of a w)
Do you use Levene’s test for trend analysis?
hell ya
Do you use Levene’s test for contrast tests?
hell ya
T or F: The grand mean is the mean of all observations across all groups.
true
It is impossible to obtain a negative value for an F-ratio
it’s very rare, but it is still possible/false
we wish to test the hypothesis of no difference between the means of two dependent samples. Tehre are 30 cases in the first sample, and 30 cases in the second. The number of degrees of freedom for this test will be
a. 28
b. 29
c. 58
d. 59
a. 28
A research report describing the results from a repeated-measures study includes the following: “The data show no significant difference between the two treatments, t(10) = 1.65, p > .05.” Based on this report, you can conclude that a total of ____ individuals participated in the research study.
a. 9
b. 10
c. 11
d. 12
d. 12
