Final Exam Practice Research I HP1 Fall 2024 PDF

12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Final exam practice Due Dec 4 at 9am Points 151 Questions 77 Available Dec 4 at 9am - Dec 8 at 11:30am Time Limit None Allowed Attempts Unlimited Instruct...

12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Final exam practice Due Dec 4 at 9am Points 151 Questions 77 Available Dec 4 at 9am - Dec 8 at 11:30am Time Limit None Allowed Attempts Unlimited Instructions These questions are for you to practice final exam. You can attempt multiple times. It will be closed on Dec 8th, 6 AM. Take the Quiz Again Attempt History Attempt Time Score KEPT Attempt 2 20 minutes 143 out of 151 LATEST Attempt 2 20 minutes 143 out of 151 Attempt 1 42 minutes 117.36 out of 151 Submitted Dec 8 at 11:27am  Question 1 2 / 2 pts Which question is NOT a foreground question that lies in the scope of PT practice? Is Berg Balance scale (BBS) or Four Square Step test (FSST) a better measure for balance test? What risk factors predict unfavorable outcomes for patients with low back pain (LBP) receiving PT? Correct! How accurate is bone scan to assist clinician in osteoporosis diagnosis ? Is high- or moderate-intensity training more effective in increasing gait speed in patients with stroke?  Question 2 2 / 2 pts Which statement is in the format of a null hypothesis We hypothesize that Otago exercise is more effective than traditional PT intervention https://sju.instructure.com/courses/43011/quizzes/135209 1/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 We hypothesize that traditional PT intervention is more effective than Otago exercise Correct! We hypothesize that there is no difference between Otago exercise and traditional PT intervention We hypothesize that the difference between Otago exercise and traditional PT intervention can't be determined  Question 3 2 / 2 pts Which statement is correct about alternative hypothesis? Also called statistic hypothesis Correct! Also called research hypothesis Assume that there is no difference between groups Can't be a directional hypothesis  Question 4 0 / 2 pts Which statement about hypothesis is wrong? Null hypothesis predicts there is no difference of outcome between two groups Correct Answer The authors predict that treatment effect is larger in group A than in group B. This is a non-directional hypothesis Research hypothesis can be directional You Answered Researchers can only reject hypothesis, but not prove alternative explanation  Question 5 2 / 2 pts Which statement is wrong about qualitative research design? It focuses on using words to describe subjects thoughts, perceptions, opinions, beliefs, attitudes and/or experiences Qualitative and quantitative study may be appropriate for different situation. Not one design is always better than the other. In a study, researchers interview patients about their opinions on PT students' cultural sensitivity, the researcher uses open-ended questions to interview patients. This is a qualitative research design. Correct! Researchers should implement control or provide intervention to make qualitative research believable  Question 6 0 / 2 pts https://sju.instructure.com/courses/43011/quizzes/135209 2/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Which statement is wrong about quantitative research design? Outcomes of interests are in numeric format, not in words Correct Answer Researchers using quantitative research design consider the truth is subjective You Answered Researchers using quantitative research design do not think their interactions with subjects will change the research outcomes Researchers using quantitative research design may provide intervention to subjects  Question 7 2 / 2 pts Which statement about research design is wrong based on time factor? A retrospective study uses historical data generated before establishing a study design Correct! A prospective study uses historical data generated before establishing a study design A cross-sectional study collects data at one time point A longitudinal study collects data at multiple time points  Question 8 0 / 2 pts Which statement is correct about level of evidence? Correct Answer A single RCT has lower level of evidence than a systematic review RCT has lower level of evidence than quasi-experimental study You Answered Observational study has higher level of evidence than experiemtanl study Meta-analysis of RCT has lower level of evidence than a single RCT.  Question 9 0 / 2 pts Which of the following statement about primary research is wrong? Researchers collect data themselves for primary research Correct Answer It is a study of studies. Review article is an example of primary research A dissertation by a Ph.D. candidate who collected data is a primary research You Answered https://sju.instructure.com/courses/43011/quizzes/135209 3/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 There is no strong need to conduct more primary research on a topic, if there are many original studies conducted on this topic with consistent findings  Question 10 2 / 2 pts Which statement about quasi-experimental design is wrong? Quasi-experimental design is an example of quantitative research A study with intervention but either control group, and/or random assignment into group are missing is a quasi- experimental research. Correct! Only study without control group can be called quasi-experimental design Quasi-experimental design has lower bias control than experimental design.  Question 11 2 / 2 pts Which statement about secondary research is correct? The researcher collects data by herself/himself RCT (Randomized clinical trial) is a type of secondary research Correct! Secondary research is "study of studies" Secondary research can only use words to summarize exisiting evidence  Question 12 2 / 2 pts Which statement is correct about secondary research design? Narrative review is a type of secondary research. It has a focused question & comprehensive literature search. Systematic review is a type of secondary research. It has a focused question & comprehensive literature search. It also uses numeric data to summarize findings from reviewed articles Meta-analysis is a type of secondary research. It has a focused question & comprehensive literature search. It uses only words to summarize findings from reviewed articles Correct! Meta-analysis is a type of secondary research. It has a focused question & comprehensive literature search. It also uses numeric data to summarize findings from reviewed articles  https://sju.instructure.com/courses/43011/quizzes/135209 4/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 13 2 / 2 pts Which type of secondary research has the highest bias control? Correct! Meta-analysis Systematic review Narrative review Scoping review  Question 14 2 / 2 pts Which type of review uses quantitative method to summarize evidence? Narrative review Systematic review Correct! Meta-analysis  Question 15 2 / 2 pts Which statement about systematic review is correct? Correct! The literature search strategy has to be systematically documented in details Only invited researchers can conduct systematic review Systematic review recommends actions that clinicians can take Systematic review is using quantitative method to summarize evidence from existing studies  Question 16 2 / 2 pts Which statement about meta-analysis is correct? The researcher should summarize all the studies found from the search, without any pre-established exclusion criteria Only invited researchers can conduct meta-analysis Meta-analysis recommends actions that clinicians can take Correct! Meta-analysis is using quantitative method to summarize evidence from existing studies  https://sju.instructure.com/courses/43011/quizzes/135209 5/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 17 2 / 2 pts Which is one of the benefits to conduct a high quality meta-analysis? Correct! High quality meta-analysis increases sample representativeness High quality meta-analysis increases effect size High quality meta-analysis increases correlation between studies High quality meta-analysis decreases statistical power  Question 18 2 / 2 pts Which statement about sensitivity and specificity is correct? Correct! Sensitivity is the capacity to correctly diagnose individuals with that condition Specificity is the capacity to correctly diagnose individuals with that condition Sensitivity is higher if true negative is higher. Specificity is higher if true positive is higher.  Question 19 2 / 2 pts Which statement is correct about Receiver Operating Characteristic Curve (ROC)? It is a graphic way to evaluate different test scores with respect to the true-positive and true-negative values. The point where the true positive value is highest should be used as cut-off point for diagnosis, even though the false positive value is very high at this cut-off point. Correct! A PT clinician reads an article with a ROC plot for 3 diagnostic tests. The clinician should choose the diagnostic test with the biggest area under the curve (AUC) if other conditions are the same (for example: test cost, test time). An area under the curve (AUC) of 0.5 in the ROC plot indicates an excellent diagnostic test.  Question 20 2 / 2 pts Calculate negative predictive value based on the following table. https://sju.instructure.com/courses/43011/quizzes/135209 6/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 94/100 96/100 94/98 Correct! 96/102  Question 21 2 / 2 pts Which one is not one of the components of a good prognosis? Correct! Estimated cost to achieve that prognosis The possible outcome (preferrably be a functional outcome) The likelihood of the outcome to occur The time frame required to achieve the outcome  Question 22 2 / 2 pts Which of the following is not a risk factor predicting higher risk of fall? Age History of falls Polypharmacy Correct! Body mass index  Question 23 2 / 2 pts Which statistic method examines if the frequency distribution is significantly different from expected distribution? Correct! Chi-square test https://sju.instructure.com/courses/43011/quizzes/135209 7/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Odds Ratio Relative risk Correlation coefficient  Question 24 2 / 2 pts Which statement is correct about Odds Ratio Odds ratio is between 0 and 1 Correct! Large Odds ratio indicates that an outcome is more likely to occur with a specific predictor Odds ratio of 0.9 is a higher risk than chance that some outcome may occur Odds ratio can be a negative number  Question 25 2 / 2 pts Which statement is correct about relative risk? Relative risk is between 0 and 1 Correct! Large relative risk indicates that an outcome is more likely to occur with a specific predictor Relative risk of 0.9 is a higher risk than chance that some outcome may occur Relative risk can be a negative number  Question 26 2 / 2 pts Calculate relative risk based on the following table. (48/12) (22/46) Correct! (48/70) (12/58) (48/22) (12/46) (48/60) (22/68)  Question 27 2 / 2 pts https://sju.instructure.com/courses/43011/quizzes/135209 8/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Which statement is correct about correlation coefficient Correlation coefficient is between 0 to 1 Correlation coefficient 0.8 indicates a stronger association than -0.8 Correct! Correlation coefficient 0.4 is a weaker association than -0.8 Correlation can explain cause-effect relationship  Question 28 2 / 2 pts A dependent variable is: Correct! The measured outcome of a study. Dependent variable: Results of experiment, that is "outcome" (O in PICO) The variable that the researchers manipulate. The cause in a cause-and-effect relationship. The unexpected finding in a research study.  Question 29 2 / 2 pts Which statement is wrong about confounders Confounders are factors that may affect study results other than your experimental manipulation (such as intervention). Correct! Researchers should control all confounders. Researchers do not exclude subjects based on confounders. Instead, they use statistical methods to analyze the effects of confounders after they collect data. An experimental study uses gait speed as treatment outcome. The researchers think baseline (before treatment) gait speed may affect treatment effect. The researchers still include subjects with a wide range of baseline gait speed. When analyzing outcome, the researchers take into consideration of baseline gait speed for the statistical analysis. This is an appropriate method to evaluate effects of confounders. Confounder is also called extraneous variable  Question 30 2 / 2 pts Which statement is correct about inclusion/exclusion criteria? Correct! https://sju.instructure.com/courses/43011/quizzes/135209 9/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Exclusion criteria are not the opposite of inclusion criteria. You should list extra characteristics than inclusion criteria to specify subject characteristics Setting up too many exclusion criteria will make the study sample less homogeneous. Setting up too many exclusion criteria increase representativeness of study sample It is desirable to conduct a study with male subjects only, even though the diagnosis of interest is equally prevalent in both genders.  Question 31 2 / 2 pts Which statement is wrong about probabilistic sampling? Correct! Probabilistic sampling is the gold standard. This sampling method should always be applied to both qualitative and quantitative studies. Probabilistic sampling is more likely to get representative sample for population. Sampling results are more likely to be similar after multiple probabilistic sampling attempts. Studies using probabilistic sampling are more likely to have better bias control, and higher external validity.  Question 32 2 / 2 pts What type of sampling is this? A researcher assigns random number to each subject, then use a computer program to draw sample one at a time. Correct! Simple random sampling Systematic sampling Stratified random sampling Cluster sampling  Question 33 1 / 1 pts What type of sampling is this? A researcher assigns random number to each subject. Then he set up a rule that he will first draw a random number from 0 to 9. Then he will add 10 to all the subsequent sample. For example, if he first draws 6, then he will keep picking 16, 26, 36 etc. https://sju.instructure.com/courses/43011/quizzes/135209 10/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Simple random sampling Correct! Systematic sampling Stratified random sampling Cluster sampling  Question 34 2 / 2 pts What type of sampling is this? A researcher draw sample based on clinic. He assign random number for each clinic. All subjects in that clinic will be sampled if that clinic gets picked. Simple random sampling Systematic sampling Stratified random sampling Correct! Cluster sampling  Question 35 2 / 2 pts Which statement is correct about the following sampling method? A researcher is interested in studying PT clinicians' lived experience as a transgender person certified and practice in woman health's physical therapy. There are limited number of PT with this specific characteristics. The researcher decided to invite certain PT with the characteristics to the study. This is non-probabilistic, convenient sampling. This is probabilistic, snowball sampling. Correct! This is non-probabilistic, purposive sampling. This is probabilistic, systematic sampling.  Question 36 2 / 2 pts Which case represents good sampling for an experimental study? A researcher wants to study physical activity level in older adults. The researcher does not have too much time, so he decided to go to a nearby senior sport club to recruit subjects. Recruited subjects all exercise moderately at least 3 times a week, while most older adults in USA do not meet recommended physical activity level. https://sju.instructure.com/courses/43011/quizzes/135209 11/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 A researcher wants to study stress level of graduate students. He decided to only invite those who do not have to work, because these students have more time to participate in his study. Correct! A researcher wants to study effect of aerobic exercise on maintaining cognitive function in older adults. She excludes patients with diagnosed dementia and MCI (mild cognitive impairment). She includes subjects with all educational level because educational level has been shown to related to cognitive function, and may affect the effect from aerobic exercise. A researcher wants to save time recruiting subjects. He asked his every subject to refer another subject to participate in his study.  Question 37 1.5 / 1.5 pts Which subject sampling method may introduce threats to internal validity of an experimental study? (more than one answer) Simple random sampling Systematic sampling Correct! Convenience sampling Correct! Snowball sampling Cluster sampling Correct! Purposive sampling  Question 38 2 / 2 pts Which statement is correct about group equivalence at baseline? Correct! Random assign subjects into intervention or control group is an attempt to achieve group equivalence at baseline Random assign subjects into intervention or control group can guarantee group equivalence at baseline Researcher should use descriptive statistics to test if groups are equivalent at baseline There is no statistical adjustment available if groups are not equivalent at baseline  Question 39 2 / 2 pts Which statement is correct about blinding? Blinding subjects from treatment grouping is always possible https://sju.instructure.com/courses/43011/quizzes/135209 12/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Blinding staffs in charge of testing research outcome is to allow testers change their testing standards according to subject's intervention grouping It is a double-blinded study if testers and intervention clinicians on the research team are blinded from group assignment Correct! It is a double-blinded study if research team and subjects are both blinded from group assignment  Question 40 2 / 2 pts Which statement is wrong about level of measurement? (see Wk 8 - Component (Sampling, level of measurement, psychometric).pdf (https://sju.instructure.com/courses/43011/files/6338537/download?download_frd=1) , slide 33) Nominal measure is basically about categories and types. There is no order of large or small number associated with it. Ordinal measure indicates order of measures. Likert scale (1: very dislike to 7; likes a lot) is an example of ordinal measure. Correct! Interval measure is discrete number without equal distance between numbers. Ratio measure is continuous number with equal distance and true zero.  Question 41 2 / 2 pts Which statement is correct about reporting outcomes at different level of measurement? Correct! Mean and dispersion should be used for reporting nominal data Mean and dispersion should be used for reporting ordinal data Mean and dispersion should be used for reporting interval data Count and frequency should be used for reporting ratio data  Question 42 2 / 2 pts Which statement is wrong about level of measurement? Correct! In ordinal measure, the difference between 1 & 2, and 2 &3 is the same. It is appropriate to use mean and standard deviation to describe outcome measures with interval and ratio data. https://sju.instructure.com/courses/43011/quizzes/135209 13/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Researchers can reduce interval data to order data. For example, researchers collect the actual weight (in lb), and later classify weight into normal, overweight, obese, morbid obese. Ratio data have the highest level of measurement precision, but it may not be the most meaningful level of measurement. For example: researchers can record gait speed from new walkers to the oldest-old. It does not make sense to compare gait speed between a new walker and a 100 year old. Therefore, the researchers classify gait speed into too slow, normal, fast categories based on age-norm.  Question 43 2 / 2 pts Which statement is correct about validity and reliability? Reliability is the ability of an outcome measure to capture what it claims to capture Validity is the stability of the assessment score of an outcome measure Correct! Validity will be low if there is high systematic errors Reliability will be high if there is high random errors  Question 44 2 / 2 pts Which statement is wrong about the relationship between reliability & validity Validity will be higher if there is no systematic error Reliability will be higher if the random errors are low If the test results cluster together but is far away from the true value, then reliability is high but validity is low Correct! If reliability is high then validity must be high  Question 45 2 / 2 pts Which statement is wrong about test reliability? Reliability refers to how repeatable test results are. High reliability means low variability in test results. Correct! If a test result is very variable, it is always the instrument or tester to be blamed for poor repeatability. Reliability is a type of psychometric properties to measure test performance  Question 46 2 / 2 pts Which statement is wrong about type of reliability? https://sju.instructure.com/courses/43011/quizzes/135209 14/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 A PT student scores patient's performance twice using the same outcome measure and video recording. This is an example of intra-rater reliability. A PT student tests one stroke patient @ day 30 and 40 after stroke using the same outcome measure. This is an example of test-retest reliability. Correct! A PT student tests a premature infant's gross motor development @ corrected age 5 month and 1 year old. This is an good example of test-retest reliability because the infant's ability does not change too much. Two PT students score patient's balance performance at the same time using the same outcome measure. This is an example of inter-rater reliability.  Question 47 2 / 2 pts Which statement is wrong about internal consistency? (You can ignore this question, it is not covered in class) Internal consistency is a type of test reliability. Correct! In a test instrument, item A can be classified to 3 subscales. This test has high internal consistency. If each factor uniquely loads onto a specific construct, then the test has high internal consistency. If all the constructs are mutually exclusive, then the test has high internal consistency.  Question 48 2 / 2 pts Which statement is wrong about split-half reliability? (You can ignore this question, it is not covered in class) A test developer build a pool of test items. Then the test developer divides all items into two parts that has similar level of difficulty. This is an example of ensuring split-half reliability. Correct! A SPT randomly chooses half of the items from an established outcome measure. This can ensure split-half reliability. Subject's performance will be similar no matter which subtest is used from a big test item pool. This is an example of split-half reliability. https://sju.instructure.com/courses/43011/quizzes/135209 15/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 A subject does not show improvement at 2nd test using a subtest with high split-half reliability. This subject probably does not have improvement.  Question 49 2 / 2 pts When a researcher chooses an outcome measure, the lowest requirement of the measure is to have what type of validity? Correct! Face validity Content validty Construct validity Concurrent validity  Question 50 2 / 2 pts An outcome measure has all relevant facets it intended to measure has high _______ validity. Correct! Content validity Construct validity Concurrent validity Convergent validity  Question 51 2 / 2 pts An outcome measure reflects the operational definition of what it intended to measure has high _______ validity. Content validity Correct! Construct validity Concurrent validity Convergent validity  Question 52 1 / 1 pts An outcome measure has score similar to established measurement is said to have high _____ validity. (You can ignore this question, it is not covered in class) https://sju.instructure.com/courses/43011/quizzes/135209 16/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Content validity Construct validity Concurrent validity Correct! Convergent validity  Question 53 2 / 2 pts Which statement is wrong about reference used to evaluate outcome measures? Norm-referenced assessments generally have database acquired from large scale epidemiological studies. Individual's outcome is compared against the "norm" from the population. Sometimes, there are gender- age- specific norms. Bayley Scale of Infant Development (BSID) is a norm-referenced assessment tool. Berg Balance Scale (BBS) is a criterion-referenced outcome measure. For each test item, point is assigned to outcome based on patient's performance in comparison to certain criteria. Researchers judge if the patient's balance is within normal limit, has high risk of falls based on the total score against pre-established criterion. Correct! For norm-referenced measurement, the norm will not change once it is established.  Question 54 1 / 1 pts The ability of an outcome measure to detect change is called: Correct! Responsiveness Minimum detectable change (MDC) Minimum clinically important difference (MCID) Validity and reliability  Question 55 2 / 2 pts You find that the outcome measure that you choose does not register a further decrease in score for the lowest scoring individuals. This is known as a: Correct! Floor effect Ceiling effect Placebo effect Cause effect  https://sju.instructure.com/courses/43011/quizzes/135209 17/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 56 2 / 2 pts Choice of outcome measure should match subjects' ability. What effect may occur in the situation below? A clinician uses a fitness measure with items that only triathlon athletes are likely to pass. The clinician uses this test in a population with poor fitness. Ceiling effect. Correct! Floor effect. Inteference effect  Question 57 2 / 2 pts Which statement is wrong about SEM (Standard error of measurement) & MDC (Minimum detectable change)? SEM is the distribution (variability) around the true test result MDC is the amount of change that just exceeds the SEM of an instrument in order to detect change Correct! It will be easier to detect true change if SEM is bigger MDC must be larger than SEM in order to detect change  Question 58 2 / 2 pts Which statement is correct about responsiveness, SEM (Standard error of measurement) and MDC (minimal detectable change)? (see Research I - 2023 - Wk 14 - Statistics.pdf, (https://sju.instructure.com/courses/43011/modules/items/1547620) and Panapto video Wk 14 - Outcome measures and statistics (https://sju.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=c8185a8d- 3b57-4b2e-a89c-b0ca01428498) ) Responsiveness of a test is its ability to measure correlation A test with high variability will have low SEM (standard error measure) A researcher can be certain that there is an intervention effect if the effect size is smaller than SEM of the chosen test Correct! If a researcher find a result significantly larger than the MDC (Minimal detectable change), then the researcher can be sure there is treatment effect https://sju.instructure.com/courses/43011/quizzes/135209 18/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024  Question 59 2 / 2 pts The smallest treatment effect that would result in a change in patient management, given its side effects, costs and inconveniences. This is a definition of: (see Research I - 2023 - Wk 14 - Statistics.pdf, (https://sju.instructure.com/courses/43011/modules/items/1547620) and Panapto video Wk 14 - Outcome measures and statistics (https://sju.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=c8185a8d- 3b57-4b2e-a89c-b0ca01428498) ) MCD: Minimal Detectable Change Content Validity Correct! MCID: Minimally Clinically Important Difference Effect Size  Question 60 2 / 2 pts Which statement is correct about effect size? Effect size larger than 0.05 is considered as a big effect Effect size can be calculated for nominal variables Effect size is calculated as effect for intervention group divided by effect for control group Correct! Standardized effect size should be used if researchers want to compare treatment effects from two different outcome measures  Question 61 2 / 2 pts In order to compare treatment effect for two similar studies Correct! Researchers should use standardized effect size Researchers should use absolute effect size An effect size of 0.5 is a large effect Different outcome measures can never be compared even if the researchers use standardized effect size  https://sju.instructure.com/courses/43011/quizzes/135209 19/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 62 2.5 / 2.5 pts Subject attrition is generally not random. Who do you think are more likely to drop out from a study? (more than one answer) Correct! Subjects with more severe physical limitations Subjects think the study will bring them benefits Subjects with high motivation Correct! Subjects do not perceive functional gain from a study Correct! Subjects who have constraints on social supports, such as trouble finding transportation to go to study site Correct! Subjects in control group who want to be in the intervention group Correct! Subjects in the intervention group who thinks the training is too difficult  Question 63 2 / 2 pts Which statement is wrong about inferential statistics It is used to make conclusion Data distribution will affect what type of statistical method to use Parametric statistical methods are appropriate for normally distributed data Correct! Quartile is a measure of inferential statistics  Question 64 2 / 2 pts Which statement is correct about inferential statistics? Correct! Can be used when need to make inference about correlation or difference between groups Only parametric statistics is inferential statistics Non-parametric statistics are a form of descriptive statistics Inferential statistics should be used to summarize findings from qualitative studies  Question 65 2 / 2 pts Which condition will require a larger sample size to detect statistical significance? https://sju.instructure.com/courses/43011/quizzes/135209 20/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 A higher setting of the alpha level (for example: researchers set an alpha level of 0.1, will need a larger sample size to detect statistical significance, than researchers select an alpha level of 0.0001) A lower requirement of statistical power (for example: researchers set power to be 0.5, will need a larger sample size to detect statistical significance, than researchers select a power of 0.95) Correct! An outcome with smaller effect size Data distribution is less variable for the study result  Question 66 2 / 2 pts For a data set that is normally distributed, About 90% of the data are within 1 standard deviation of the mean Correct! About 95% of the data are within 2 standard deviations of the mean The total area under the bell curve is unknown About 34.1% of the data are between 3 and 4 standard deviations of the mean  Question 67 2 / 2 pts Which statement is correct about confidence interval (CI) Correct! For the same data set, the data range is bigger if a researcher specified a higher % of CI (example: range for 99% CI is larger than range for 90% CI) If I want to know the 95% CI of a normally distributed data set, I should calculate the CI as mean - 1 standard deviation, to mean + 1 standard deviation CI is a range of values within which the true value is unlikely to lie within a specified probability If the 99% CI of two studies do not overlap, the probability is smaller than 1% that these two results are statistically different.  Question 68 2 / 2 pts For the following scenario with different effect size and standard deviation, which one is least likely to detect a group difference? Big effect size with big standard deviation Big effect size with small standard deviation https://sju.instructure.com/courses/43011/quizzes/135209 21/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Correct! Small effect size with big standard deviation Small effect size with small standard deviation  Question 69 2 / 2 pts For the following scenario with different effect size and sample size, which one is least likely to detect a group difference? Big effect size with large sample size Big effect size with small sample size Small effect size with large sample size Correct! Small effect size with small sample size  Question 70 2 / 2 pts Which statement is correct about interprting results based on CI? A study result shows 99% CI of the effect size includes 0 in the range. The effect size is statistically significant and there is a treatment effect. A study result shows 99% CI of the correlation includes 0 in the range. The correlation is statistically significant A study result shows 99% CI of the odds ratio includes 1 in the range. The correlation is statistically significant Correct! A study result shows a correlation coefficient of -0.9 with p=0.5. The correlation coefficient is large but not statistically significant.  Question 71 2 / 2 pts Which statement is correct about alpha level? Correct! Alpha level is an arbitrary threshold used to detect statistical significance. It is also called significance level. Alpha level is commonly set at 0.001 to decide if there is a statistical significant result. It is harder to detect significant findings for an alpha level of 0.1 then 0.05 If a researcher sets an alpha level at 0.01, then the researcher should accept the null hypothesis if the calculated p value is 0.0001 for the data set  https://sju.instructure.com/courses/43011/quizzes/135209 22/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 72 2 / 2 pts When alpha level is set at 0.05, and the p value of the statistical test is 0.03. The researcher should reject the null hypothesis, and claim there is a significant difference. Correct! True False  Question 73 2 / 2 pts Which statement is correct about interpreting p value based on set alpha level? If a researcher set an alpha level of 0.01, then 99.99% of confidence intervanl should be used to judge if null hypothesis should be rejected An alpha level of 0.10 is a more stringent criteria than 0.01 to judge statistical significance For hypothesis testing, type I error occurs when there is a false negative finding. Correct! A researcher set an alpha level of 0.01 to detect statistical significance. The calculated p value is 0.0001. The researcher should reject the null hypothesis and declare there is statistical significant difference among groups.  Question 74 2 / 2 pts Which power level is more likely to find a true difference? A study with power of 0.6 Correct! A study with power of 0.8 A study result with small effect size A study result with big standard deviation  Question 75 2 / 2 pts If a researcher set a higher power level for her study, then she needs to test more subject. Correct! True False No answer text provided. No answer text provided.  https://sju.instructure.com/courses/43011/quizzes/135209 23/24 12/8/24, 1:44 PM Final exam practice: Research I HP1 Fall 2024 Question 76 2 / 2 pts Which statement is correct about statistical power? Correct! Statistical power is the probability that a statistical test will detect difference between groups, if there is a difference. Statistical power is generally set at 0.01. A study with high power will also have a big effect size. A study will require more subjects if the statistical power is set a lower level.  Question 77 2 / 2 pts Which factor will affect decision on choise of statistical method? (more than one answer) Correct! To test relationship or difference Correct! Level of measurement of data (nominal, ordinal, interval or ratio) Correct! Data distribution Correct! If normality assumption is met Correct! Independent or dependent data Correct! Within-subject or between-subject design Correct! Number of groups https://sju.instructure.com/courses/43011/quizzes/135209 24/24