EBP 2 - EXAM 1 2024/2025 WITH 100% ACCURATE SOLUTIONS, Exams of Nursing

Download EBP 2 - EXAM 1 2024/2025 WITH 100% ACCURATE SOLUTIONS and more Exams Nursing in PDF only on Docsity!

EBP 2 - EXAM 1 2024/2025 WITH 100%

ACCURATE SOLUTIONS

inferential statistic

• p value
• mean difference
• confidence interval - Precise Answer ✔✔Used to make inference about the population; provide examples descriptive statistic
• mean
• standard deviation
• range - Precise Answer ✔✔Used to describe the sample; provide examples probability and sampling error - Precise Answer ✔✔What two things are statistical inference assumptions based off? 68% - Precise Answer ✔✔What is the known probability between -1 SD and +1 SD? 95% - Precise Answer ✔✔What is the known probability between -2 SD and +2 SD? 99% - Precise Answer ✔✔What is the known probability between -3 SD and +3 SD? z score - Precise Answer ✔✔What term is basically the same as a standard deviation? sampling error - Precise Answer ✔✔Tendency for sample values (mean) to differ from population values (mean)

standard error of the mean - Precise Answer ✔✔the standard deviation of a sampling distribution; allows us to estimate population parameters (like confidence intervals) sampling distribution - Precise Answer ✔✔distribution of a statistic across an infinite number of samples central limit theorem - Precise Answer ✔✔The sampling distribution of means will approach normal as N increases (even if the underlying population is not normal)

• Smaller sample standard deviation (s)
• Larger n (# of people in sample) --> SEM = s / sqrt(n) - Precise Answer ✔✔What makes the standard error of the mean (SEM) smaller? point estimate - Precise Answer ✔✔a single value that represents the best estimate of the population value; always falls within the confidence interval confidence interval - Precise Answer ✔✔range of values that we are confident contains the population parameter precision - Precise Answer ✔✔The width of a confidence interval concerns the ______ of the estimate. The numbers closest to the point estimate - Precise Answer ✔✔Which number is a confidence interval is the most probable value?

1. larger sample size
2. less variance
3. lower selected level of confidence (like 90% CI instead of 95% - Precise Answer ✔✔How do you increase precision and thus decrease the width (narrow) the confidence interval? 1.96 - Precise Answer ✔✔What is the Z value for the 95% confidence interval?

1. Does this difference represent a real difference in the population?
2. Is this difference due to sampling error? - Precise Answer ✔✔What two questions are key thoughts for hypothesis testing? Null Hypothesis (H0) - Precise Answer ✔✔Hypothesis testing: differences represents a sampling error; states that the population means are equal (no true differences) Alternative Hypothesis (Ha) - Precise Answer ✔✔Hypothesis testing: difference represents a true and real difference Support the null hypothesis (no true difference) / fail to reject the null - Precise Answer ✔✔p value > alpha Reject the null (true difference) - Precise Answer ✔✔p value < alpha type 1 error (liar) - there is no difference but we said there was - Precise Answer ✔✔When the null is true but you reject the null AKA when there is no difference but you say there is alpha error / false positive - Precise Answer ✔✔another name for a type 1 error beta error / false negative - Precise Answer ✔✔another name for a type 2 error type 2 error (blind) - there's a difference but we failed to find it - Precise Answer ✔✔When the null is false but you don't reject it AKA when there is a real difference but you accept the null (say there isn't a true difference) alpha - Precise Answer ✔✔maximum probability of a type 1 error that is set by the researcher before running statistics Max chance of type 1 error (false positive) is 5% - Precise Answer ✔✔What does an alpha value of 0. mean?

p value - Precise Answer ✔✔probability of a type 1 error if the null hypothesis is true; probability of observing a value more extreme than the actual value observed if then null is true Alpha is set before the research while p value is calculated after and is compared to alpha - Precise Answer ✔✔What is the difference between a p value and alpha? TRUE - Precise Answer ✔✔T/F: Both alpha and p value are related to type I errors. sampling error - Precise Answer ✔✔If we fail to reject the null hypothesis, we attribute any observed difference to ____________ only. FALSE

• We don't interpret non-significant differences as real and shouldn't even consider them trends.
• Non-significant differences are only attributable to change.
• Don't use magnitude of p as a indication of validity (no such thing as highly significant) - Precise Answer ✔✔T/F: Even if the p value isn't less than alpha, we can still detect differences that are real trends in the data. Non-significant is the confidence interval contains 0 - Precise Answer ✔✔How can you use confidence intervals to determine if something is a significant difference? Statistical significance --> CIs and p value Effect size --> CIs only

• Neither tell us clinical significance! - Precise Answer ✔✔P values and confidence intervals both tell us about __________ while only confidence intervals give us an estimate of __________. statistical power - Precise Answer ✔✔The probability of finding a statistically significant difference if such difference exists in the real world 1 - type II error

AKA

1 - Beta - Precise Answer ✔✔Power =? Less statistical powerful - Precise Answer ✔✔More type II error means what?

1. alpha
2. effect size
3. variance
4. sample size - Precise Answer ✔✔What are the four pillars of power? increase sample size - Precise Answer ✔✔What is the best way to improve statistical power?

• Increasing alpha, effect size, and sample size
• Decreasing variance - Precise Answer ✔✔What three things increase power when increased? What thing increases power when decreased? power a priori - Precise Answer ✔✔power analysis prior to collecting data to ensure that the research design is powerful enough calculate one based on MCID and actual estimate of variance (way better than using standard effect sizes) - Precise Answer ✔✔What is the best way to perform a priori analysis? fail to reject the null AKA when you don't find a difference - Precise Answer ✔✔When do you perform a power post-hoc?

1. traditional cohen approach

2. confidence interval analysis of effect size (better way!) - Precise Answer ✔✔What are two ways to perform a post hoc power analysis? Which is better? Possibility that there is a true difference and you are lacking power so need to decrease type II error --> Not definitive and underpowered - Precise Answer ✔✔When performing a post-hoc power analysis, you noticed that the MCID falls within the upper bound of the confidence interval despite failing to reject the null, what is your conclusion? There definitely isn't a true difference ---> Definitely negative and adequately powered - Precise Answer ✔✔When performing a post-hoc power analysis, you noticed that the MCID falls outside the confidence interval despite failing to reject the null, what is your conclusion? adequate - Precise Answer ✔✔If MCID is excluded from confidence interval, then there is __________ power in a post hoc analysis. 2x3 mixed model design - Precise Answer ✔✔What do we call a study that has two independent variables one with 2 levels and the other with 3 levels? Alternate hypothesis (reject the null) - Precise Answer ✔✔"Is it likely real in the population?" Null hypothesis - Precise Answer ✔✔"Is it likely just sampling error?"

• effect size
• alpha
• sample size
• variance - Precise Answer ✔✔What information is used during the a-priori power analysis? t-test - Precise Answer ✔✔statistical method to decide whether an observed difference in sample scores represents a "real" difference in the population vs. just sampling error 2 groups (AKA 2 levels of 1 IV) - Precise Answer ✔✔How many groups is a t-test designed for?

All sources of variability within a set of data that cannot be explained by the independent variable (basically just variability between groups)

• does NOT mean mistakes or miscalculation - Precise Answer ✔✔What does 'error' in a t-test mean? within group variability - Precise Answer ✔✔portion of total variance that cannot be explained by research design (mean square for error) parametric statistics - Precise Answer ✔✔branch of statistics that assumes that sampel data comes from a population that follows a probability distribution based on a fixed set of parameters t-test ANOVA ANCOVA Regression Correlation - Precise Answer ✔✔What are some examples of parametric statistics?

1. Samples are randomly drawn from populations
2. Population is normally distributed
3. Homogeneity of variance (spread of scores within groups)
4. Data from ratio or interval (continuous) scales - Precise Answer ✔✔What are the four basic assumptions for parametric tests? Random sampling - Precise Answer ✔✔Which of the four basic assumptions of parametric tests rarely happens? Levene's test for t-test equal variance --> most important with unequal group sizes - Precise Answer ✔✔What is usually used to statistically test the homogeneity of variance for parametric testing? When is this most important? TRUE - Precise Answer ✔✔T/F: Using NPRS data is okay for parametric testing despite being ordinal, it's just not ideal. On the other hand, MMT scores should not be used because it's too ordinal.

Nondirectional "Group A will score different from Group B" - Precise Answer ✔✔A two-tailed t-test is a __________ hypothesis. Directional "Group A will score better than Group B" - Precise Answer ✔✔A one-tailed t-test is a ______ hypothesis. One tailed t-test (directional hypothesis)

• as long as the difference is in the hypothesized direction - Precise Answer ✔✔Which type of t-test (one or two tailed) has more statistical power? Between subject design for two groups - Precise Answer ✔✔When do you use an independent (also known as unpaired t test)? Mean change - Precise Answer ✔✔How do you eliminate time as an independent variable during t testing to ensure you're comparing two means only? p value < alpha (usually 0.05) - Precise Answer ✔✔When doing Levene's test for equality of variances, what p value would indicate a difference between groups? Greater - Precise Answer ✔✔If calculated t value is _____ than the critical t value of alpha then the the results are statistically significant. When there is one group (within subjects design) - Precise Answer ✔✔When do you use an paired t test? Paired - less error variance than in independent groups (If there's a difference we're more likely to find it) - Precise Answer ✔✔Which is more statistically powerful, paired or unpaired t-test?

Homogeneity of variance --> variance is assumed same since it's the same people (within groups) - Precise Answer ✔✔Which of the four main assumptions for parametric tests do you not have to worry about as much with a paired t-test? Compares 3 or more groups (levels of 1 IV) OR 2 or more IVs - Precise Answer ✔✔ANOVA is basically a t-test but with what difference? Are observed differences in whole set of means greater than would be expected by chance alone? - Precise Answer ✔✔What is the question that ANOVA testing aims to answer? F statistic - Precise Answer ✔✔What is the 't statistic' equivalent of ANOVA? How many independent variables there are Ex: one-way ANOVA has one independent variable while two-way has two, etc. - Precise Answer ✔✔One way vs. 2-way vs. 3-way ANOVA just concerns what? Repeated measures --> within subject Independent groups --> between subject - Precise Answer ✔✔A repeated measures ANOVA just means _____ subject while an independent groups ANOVA just means ______ subject. At least one IV is within subject and one is between - Precise Answer ✔✔What is meant by a mixed model ANOVA? F statistic - Precise Answer ✔✔? = (variance between grand mean and each group mean) / (variance within each group) 2-way ANOVA

• Factorial means multiple independent variables, in this case two so 2 way ANOVA - Precise Answer ✔✔What type of test would you run on a factorial (two) design experiment?

• Main effect of A
• Main effect of B
• Interaction of A and B - Precise Answer ✔✔What are the findings from a 2x3 ANOVA independent samples? Line parallel (no crossing) --> No interaction AKA "Scores across levels of factor A do not depend on the levels of factor B" - Precise Answer ✔✔What do parallel lines signify in a independent sample ANOVA? disordinal interaction - Precise Answer ✔✔When the lines cross and we cannot interpret significant main effects, what type of interaction is this? ordinal interaction - Precise Answer ✔✔When the lines don't cross and we can interpret significant main effects, what do we call this interaction? disordinal - Precise Answer ✔✔"Flexion vs. extension is better depending upon stretch type" is an example of what type of interaction? ordinal - Precise Answer ✔✔"Prolonged stretching is best regardless of position of the knee" is an example of what type of interaction? Like a paired t-test - Precise Answer ✔✔A one-way ANOVA with repeated measures samples is similar to what type of test but instead it has more than two levels of time? Repeated measures - Precise Answer ✔✔Which ANOVA is more powerful: repeated or independent samples? sphericity - Precise Answer ✔✔Homogeneity of variance of difference

No difference in variance for ANOVA testing

• If there is a significant finding, just need to adjust p value. - Precise Answer ✔✔If there is a non- significant finding with Mauchly's test of sphericity, what does this mean? 2x2 ANOVA Mixed Design
• At least 1 between-subjects IV and at least 1 within-subjects IV (usually time) - Precise Answer ✔✔What is the principal statistical analysis in a RCT? multiple comparison tests - Precise Answer ✔✔What do you use to make pairwise comparisons, or in other words determine where the difference is?

1. Post-hoc
2. Planned comparisons (a priori) - Precise Answer ✔✔What are the two different strategies for multiple comparisons tests? Multiply alpha by the number of post-hoc tests - Precise Answer ✔✔How do you find the family-wise type 1 error rate? Decrease family-wise error rate --> Bonferroni Correction - Precise Answer ✔✔What is the goal of multiple comparison tests? What's the solution? Divide alpha by the number of statistical tests and use that alpha value for each post-hoc test - Precise Answer ✔✔How do you perform a Bonferroni Correction? Too conservative (when you decrease the risk of type 1 error you increase the risk of type 2 error) - Precise Answer ✔✔What is the downside to using Bonferroni correction? Tukey's Honestly Significant Difference or Sidak - Precise Answer ✔✔Which Multiple comparison test is the middle of the road when it comes to risk of type 1 vs. 2 error and most commonly used test?

Fisher's Least Significant Difference (LSD) - basically unadjusted t tests - Precise Answer ✔✔Which multiple comparison test is the most powerful (and has the greatest chance of type 1 error)? Bonferroni t test - Precise Answer ✔✔Which multiple comparison test is the least powerful and has the largest chance for type 2 error? No! Only two levels so we know where the difference is between - Precise Answer ✔✔When there are two levels of an IV, do we need to perform post-hoc testing? omnibus test - Precise Answer ✔✔The main statistical test used to analyze data in a study is called the what? 1 tailed repeated measures ANOVA - Precise Answer ✔✔Which type of ANOVA has the greatest statistical power? Percentage chance of type 1 error - Precise Answer ✔✔What does alpha tell you? When the assumptions for parametric tests are not met ex: skewed distributions (not normal distribution), heterogenous variance between groups, non- continuous data (ordinal or nominal data) - Precise Answer ✔✔When do you use a nonparametric test? ordinal - Precise Answer ✔✔rank, order level of measurement nominal data - Precise Answer ✔✔data of categories only. Data cannot be arranged in an ordering scheme. Parametric tests

• This is because nonparametric tests lose some valuable data from switching measurements into ranks / counts - Precise Answer ✔✔Which type of tests are more powerful: nonparametric or parametric?

Only 65-95% as powerful - Precise Answer ✔✔How powerful are nonparametric tests compared to parametric tests?

• Comparison of ranks of scores
• Comparison of counts or signs of scores - Precise Answer ✔✔What are nonparametric tests based off? P: Unpaired t-test NP: Mann-Whitney U test - Precise Answer ✔✔When there are two independent groups that need to be compared what is the parametric and nonparametric tests that could be used to analyze the data? P: Paired t-test NP: Sign test / Wilcoxon signed-ranks test - Precise Answer ✔✔When there are two related scores that need to be compared what is the parametric and nonparametric tests that could be used to analyze the data? P: One-way ANOVA NP: Kruskal-Wallis ANOVA by ranks - Precise Answer ✔✔When there are three or more independent groups that need to be compared what is the parametric and nonparametric tests that could be used to analyze the data? P: One-way repeated measures ANOVA NP: Friedman two-way ANOVA by ranks - Precise Answer ✔✔When there are three or more related scores that need to be compared what is the parametric and nonparametric tests that could be used to analyze the data? Unable to perform on more complex designs (like a mixed 2x3 study design) - Precise Answer ✔✔What is one of the downsides of nonparametric testing in regard to research design?

Unpaired t-test

• ANOVA is used for 3 more groups - Precise Answer ✔✔If you have one independent variable with two levels and are using a between groups design, what type of statistical test can you use assuming it meets the requirements for a parametric test (ex: normal distribution, continuous data, etc.)? Mann-Whitney U test - Precise Answer ✔✔If you have one independent variable with two levels and are using a between groups design, what type of statistical test can you use assuming it does not meet the requirements for a parametric test (ex: normal distribution, continuous data, etc.)? Paired: within group ("Is there a difference between conditions in the same person?" Unpaired: between group ("Is there a difference between groups?") - Precise Answer ✔✔When do you use paired vs. unpaired t test? When there is a confounding variable (Covariate) - Precise Answer ✔✔When do you choose ANCOVA over a one-way ANOVA? Kruskal-Wallis ANOVA - Precise Answer ✔✔The research study has more than 3 independent groups ( IV with 3 levels) and doesn't meet the requirements to be a one-way ANOVA. What would the nonparametric test equivalent be? Friedman ANOVA - Precise Answer ✔✔A study doesn't meet the requirements to be a one-way repeated measures ANOVA but had one group with repeated measures. What is the nonparametric equivalent? Ranks - each score is converted to a rank "Is the difference in ranks larger than what would be expected by chance alone? - Precise Answer ✔✔The nonparametric equivalent of an independent t-test is based on what? Signs - Each score is converted to a sign or other dichotomous decision (ex: smaller, larger, or no difference)

"Is the difference in sign frequencies larger than would be expected by chance alone?" - Precise Answer ✔✔The nonparametric equivalent of a paired t t-test is based on what? IV: nominal DV: continuous - Precise Answer ✔✔What are the IV and DV levels of measurement for both t-tests and ANOVA? IV: Nominal DV: ordinal - Precise Answer ✔✔What are the IV and DV levels of measurement for nonparametric tests? Both IV and DV are continuous - Precise Answer ✔✔What are the IV and DV levels of measurement for correlation and regression? What is the strength of the association? - Precise Answer ✔✔What question does correlation aim to address? What is the strength of prediction? - Precise Answer ✔✔What question does regression aim to address? What is the difference between means? - Precise Answer ✔✔What question do both t-tests and ANOVA aim to address? Are the ranks different? - Precise Answer ✔✔What question does nonparametric testing aim to address? covary - Precise Answer ✔✔when change in one variable is associated with change in another variable; when variables vary together Linear relationships only

• since based off equation for a straight line
• this means there could be a trend in the data but unless it is linear, the correlation tests won't catch it (for example, this is why plotting the points is so critical. You could find a curvilinear relatioinship) - Precise Answer ✔✔Correlations quantify strength of what? -1 to +1 - Precise Answer ✔✔What is the range for the correlation coefficient (r)? Direction of relationship
• positive vs. negative correlation - Precise Answer ✔✔What does the sign of the correlation coefficient indicate? Closer to -1 or +1 - Precise Answer ✔✔What correlation coefficient indicates a stronger relationship? Higher correlation coefficient - Precise Answer ✔✔Tighter grouping (or a more defined clear line on a scatter plot) indicates what? 0 - 0.25 - Precise Answer ✔✔What r value range indicates little to no relationship? 0.26 - 0.5 - Precise Answer ✔✔What r value range indicates a fair relationship? 0.51 - 0.75 - Precise Answer ✔✔What r value range indicates a moderate to good relationship? 0.75 - 1 - Precise Answer ✔✔What r value range indicates a good to excellent correlation? Depends on what exactly you're looking at
• Ex: 0.2 correlation can be important! It's very rare in research to be above 0.5 correlate - Precise Answer ✔✔Although there are general guidelines for interpreting the r correlation value, what is most important to take into consideration when determining if it is important? coefficient of determination - Precise Answer ✔✔The percent of variance in y that is explained or accounted for by x; or simply the amount of overlap

The coefficient of determination - more directly interpretable than simply r value because tells you the percent of variance that is explained between two variables - Precise Answer ✔✔What is the square of the correlation coefficient (aka r^2)? correlation matrix

• AKA table of correlation values - Precise Answer ✔✔What is an efficient way to look at a bunch of data and run correlation tests? The correlation between variable x and variable y is not significantly different from zero AKA the correlation r is 0 (not correlated) - Precise Answer ✔✔What is the null hypothesis in correlation testing? Sample size
• Trivial r coefficient are often statistically significant if the sample is large enough - Precise Answer ✔✔The r correlation coefficient is very sensitive to what? Pearson Product-Moment Correlation Coefficient (r) - Precise Answer ✔✔Which correlation coefficient is used when both variables are continuous and is the most common? Spearman Rank (rho) Correlation Coefficient (rs) - Precise Answer ✔✔Which correlation coefficient is used as a non-parametric analog of Pearson r consisting of at least one ordinal variable? Point biserial correlation (r.pb)
• ex: gender vs. height - Precise Answer ✔✔Which correlation coefficient is used when one variable is dichotomous and the other is continuous (interval or ratio)? The correlation gives us strength of relationship while t-test only gives statistical significance - Precise Answer ✔✔What is the difference between a t test and a point biserial correlation? Rank biserial correlation (r.rb)
• ex: MMT (ordinal) vs. gender (nominal) - Precise Answer ✔✔Which correlation coefficient is used when one variable is dichotomous (nominal) and the other is ordinal?

Phi coefficient

• ex: gender and group - Precise Answer ✔✔What is computationally the same as Pearson's r but both variables are dichotomous? There is none! Data is completely dichotomous so all the points would stack on top of eachother - Precise Answer ✔✔What is the purpose behind using a scatter plot when calculating the phi coefficient? Chi square test - Precise Answer ✔✔What is the phi coefficient similar to? Pearson r - Precise Answer ✔✔Which correlation coefficient to use: ratio vs. ratio data? Spearman rho - Precise Answer ✔✔Which correlation coefficient to use: ordinal vs. ratio data? Spearman rho - Precise Answer ✔✔Which correlation coefficient to use: ordinal vs. ordinal? point biserial - Precise Answer ✔✔Which correlation coefficient to use: nominal vs. ratio? point biserial - Precise Answer ✔✔Which correlation coefficient to use: ordinal vs. nominal? phi coefficient - Precise Answer ✔✔Which correlation coefficient to use: nominal vs. nominal? covariance - Precise Answer ✔✔As one variable changes, the other variable also changes Yes! That is why you need to be cautious with correlation interpretation
• correlation does not assess differences or agreement (that's what ICCs do!) - Precise Answer ✔✔Is it possible to have a statistically significant difference AND excellent correlation coefficient?

RCTs - Precise Answer ✔✔What type of experiment allows you to make causal statements?

1. correlation vs. agreement
2. correlation vs. causation
3. extreme outliers
4. limits in range of score (AKA can't generalize beyond range of scores) - Precise Answer ✔✔What are four cautions of interpretation when thinking about correlation coefficients?

• can create inflated correlation with only a few extreme data points --> Which is why scatterplots are helpful to see outliers - Precise Answer ✔✔What is the problem with extreme outliers when it comes to correlation? What is one way we can address this? Limited range of scores in sample - Precise Answer ✔✔What is one reason that something might have a low correlation despite there being a true correlation? Used to analyze categorical data (nominal data) AKA both the IV and DV are nominal - Precise Answer ✔✔When is a chi square test used?

1. Goodness of fit
2. Tests of independence (association) - Precise Answer ✔✔What are the two types of chi square? Goodness of fit compares to uniform frequencies while association compares to a contingency table

• Association is more common - Precise Answer ✔✔What's the difference between goodness of fit and tests of independence (association) chi square tests? reliability

• expected to repeat the same score on two different occasions provided that the characteristic of interest doesn't change
• tied to concept of measurement error - Precise Answer ✔✔extent to which a measurement is consistent and free from error

• pearson correlation (r)
• intraclass correlation coefficient (ICC) - Precise Answer ✔✔What are the reliability coefficients for continuous (interval/ratio) data?
• percent agreement
• kappa - Precise Answer ✔✔What are the reliability coefficients for discrete (ordinal/nominal) data?

1. Assesses relationship and not agreement
2. Only two raters or occasions could be compared - Precise Answer ✔✔What are the two main problems with the pearson r coefficient and why is it considered the old approach? ICC and kappa

• Both give single indicators of reliability that capture strength of relationship plus agreement in a single value - Precise Answer ✔✔What have replaced pearson r to quantify reliability? ICC --> Continuous scale scores Kappa --> categorical scores - Precise Answer ✔✔When do you use ICC to quantify reliability? What about kappa? Range from 0 (no reliability) to 1 (perfect reliability) - Precise Answer ✔✔How are reliability coefficients quantified? reliability - Precise Answer ✔✔true score variability / (tru score variability + error variability) =? larger --> Larger between subject variability (true variability), higher reliability coefficient - Precise Answer ✔✔Reliability will be ___________ when true variance is greater. ICCs give a standardized estimate of reliability without units whereas SEM is unstandardized is in the units of measurement. - Precise Answer ✔✔What is the difference between ICCs and SEM?

ICC (Model #, form #) - Precise Answer ✔✔The first number in the ICC type parenthesis indicates the ______ while the second indicates the ______. Model 1 - Precise Answer ✔✔ICC model where each subject is measured by a different set of raters who are randomly chosen; most conservative and rarely used in clinical research Model 2 - Precise Answer ✔✔ICC model where each subject is measured by the same raters; raters are randomly chosen and representative of the rater population; results are generalizable Model 3 - Precise Answer ✔✔ICC model where each subject is measured by the same rater and the raters are the only ones of interest so the results are not generalizable; less conservative Model 2 - Precise Answer ✔✔Which ICC model is most common for inter-rated reliability or test-retest reliability? Model 3 - Precise Answer ✔✔Which ICC model is most common for intra-rater reliability? Number of observations used to obtain a reliability estimate - Precise Answer ✔✔What does the second number AKA the ICC form number represent?

.90 - Precise Answer ✔✔What ICC scores are considered ideal for clinical measurements? < 0.75 - Precise Answer ✔✔What ICC scores are considered poor to moderate reliability? On the diagonal - Precise Answer ✔✔Where are the agreements located in a frequency table for reliability? Percent agreement - Precise Answer ✔✔Reliability for categorical scales that simply shows how often the raters agree ranging from 0-100%

• Doesn't account for agreement due to chance

• Tends to overestimate reliability - Precise Answer ✔✔What is the problem with percent agreement? kappa coefficient - Precise Answer ✔✔Reliability measurement for categorical scales when the proportion of agreement between raters after chance agreement has been removed weighted kappa - Precise Answer ✔✔takes into account the consequence of disagreements between examiners, used when all disagreements are not equally consequential when assessing reliability for categorical scales 0.8 - 1 - Precise Answer ✔✔What kappa scores indicates excellent agreement beyond chance? internal consistency - Precise Answer ✔✔used to construct and evaluate scales and questionnaires reliability; estimates how well the items that reflect the same construct yield similar results cronbach's alpha (a) - Precise Answer ✔✔represents correlation among items and correlation of each individual item with the total score during internal consistency analysis 0.7 - 0.9 Helps eliminate items from test that are not homogenous to the set or are not contributing unique information - Precise Answer ✔✔What is the recommended value for cronbach's alpha during internal consistency testing? response stability (basically the same as test-retest reliability0 - Precise Answer ✔✔Stability of repeated measures over time SEM MDC CV - Precise Answer ✔✔What are the three statistical methods for response stability? SEM

• remember that ICC and kappa are relative measures of reliability since unitless - Precise Answer ✔✔Absolute measure of reliability minimal detectable difference / minimal detectable change (MDC)
• in units of measurement as variable - Precise Answer ✔✔amount of change in a variable that must be achieved to reflect a true change or difference coefficient of variation (CV) - Precise Answer ✔✔ratio of SD to mean as expressed as a percentage; compares variability between two distributions on different scales by eliminating units alternate forms ex: comparing a traditional goniometer to a inclinometer - Precise Answer ✔✔comparing different methods of testing the same phenomenon with different instruments limits of agreement - Precise Answer ✔✔range that includes approximately 95% of differences Level of agreement and bias
• bias if all above or below the mean of dots (bias line) - Precise Answer ✔✔A Bland-Altman plot can show what? B. that we are more confident that our results are not due to sampling error - Precise Answer ✔✔An appropriate conclusion regarding a p-value of 0.001, compared to a p-value of 0.04, is that our result is.... A. clinically important that our effect size is larger B. that we are more confident that our results are not due to sampling error C. that our effect size is larger The sampling distribution of the mean - Precise Answer ✔✔If we draw an infinite number of samples from a population and then plot their means, we get:

False

• If MCID is within the CI, then inadequately powered. - Precise Answer ✔✔T/F: When doing a post-hoc power analysis, if the MCID is included within the confidence interval, you had adequate power. C. both - Precise Answer ✔✔Which of the following can be used to test whether a difference in a sample likely represents more than just chance (i.e. sampling error)? A. P values B. Confidence intervals C. Both D. Neither False - Precise Answer ✔✔T/F: A 1-way ANOVA is generally more powerful than a RM ANOVA. 2 main effects - Precise Answer ✔✔How many possible main effects exist for a 2x4 ANOVA? 1-way RM ANOVA - Precise Answer ✔✔You read a study that includes 1 group of athletes and tests their running performance 3 times- at baseline, after fasting for 12 hours, and immediately after a "carbo- loaded" meal. Which single statistical test would be able to appropriately analyze their data? 2-way mixed model ANOVA
• Two IV
• One IV is within-subject (time) and one is between (performance) - Precise Answer ✔✔You read a study that includes 2 groups of athletes- one group of breakfast eaters and another group of non- breakfast eaters. Running performance of each group is tested 3 times- at baseline, after fasting for 12 hours, and immediately after a "carbo-loaded" meal. Which single statistical test would be able to appropriately analyze their data?

Less within group variance - Precise Answer ✔✔Since an ANOVA is essentially the ratio of "(variance between the grand mean and group means) / variance within each group", which of the following make it easier to find a statistical difference? False

• Random sampling is rarely met - Precise Answer ✔✔T/F: t-tests, like all parametric tests, are highly sensitive to even slight violations to assumptions (i.e. should not be used when any assumptions are violated). Failed homogeneity of variance (they are different at baseline) so need to use the adjusted t test output line in SPSS - Precise Answer ✔✔What is your conclusion if you calculate a Levene's test p value of 0.03 (assuming alpha is 0.05)? D. They are performed instead of an ANOVA - Precise Answer ✔✔Which of the following are true regarding planned comparisons: A. They always test for every possible difference B. They are performed after an ANOVA C. They are used more commonly than post-hoc tests D. They are performed instead of an ANOVA