Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
58 Cards in this Set
- Front
- Back
|
Spearman correlation and Pearson chi-square are two examples of …
|
non-parametric statistics
|
|
|
Non-parametric statistics should be typically used when the dependent variable is ____
|
ordinal with 4 or fewer points; ranked data; nominal; highly skewed (>2.0) (4 marks)
|
|
|
Non-parametric tests are not as _____ as parametric tests.
|
strict
|
|
|
Non-parametric tests do not test hypotheses about …
|
mean differences.
|
|
|
Technically, in the context of group differences, the non-parametric statistic null hypothesis is that the data have been sampled from …
|
identical populations
|
|
|
Non-parametric statistics assume that the _____ of the distributions are homogeneous.
|
shapes |
|
|
Kruskal-Wallis assumes _____ of variance.
|
homogeneity
|
|
|
In the context of non-parametric statistics, there is a _____version of Levene’s test of homogeneity of variance.
|
median-based
|
|
|
Technically, we should not apply parametric statistics on data scored on _____ scales.
|
Likert |
|
|
If data are Likert in nature and have ______ points in the scale, it may be argued to be justifiable to apply parametric statistics to them.
|
5 or more
|
|
|
Non-parametric statistics should always be applied to ____ data.
|
ranked |
|
|
Although simulation research shows that the accuracy of parametric statistics is fairly good in cases where the ordinal scale has 5 or more points, this research assumes that _____, which is ___ in practice.
|
all points in the scale are used; often not the case (2 marks)
|
|
|
The Kruskal-Wallis test is based on transforming the original dependent variable data into ______.
|
ranks.
|
|
|
Strictly speaking, the Kruskal-Wallis null hypothesis is that the groups have …
|
identical population distributions
|
|
|
Differences in group ______ may be implied by a statistically significant Kruskal-Wallis statistic.
|
mean ranks |
|
|
The notion that a difference in group medians may be implied by a statistically significant Kruskal-Walliss statistic is based on …
|
strict assumptions being met.
|
|
|
In order to test the assumption of normally distributed data, many people use a statistic called …
|
Shapiro-Wilks.
|
|
|
The problem with statistics such as Shapiro-Wilks for testing distributions for “sufficient normality” is that they are …
|
excessively powerful
|
|
|
In order to get SPSS to conduct the median-based version of Levene’s test of homogeneity of variance, one must use the ____ utility and then select the _____ option.
|
Descriptives --> Explore; power estimation within the Plots window (2 marks)
|
|
|
Kruskal-Wallis is essentially based on ____ and then _____the summed ranks within each group.
|
summing; squaring (2 marks)
|
|
|
What effect size estimate does SPSS calculate for Kruskal-Wallis?
|
None
|
|
|
Effect size for Kruskal-Wallis can be calculated by…
|
dividing the chi-square value by N -1 (2 marks) |
|
|
Like ANOVA, Kruskal-Wallis is an ______ statistic.
|
omnibus |
|
|
Typically, if a researcher obtains a statistically significant Kruskal-Wallis, the researcher will follow-up with a series of …
|
Mann-Whitney tests.
|
|
|
The Mann-Whitney test should be ____, given the existence of _____.
|
dead; Kruskal-Wallis
|
|
|
The Mann-Whitney test is a non-parametric equivalent of the …
|
independent samples t-test.
|
|
|
The Mann-Whitney test tests the null hypothesis that the _____ are equal.
|
mean ranks
|
|
What is the Mann-Whitney value associated with the SPSS table below?
|
6.00
|
|
From what analysis is the below window in SPSS?
|
Mann-Whitney
|
|
From what analysis the below window in SPSS?
|
Kruskal-Wallis
|
|
|
What estimate of effect size does SPSS calculate for Mann-Whitney test?
|
none. |
|
|
What is the eta squared formula for the Mann-Whitney test?
|
Z squared divided by N-1 (2 marks)
|
|
|
An alternative to conducting a series of Mann-Whitney U tests as a follow-up to a significant Kruskal-Wallis is to conduct a …
|
Jonckheere-Terpstra test
|
|
|
The Jonckheere-Terpstra test is really just a…
|
trend analysis for mean ranks.
|
|
In the SPSS window below, what two tests are missing from the ‘Test Type’ box?
|
Kruskal-Wallis; Jonckheere-Terpstra (2 marks)
|
|
|
The effect size estimate reported by SPSS for the Jonckheere-Terpstra test is referred to as…
|
there is none.
|
|
|
An effect size statistic that can be calculated by hand based on the Jonckheere-Terpstra statistic is known as….
|
eta squared
|
|
|
The following formulate can be used to calculate effect size for the Jonchkeere-Terpstra statistic….
|
Z2/(N-1)
|
|
|
Friedman’s ANOVA does not assume ______ data.
|
normally distributed |
|
|
Friedman’s ANOVA can be used on data that conform to a ____ factor.
|
within-subjects |
|
|
One of the limitations of the Friedman ANOVA test is that there is no _____ version of it.
|
factorial |
|
|
Given the existence of Friedman’s ANOVA, _____ should be dead.
|
Cochran's Q |
|
|
Referring to Friedman’s ANVOA as an ANOVA is a …
|
misnomer. |
|
|
The Friedman ANOVA statistic converts the dependent variable into …
|
ranks. |
|
|
When you really get down to it, Friedman’s ANOVA may be considered a test of ...
|
consistency. |
|
In the SPSS window below, what two tests are missing from the ‘Test Type’ box?
|
Friedman’s ANOVA; Cochran’s Q (2 marks)
|
|
|
The formula to estimate effect size in the Friedman ANVOA case is…
|
chi-square/((N)(k-1)) (2 marks)
|
|
|
In the context of conducting a Friedman’s ANOVA, the effect size estimate is known as ….
|
Kendall’s W
|
|
|
Given that _______ is a measure of effect size, and you can get exactly the same p value for this effect size in SPSS, there is really no need to conduct a …
|
Kendall’s W; Friedman’s ANOVA (1 mark; yes, 1 mark)
|
|
|
The ______ is a non-parametric equivalent of the dependent samples t-test.
|
Wilcoxon-Signed Rank Test
|
|
|
The ______ could be used as a post-hoc test in the Friedman omnibus statistic scenario.
|
Wilcoxon-Signed Rank Test
|
|
|
Given the existence of Cochran’s Q, ____ should be dead.
|
McNemar chi-square
|
|
|
Given the existence of Friedman’s ANOVA, _____ should be dead.
|
Cochran’s Q.
|
|
|
Many non-parametric statistics do in fact assume …
|
homogeneity of variance.
|
|
|
Non-parametric statistical analyses are often _____ than the corresponding parametric statistical analyses.
|
less powerful
|
|
|
One of the main limitations associated with non-parametric statistics is that you can’t use them to test …
|
interactions.
|
|
|
In practice, all too often, non-parametric statistics are not actually the solution to our …
|
statistical problems.
|
|
|
Between-groups ANOVA is to Kruskal-Wallis as repeated measures ANOVA is to…
|
Friedman ANOVA
|
