Stats

obacco smoke exposure in infancy was reported to be a risk factor for (i.e. have a positive association with) childhood asthma, with a p-value of 0.01. Which 95% confidence interval would make sense in this scenario? 0.5 to 2.1 0.9 to 1.1 1.2 to 1.5

((NOT)) 0.9 to 1.1

A survey of holiday travelers is conducted at various arrival terminals of airports. The relationship between sleep-amount in the past 24 hours and patience-level (on a scale of 1 to 100) is assessed using a linear regression model, using sleep-amount in hour increments as the independent variable. Statistical analysis yields a coefficient of 7.1 for sleep-amount, and a constant term of 2.3. What would the predicted patience level be for a traveler reporting 8 hours of sleep in the last 24 hours? 59.1 56.8 7.1 2.3

((NOT)) 56.8

A research team wants to know if depression scores taken at baseline in a study predicts weight change over the course of the study. Weight change is measured by subtracting baseline weight from the end-of-study weight, and is normally distributed. You may assume that depression score is a symmetrically distributed continuous variable. paired-samples t-test One-way ANOVA Linear regression Logistic regression A test designed for repeated measures, more than two timepoints Kruskal-Wallis test

((NOT)) A test designed for repeated measures, more than two timepoints

Which of the following statements is true regarding correlation and regression coefficients? Both correlation and regression coefficients are sufficient for establishing a causal relationship between two variables, but only if they are statistically significant. Correlation coefficients can tell us only about the direction of the relationship between two variables, whereas a positive regression coefficient establishes causality. Regression coefficients can tell us only about the direction of the relationship between two variables, whereas a positive correlation coefficient establishes causality. Neither correlation nor regression coefficients, on their own, can establish a causal relationship between two variables.

((NOT)) Both correlation and regression coefficients are sufficient for establishing a causal relationship between two variables, but only if they are statistically significant.

The initial studies establishing maternal diethylstilbesterol (DES) intake as a cause of vaginal adenocarcinoma in female offspring were case-control studies. This was probably largely because: Cohort studies had not yet been invented. The disease outcome is rare. A woman taking DES was always rare. The investigators had probably just happened to have a number of cases in their practices.

((NOT)) Cohort studies had not yet been invented. ((NOT)) A woman taking DES was always rare.

A study reports an association between a predictor X and an outcome Y with an odds ratio of 1.2 (95% confidence interval: 1.1 to 1.3). Which of the following interpretations is correct? Exposure to X was associated with increased risk for Y, but this association was not statistically significant. Exposure to X was associated with decreased risk for Y, but this association was not statistically significant Exposure to X was significantly associated with increased risk for Y. Exposure to X was significantly associated with decreased risk for Y.

((NOT)) Exposure to X was associated with decreased risk for Y, but this association was not statistically significant

A confidence interval: None of the answers, above or below, are correct. Is a range of possible values based on the opinions of at least two experts. Is sometimes reported in place of, or in addition to, a point estimate. Should not be used in a research article because it can mislead the reader.

((NOT)) None of the answers, above or below, are correct

Which of the following statements is true? ORs can be used to assess exposure effects, but RRs cannot. ORs can be used to assess treatment effects, but RRs cannot. ORs and RRs are calculated using the same formula; they differ only in the type of data they describe. ORs and RRs are interpreted similarly with respect to whether they are less than, equal to or greater than 1.

((NOT)) ORs and RRs are calculated using the same formula; they differ only in the type of data they describe.

An investigator would like to examine the association between depression and stress scores in a subset of individuals with a history of psychiatric treatment. He notes that in this subset of individuals, the depression scores were highly left skewed. That is, the majority of subjects clustered at the high end of the scale, but there were a few that had very low scores. Stress scores for this subset were normally distributed. The sample size is relatively small. The investigator has no hypothesis regarding which trait influences the other. What test should he use? One-way ANOVA Mann-Whitney Rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Spearman correlation coefficient Logistic regression

((NOT)) One-way ANOVA

Which of the following statements regarding z-tests and t-tests is true? For large sample sizes, the T-test yields similar results to the Z-test. There is one single Z-distribution, but many T-distributions which vary by sample size. All of the statements, above or below, are true. The t-test is more conservative than the Z-test (i.e. an investigator is less likely to reject the null hypothesis using the t-test), because a t-distribution has heavier tails.

((NOT)) The t-test is more conservative than the Z-test (i.e. an investigator is less likely to reject the null hypothesis using the t-test), because a t-distribution has heavier tails.

A statistician tells a researcher that the correlation coefficient between two variables is 0.3, and that the linear regression coefficient is 5.2. Which of the following statements regarding this scenario is true? This scenario is not possible because a regression coefficient is constrained to lie between -1 and 1. This is a possible scenario because correlation and regression coefficients frequently conflict with respect to the direction of a relationship. This is a possible scenario because both coefficients indicate a positive relationship between variables, and the regression coefficient is not constrained to lie between -1 and 1. This scenario is not possible because the regression and correlation coefficients conflict with respect to the direction of the relationship.

((NOT)) This scenario is not possible because a regression coefficient is constrained to lie between -1 and 1.

A paired t-test is essentially the same as a single sample t-test because: we are testing the mean from a single column of data, generated by averaging the two paired observations. we are testing the mean from a single column of data, generated by subtracting one of the paired observations from the other. given a pair of observations, we need only test the second to see if it differs from 0. given a pair of observations, we need only test the first to see if it differs from 0.

((NOT)) given a pair of observations, we need only test the first to see if it differs from 0.

A relative risk of -1.5 means that the risk of an event is the same in both the exposed and unexposed groups. greater in the exposed group compared to the unexposed group. greater in the unexposed group compared to the exposed group. This situation is impossible. A relative risk can never be less than 0.

((NOT)) greater in the unexposed group compared to the exposed group.

A research team wants to know if depression scores taken at baseline in a study predicts weight change over the course of the study. Weight change is measured by subtracting baseline weight from the end-of-study weight, and is normally distributed. You may assume that depression score is a symmetrically distributed continuous variable. paired-samples t-test One-way ANOVA Linear regression Logistic regression A test designed for repeated measures, more than two timepoints Kruskal-Wallis test

((NOT)) paired-samples t-test

The correlation coefficient: should be expressed in both units of the variables being compared, e.g. cm-years for a correlation between height and age. provides a unit-free measure of association between two variables. is dependent upon the units of measurement used for each variable, i.e. will change if inches are converted to centimeters.

((NOT)) should be expressed in both units of the variables being compared, e.g. cm-years for a correlation between height and age. ((NOT)) All responses, above or below, are correct.

In a normal distribution, 95% of all observations fall symmetrically between what two z-scores? -1.00 and 1.00 -2.58 and 2.58 - 1.96 and 1.96 -1.64 and 1.64

- 1.96 and 1.96

In a normal distribution, 95% of all observations fall symmetrically between what two z-scores? -2.58 and 2.58 -1.00 and 1.00 - 1.96 and 1.96 -1.64 and 1.64

- 1.96 and 1.96

In a normal distribution, 68% of all observations fall symmetrically between what two z-scores? - 1.96 and 1.96 -2.58 and 2.58 -1.00 and 1.00 -1.64 and 1.64

-1.00 and 1.00

A researcher reports an odds ratio of 1.8 and a 95% confidence interval of 1.1 to 3.5. Which p-value below would make sense in this scenario? 0.16 1.25 0.04

0.04

A researcher reports an odds ratio of 1.8 and a 95% confidence interval 0.8 to 2.5. Which p-value below would make sense in this scenario? 0.001 0.01 0.16 1.25

0.16

A researcher reports an odds ratio of 1.8 and a 95% confidence interval 0.8 to 2.5. Which p-value below would make sense in this scenario? 0.01 0.16 0.001

0.16

Day-care exposure in infancy was reported to have a protective effect on (i.e. have an inverse association with) childhood asthma, with a p-value of 0.01. Which 95% confidence interval would make sense in this scenario? 0.9 to 1.1 0.5 to 2.1 0.5 to 0.9

0.5 to 0.9

Exposure to peanut butter in infancy was reported to have no statistically significant association with asthma in the child, with a p-value of 0.91. Which 95% confidence interval would make sense in this scenario? 1.2 to 1.5 0.5 to 0.9 0.5 to 2.1

0.5 to 2.1

The table below contains scores from six people scales measuring perceived friendliness of staff and satisfaction with service. Based on intuition (rather than computation), which of the following makes the most sense as a correlation coefficient between the two scales? Perceived friendliness of staff Satisfaction with service 3 2 7 5 4 3 1 1 8 7 6 7 Hint: Make a quick scatter plot of this data. 0.92 -0.92 9.2 0.1

0.92

Tobacco smoke exposure in infancy was reported to be a risk factor for (i.e. have a positive association with) childhood asthma, with a p-value of 0.01. Which 95% confidence interval would make sense in this scenario? 0.9 to 1.1 0.5 to 2.1 1.2 to 1.5

1.2 to 1.5

For the data set below, which value is the best estimate of the standard deviation? 0, 1, 2, 3, 4 1.5 -1.3 4.0 0.25

1.5

Scores from a commonly used psychological instrument follow a normally distribution with a mean of 100 and standard deviation of 25. What is the z-score associated with a score of 150? 25 50 -2 2

2

A survey of holiday travelers is conducted at various arrival terminals of airports. The relationship between sleep-amount in the past 24 hours and patience-level (on a scale of 1 to 100) is assessed using a linear regression model, using sleep-amount in hour increments as the independent variable. Statistical analysis yields a coefficient of 7.1 for sleep-amount, and a constant term of 2.3. What would the predicted patience level be for a traveler reporting no sleep in the last 24 hours? 7.1 56.8 2.3

2.3

A psychiatrist devised a short screening test for depression. An independent blind comparison was made with a gold standard for diagnosis of depression among 200 psychiatric outpatients. Among the 50 outpatients found to be depressed according to the gold standard, 35 patients were positive for the test. Among 150 patients found not to be depressed according to the gold standard, 30 patients were found to be positive for the test. What is the prevalence of depression among the 200 outpatients? 54% 70% 25% 80%

25%

Birthweights for newborns follow a normal distribution with a mean of 3.5kg and a standard deviation of 0.5 kg. A newborn from this population has a Z-score of -1. What is her weight? 4.0 kg 2.5 kg 3.0 kg 3.4 kg

3.0 kg

One step in conducting a chi-square test is the calculation of the observed count minus the expected count for each cell in a two-way table. What is the minimum expected count required in each cell for the chi-square test to be valid? 1 25 5 10

5

Birthweights for newborns follow a normal distribution with a mean of 3.5 kg and a standard deviation of 0.5 kg. A newborn from this population has a Z-score of 3. What is her weight? 3.8 kg 3.0 kg 2.0 kg 5.0 kg

5.0 kg

A psychiatrist devised a short screening test for depression. An independent blind comparison was made with a gold standard for diagnosis of depression among 200 psychiatric outpatients. Among the 50 outpatients found to be depressed according to the gold standard, 35 patients were positive for the test. Among 150 patients found not to be depressed according to the gold standard, 30 patients were found to be positive for the test. What is the positive predictive value of the psychiatrist's screening test? 25% 70% 80% 54%

54%

What percent of participants had a full-time job? [tab work] 435 15.02 22.75 62.23

62.23

Perform a t test for related samples (paired t-test), to answer the question on whether students' satisfaction with weight is significantly different from current weight (satcurwt) and weight at age 18 (satwt18), be careful to use the correct variables [ttest satwt18r = satwtrec]. What is the t statistic? 0.69 0.53 7.94 696

7.94

A psychiatrist devised a short screening test for depression. An independent blind comparison was made with a gold standard for diagnosis of depression among 200 psychiatric outpatients. Among the 50 outpatients found to be depressed according to the gold standard, 35 patients were positive for the test. Among 150 patients found not to be depressed according to the gold standard, 30 patients were found to be positive for the test. What is the sensitivity of the psychiatrist's screening test? 70% 89% 25% 54%

70%

To evaluate the performance of a new diagnostic test, the developer checks it out on 100 known cases of the disease for which the test was designed, and on 200 controls known to be free of the disease. Ninety of the cases yield positive tests, as do 30 of the controls. Based on these data, the positive predictive value and false discovery rate (false positives among those who test positive), respectively, are: 75% and 25% 85% and 25% 75% and 15% 85% and 15%

75% and 25%

The variance of a sample of 81 observations equals 64. The standard deviation of the sample equals: 9 6,561 4096 8

8

A psychiatrist devised a short screening test for depression. An independent blind comparison was made with a gold standard for diagnosis of depression among 200 psychiatric outpatients. Among the 50 outpatients found to be depressed according to the gold standard, 35 patients were positive for the test. Among 150 patients found not to be depressed according to the gold standard, 30 patients were found to be positive for the test. What is the specificity of the psychiatrist's screening test? 80% 25% 70% 54%

80%

To evaluate the performance of a new diagnostic test, the developer checks it out on 100 known cases of the disease for which the test was designed, and on 200 controls known to be free of the disease. Ninety of the cases yield positive tests, as do 30 of the controls. Based on these data, the specificity and false positive rate (proportion testing positive among controls), respectively, are: 85% and 15% 85% and 25% 75% and 15% 75% and 25%

85% and 15%

What is the p-value, and is the p-value significant? 0.156, not significant 7.9, not significant <0.00001, significant None of the above

<0.00001, significant

A survey of holiday travelers is conducted at various arrival terminals of airports. The relationship between sleep-amount in the past 24 hours and patience-level (on a scale of 1 to 100) is assessed using a linear regression model, using sleep-amount in hour increments as the independent variable. Statistical analysis yields a coefficient of 7.1 for sleep-amount, and a p-value of 0.216. Which of the following statements represents an accurate conclusion? A 1 hour increase in amount slept in the past 24 hours was associated with an increase of 7.1 units on the patience scale, but this increase was not statistically significant. Patience level was significantly negatively associated with amount of sleep in the past 24 hours (p = 0.216). There was an error in this calculation, because regression coefficients must fall between -1 and 1.

A 1 hour increase in amount slept in the past 24 hours was associated with an increase of 7.1 units on the patience scale, but this increase was not statistically significant.

Discuss why the larger t statistics are more likely to be statistically significant. A larger t statistic indicates a smaller mean difference between timepoints (or between paired subjects). Larger t statistics have larger observed p values. There is no difference in statistical significance between large or small t statistics. A larger t statistic indicates a larger mean difference between timepoints (or between paired subjects). Larger t statistics have smaller observed p values.

A larger t statistic indicates a larger mean difference between timepoints (or between paired subjects). Larger t statistics have smaller observed p values.

"Children can learn a second language faster before the age of 7 than after the age of 7." This statement is: A one-tailed hypothesis A two-tailed hypothesis A null hypothesis A non-scientific statement

A one-tailed hypothesis

For a group of 30 volunteers, a researcher calculates a mean systolic blood pressure (SBP) before and after a meditation intervention of 115 and 110, respectively. She then conducts a paired t-test and obtains a p-value of 0.03. Out of curiosity, she next conducts an independent t-test on the same data and obtains a p-value of 0.09. Which statement below best explains why this would happen? The difference in means would be considerably smaller if it were calculated by subtracting the mean pre-SBP from the mean post-SBP (as opposed to taking the mean of the difference scores), and a smaller difference in means would yield less power. A paired t-test is more powerful than an independent t-test because it considers the variation of a single distribution, the difference scores, as opposed to two separate distributions. In other words, the signal is essentially the same, but there is considerably less noise. None of these statements, above or below, are correct. The p-value is unrelated to power. The difference in means would be considerably larger if it were calculated by subtracting the mean pre-SBP from the mean post-SBP (as opposed to taking the mean of the difference scores). However, the variation would also increase, and thus no rule of thumb can be drawn regarding increased or decreased power.

A paired t-test is more powerful than an independent t-test because it considers the variation of a single distribution, the difference scores, as opposed to two separate distributions. In other words, the signal is essentially the same, but there is considerably less noise.

Which of the following scenarios does not require an analysis technique that accounts for correlation between observations? A researcher compares blood pressure between subjects randomized to treatment vs placebo. Blood pressure data is collected monthly over the course of a year. A researcher collects medical records on 10000 adults and compares length of hospital stay between men and women. A researcher runs a city-wide intervention program in 5 cities, and compares hospital-level outcomes to those of 5 control cities. A researcher examines the effect of a self-esteem program among adolescents. To obtain a sufficient number of subjects he recruits students from three different schools.

A researcher collects medical records on 10000 adults and compares length of hospital stay between men and women.

Which scenario below describes dependent samples (statistically speaking)? A researcher collects vision scores for 100 children, and then compares the scores of boys and girls. A researcher allows subjects to choose to spend an hour either watching an educational video or participating in a meditation activity, then compares blood pressures between the two groups. A researcher gives an experimental drug to one group of patients and a placebo to a second group of patients, and then compares cholesterol levels between groups at 1 month follow up. A researcher collects taste test data on a group of subjects. All of the subjects cleanse their palate thoroughly, drink a sweet beverage, and complete a second taste test to be compared to the first.

A researcher collects taste test data on a group of subjects. All of the subjects cleanse their palate thoroughly, drink a sweet beverage, and complete a second taste test to be compared to the first.

A researcher would like to investigate whether stress levels change over time for a group of adults entering an exercise program. He compares scores from a life-stress measurement instrument given to each participant at baseline, 6 weeks, and 12 weeks. What test should he use? paired-samples t-test Mann-Whitney Rank-sum test Wilcoxon signed-rank test Chi-square test A test designed for repeated measures, more than two timepoints Linear regression

A test designed for repeated measures, more than two timepoints

A reviewer tells a research team that they have insufficient power to detect an effect in their study, and recommends an increased sample size. Why might this be problematic for the researchers? All answers, above or below, are correct. Increasing sample size may require expanding the study to multiple sites, which can strain resources and complicate analysis. Increasing sample size often requires increasing the budget, which is almost always limited. Increasing sample size may require increased recruitment time, which may be limited.

All answers, above or below, are correct.

A spreadsheet has a variable of "female," includes a dropdown menu for data entry, with choices "yes," and, "no." What type of variable might "female" be considered? categorical All answers, above or below, are correct. dichotomous (i.e. exactly two categories) numeric, if "yes" responses are assigned a 1, and "no" responses are assigned 0.

All answers, above or below, are correct.

In the case of a statistically significant result from a t-test, why should the observed difference in means be reported in addition to a p-value? To show the magnitude of the difference. To show the direction of the difference. To help the reader assess whether the difference is clinically meaningful. All answers, above or below, are correct.

All answers, above or below, are correct.

Which of the following statements regarding the selection of the selection of the significance level, alpha, is true? Alpha should be selected with thought given to which type of error (Type I or Type II) has more serious consequences. Good science requires that alpha always be selected a priori, and never changed upon seeing the results of a statistical test. Alpha should not be arbitrary. If a research team is not sure of what value to set alpha, they should consider norms for their field of research. All answers, above or below, are correct.

All answers, above or below, are correct.

A survey will be given to 100 students randomly selected from the freshman class at the UA. What is the population? All freshmen college students across the US All students at the UA All freshmen at the UA The 100 selected students

All freshmen at the UA

A non-parametric test should be used when: The outcome data is highly skewed or has extreme outliers that are plausible values. The outcome data is ordinal, as opposed to continuous. All of these scenarios warrant the use of a non-parametric test.

All of these scenarios warrant the use of a non-parametric test.

A non-parametric test should be used when: The outcome data is ordinal, as opposed to continuous. The outcome data is highly skewed or has extreme outliers that are plausible values. All of these scenarios warrant the use of a non-parametric test. The sample size is too small to expect a normally distributed sampling distribution.

All of these scenarios warrant the use of a non-parametric test.

Which of the following non-parametric tests does not rely on the calculation of ranks? Wilcoxon signed-rank test All of these tests rely on the calculation of ranks. Mann Whitney rank-sum test Kruskal-Wallis test

All of these tests rely on the calculation of ranks.

Regression models differ from correlation models in that: Regression models include an intercept term, whereas correlation models do not. All responses, above or below are correct. Correlation coefficients always fall between -1 and 1, whereas regression coefficients are not constrained in this manner.

All responses, above or below are correct.

When interpreting a correlation coefficient, it is important to look at: The magnitude of the correlation coefficient. The significance of the correlation coefficient. All responses, above or below, are correct. The direction (+/- sign) of the correlation coefficient.

All responses, above or below, are correct.

Which of the following is true regarding when to use an independent samples T-test or an ANOVA? An ANOVA is not typically used to compare means between only two groups, but if this is done, it will yield the same p-value as an independent samples t-test that assumes equal variances. A t-test can never be used to simultaneously compare means between more than two groups. An independent samples t-test is used to compare means between two groups, while the ANOVA can compare means for 3 or more groups simultaneously. All responses, above or below, are correct.

All responses, above or below, are correct.

If my null hypothesis is, "Dutch people do not differ from English people in height," what is my alternative hypothesis? Dutch people are taller than English people. All statements above or below are plausible alternative hypotheses. Dutch people differ in height from English people. English people are taller than Dutch people.

All statements above or below are plausible alternative hypotheses.

Which of the following statements is an accurate description of logistic regression and linear regression? Logistic regression is used for dichotomous outcomes, whereas linear regression is used for continuous outcomes. All statements, above or below are correct. Both linear and logistic regression models may be used to describe the effect of a predictor on an outcome while controlling for other variables. Both linear and logistic regression models assume that a linear (or log-linear for logistic models) relationship exists between the predictor and outcome.

All statements, above or below are correct.

Which of the following statements regarding standard errors and standard deviations is true? All statements, above or below, are true. Standard errors describe the spread of a sampling distribution. Assuming a sample size greater than 1, the standard error will always be smaller than the standard deviation. Standard deviations describe the spread of a population distribution.

All statements, above or below, are true.

You want to run a regression model using total IPA as the dependent variable and self-confidence as the independent variable. Does self-confidence predict total IPA [regress total confid]? Select all that apply. The model is significant, F (1, 659) < 0.00001, which means that self-confidence is a significant predictor for total IPA. Approximately 84% of the variability in total IPA is explained by self-confidence in the model. The coefficient for self-confidence is 1.98, t = 59.11, p-value < 0.0001, this means the slope is significantly different from zero, for every unit increase in self-confidence, total IPA increases by 1.98 units. The constant coefficient is 27.81, therefore the regression model can be expressed as total IPA = 27.81 + 1.98 (self-confidence). All the statements are true.

All the statements are true.

Which of the following statements is true regarding the relationship between the p-value and alpha? The p-value and alpha are the same thing. Alpha is fixed prior to conducting a statistical test, whereas the p-value is reported upon completion of the statistical test. If the p-value is less than alpha, then the null hypothesis is accepted (not rejected). If the p-value is greater than alpha, then the null hypothesis is rejected.

Alpha is fixed prior to conducting a statistical test, whereas the p-value is reported upon completion of the statistical test.

Fisher's exact test should be used in place of the chi-square test when: An expected cell count is less than 5. When using continuous as opposed to categorical variables. When there are more than 5 categories in each variable being assessed. Fisher's exact test is a shortcut from pre-computer days that is always inferior to chi-square.

An expected cell count is less than 5.

Which of the following research methods would provide the best evidence of causality? An observational study in which correlations between many variables are taken into account. An experimental study in which a variable is systematically manipulated to observe its effect on an outcome. A cross-sectional study in which extensive descriptive data is collected at a single time-point. An ecological study in which population-wide characteristics are compared across many countries.

An experimental study in which a variable is systematically manipulated to observe its effect on an outcome.

Which of the following is true about a 95% confidence interval of the mean? Approximately 95% of population means will fall within the limits of the confidence interval. Approximately 95 out of 100 confidence intervals will contain the population mean. Approximately 95 out of 100 sample means will fall within the limits of the confidence interval. There is a 95% probability that the population mean falls within the limits of the confidence interval.

Approximately 95 out of 100 confidence intervals will contain the population mean.

Confidence intervals: Should rarely be used in research articles because they can mislead the reader. Are constructed using subjective evaluations of confidence. None of the answers, above or below, are correct. Are sometimes reported in place of, or in addition to, point estimates.

Are sometimes reported in place of, or in addition to, point estimates.

Why is a more stringent alpha level used for post-hoc tests? Because running multiple significance tests increases the likelihood of a type I error. None of these are correct. A less stringent alpha level should be used on post-hoc tests. Because running multiple significance tests increases the likelihood of finding no significant association. Because running multiple significance tests increases the likelihood of type II error.

Because running multiple significance tests increases the likelihood of a type I error.

Why is it important to conduct post-hoc tests after a global test, such as ANOVA, yields a statistically significant result? Because the global test does not tell us between which groups, or conditions, the significant differences lie. None of these are correct. It is only necessary to conduct post-hoc tests after a non-significant result on a global test. Because post-hoc tests are always necessary, after one achieves either significant and non-significant results on the global test. Because the global test does not give a real p-value.

Because the global test does not tell us between which groups, or conditions, the significant differences lie.

A researcher creates a histogram of the heights of all undergraduate students. The distribution is likely to be: Bi-modal, with distinct peaks at the mean height for females and mean height for males. Highly left skewed, because there is an obvious maximum height, but no limit to how short a student can be. Highly right skewed, because there is an obvious minimum height, but no limit to how tall a student can be. Uniform (i.e. all heights equally represented)

Bi-modal, with distinct peaks at the mean height for females and mean height for males.

Which of the following statements is an accurate comparison between logistic regression and linear regression? Both are tools that model relationships for a continuous outcome. Both are tools that model relationships for a categorical outcome. Both are tools that model relationships between a predictor and an outcome. Both are tools that model relationships for a dichotomous outcome.

Both are tools that model relationships between a predictor and an outcome.

Regression models are similar to correlation models in that: All responses, above or below are correct. Both regression and correlation coefficients tell us the direction of the relationship. Both regression models and correlation models are symmetric, i.e. neither depend on which variable is dependent or independent.

Both regression and correlation coefficients tell us the direction of the relationship.

What type of data must the exposure and outcome variable be in order to calculate either an OR or RR (as covered in week 3 content)? Both variables must be categorical, with no limit to the number of categories. The exposure variable must be dichotomous (binary), but the outcome can be any type of variable. Both variables must be dichotomous (binary). The outcome variable must be continuous, but the exposure can be any type of variable.

Both variables must be dichotomous (binary).

A reviewer tells a research team that they have insufficient power to detect an effect in their experimental study of a new treatment, and recommends finding a way to increase the effect size. How might the investigators achieve this? By selecting a more heterogeneous population. None of the responses, above or below, are correct. There is no realistic way to increase effect size. By decreasing the treatment dosage, if ethically possible. By increasing the treatment dosage, if ethically possible.

By increasing the treatment dosage, if ethically possible.

A reviewer tells a research team that they have insufficient power to detect an effect in their study, and recommends finding a way to decrease the variance in the predictor of interest. How might the investigators achieve this? By decreasing the sample size. None of the responses, above or below, are correct. There is no realistic way to decrease variance. By selecting a more heterogeneous population. By selecting a more homogeneous population.

By selecting a more homogeneous population.

A research team wants to know if the type of pet ownership (yes/no) is associated with political party affiliation (republican, democrat, independent, other, none). What test would work best for assessing this relationship? paired-samples t-test Kruskal-Wallis test Chi-square test Linear regression

Chi-square test

A researcher conducts an ANOVA to compare mean weight gain between three intervention groups: A, B and C. The resulting p-value is 0.003. Assuming an alpha of 0.05, what should be the researchers next step? Conduct post hoc tests by creating new interventions to try with the subjects. Post hoc tests would not be appropriate in this scenario. Conclude simply that there is no difference in weight gain between the three groups. Post hoc tests would not be appropriate in this scenario. Conclude simply that there is a statistically significant difference in weight gain between the three groups. Conduct post hoc tests to determine if significant differences occurred between groups A and B, A and C, or B and C.

Conduct post hoc tests to determine if significant differences occurred between groups A and B, A and C, or B and C.

In a linear regression model, the intercept term is often referred to as the ______. Constant Significance Slope Standard Error

Constant

Which of the following is an example of a one-sided alternative hypothesis? Dogs are smarter than cats. There is an association between caloric intake and political affiliation. Robins lay a different number of eggs per year, on average, than woodpeckers. There is no association between physical activity level and academic achievement.

Dogs are smarter than cats.

An investigator wants to see if there is a relationship between dog ownership (yes/no) and Type 1 diabetes (yes/no). Which of the following statements regarding the choice of a chi-square test vs a two-sample test of proportions is true? Only the chi-square test would be appropriate, because there is no way to express diabetes prevalence as a proportion. Neither the chi-square nor two-sample test of proportions would be appropriate because it is unlikely that these variables are related. Either test could be used, because both can assess the association between two dichotomous (binary) variables. Only the two-sample test of proportions would be appropriate, because there would be no way to create a two-way contingency table for these data.

Either test could be used, because both can assess the association between two dichotomous (binary) variables.

The positive predictive value of a clinical test: Refers to the ability of a test to identify patients without the disease. Equals (true positives)/(true positives + false positives) Does not depend on the prevalence of the disease in the population. Refers to the ability of a test to identify patients with the disease.

Equals (true positives)/(true positives + false positives)

A study reports an association between a predictor X and an outcome Y with an odds ratio of 0.7 (95% confidence interval: 0.3 to 1.1). Which of the following interpretations is correct? Exposure to X was associated with increased risk for Y, but this association was not statistically significant. Exposure to X was associated with decreased risk for Y, but this association was not statistically significant Exposure to X was significantly associated with increased risk for Y. Exposure to X was significantly associated with decreased risk for Y.

Exposure to X was associated with decreased risk for Y, but this association was not statistically significant

A study reports an association between a predictor X and an outcome Y with an odds ratio of 1.2 (95% confidence interval: 0.9 to 1.4). Which of the following interpretations is correct? Exposure to X was associated with increased risk for Y, but this association was not statistically significant. Exposure to X was associated with decreased risk for Y, but this association was not statistically significant Exposure to X was significantly associated with increased risk for Y. Exposure to X was significantly associated with decreased risk for Y.

Exposure to X was associated with increased risk for Y, but this association was not statistically significant.

A study reports an association between a predictor X and an outcome Y with an odds ratio of 0.7 (95% confidence interval: 0.5 to 0.9). Which of the following interpretations is correct? Exposure to X was associated with increased risk for Y, but this association was not statistically significant. Exposure to X was associated with decreased risk for Y, but this association was not statistically significant Exposure to X was significantly associated with increased risk for Y. Exposure to X was significantly associated with decreased risk for Y.

Exposure to X was significantly associated with decreased risk for Y.

A study reports an association between a predictor X and an outcome Y with an odds ratio of 1.2 (95% confidence interval: 1.1 to 1.3). Which of the following interpretations is correct? Exposure to X was associated with increased risk for Y, but this association was not statistically significant. Exposure to X was associated with decreased risk for Y, but this association was not statistically significant Exposure to X was significantly associated with increased risk for Y. Exposure to X was significantly associated with decreased risk for Y.

Exposure to X was significantly associated with increased risk for Y.

Results from a meta-analysis provide less evidence for the presence or absence of an effect, than the results from any single study within that meta-analysis. True False

False

Which of the following makes sense as an alternative hypothesis? There is no difference in gut microflora between people who consume pickles and those who do not. Horses and cows have the same heart rates. Fathers' heights are different than their sons' heights. Female graduate school applicants have the same GRE scores, on average, as male applicants.

Fathers' heights are different than their sons' heights.

One of the main differences between z and t-tests is that: For the z-test the population standard deviation is known. The z-test is appropriate only for categorical data. For the z-test the population standard deviation is estimated by the sample standard deviation. The z-test and t-test are actually identical, with preference for usage depending on the field of research.

For the z-test the population standard deviation is known.

Run a simple logistic regression predicting lymph node involvement by x-ray. What can you say about the model results? Select all that apply [logistic nodes xray]. For those with a positive x-ray the odds of having lymph node involvement increase by 8.86, compared to those with negative x-ray. X-ray significantly predicts lymph node involvement, the z = 3.13 and p-value = 0.002. The p-value is less than 0.05 and therefore we can reject the null hypothesis. The 95% confidence interval does not include the null value of 1, and therefore x-ray is a significant predictor of lymph node involvement. None are correct

For those with a positive x-ray the odds of having lymph node involvement increase by 8.86, compared to those with negative x-ray. X-ray significantly predicts lymph node involvement, the z = 3.13 and p-value = 0.002. The p-value is less than 0.05 and therefore we can reject the null hypothesis. The 95% confidence interval does not include the null value of 1, and therefore x-ray is a significant predictor of lymph node involvement.

A student learns that her height falls at one standard deviation below the mean for adult women. Assuming that the height of adult women follows a normal distribution, this means: This woman is shorter than 97.7% of women. Her height has a z-score of 1. Her height has a z-score of -1. She is taller than the majority of women.

Her height has a z-score of -1.

You have sampled 25 students randomly selected from the UA nursing program to find the mean daily sleep time. A 95% confidence interval for the mean sleep time is 5.2 to 7.8 hours. Which of the following statements gives a valid interpretation of this interval? 95% of the 25 nursing students sleep between 5.2 and 7.8 hours, on average. If this procedure were repeated many times, 95% of the sample means would fall between 5.2 and 7.8 hours. 95% of UA nursing students sleep between 5.2 and 7.8 hours, on average. If this procedure were repeated many times, 95% of the resulting confidence intervals would contain the true mean sleep time of UA nursing students.

If this procedure were repeated many times, 95% of the resulting confidence intervals would contain the true mean sleep time of UA nursing students.

researcher finds that people with pets have 5 mm Hg lower systolic blood pressure than people without pets, on average, and reports a p-value of 0.02. Which of the following represents a correct interpretation of this p-value? There is a 2% chance that the 5 mm Hg difference between groups is clinically relevant. There is a 2% chance that the researcher has correctly rejected the null hypothesis in this study. If, in fact, there is no association between pets and blood pressure, the probability of finding the difference of at least 5 mm Hg, given the study sample size and standard deviation, is 2%. If, in fact, there is no association between pets and blood pressure, the probability of finding the difference of at least 5 mm Hg, given the study sample size and standard deviation, is 98%.

If, in fact, there is no association between pets and blood pressure, the probability of finding the difference of at least 5 mm Hg, given the study sample size and standard deviation, is 2%.

A correlation of -0.5 describes a scatter plot in which the slope: Is downwards (from the top left corner to the bottom right corner of the graph). Is upwards (from the bottom left corner to the top right corner of the graph). Is vertical.

Is downwards (from the top left corner to the bottom right corner of the graph).

A confidence interval: Is a range of possible values based on the opinions of at least two experts. Is sometimes reported in place of, or in addition to, a point estimate. None of the answers, above or below, are correct. Should not be used in a research article because it can mislead the reader.

Is sometimes reported in place of, or in addition to, a point estimate.

A correlation of 0.5 describes a scatter plot in which the slope: Is downwards (from the top left corner to the bottom right corner of the graph). Is upwards (from the bottom left corner to the top right corner of the graph).

Is upwards (from the bottom left corner to the top right corner of the graph).

How does the shape of the t-distribution change as the sample size increases? It becomes left skewed. It approaches the normal distribution. It becomes heavier tailed. It becomes right skewed.

It approaches the normal distribution.

First generate a histogram of life purpose and satisfaction [histogram life]. What can you say about the distribution? It is a normally distributed distribution It is skewed to the left It is skewed to the right

It is skewed to the left

A random sample of army inductees was selected and divided into four equal groups. An actor then gave a speech extolling the virtues of army life to each group, but dressed differently for each, as follows: 1) as a private; 2) as a sergeant; 3) as a captain; and 4) as a colonel. The researcher would like to test the hypothesis that perceived status of a speaker is associated with level of enthusiasm generated on the topic. As part of the follow-up questionnaire, inductees were asked to rate their enthusiasm toward army life on a 5 point scale. Scales less than 7 points are typically considered ordinal. What test should the researcher use? paired-samples t-test Mann-Whitney Rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Spearman correlation coefficient

Kruskal-Wallis test

A researcher randomizes 45 participants to one of three different exposures to animals. Group 1 pets a dog, group 2 watches fish in an aquarium, and group 3 looks at cute images of kittens. He then measures cortisol levels for each participant during the exposure. The resulting data were highly skewed. What test should be used to analyse the data? Mann Whitney rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Friedman's ANOVA

Kruskal-Wallis test

An investigator administers a 50 point survey to 20 students from each of four Health Science Colleges: Nursing, Medicine, Pharmacy and Public Health. What test should she apply to determine if attitudes toward using lab animals for research differs between students in each of these colleges. The survey scores for these students are highly left skewed. paired-samples t-test Mann-Whitney Rank-sum test Kruskal-Wallis test Logistic regression

Kruskal-Wallis test

A researcher is interested in understanding the odds ratios for different factors associated with whether or not a patient will fall while in the hospital. What test should he use? independent samples t-test paired-samples t-test One-way ANOVA Linear regression Logistic regression Wilcoxon signed-rank test

Logistic regression

A researcher measured cortisol levels among 40 individuals while they read political news stories. He split the data into two groups: 20 males and 20 females. The resulting data were highly skewed. What test should be used to analyse the data? Mann Whitney rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Friedman's ANOVA

Mann Whitney rank-sum test

A researcher conducts a survey of 25 male and 25 female randomly selected nurses working in a variety of practice settings. What statistical test should he apply to determine if income among nurses differs by gender? Income is known to be positively skewed and the researcher cannot afford to collect large samples. paired-samples t-test One-way ANOVA Mann-Whitney Rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Chi-square test Spearman correlation coefficient

Mann-Whitney Rank-sum test

An investigator is interested in gender differences in students' attitudes toward using lab animals for research. A 50 point survey is given to 10 male students and 10 female students. The survey is designed so that high scores indicate favorable attitudes toward animal research. After completion, researchers found that the distribution of responses to the survey were highly left skewed. That is, the majority of students had favorable attitudes toward research participation, but a few students were adamantly opposed and scored near 0. Which statistical test should the investigator use to compare the groups? One-way ANOVA Mann-Whitney Rank-sum test Wilcoxon signed-rank test Chi-square test Linear regression

Mann-Whitney Rank-sum test

Which of the following tests would be most appropriate to determine differences between two unpaired groups when the dependent variable is measured using an ordinal scale? paired-samples t-test Mann-Whitney Rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test

Mann-Whitney Rank-sum test

Which of the following is not a way of describing the dispersion of a set of observations? mode interquartile range range standard deviation

Mode

For a group of 30 volunteers, a researcher calculates a mean systolic blood pressure (SBP) before and after a meditation intervention of 115 and 110, respectively. She then conducts a paired t-test and obtains a p-value of 0.03. Which statement would be the best interpretation of these results? Assume an alpha of 0.05. Each participant had a decrease of 5mmHg in SBP following meditation, a difference that was statistically significant. ) Each participant had a decrease of 5mmHg in SBP following meditation, but this difference was not statistically significant. On average, participants experienced a decrease of 5 mmHg in SBP following meditation, a difference that was statistically significant. On average, participants experienced a decrease of 5 mmHg in SBP following meditation, but this difference was not statistically significant.

On average, participants experienced a decrease of 5 mmHg in SBP following meditation, a difference that was statistically significant.

A cafeteria manager is dismayed by the amount of food that is left on plates. One way to reduce waste is to prepare the food using a tastier recipe. Suppose that three unique recipes for an entrée are served on randomly selected days in a hospital cafeteria. As the plates are cleaned in the kitchen, the weight of entrée left on the plates is recorded. How might the manager gauge the preference of the recipes? Assume entrée weights are normally distributed. independent samples t-test paired-samples t-test One-way ANOVA Wilcoxon signed-rank test Linear regression

One-way ANOVA

A researcher tests the hypothesis that perception of a person's height depends on his/her perceived status. A random sample of army inductees was selected and divided into four equal groups. An actor then gave a speech extolling the virtues of army life to each group, but dressed differently for each, as follows: 1) as a private; 2) as a sergeant; 3) as a captain; and 4) as a colonel. The inductees completed questionnaires evaluating the speech. As part of the questionnaire, they were asked to evaluate the speaker's height. Assuming normally distributed responses of height, what statistical test should the researcher apply? one sample t-test independent samples t-test paired-samples t-test One-way ANOVA Chi-square test Logistic regression

One-way ANOVA

If we were to remove the 10 in this data set: 0, 1, 1, 2, 2, 2, 3, 3, 5, 10 Only the mode would decrease All measures of central tendency would decrease Only the mean would decrease Only the median would decrease

Only the mean would decrease

Which of the following statements is true regarding parametric and non-parametric tests conducted on the same data? Parametric tests have identical power to non-parametric tests. Parametric tests always have less power than non-parametric tests. Parametric tests usually have more power than non-parametric tests.

Parametric tests usually have more power than non-parametric tests.

An investigator would like to know how depression and stress scores are associated. Both are measured on 100 point scales and both scores tend to be normally distributed. The researcher has no hypothesis regarding which trait influences the other. Which test should he conduct? independent samples t-test paired-samples t-test One-way ANOVA Pearson correlation coefficient Linear regression

Pearson correlation coefficient

A researcher conducts an ANOVA to compare depression levels between three intervention groups: A, B and C. The resulting p-value is 0.147. Assuming an alpha of 0.05, what should be the researcher's next step? Conduct post hoc tests by creating new interventions to try with the subjects. Post hoc tests would not be appropriate in this scenario. Conclude simply that there is a statistically significant difference in depression level between the three groups. Post hoc tests would not be appropriate in this scenario. Conclude simply that there is no significant difference in depression levels between the three groups. Conduct post hoc tests to determine if significant differences occurred between groups A and B, A and C, or B and C.

Post hoc tests would not be appropriate in this scenario. Conclude simply that there is no significant difference in depression levels between the three groups.

The sensitivity of a clinical test: Refers to the ability of a test to identify patients without the disease. Equals (true positives)/(true positives + false positives) Depends on the prevalence of the disease in the population. Refers to the ability of a test to identify patients with the disease.

Refers to the ability of a test to identify patients with the disease.

The specificity of a clinical test: Refers to the ability of a test to identify patients without the disease. Equals (true positives)/(true positives + false positives) Depends on the prevalence of the disease in the population. Refers to the ability of a test to identify patients with the disease.

Refers to the ability of a test to identify patients without the disease.

A researcher testing an hypothesis regarding a population proportion obtains a p-value of 0.044. What is the appropriate conclusion at alpha levels of 0.10, 0.05, and 0.01? Reject H0 at alpha = 0.10, but not at alpha = 0.05 or 0.01. Reject H0 at alpha = 0.10 or 0.05, but not at alpha = 0.01. Do not reject H0 at any of these alpha levels. Reject H0 at all three alpha levels.

Reject H0 at alpha = 0.10 or 0.05, but not at alpha = 0.01.

A Mann-Whitney rank-sum test is used to compare satisfaction ratings between new customers and returning customers. The median rating for new customers was 4 and the median rating for returning customers was 5. The statistical test yields a p-value of 0.091. Which of the following is an appropriate manner of reporting these results? Assume an alpha level of 0.05. Returning customers gave higher ratings than they originally did as new customers (medians: 5 and 4, respectively), but this difference was not statistically significant (p = 0.091). None of these are correct. Means should be reported as opposed to medians when conducting a Mann-Whitney rank-sum test. Returning customers gave slightly higher ratings than new customers (medians: 5 and 4, respectively), but this difference was not statistically significant (p = 0.091).

Returning customers gave slightly higher ratings than new customers (medians: 5 and 4, respectively), but this difference was not statistically significant (p = 0.091).

Distributions of household income are typically: Uniform, because all income levels are often equally represented in a population. Symmetric, because incomes typically vary from the mean in a symmetric manner, and there is no lower or upper limit to income. Left skewed, because there are often a few households that have much lower incomes than the majority. Right skewed, because there are often a few households that have much higher incomes than the majority.

Right skewed, because there are often a few households that have much higher incomes than the majority.

What is the mean and standard deviation for satisfaction with current weight (satcurwt) and satisfaction with weight at age 18 (satwt18)? [sum satcurwt], and [sum satwt18]. Satisfaction with current weight mean = 699, s.d. = 1 to 10, satisfaction with weight at age 18 mean = 697 and s.d. = 1 to 10 Satisfaction with current weight mean = 5.95, s.d. = 2.68, satisfaction with weight at age 18 mean = 7.07 and s.d. = 2.58 None of the above

Satisfaction with current weight mean = 5.95, s.d. = 2.68, satisfaction with weight at age 18 mean = 7.07 and s.d. = 2.58

What can you conclude about the t test? Since the p-value is less than 0.05 we can reject the alternative hypothesis of difference among the means and conclude that there is no difference between them. The mean satisfaction with weight at age 18 is the same as the mean satisfaction with weight currently. Since the p-value is not less than 0.05 we cannot reject the null hypothesis of no difference among the means and conclude that there is no difference between them. Satisfaction with weight is not higher for age 18 than currently. Since the p-value is less than 0.05 we can reject the null hypothesis of no difference among the means and conclude that there is a difference between them. Satisfaction with weight is higher for age 18 (mean = 0.69) than currently (mean = 0.53).

Since the p-value is less than 0.05 we can reject the null hypothesis of no difference among the means and conclude that there is a difference between them. Satisfaction with weight is higher for age 18 (mean = 0.69) than currently (mean = 0.53).

When a researcher wants to know if a relationship between two continuous variables is positive or negative, the part of the linear regression model she is most concerned with is the ______. Constant Standard error Slope

Slope

What does a significant test statistic tell us? The null hypothesis is false. There is an important effect. That the effect seen in a sample is larger than we would expect to occur by chance if, in fact, there were no effect in the population. All of the above.

That the effect seen in a sample is larger than we would expect to occur by chance if, in fact, there were no effect in the population.

The correlation between two variables A and B is 0.12 with significance of p < 0.05. What can we conclude? That there is a small relationship between A and B. That there is a substantial relationship between A and B, though it is not statistically significant. That either A causes B or B causes A.

That there is a small relationship between A and B.

In reporting differences in means between two experimental groups, a journal article reports t-statistics in lieu of p-values. All comparisons involve the same two groups (i.e. sample sizes are fixed). Which statement is true with respect to how these t-statistics can be interpreted? The further the t-statistic falls from 0, the smaller the associated p-value. Nothing can be deduced regarding associated p-values, because t-statistics and p-values are not related. The further the t-statistic falls from 0, the larger the associated p-value. Only positive t-statistics can be used to determine a p-value.

The further the t-statistic falls from 0, the smaller the associated p-value.

Using the Munro data, you want to test the hypothesis that total inventory of personal attitudes (IPA) (total) is correlated with self-confidence (confid). You first run histograms of these two variables. What can you say about the distributions of total IPA and self-confidence [histogram total] and [histogram confid]? Both histograms are normal The histogram for total is skewed to left and the histogram for self-confidence is considered normally distributed. Both histograms are skewed to the right The data is dichotomous, therefore the best graphs to display this data are bar graphs.

The histogram for total is skewed to left and the histogram for self-confidence is considered normally distributed.

If a correlation coefficient has an associated probability value of .02 then: There is only a 2% chance that we would get a correlation coefficient this big (or bigger) if the null hypothesis were true. The results are important. We should accept the null hypothesis. The hypothesis has been proven.

The hypothesis has been proven.

Which statement is true about the following pair of linear equations? y = 1 + 3X and y = 2 +3X. The lines have negative slopes. The lines are parallel. The lines are perpendicular.

The lines are parallel.

What is the mean and standard deviation for life purpose and satisfaction? [sum life]. The mean = 15.02 and the standard deviation = 105 The mean = 89.75 and the standard deviation = 23 to 119 The mean = 89.7 and the standard deviation = 17.08

The mean = 89.7 and the standard deviation = 17.08

Which of the following is an alternative (as opposed to a null) hypothesis? The mean heights for groups A and B are the same. The mean height for group A is greater than the mean for group B. The mean height for group A is equal to the mean height for group B. There is no association between group membership and height.

The mean height for group A is greater than the mean for group B.

In a symmetric distribution: The mean is equal to the median. The mean is less than the median. The mean is greater than the median. The tail points toward the left.

The mean is equal to the median.

In a positively skewed distribution: The mean is equal to the median. The tail points toward the left. The mean is less than the median. The mean is greater than the median.

The mean is greater than the median.

In a negatively skewed distribution: There can be at most 1 mode. The mean is less than the median. The mean is equal to the median. The tail points toward the right.

The mean is less than the median.

A researcher finds a significant difference between two groups using a non-parametric test. At minimum, she should report: The F statistic The median for each group The mean for each group The standard deviation for each group

The median for each group

What can you say about the STATA output and the Wilcoxon's Matched-Pairs Signed-Ranks Test? Select all that apply. The p-value is -1.94, this is not significant. The null hypothesis is that there is no difference in the number of hypotheses residents wrote before and after their residence. The p-value for the signed-rank test is 0.0519 which is not less than 0.05 and therefore we cannot reject the null hypothesis and conclude that there is no difference in the number of hypotheses residents wrote before and after their residence. All are true

The null hypothesis is that there is no difference in the number of hypotheses residents wrote before and after their residence. The p-value for the signed-rank test is 0.0519 which is not less than 0.05 and therefore we cannot reject the null hypothesis and conclude that there is no difference in the number of hypotheses residents wrote before and after their residence.

State the null and alternative hypotheses in words and equation format for the paired sample t-test. The null hypothesis states that there is no difference in mean satisfaction with weight between current weight and weight at age 18. H0: δ = 0. The alternative hypothesis states that there is a difference in mean satisfaction with weight between current weight and weight at age 18. The difference is non-zero, HA: δ ≠0. The null hypothesis states that there is a difference in mean satisfaction with weight between current weight and weight at age 18. The difference is non-zero, HA: δ ≠0. The alternative hypothesis states that there is no difference in mean satisfaction with weight between current weight and weight at age 18. H0: δ = 0.

The null hypothesis states that there is no difference in mean satisfaction with weight between current weight and weight at age 18. H0: δ = 0. The alternative hypothesis states that there is a difference in mean satisfaction with weight between current weight and weight at age 18. The difference is non-zero, HA: δ ≠0.

Run a simple logistic regression predicting lymph node involvement by acid phosphatase. What can you say about the model results? Select all that apply [logistic nodes acid]. The odds ratio indicates that for every unit increase in acid phosphatase the odds of having lymph node involvement increase by 1.02. Acid phosphatase does not significantly predict lymph node involvement, the z = 1.62 and p-value = 0.104. The p-value is not less than 0.05 and therefore we cannot reject the null hypothesis. The 95% confidence interval includes the null value of 1, and therefore acid phosphatase is not a significant predictor of lymph node involvement. None are correct

The odds ratio indicates that for every unit increase in acid phosphatase the odds of having lymph node involvement increase by 1.02. Acid phosphatase does not significantly predict lymph node involvement, the z = 1.62 and p-value = 0.104. The p-value is not less than 0.05 and therefore we cannot reject the null hypothesis. The 95% confidence interval includes the null value of 1, and therefore acid phosphatase is not a significant predictor of lymph node involvement.

Run a simple logistic regression predicting lymph node involvement by stage of disease. What can you say about the model results? Select all that apply [logistic nodes stage]. The odds ratio indicates that for those with an advanced stage of disease the odds of having lymph node involvement increase by 5.25. Stage of disease significantly predicts lymph node involvement, the z = 2.63 and p-value = 0.009. The p-value is less than 0.05 and therefore we can reject the null hypothesis. The 95% confidence interval does not include the null value of 1, and therefore stage of disease is a significant predictor of lymph node involvement. None are correct

The odds ratio indicates that for those with an advanced stage of disease the odds of having lymph node involvement increase by 5.25. Stage of disease significantly predicts lymph node involvement, the z = 2.63 and p-value = 0.009. The p-value is less than 0.05 and therefore we can reject the null hypothesis. The 95% confidence interval does not include the null value of 1, and therefore stage of disease is a significant predictor of lymph node involvement.

Which of the following statements is false regarding the relationship between the p-value and alpha? Alpha is fixed prior to conducting a statistical test, whereas the p-value is reported upon completion of the statistical test. The p-value and alpha are the same thing. If the p-value is greater than alpha, then the null hypothesis is accepted (not rejected). If the p-value is less than alpha, then the null hypothesis is rejected.

The p-value and alpha are the same thing.

Produce a chi-square test to test the hypothesis that males and females differ in their work status [tab work gender, chi col]. What is the p-value for the chi-square test and is the p-value significant? The p-value is 51.77 this is not less than 0.05 and therefore it is not significant The p-value is <0.0001 this is less than 0.05 and therefore it is significant The p-value is 79.53 this is not less than 0.05 and therefore it is not significant

The p-value is <0.0001 this is less than 0.05 and therefore it is significant

Which of the following does not impact the power of a statistical test? The practical relevance of the research question Choice of alpha The effect size (e.g. difference in means) Sample size

The practical relevance of the research question

Which of the following is a null (as opposed to an alternative) hypothesis? The proportion of asthma for groups A and B are not the same. There is a relationship between group membership and asthma. The proportion of asthma is the same in groups A and B. The proportion of asthma among group A is greater than the proportion for group B.

The proportion of asthma is the same in groups A and B.

The R-squared value in either a correlation or regression model tells us: How many variables are included in the model The proportion of variance in the outcome variable that can be explained by the model. All of these responses, above or below, are correct.

The proportion of variance in the outcome variable that can be explained by the model.

Given a sample size greater than 1, which of the following statements regarding a sampling distribution and its underlying population distribution is true? The sampling distribution always resembles the shape of the population distribution. Because the samples are drawn at random from the population, no generalizations can be made regarding its properties. The sampling distribution is always narrower than the population distribution. The sampling distribution is always wider than the population distribution.

The sampling distribution is always narrower than the population distribution.

If we were to remove the 10 in this data set: 0, 1, 1, 2, 2, 2, 3, 3, 5, 10 The standard deviation would decrease. None of the statements, above or below, can be deduced. The standard deviation would increase The standard deviation would not be affected.

The standard deviation would decrease.

What is the relationship between sample size and the standard error of the mean? The standard error increases as the sample size increases. The standard error is unaffected by the sample size. The standard error decreases as the sample size decreases. The standard error decreases as the sample size increases.

The standard error decreases as the sample size increases.

A Wilcoxon signed-rank test comparing pre and post test scores for 20 subjects yields a p-value of 0.003. The median score on the 1st test was 10. The median score on the 2nd test was 15. Which of the following is an appropriate manner of reporting these results? Assume an alpha level of 0.05. The subjects scored higher on the post-test (median = 15) as compared to the pre-test (median = 10), however this difference was not statistically significant (p = 0.003). The subjects scored higher on the post-test (median = 15) as compared to the pre-test (median = 10). This difference was statistically significant (p = 0.003). None of these are correct.

The subjects scored higher on the post-test (median = 15) as compared to the pre-test (median = 10). This difference was statistically significant (p = 0.003).

An investigator accidentally runs an ANOVA when comparing means between only two groups (as opposed to 3 or more groups). He then runs a t-test that assumes equal variances on the same data. Which of the following statements best describes what he will find? The t-test will yield a different p-value than the ANOVA, but it is impossible to tell whether it will be smaller or larger. The t-test will yield a larger p-value than the ANOVA. The t-test will yield the same p-value as the ANOVA. The t-test will yield a smaller p-value than the ANOVA.

The t-test will yield the same p-value as the ANOVA.

What is reasonable explanation for a set of scores in which the mean is 81 and the median is 68? There are many typical scores. There are too many scores. There are two extremely high scores. There is one extremely low score.

There are two extremely high scores.

What can you conclude from the chi-square test results? There is a significant difference in work status between males and females. A larger proportion of males have full-time work (79.5%) compared to females (52.27%). There is no significant difference in work status between males and females. Males and females have the same work proportions.

There is a significant difference in work status between males and females. A larger proportion of males have full-time work (79.5%) compared to females (52.27%).

You want to determine the correlation coefficient and p-value for the association between total IPA and self-confidence, what are your findings [pwcorr total confid, sig]? There is no significant correlation between total IPA and self-confidence, r = 0.0000, p-value is 0.917. There is a strong positive correlation between total IPA and self-confidence, r = 0.91, p-value is <0.00001. There is a strong negative correlation between total IPA and self-confidence, r = -0.917, p-value is <0.00001. There is a strong positive correlation between total IPA and self-confidence, r < 0.00001, p-value = 0.91.

There is a strong positive correlation between total IPA and self-confidence, r = 0.91, p-value is <0.00001.

You want to investigate the relationship between total IPA and self-confidence, you run a scatterplot. What can you say about the relationship between total IPA and self-confidence [scatter total confid, ||lfit total confid]? There is a strong negative relationship between total IPA and self-confidence, higher levels of self-confidence correspond to lower levels of total IPA. There is no relationship between total IPA and self-confidence, higher levels of self-confidence correspond to no changes in total IPA. There is a strong positive relationship between total IPA and self-confidence, higher levels of self-confidence correspond to higher levels of total IPA.

There is a strong positive relationship between total IPA and self-confidence, higher levels of self-confidence correspond to higher levels of total IPA.

If a correlation coefficient has an associated probability value of .02 then: The hypothesis has been proven. There is only a 2% chance that we would get a correlation coefficient this big (or bigger) if the null hypothesis were true. The results are important. We should accept the null hypothesis.

There is only a 2% chance that we would get a correlation coefficient this big (or bigger) if the null hypothesis were true.

A researcher measures heartrate before and after a relaxation intervention on a group of 100 students randomly selected from the UA undergraduate program. The average reduction in heart rate was 2.5 beats per minute, with a confidence interval of -1 to 6. What can we conclude? This intervention will reduce heartrate among UA students, on average, because the confidence interval does not contain 0. This intervention will reduce heartrate among UA students, on average, because the mean reduction in heartrate for the sample was 2.5 beats per minute. This intervention may not reduce heartrate among UA students, on average, because the confidence interval contains 0. There was not really a mean reduction of 2.5 beats per minute in this sample because the confidence interval contains 0.

This intervention may not reduce heartrate among UA students, on average, because the confidence interval contains 0.

A meta-analysis involves computing effect sizes for a series of studies that investigated the same research question, and taking a weighted average of those effect sizes. True False

True

A type I error occurs when: We conclude that there is not an effect in the population when in fact there is. The data we type into a spreadsheet is different from the data collected. We conclude that the test statistic is significant when in fact it is not. We conclude that there is an effect in the population when in fact there is not.

We conclude that there is an effect in the population when in fact there is not.

A type II error occurs when: We conclude that there is not an effect in the population when in fact there is. We conclude that the test statistic is significant when in fact it is not. The data we type into a spreadsheet is different from the data collected. error. We conclude that there is an effect in the population when in fact there is not.

We conclude that there is not an effect in the population when in fact there is.

What is one reason that the t-test more commonly used than the Z-test? We seldom know the population standard deviation, which is required for the Z-test. When a t-test and Z-test are performed using the same data, the t-test yields a smaller p-value. We usually know the population standard deviation, which is required for a t-test. A z-test is only preferred when the sample size is less than 30.

We seldom know the population standard deviation, which is required for the Z-test.

When might the mean be a better measure of central tendency than the median? When it is desirable that all observations are taken into account through inclusion in the calculation. When the distribution is multimodal. When the distribution is asymmetric. Never. The median is always a superior measure of central tendency compared to the mean.

When it is desirable that all observations are taken into account through inclusion in the calculation.

When might the median be a better measure of central tendency than the mean? When the data include extreme values that might skew, or pull, the mean to one side or the other. Never. The mean is always a superior measure of central tendency compared to the median. When the data have a single mode. When the data are symmetrically distributed.

When the data include extreme values that might skew, or pull, the mean to one side or the other.

A polling agency reports that a 30% approval rating of congress, with a margin of error of 5%. Which interpretation of this is incorrect? 30% of survey respondents indicated that they approve of congress, but the actual percentage in the population might be a bit more or a bit less. The confidence interval for approval of congress is 25% to 35%. While 30% of the sample indicated approval, it is more likely that only 5% approve of congress. If this poll were conducted thousands of time, it would be expected that the actual approval rate in the population would be captured 95% of the time.

While 30% of the sample indicated approval, it is more likely that only 5% approve of congress.

The _________test is calculated by ranking all of the scores from two independent groups from lowest to highest and adding up the ranks separately for each group. Wilcoxon signed-rank Wilcoxon rank-sum Kruskal-Wallis test Chi-square test

Wilcoxon rank-sum

The _________ test is based on calculating the difference between two sets of scores, making a note of the sign of the difference (positive or negative), and then ranking the differences from lowest to highest. Wilcoxon rank-sum Chi-square test Wilcoxon signed-rank Kruskal-Wallis test

Wilcoxon signed-rank

A group of 20 mice twins was selected at birth. One mouse from each twin set was randomly assigned to either the experimental or control group, and its sibling was placed in the opposite group. The mice were thus paired on the basis of genetic sameness. The experimental group was raised in a shared environment with toys and were allowed out of their cages daily to explore. The control group spent the entire time alone in dimly lit cages with no toys. Each mouse was then put through a series of resilience tests and assigned an overall score for resilience. It was noted that for this particular sample of mice, resilience scores were highly right skewed. That is, most mice scored within a small range, but a few mice had much higher scores. What test should the researcher use to test for a difference in resilience between experimental and control mice? independent samples t-test paired-samples t-test One-way ANOVA Wilcoxon signed-rank test Pearson correlation coefficient

Wilcoxon signed-rank test

A researcher wants to determine whether people's ability to identify objects with their right eye differs from their ability with their left eye. 12 subjects were presented with a series of images and were scored on their abilities to identify objects which each eye. Identification scores ranged from 10 to 50 and were left skewed. What test should be used? One-way ANOVA Mann-Whitney Rank-sum test Wilcoxon signed-rank test Chi-square test Pearson correlation coefficient

Wilcoxon signed-rank test

A researcher wants to understand people's physiological reactions to horror films. She measures blood pressure on 20 subjects both before and after viewing a horror film. She notices at least one considerably large outlier at both time points. What test should be used to analyse the data? Mann Whitney rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Friedman's ANOVA

Wilcoxon signed-rank test

Run a multiple logistic regression model predicting lymph node involvement including the three predictor variables in the model, serum acid phosphatase, X-ray results, and stage of disease. Interpret the results and determine which is the strongest predictor for lymph node involvement. Select all that apply [logistic nodes acid xray stage]. X-ray (z=2.65, p-value = 0.008) and stage of disease (z=2.38, p-value = 0.018) significantly predict lymph node involvement. Acid phosphatase (z=1.63, p-value = 0.103) is not a significant predictor of lymph node involvement. X-ray (odds ratio = 7.86) has the highest odds of predicting lymph node involvement for this model. None are correct

X-ray (z=2.65, p-value = 0.008) and stage of disease (z=2.38, p-value = 0.018) significantly predict lymph node involvement. Acid phosphatase (z=1.63, p-value = 0.103) is not a significant predictor of lymph node involvement. X-ray (odds ratio = 7.86) has the highest odds of predicting lymph node involvement for this model.

In a case-control study of recreational drug use and depression, cases and matched controls are each interviewed by interviewers who are not blinded as to whether the subject is a case or a control. Many of the interviewers are in fact convinced that recreational drug use is a cause of depression. Is this likely to represent a bias? No, because the interviewers cannot affect whether the subjects are considered cases or controls; that has already been decided. Yes, a bias toward finding a positive relationship between drug use and depression. Yes, a bias toward finding no relationship between drug use and depression. Yes, but the direction of bias cannot be predicted.

Yes, a bias toward finding a positive relationship between drug use and depression.

Run a one-way ANOVA using life purpose and satisfaction (life) as the dependent variable and work (work) as the independent variable. Does mean life purpose and satisfaction differ by work status? [oneway life work, tab bonferroni] No, the p-value is 5.26 which is not less than 0.05 and therefore we cannot reject the null hypothesis and conclude that there is no difference in mean life purpose and satisfaction and the categories of work. Yes, the p-value is 0.0049 which is less than 0.05 and therefore we can reject the null hypothesis and conclude that at least one mean is different.

Yes, the p-value is 0.0049 which is less than 0.05 and therefore we can reject the null hypothesis and conclude that at least one mean is different.

According to the Bonferroni test results, is there a significance difference in life purpose and satisfaction between those who are unemployed and those who are employed full-time? Yes, the p-value is 0.012 which is less than 0.05, and therefore we can conclude that there is a difference in means. No, the p-value is 5.42 which is not less than 0.05, and therefore we can conclude that there is no difference in means.

Yes, the p-value is 0.012 which is less than 0.05, and therefore we can conclude that there is a difference in means.

When conducting an hypothesis test to assess the difference between two proportions, the ___ distribution is commonly used: p F Z t

Z

It is known that for right-handed people, the dominant (right) hand tends to be stronger. For left-handed people who live in a world designed for right-handed people, the same may not be true. To test this, muscle strength was measured on the right and left hands of a random sample of 15 left-handed men and the difference (left - right) was found. The best analysis tool for these data would be: an independent samples t-test a paired t-test This is a qualitative study and does not lend itself to statistical analysis. a two sample test of proportions

a paired t-test

A researcher conducts a statistical test from her sample data, from which she concludes there is an association between sugar consumption and depression. If, in fact, there is not really an association between these two variables in the larger population, her error would best be described as: a type IV error a type II error a type III error a type I error

a type I error

A researcher conducts a statistical test from his sample data, from which he concludes there is no association between exercise level and credit card debt. If, in fact, there really is an association between these two variables in the larger population, his error would best be described as: a type I error a type II error a type IV error a type III error

a type II error

A polling agency reports that candidate A is expected to receive 47% of the vote, with a margin of error of 2%. This margin of error: can be interpreted as a confidence interval that is 4 percentage points wide. provides an interval estimate for the population proportion of 45% to 49%. all answers, above or below are correct. suggests that the true population proportion is estimated to fall within two percentage points of 47%.

all answers, above or below are correct.

A grad student has heard that pets improve health, and wants to test the hypothesis that the type of pet matters. She recruits participants from a free health screening booth at her college, measures their blood pressure and asks them about their pet ownership. She then runs a comparison of mean systolic blood pressure on all single pet owners, making a comparison those whose pet is a mammal vs. those who have a non-mammal pet. The best analysis tool for these data would be: a two sample test of proportions an independent samples t-test a paired t-test This is a qualitative study and does not lend itself to statistical analysis.

an independent samples t-test

A researcher uses zip-codes from a patient database to determine whether study subjects reside in a rural or urban setting. Zip code and residential setting are what types of data? continuous and categorical, respectively ordinal and continuous, respectively. both are categorical both are ordinal.

both are categorical

A chi-square test can be used when both the independent (predictor) and dependent (outcome) variable are: both categorical categorical and continuous, respectively continuous and categorical, respectively both continuous

both categorical

If you increase the sample size and confidence level at the same time, what will happen to the length of your confidence interval? make it smaller cannot be determined from the given information make it bigger it will stay the same

cannot be determined from the given information

A clinic collects data on preferred means of follow-up communication, as in phone call, email, letter sent to physical address. What type of data is this? ordinal categorical cannot be determined from information provided. continuous

categorical

Monthly income, to the nearest dollar, can be considered what type of variable? ordinal categorical cannot be determined from information provided. continuous

categorical

The amount of time needed to complete this quiz is an example of which type of variable? dichotomous continuous ordinal categorical

continuous

A researcher conducts a single survey in which participants are asked about dietary and exercise habits as well as current health status. This survey would best be described as: cross-sectional experimental ecologic longitudinal

cross-sectional

A 95% confidence interval is a range of values that _____ contains the true value of a population parameter. rarely never frequently always

frequently

If we calculate both an odds ratio and a relative risk from the same contingency table, the odds ratio will always be: the same as the relative risk. half of the relative risk. further from 1.0 than the relative risk. closer to 1.0 than the relative risk.

further from 1.0 than the relative risk.

A relative risk greater than 1.0 means that the risk of an event is the same in both the exposed and unexposed groups. greater in the exposed group compared to the unexposed group. greater in the unexposed group compared to the exposed group. This situation is impossible. A relative risk can never be greater than 1.0.

greater in the exposed group compared to the unexposed group.

A relative risk less than 1.0 means that the risk of an event is the same in both the exposed and unexposed groups. greater in the exposed group compared to the unexposed group. greater in the unexposed group compared to the exposed group. This situation is impossible. A relative risk can never be less than 1.0.

greater in the unexposed group compared to the exposed group.

A type II error can occur when the null hypothesis is: correct incorrect neither correct nor incorrect either correct or incorrect

incorrect

A consumer testing group compared two leading laundry detergents to determine which got laundry whiter. White towels that had been subjected to a variety of filthy treatments were randomly assigned to one of the two detergents. After washing, each set of towels was tested with a photometer for the amount of light reflected (testing their cleanliness). What is the statistical test that should be used to compare the two detergents? Assume the amount of light reflected is symmetrically distributed. independent samples t-test paired-samples t-test One-way ANOVA Wilcoxon signed-rank test Pearson correlation coefficient

independent samples t-test

In the linear equation y = b0 + b1X1, b0 represents the _______ and b1 represents the _______. intercept, slope slope, gradient residual, intercept

intercept, slope

How would you describe the distribution of the following data set? 1, 5, 10, 11, 11, 12, 12, 12, 13, 13, 14 left skewed right-skewed multimodal symmetric

left skewed

All of the following would be best considered as categorical variables, EXCEPT: marital status letter grade in statistics course type of car city of birth

letter grade in statistics course

Increasing the confidence level, while holding the sample size the same, will do what to the length of your confidence interval? make it bigger cannot be determined from the given information make it smaller it will stay the same

make it bigger

Increasing the sample size, while holding the confidence level the same, will do what to the length of your confidence interval? make it bigger make it smaller it will stay the same cannot be determined from the given information

make it smaller

For the data set below, calculate the mean, median, and mode. 0, 1, 1, 2, 2, 2, 3, 3, 5, 10 mean: 2; median: 2; mode: 2 mean: 2.9; median: 2.5; mode: 3 mean: 3; median: 2.5; mode: 3 mean: 2.9; median: 2; mode: 2

mean: 2.9; median: 2; mode: 2

Given that the sample size is sufficiently large, the sampling distribution of the mean will be normally distributed with a mean equal to the population mean. For this distribution, a Z-score of 1 represents: one standard above the population mean. one percentile above the population mean. one standard error above the population mean. one parameter above the population mean.

one standard error above the population mean.

College students are surveyed to rate their most recent interaction with the instructor, given choices of unsatisfactory, satisfactory, or exceeded expectations. What type of data is this? categorical continuous ordinal cannot be determined from information provided.

ordinal

College students are surveyed to rate their most recent interaction with the instructor, given choices of unsatisfactory, satisfactory, or exceeded expectations. What type of data is this? continuous categorical ordinal cannot be determined from information provided.

ordinal

Of what is p the probability? p is the probability that the results would be replicated if the experiment were conducted a second time. p is the probability that the null hypothesis (H0) is false. p is the probability of observing an effect at least as large as that found in the sample, if in fact, there were no effect in the population. p is the probability that the null hypothesis (H0) is true.

p is the probability of observing an effect at least as large as that found in the sample, if in fact, there were no effect in the population.

A company would like to determine if pay increases result in increased productivity. The company captures productivity through a measure of work units, which tends to be normally distributed. The number of work units is measured both before and after a pay increase for a random sample of employees. What statistical test should the company use to determine if the number of work units increases following a pay raise? paired-samples t-test One-way ANOVA Mann-Whitney Rank-sum test Chi-square test Pearson correlation coefficient

paired-samples t-test

A mouse study was conducted to test whether environmental stimuli affect brain growth. A group of 20 mice twins was selected at birth. One mouse from each twin set was randomly assigned to either the experimental or control group, and its sibling was placed in the opposite group. The mice were thus paired on the basis of genetic sameness. The experimental group was raised in a shared environment with toys and were allowed out of their cages daily to explore. The control group spent the entire time alone in dimly lit cages with no toys. After 3 months, the mice were sacrificed and the weight of the cortex was obtained and recorded in milligrams. Assuming normal weight distribution, what statistical test should be used to compare cortical weight between the experimental and control mice? independent samples t-test paired-samples t-test One-way ANOVA Logistic regression

paired-samples t-test

Which of the following tests would be best for determining if there is a difference in mean weight for a group of individuals before and after a diet program? Assume both weight and weight-change are normally distributed for this group. paired-samples t-test Mann-Whitney Rank-sum test Kruskal-Wallis test Chi-square test Spearman correlation coefficient

paired-samples t-test

Which of the following tests would be most appropriate to determine the difference in an individual's weight before and after a diet program? one sample t-test independent samples t-test paired-samples t-test One-way ANOVAMann-Whitney Rank-sum test Wilcoxon signed-rank test Kruskal-Wallis test Chi-square test Pearson correlation coefficient Spearman correlation coefficient Linear regressionLogistic regression

paired-samples t-test

A mean is known as a parameter if it is computed from the: Z-distribution statistics population sample

population

The binomial distribution is appropriate for constructing a confidence interval around a sample: median proportion variance mode

proportion

An outbreak of gastrointestinal illness occurs following a large community potluck picnic. In an effort to identify which food caused the outbreak, health workers conduct interviews among 60 picnic-goers, 20 of whom fell ill and 40 who did not. For each picnic-goer, the health workers record which foods were placed on the plate, and the amount of each consumed. This type of study would best be described as a: qualitative study prospective cohort study retrospective case-control study single blinded clinical trial

retrospective case-control study

How would you describe the distribution of the following data set? 0, 1, 1, 2, 2, 2, 3, 3, 5, 10 right skewed left skewed multimodal symmetric

right skewed

With respect to symbols representing the standard deviation, the difference between s and σ (sigma) is that: s is for a population and σ (sigma) is for a sample s is a parameter and σ (sigma) is not s is for a sample and σ (sigma) is for a population There is no difference. These symbols can be used interchangeably.

s is for a sample and σ (sigma) is for a population

A mean is known as a statistic if it is computed from the: Z-distribution parameters population sample

sample

All of the following would best be considered as continuous variables, EXCEPT: shirt size grade point average weight height

shirt size

How would you describe the distribution of the following data set? 0, 0, 1, 1, 2, 2, 3, 3, 4, 4 bell-shaped multimodal symmetric left skewed

symmetric

Non-parametric tests should be used when: you have symmetric distributions and large samples sizes for all data being assessed. you want to increase the power of your experiment. the assumptions of parametric tests have not been met.

the assumptions of parametric tests have not been met.

If the variance of a data set is correctly computed with the formula using n - 1 in the denominator, which of the following is true? the data set is a population the data set could be either a sample or a population the data set is from a census the data set is a sample

the data set is a sample

One problem with hypothesis testing is that a real effect may not be detected. This problem is least likely to occur when the effect is small and the sample size is small. the effect is small and the sample size is large. the effect is large and the sample size is large. the effect is large and the sample size is small.

the effect is large and the sample size is large.

One problem with hypothesis testing is that a real effect may not be detected. This problem is most likely to occur when the effect is large and the sample size is small. the effect is large and the sample size is large. the effect is small and the sample size is small. the effect is small and the sample size is large.

the effect is small and the sample size is small.

A variable is a characteristic or attribute that can assume different values. Based on this definition, all of the following are examples of variables, EXCEPT: student weights the height of the College of Nursing building temperature the height of maple trees

the height of the College of Nursing building

In a five number summary (i.e. the values used in creating a boxplot), which of the following is not used for data summarization? the median the largest value the 25th percentile the mean

the mean

A reviewer tells a research team that their choice of an independent t-test is not valid for their hypothesis because the two samples being compared are dependent. This means the observations in one sample are not correlated with the observations in the other sample. the observations in one sample reveal no information about those in the other sample. None of these responses is correct. Dependence of samples is unrelated to observations. the observations in one sample are related to the observations in the other sample.

the observations in one sample are related to the observations in the other sample.

Unlike a parametric test, a non-parametric test prevents an extreme outlier from being more influential than other data points because: the outlier receives a ranking that is just one unit different than the next largest (or smallest) observation. the outlier is eliminated from the data. None of these are correct. An outlier is equally influential when data are assessed by non-parametric and parametric methods.

the outlier receives a ranking that is just one unit different than the next largest (or smallest) observation.

Both the Chi-Square and Fisher's exact test measure the degree of imbalance between two categorical variables. If there is no imbalance (i.e. if cells are roughly balanced), then: All choices, above or below, are correct. we can conclude that there is a statistically significant association between the two variables. we will reject the null hypothesis. the p-value will be greater than 0.05.

the p-value will be greater than 0.05.

An odds ratio of 1.0 means that the odds of an event is _________ 100 times greater in the unexposed group compared to the exposed group. 100 times greater in the exposed group compared to the unexposed group. the same in both the exposed and unexposed groups. 10 times greater in the exposed group compared to the unexposed group.

the same in both the exposed and unexposed groups.

Fisher's exact test can become computationally complicated and may need to be avoided when: there are many categories in either or both of the variables being assessed. None of these responses, above or below, are correct. There are no caveats to the use of Fisher's exact test. an expected cell count is less than 5. both the independent (predictor) and dependent (outcome) variables are dichotomous.

there are many categories in either or both of the variables being assessed.

Which of the following is not a measure of central tendency? variance mode mean median

variance

All of the following would best be considered as ordinal variables, EXCEPT: weight in kilograms letter grade in a statistics course shirt size education experience (elementary school, high school, some college, college graduate)

weight in kilograms

A paired hypothesis test is not appropriate: when comparing scores for two mutually exclusive groups. when comparing scores for one group of individuals across two timepoints. when comparing pre and post intervention scores for a group of study participants. when comparing scores between cases and controls that have been matched according to age and other demographic factors.

when comparing scores for two mutually exclusive groups.

Your professor determines that exam scores for your entire class were normally distributed, and thus, gives the students results in the form of Z-scores. You receive a Z-score of 0. This indicates that: everyone in the sample had a higher score. you had an above average score. you had no answers correct on the exam. you had an average score.

you had an average score.

our professor determines that exam scores for your entire class were normally distributed, and thus, gives the students results in the form of Z-scores. You receive a positive Z-score. This indicates that: your score is in the upper half of the distribution. your professor curved the scores. you are below the mean on most tests, but you were not on this one. everyone in the sample had a lower score.

your score is in the upper half of the distribution.