Exam 2 Study Set

What would this same result for the study described in #8 be if a one-tailed probability were required?

0.045

Which of the following is the conventional level of statistical power that we strive for:

0.80

Which of the following values for an estimation of sphericity would indicate your data is perfectly spherical?

1.0

You conduct an experiment where 20 participants are subjected to all 3 levels of a treatment (placebo, standard of care, experimental treatment) to see if there is any difference in function. What is the dfM ?

2

How many groups were compared in this analysis? Sum of Squares Between Groups: 1582.858 Sum of Squares Within Groups: 2142.488 Total Sum of Squares: 3725.347 df Between Groups: 2 df Within Groups: 45 df Total: 47 Mean Square Between Groups: 791.429 Mean Square Within Groups: 47.611 F: 16.62

3

If you are running an ANOVA to test for differences between 4 independent groups, how many degrees of freedom will the model sum of squares have?

3

In order for sphericity to be an issue in a repeated measures design, what is the minimum number of conditions required?

3

What is the dfR ? (This is also called the residual or error df)

38

What is the total number of subjects in this dataset (ie. N)? Sum of Squares Between Groups: 1582.858 Sum of Squares Within Groups: 2142.488 Total Sum of Squares: 3725.347 df Between Groups: 2 df Within Groups: 45 df Total: 47 Mean Square Between Groups: 791.429 Mean Square Within Groups: 47.611 F: 16.62

47

Which of the following scenarios is most likely to give us confidence that the observed difference in our sample is a true difference and not due to chance?

A large average difference between means with a small standard deviation of differences

When the assumption of sphericity is violated, you should:

Adjust the model and error degrees of freedom accordingly

A t-test can be used for which of the following research design?

All of the above (between group design, within subjects design, and repeated measures design)

A dependent or related samples t-test is used to compare:

All of the above (means from the same group on a measure obtained at two different points in time, means from two different groups that are not independent, and means from two groups where participants are matched on a particular variable)

Once you have calculated your t-statistic, you must compare your statistic to a critical value of the t-distribution. Which of the following is/are needed to determine whether your calculated t-statistic is significant?

All of the above (number of degrees of freedom, whether you tested a directional or nondirectional hypothesis, and your predetermined alpha level)

Which of the following is an advantage gained when using a repeated measures ANOVA as opposed to an independent groups ANOVA?

All of the above (systematic variation is reduced, between group variance is increased, and there is increased sensitivity for det4ecting experimental effects)

When the assumption of sphericity is violated, which of the following are options to deal with the violation?

All of the above are options (adjust the degrees of freedom using Greenhouse-Geisser or Huynh-Felt estimates of sphericity, use the non-parametric Friedman's ANOVA, and use a multivariate test statistic (MANOVA) that does not assume sphericity)

In SPSS, when you find a significant Kruskal-Wallis test, you may ask SPSS to conduct pairwise comparisons in one of two ways. Which of the following options will result in the largest inflation of familywise error rate?

All pairwise

In this output, which of the following terms represents SSM? Sum of Squares Between Groups: 1582.858 Sum of Squares Within Groups: 2142.488 Total Sum of Squares: 3725.347 df Between Groups: 2 df Within Groups: 45 df Total: 47 Mean Square Between Groups: 791.429 Mean Square Within Groups: 47.611 F: 16.62

Between groups SS

In regards to the SPSS output in #9, you conclude that the overall ANOVA is significant and that you now must conduct follow-up pairwise comparisons to determine which group differences are significant (although you have a pretty good idea from just looking at the plots!). Although not shown in the output above, you know that sphericity was violated and you dealt with that issue already. Which of the following post-hoc tests should you use to control for Type I error but still preserve power adequately?

Bonferroni

The test statistic for a Friedman's ANOVA approximates which of the following distributions?

Chi-square distribution

The test statistic that is calculated for a Kruskal-Wallis has a distribution from the family of _____?

Chi-square distributions

Which of the following would be an appropriate use of t-tests for analyzing significance of differences?

Comparison between 2 groups on an interval/ratio measure

Which of the following assumptions must be met to use the paired sample t-test?

Differences between paired scores are normally distributed

The assumption of sphericity that must be met for repeated measures ANOVA is analogous to the homogeneity of variance assumption EXCEPT that HOV relates to equality of variances between groups while assumption of sphericity relates to:

Equal variances between pairs of treatment levels

To assess the meaning of a calculated F-ratio, the value is compared to a critical value from which of the following distributions?

F-distribution

What is the test statistic that is calculated for a Friedman's ANOVA?

Fr

Adjustments designed to correct for violations in sphericity, will not affect the F-ratio but can yield a different p-value that makes the test more conservative. (See Jane Superbrain 15.3) Which of the following corrections will yield the most conservative results in a repeated measures analysis?

Greenhouse-Geisser correction

The test statistic that is calculated for a Kruskal-Wallis test is which of the following?

H-statistic

A therapist was looking at the effects of an intervention on FIM levels. Three groups were used: -control -treatment 1 -treatment 2 Based on the output below, what should the researcher conclude? Questions 5 and 6 relate to this output Sum of Squares Between Groups: 529.437 Sum of Squares Within Groups: 2331.135 Total Sum of Squares: 2860.572 df Between Groups: 2 df Within Groups: 45 df Total: 47 Mean Square Between Groups: 264.719 Mean Square Within Groups: 51.803 F: 5.110 Sig: 0.010

Homogeneity of variance assumption was met

Before running your analysis of differences between 3 independent groups, you test whether variances across groups in the overall analysis are equal. When you find that homogeneity of variance has been violated, all of the following provide an appropriate course of action EXCEPT:

Ignore this finding if your sample size is greater than 30 and proceed with t-test post hoc analyses

An orthogonal contrast means that the contrasts that you choose to do following a significant ANOVA are:

Independent of each other

Recalling the important relationship that we studied early in the semester: Test stastic= variance explained by the model / variance not explained by the model, which of the following would you predict would define our F- statistic?

MSm/MSr

The non-parametric equivalent of the independent t-test is the:

Mann-Whitney test

When a significant Kruskal-Wallis test is found, then you must proceed with pairwise comparisons to determine which group mean differences are significant. The test for significance of differences for these pairwise comparisons is performed using which of the following tests?

Mann-Whitney tests

Which of the following tests is used to assess whether the assumption of sphericity has been met?

Mauchly's test

Most statistical tests can be broadly defined as the ratio between variance explained by the model we are using (systematic variation) to the variance that cannot be explained by the model (unsystematic variation). When calculating a related samples t-test, the systematic variation is:

Mean difference observed between pairs in our sample

If the null hypothesis is true and we run the Mann-Whitney test on data, the expectation is that the _____ in the two groups will be approximately equal.

Mean ranks

In a study investigating whether there is a difference in age at diagnosis of type 2 diabetes for young adult men and women, the result obtained for a two-tailed probability is U =3, p = .09). What does this result signify?

No real evidence that age at diagnosis is different for young adult men and women

As long as parametric assumptions are met, you prefer to analyze your data using parametric tests versus non-parametric tests because of which of the following reasons?

Non-parametric tests in general are less powerful

Since the F-ratio is a ratio of systematic variation to unsystematic variation, a calculated F-statistic of less than 1 automatically signifies:

Non-significant effect

Similar to other non-parametric tests, the Friedman's ANOVA is based on ranked data versus the actual scores. In our example, we have 18 participants who have measurements of grip strength for 3 different conditions. Which of the following best describes how you would rank the data?

Rank the data across the three conditions for each person then sum the ranks for each condition

Designing your study such that you can use a related samples t-test rather than an independent t-test improves your statistical power. The best explanation for this fact is that a related samples test:

Reduces the unsystematic variation

When testing assumptions to compare means for several independent groups using an ANOVA, it is important to test normality _____

Separately for each group

When calculating an independent t-test, the numerator is the difference between group means and the denominator is:

Standard error of the differences between the group means

When comparing the same group of individuals under different treatment conditions to determine whether there are significant differences between the treatments, a one-way ANOVA is NOT appropriate to use because:

The assumption of sphericity is violated

What does the term familywise error rate mean?

The error rate across statistical tests conducted on one dataset to answer the same research question

The degrees of freedom for the Kruskal-Wallis test statistic are determined by________?

The number of groups being compared minus 1

Model sum of squares (SSM) is the amount of variation explained by our model. Which of the following best describes the model sum of squares when comparing independent groups?

The sum of squared differences between each group mean and the grand mean

SST is the total sum of squares or total amount of variation in our dataset. Which of the following best describes total sum of squares when comparing independent groups?

The sum of squared differences between each individual group mean

The residual sum of squares (SSR) is the variation that cannot be explained by our model. Which of the following best describes SSR when comparing independent groups?

The sum of squared differences between each individual score and the group mean to which that person belongs to

Levene's test is used to test whether:

The variances in different groups are equal

What should the researcher conclude about the BDIDIF ANOVA output?

There is a significant difference somewhere between the control, treatment 1, and treatment 2

The Kruskal-Wallis test is appropriate when you wish to compare_____?

Three or more independent samples

A Bonferroni correction is when:

You apply a criterion for significance based on the usual criterion for significance (.05) divided by the number of tests performed

After running a Mann-Whitney test to determine whether differences found are significant, you wish to calculate effect size. Which of the following will you need to compute effect size?

Z-score of the test statistic and sample size

How are degrees of freedom determined for the total sum of squares (SST)?

Total number of people in the sample minus 1 (N-1)

What symbol represents the test statistic for the Mann-Whitney test?

U

When conducting an ANOVA, it is important to test for homogeneity of variance (HOV) between the groups being compared. If your Levene's test shows that the HOV is violated, which of the following typically makes the best adjustment to the F-ratio in terms of preserving statistical power?

Welch test

Under what conditions would planned comparisons be used?

When specific hypotheses have been generated prior to the start of the experiment

Post hoc tests are used under which of the following conditions?

When there are no specific hypotheses about specific comparisons decided before the experiment

A researcher was interested in stress levels of lecturers during lectures. She took the same group of eight lecturers and measured their anxiety (out of 15) during a normal lecture and again in a lecture in which she had paid students to be disruptive and misbehave. The data were not normally distributed. Which test should she use to compare her experimental conditions?

Wilcoxon signed-rank test

Ten visitors to an art gallery are asked to rate each of two sculptures for aesthetic value on a scale of 1 (low value) to 5 (high value) What procedure would be used to test the hypothesis that the ratings on each of the sculptures are not significantly different?

Wilcoxon signed-rank test

The non-parametric equivalent of the paired-samples t-test is the:

Wilcoxon signed-rank test

In this output, which of the following terms represents SSr? Sum of Squares Between Groups: 1582.858 Sum of Squares Within Groups: 2142.488 Total Sum of Squares: 3725.347 df Between Groups: 2 df Within Groups: 45 df Total: 47 Mean Square Between Groups: 791.429 Mean Square Within Groups: 47.611 F: 16.62

Within Groups SS