The flashcards below were created by user
emm64
on FreezingBlue Flashcards.

Standard Error (SE)
estimate of the precision of parameter estimates

Kurtosis
 peakedness or flatness of a distribution
 (2 to infinity)

Skewness
 Asymmetry of a distribution
 (3 to 3)

Univariate
 Research Level of finding:
 Distribution of individual measures (including outcomes, predictors and covariates)

Bivariate
 Research Level of finding:
 Relationship between outcomes, predictors & covariates

Multivariate
 Research Level of finding:
 Model outcomes as functions of predictors, controlling for covariates

Clinical Trial (CT)
 Test how well new medical approaches work in human.
 These studies test new methods of screening, prevention, diagnosis, or treatment of a disease.
 The safest and fastest way to find treatments that work in human.

Randomized Clinical Trial (RCT)
 Randomize subjects to treatment and control arm
 Can get rid of initial differences between arms

Observational Study
 Individuals are observed and outcomes are
 measured.
 No attempt is made to affect the outcome.
 The researcher doesn’t control over the experiment.

Prospective
watch for outcomes during the study period.

Retrospective
 look backwards and examine exposures to suspected risk or protection factors in relation to an outcome that is already established at
 the start of the study

Crosssectional
collect data at one time point only

Longitudinal
collect data from multiple time points

Sampling unit
 Divide population into sampling units
 (e.g., individual person, household,
 hospital, etc.)
 Must not overlap and cover the whole population

Sampling frame
List of all sampling units

Simple random sample
Probability sample where every unit has an equal chance to be selected (k/N)

Stratified random sample
Probability sample where sampling frame is divided into strata and random sample is drawn within each stratum.

Cluster sample
 Probability sample where clusters (primary sampling units) are randomly
 selected and individual units are nested and selected within a cluster.

Convenience sample
 Nonprobability
 get what is convenient and available as the sample.

Systematic sample
 Nonprobability
 (frame is usually not randomly ordered) 
 get sample in a systematic way (e.g., draw every kth unit)

Clinical examination
 Data are collected through clinical/physical exam (e.g., clinical
 checkup tooth cavity, etc.)
 objective & treated as gold standard

Laboratory test
 data are collected through lab test (e.g., biopsy for oral lesion,
 etc.)
 objective & treated as gold standard

Survey
 generate selfreported data
 efficient for large population study
 usually not treated as gold standard

Medical record abstraction
 data are collected from chart review
 also treated as gold standard, but less accurate than
 clinical exam and lab test due to time lag and others

Claims
 claims data from health insurance/HMO
 large amount of data but often incomplete/inaccurate

Effect size
 Difference between parameters to be tested
 conveys the estimated magnitude of a relationship without making any statement about whether the apparent relationship in the data reflects a true relationship

Which of the following statements is NOT true?
 a. The setting of the study is generally mentioned in the Method section.
 b. Sampling frame is the list of all sampling units.
 c. The target population is usually a part of the study sample.
 d. Statistical hypothesis is stated in terms of statistical parameters.
C

The key characteristics of longitudinal study is that,
 a. The subjects are randomized to treatment and control arm.
 b. The outcome measure has already been observed at the beginning of the study.
 c. No attempt will be made to affect the outcome.
 d. The data are collected from multiple time points over time
D

We commit a type I error when we,
 a. Reject a false null hypothesis.
 b. Reject a true null hypothesis.
 c. Accept a false null hypothesis.
 d. Accept a true null hypothesis.
B

Each item or person in the population of interest has an equal and independent chance of being selected. What is this sample called?
 a. Convenience Sample
 b. Stratified Sample
 c. Simple Random Sample
 d. Systematic Sample
C

Which of the following study can reduce the initial differences between the treatment and control arms?
 a. Randomized Clinical Trial (RCT)
 b. Observational study
 c. Any prospective study
 d. Any retrospective study
A

The statement “restored teeth after incomplete excavation have the same fracture strengths as restored teeth after complete excavation” is a
 a. Setting of the study
 b. Statistical hypothesis
 c. Null hypothesis
 d. Alternative hypothesis
C

Suppose a researcher conducts an experiment to test a hypothesis. If she doubles her sample size, which of the following will increase?
 a. The power of the hypothesis test
 b. Type I error
 c. Critical value
 d. Type II error
A

Which of the following statements about retrospective study is true?
 a. Retrospective study is a type of study in which individuals are observed and certain outcomes are measured during the study.
 b. Retrospective study can never get rid initial difference in patient characteristics between groups.
 c. Retrospective study looks backwards and examines exposures to suspected risk in relation to the already established outcome.
 d. Retrospective study is the safest and fastest way to find treatments that work in human.
C

Which of the following statements is true?
 a. The self reported data generated from survey are usually treated as gold standard.
 b. The national health data are sampled by simple random sample across the U.S.
 c. The national oral health data can be obtained from web site.
C

The Bivariate Analysis is used to
 a. Describe the measurement and distribution of individual measures.
 b. Describe the relationship and association between two measures.
 c. Model a binary outcome as a function of a number of predictors.
 d. Model a continuous outcome as a function of a number of predictors.
B

The pulp status (normal, mildly necrotic, moderately necrotic, severely necrotic) is which type of measurements,
 a. Nominal
 b. Ordinal
 c. Ratio
 d. Interval
B

The number of Decayed Missing and Filled Teeth (DMFT) is which type of measurements,
 a. Nominal
 b. Ordinal
 c. Ratio
 d. Interval
C

Which of the following is non parametric test?
 a. Two sample t test
 b. Chisquare test
 c. Analysis of variance (ANOVA)
 d. Sign test
D

Which of the following statements about Sign Test is NOT true?
 a. Sign test will test if there is a difference between paired samples.
 b. Sign test required the outcome has a normal distribution.
 c. It is a nonparametric test and the ties will be ignored in the final calculation.
 d. The null hypothesis could be that the number of positive difference and the number of negative difference are the same.

The mean and variance is used to describe,
 a. The central location of the data
 b. The dispersion of the data
 c. The central location and the dispersion of the data
 d. The dispersion and the central location of the data
C

Which of the following measures the relative peakedness of a distribution?
 a. Variance
 b. Mode
 c. Skewness
 d. Kurtosis
D

Which of the following statements about Analysis of Variance (ANOVA) is NOT true?
 a. ANOVA requires the outcome has a normal distribution. It is a generalization of two sample ttest.
 b. ANOVA requires the data to be independent within and between groups, but the variance for each group can be different.
 c. When the Fstatistics exceeds the critical value for 5% level with a pvalue <0.05, we will conclude that the means of the groups are different.
 d. ANOVA is equivalent to ttest if comparing two groups

Which of the following is NOT true about hypothesis testing?
 a. Hypothesis testing is used to judge the evidence for hypothesis from the data.
 b. Hypothesis testing is conservative in the usual setting with α=.05 and β=0.20 .
 c. Hypothesis testing can only be used for continuous and normal distributed outcomes.
C

Which of the following statements is NOT true?
 a. In a design with α=.05 and β=0.20, then we think that mistaken rejection of null hypotheses is more serious than mistaken acceptance.
 b. Pvalue is the probability that the test statistic assumes a value as extreme as, or more extreme than, that observed, given the null hypothesis is true.
 c. The Chisquare test is used to compare observed data with the data we would expect to obtain according to a specific hypothesis.
 d. The type I error is committed if we accept a false null hypothesis.
D, type I is Rejecting True Null

Dr. Helen Sweet, practicing dentistry around Houma, Louisiana, used periodontal index scores to classify her patients into three groups: Group I scored from 1.00 to 2.10, Group II scored from 2.11 to 2.83, Group III scored greater than 2.83. The three groups should, she said, have different blood glucose level. Assuming periodontal index score was normally distributed, what statistical test would you use to formally assess the null hypothesis?
 a. Ttest
 b. Signtest
 c. Chisquare test
 d. ANOVA
D

1. Find the true statement about Research Design:
 a. Research question is the issue that leads to the need for your study;
 b. The hypothesis proposes explanation for an observable phenomenon;
 c. Study sample is usually smaller than the target population;
 d. a and b
 e. a, b, and c
e

The statement that “There is no difference of bone loss affected by periodontitis between the control and the experimental groups” fits better for:
 a. Setting of a study
 b. Null hypothesis
 c. Alternative hypothesis
B

The key difference between observational study and interventional study is:
 a. Whether the data are collect at one time point only or from multiple time points
 over time
 b. Whether the study look backward and examine exposures to suspected risk
 c. Whether the researcher intervenes the process
 d. a, b, and c
C

A sample of dental students is needed to evaluate National Board examine. You draw at random 1/4 of the students from each of the accredited programs. What is this sample called?
 a. Simple Random Sample
 b. Stratified Sample
 c. Systematic Sample
 d. Convenience Sample
B

Sampling frame is a list of all sampling units that cover the whole population and must not overlap.

Suppose a researcher conducts an experiment to test a hypothesis. If he doubles his sample size, which of the following will increase?
 I. The power of the test.
 II. The effect size of the test.
 III. The probability of making a type II error.
 a. I only
 b. II only
 c. III only
 d. All of the above
 e. None of above


Significance level or the type I error is the probability that we reject a null hypothesis when it is false.

Find the true statement about data collection:
 a. Data collection is the foundation for all findings, conclusions & inferences
 b. The primary data collection can be done through clinical examination, laboratory test, patient survey, and medical record abstraction.
 c. Secondary data collection collects the data that have already been collected by someone else and exist somewhere.
 d. a, b
 e. a, b, c
E

Logistic regression is used,
 a. To model log odds of a binomial outcome measure as functions of some covariates
 b. To compare means across multiple groups
 c. To model a continuous outcome measure as functions of some covariates
 d. To conduct nonparametric analysis
A

A study with low statistical power has a high probability of rejecting the null hypothesis when it is false.

Power (1β) is the probability that you reject the null hypothesis, given that the alternative hypothesis is true.

Skewness is
 a. A measure that characterizes the relative peakedness or flatness of a distribution
 b. A measure of the asymmetry of the probability distribution of a realvalued random
 variable
 c. A measure that ranges from 3 to +3
 d. A measure that ranges from 2 to infinity
 e. a and c
 f. a and d
 g. b and c
 h. b and d
H

A study was conducted to determine the relationship between flossing and the number of sites that bled on periodontal probing (BOP). Flossing was classified as daily, less than daily, and never. Assuming that number of site of BOP was normally distributed, what statistical method fits the best to formally assess the null hypothesis that daily flosser, less than daily flosser, and never flosser have the same mean number of BOP?
 a. Ttest
 b. Chisquare test
 c. Signtest
 d. ANOVA
D

The pulp status (normal, mildly necrotic, moderately necrotic, severely necrotic) is of which following types of measurement?
 a. Nominal
 b. Ordinal
 c. Ratio
 d. Interval
B

Which of the following is an estimate of the precision of parameter estimates due to sampling?
 a. Variance
 b. Standard Deviation (SD)
 c. Mode
 d. Standard Error (SE)
 e. Mean
D

