Stats Final Multiple Choice

The Odds Ratio is the most common and useful measure of effect size when evaluating categorical data. A. True B. False

A. True

True or False: Loglinear analysis is an extension of the chi-square test, thus has similar assumptions A. True B. False

A. True

Which test would you use to look at relationships between categorical variables? a. Chi-square test b. T-test c. ANOVA d. All of the above

a. Chi-square test

In the results of a chi-square analysis, the odds ratio is used as a measure of: a. Effect size b. p value c. Beta value d. None of the above

a. Effect size

What test statistic does an ANOVA analysis produce? a. F-statistic b. M-statistic c. B-statistic d. D-statistic

a. F-statistic

When residuals at each level of the predictors have the same variance (in other words, there is constant variance of the residuals), this is: a. Homoscedasticity b. Heteroscedasticity c. Correlation d. Multicollinearity

a. Homoscedasticity

This test is used when there are two experimental conditions and different participants were assigned to each condition a. Independent samples t-test b. Paired samples t-test c. Bivariate correlations d. Matched-pairs t-test

a. Independent samples t-test

For a non-parametric data set, ____________ is used instead of ______________ to calculate correlations when you have a small data set with a large number of data sets. a. Kendall's Tau & Spearman's Coefficient b. Kendall's Tau & Pearson's R c. Spearman's Coefficient & d. Pearson's R e. Pearson's R & Kendall's Tau

a. Kendall's Tau & Spearman's Coefficient

The likelihood ratio statistic is an alternative to the: a. Pearson's chi-square b. Yates's continuity correction c. Fisher's exact test d. ANOVA

a. Pearson's chi-square

The difference between simple regression and multiple regression is: a. Simple regression predicts one outcome variable from one predictor variable while multiple regression uses several predictor variables. b. Simple regression calculates one regression value between a pair of outcome and predictor variables while multiple regression calculates multiple regression values for the pair c. Simple regression uses one predictor variable to predict one outcome variable while multiple regression uses several predictor variables to predict several outcome variables d. There is no difference, simple and multiple regression are the same statistical processes

a. Simple regression predicts one outcome variable from one predictor variable while multiple regression uses several predictor variables.

A correlation coefficient is an effect size a. True b. False

a. True

A familywise error would occur if multiple t-tests are used to compare differences between 3 or more means. a. True b. False

a. True

A partial correlation controls for the effects a third variable has on both variable in the correlation, while a semi-partial correlation controls for the effect a third variable has on only one of the variables in the correlation. a. True b. False

a. True

An error rate across statistical tests conducted on the same experimental data is known as the familywise error rate. a. True b. False

a. True

Both the independent t-test and dependent t-test are parametric tests based on the normal distribution: a. True b. False

a. True

If the value of R2 is .335 then that tells us that the predictor accounts for 33.5% of the variation a. True b. False

a. True

In a saturated model, all of the standard errors are 0 (or very, very close to 0). a. True b. False

a. True

In a t-test, we are comparing the obtained value of t against what we would expect to get by chance. a. True b. False

a. True

Pearson's chi-square test examines whether there is an associated between two categorical variables. a. True b. False

a. True

True or False The point-biserial correlation coefficient (rpb) is used when one variable is a discrete dichotomy (e.g., pregnancy), whereas the biserial correlation coefficient (rb) is used when one variable is a continuous dichotomy (e.g., passing or failing an exam). a. True b. False

a. True

True or False: The loglinear analysis starts with a model that contains all the main effects. Interactions and terms are removed in a hierarchical manner, starting with the highest-order interaction. a. True b. False

a. True

When doing a hierarchical regression, you select predictors based on previous literature and you decide which order they're entered into the model. a. True b. False

a. True

When we are using regression to understand our data, we are looking to find the line that best describes our data, and then estimate the slope and intercept of the line. a. True b. False

a. True

A simple ANOVA aov(), tells us both the result of the variance between groups and how the groups differed. a. True b. False

b. False

If Levene's test is significant (p-value <.05), we can do an ANOVA without taking steps to rectify the matter a. True b. False

b. False

In a dependent t-test, the sampling distribution of scores should be normal. a. True b. False

b. False

Independent study designs tend to have more power than repeated-measure designs. a. True b. False

b. False

The Welch's independent t-test assumes homogeneity of variance. a. True b. False

b. False

True or false: You can use both Spearman's correlation coefficient and Pearson's r with ordinal data. a. True b. False

b. False

Variance of Sum Law states that the variance of a difference between two independent variables is not equal to the sum of their variances. a. True b. False

b. False

We can tell the direction of hypotheses just by looking at the results of the Benjamini-Hochberg post-hoc tests. a. True b. False

b. False

When we assess correlation amongst our given variables, we are also determining causation. a. True b. False

b. False

Anova is a _____ test a. Post-hoc b. Omnibus c. Correlational d. Categorical data

b. Omnibus

Assumptions for regression analysis include all of the following EXCEPT: a. Non-zero variance b. Perfect multicollinearity c. Homoscedasticity d. Linerality

b. Perfect multicollinearity

Which of the following is true? a. Between measures designs have more power than repeated measures designs. b. Repeated measures designs have more power than between measures designs. c. Between measures and repeated measures designs have the same amount of power. d. Whether an experimental design is between measures or repeated measures does not provide any information about its relative power

b. Repeated measures designs have more power than between measures designs.

An R^2 value of .45 can tell us: a. There is a weak linear relationship between the independent and dependent variables b. The independent variable X accounts for 45% of the variation seen in the outcome c. There is an effect size of .45 d. There is a mean difference of .45 between groups

b. The independent variable X accounts for 45% of the variation seen in the outcome

A correction to the Pearson chi-square is referred to as what? a. Fisher's exact test b. Yate's continuity correction c. Likelihood ratio d. Effect size

b. Yate's continuity correction

To do a t-test in R, we use the function: a. anova() b. t.test() c. stat.summary() d. sqrt()

b. t.test()

Which of the following functions produces a contingency table in R? a. CrossTable() b. xtabs() c. subset() d. loglm()

b. xtabs()

All of the following are assumptions of a t-test EXCEPT: a. The data are normally distributed b. Data are measured at least at the interval level c. All t-tests assume homogeneity of variance d. All of the above are assumptions of t-tests

c. All t-tests assume homogeneity of variance

You _____ use a Welch's F ratio when the Levene's test is not significant: a. Always b. Never c. Don't have to

c. Don't have to

Select the response that is an example of a negative correlation: a. Higher IQ scores are associated with better academic achievement b. Less stress is associated with lower risk for heart problems c. Higher anxiety about a test is associated with less time spent studying d. Less stress is associated with fewer classes during a PhD program

c. Higher anxiety about a test is associated with less time spent studying

After running your ANOVA and finding it significant, you must do follow up analyses to determine which groups are significantly different. If you have specific hypotheses you would use ___________ and if you do not have specific hypotheses you would use________. a. Post hoc tests, planned comparisons b. Planned comparisons, bootstrapping c. Planned comparisons, post hoc tests d. Bonferroni, post hoc tests

c. Planned comparisons, post hoc tests

Which of the following multiple regression methods are not recommended by statisticians: a. Hierarchical regression b. Forced entry c. Stepwise methods d. Statisticians recommend any of these methods, depending on your situation

c. Stepwise methods

What is the effect size of a t-test? a. T-statistic b. Z-score c. r d. Odds ratio

c. r

If you have a contingency table with 3 rows and 2 columns, how many degrees of freedom should you use for your chi-square test? a. 6 b. 1 c. 3 d. 2

d. 2

Which of the following are post-hoc tests? a. Bonferroni b. F-ratio c. Tukey d. A and C only e. All of the above

d. A and C only

If categorical variable expected frequencies in each cell are not greater than 5, what is a way to compute the exact probability of the chi-square statistic? a. Likelihood ratio statistic b. t-test c. One way ANOVA d. Fisher's exact test

d. Fisher's exact test

Which of the following is NOT an assumption of the chi-square test: a. The frequency data is independent b. The frequency data is normally distributed c. Expected frequencies should be greater than 5 d. None of the above

d. None of the above

The Bonferroni correction minimizes ____ and increases _____: a. Type II error/Type I error b. Type I error/P-values c. P-values/Type II error d. Type I error/Type II error

d. Type I error/Type II error

A ______ correlation is a correlation between two variables while a _____ correlation looks at the relationship between two variables while controlling the effect of one or more additional variables: a. point-biserial/partial b. partial/bivariate c. bivariate/biserial d. bivariate/partial

d. bivariate/partial

What does an F ratio less than 1 tell you? a. Your F statistic is insignificant. b. Your F statistic is significant. c. The amount of unsystematic variance is greater than the amount of systematic variance in your model. d. The amount of systematic variance is greater than the amount of unsystematic variance in your model. e. A and C f. B and D

e. A and C

What is the difference between simple linear regression and multiple regression? a. Simple linear regression is when you have one predictor variable and one outcome variable b. Multiple regression is when you have one predictor variable and multiple outcome variables c. Multiple regression is when you have multiple predictor variables and one outcome variable d. Simple linear regression is the same as multiple regression but a simpler statistical model e. A and C f. D and B

e. A and C

What is the difference between an independent and a dependent t-test? a. A dependent T test depends on specific rules about effect sizes b. A dependent T test can test the change or difference between two means coming from the same group at different times or under different conditions c. You need a larger sample size to do a dependent T test d. Independent T tests compares means of different groups in different conditions e. B and D f. A and C

e. B and D