Quiz MCQ

Question 1

The subject doesn’t want to admit to the interviewer the amount of alcohol consumption per week, so admit just half of it.

a. Recall Bias
b. Random Error
c. Response Bias
d. Selection Bias

Answer 1

c. Response Bias

Question 2

What are the properties of the cohort in a cohort study?

a. Have the risk of the end point.
b. Made by carefully selected cases.
c. Free from the endpoint we observe.
d. Get no explosion during the study.

Answer 2

a. Have the risk of the end point.

c. Free from the endpoint we observe

Question 3

What type would you choose to examine a rare endpoint?

a. Case control study
b. Cohort study

Answer 3

a. Case control study

Question 4

Please select the correct answer!
Select one or more:
a. Statistics is a collection of mathematical techniques that help to analyze and present data.
b. The lack of information leads to randomness.
c. Physical laws always describe deterministic processes.
d. Everyday thinking estimates probabilities and risks surprisingly well

Answer 4

a. Statistics is a collection of mathematical techniques that help to analyze and present data.
b. The lack of information leads to randomness.

Question 5

Select one or more:

a. Computers cannot generate random numbers.
b. A medical prognosis is essentially a probability statement.
c. Probability can be interpreted as our belief that an event (like rain falls tomorrow) will occur.
d. Data collected from the populations helps to make individual decisions.

Answer 5

b. A medical prognosis is essentially a probability statement.
c. Probability can be interpreted as our belief that an event (like rain falls tomorrow) will occur.
d. Data collected from the populations helps to make individual decisions.

Question 6

Which statement(s) are true to the Wilcoxon signed rank test?
Select one:
a. Normality assumption has to be fullfilled.
b. The null hypothesis could be that the difference of the medians of the paired samples is 0, assuming that the distribution of the difference between the samples is symmetrical to the median.
c. The null hypothesis is the equality of the median of the paired samples.
d. The null hypothesis could be that the median of the difference between the paired samples is 0, assuming that the distribution of the difference between the samples is symmetrical to the median.

Answer 6

d. The null hypothesis could be that the median of the difference between the paired samples is 0, assuming that the distribution of the difference between the samples is symmetrical to the median.

Question 7

The meaning of p-value in a hypothesis test:
Select one:
a. the probability of having a corresponding (to the sample) or more extreme if the alternative hypothesis is true
b. the probability of having a corresponding (to the sample) value if the alternative hypothesis is true
c. the probability of having a corresponding (to the sample) value if the null hypothesis is true
d. the probability of having a corresponding (to the sample) or more extreme if the null hypothesis is true

Answer 7

d. the probability of having a corresponding (to the sample) or more extreme if the null hypothesis is true

Question 8

The aim of a hypothesis test is:
Select one:
a. to give a statistical answer on yes/no question.
b. to answer whether the effect is relevant or not.
c. to determine an error probability.
d. to calculate the significance level.

Answer 8

a. to give a statistical answer on yes/no question.

Question 9

You are analyzing the results of a cohort study. At the baseline examination you assessed cardiovascular risk factors on the enrolled subjects. During the 10-year follow-up you recorded the occurrence of several outcomes (death, cardiovascular diseases).
Suppose you want to examine the association between diabetes at the baseline examination and the incidence of death during the follow-up. There were no losses to follow-up for the outcome of death in this study; mortality status was obtained for everyone. Which of the following measures can you calculate validly from the data?

a. risk ratio
b. incidence rate ratio
c. none of them
d. odds ratio

Answer 9

a. risk ratio
b. incidence rate ratio
d. odds ratio

Question 10

Select the correct statements about ’prevalence’.
Select one or more:
a. Prevalence ratio can be calculated as the number of new cases developing during the follow-up period divided by the total number at risk at the beginning of the study.
b. One of the most frequently used measures to describe causality between exposure and outcome.
c. Its value is highly influenced by the incidence and the duration of the disease.
d. It describes the existence of a disease at some point in time.

Answer 10

c. Its value is highly influenced by the incidence and the duration of the disease.
d. It describes the existence of a disease at some point in time.

Question 11

Which are the methods what we can use to address confounding in the design phase?

Select one or more:

a. Matching
b. We enroll at least twice as many exposed subjects to the study as control subjects.
c. There is no way to address confounding in the design phase of cohort studies.
d. Randomization

Answer 11

a. Matching

d. Randomization

Question 12

The members of the Ethical Committee, EXCEPT:
Select one:
a. independent physician
b. head of department 
c. lay member
d. theologian

Answer 12

b. head of department

Question 13

Which is not necessary in a good inform consent?
Select one:
a. the number of enrolled patients
b. insurance
c. the experimental character
d. the end day of the study

Answer 13

d. the end day of the study

Question 14

Which is not the principle of the human research according to the Belmont
Report?
Select one:
a. respect for the person
b. justice
c. equity 
d. beneficiency

Answer 14

c. equity

Question 15

Which of the following can mean in ideal case the sure solution of confounding?
Select one:
a. Randomization 
b. Standardization
c. Matching
d. Stratification

Answer 15

a. Randomization

Question 16

In a linear regression, what do we assume about the nature of the response variable? It is…
Select one:
a. Binary
b. Continuous 
c. Ordinal
d. Time-to-event

Answer 16

b. Continuous

Question 17

What is a dummy variable?
Select one:
a. A continuous variable
b. A categorical variable
c. A variable that can take more than two values
d. A variable that can take only two values

Answer 17

d. A variable that can take only two values

Question 18

What is multicollinearity?
Select one:
a. If explanatory variables explain the response variable
b. If explanatory variables explain each other
c. If the response variable explains the explanatory variables
d. If the response variable explains itself

Answer 18

c. If the response variable explains the explanatory variables

Question 19

Which of the following is the type of the test that tests the significance of a single variable?
Select one:
a. t-test 
b. F-test
c. χ 2 -test
d. z-test

Answer 19

a. t-test

Question 20

Which of the following violate linearity as a model assumption?
Select one or more:
a. If a variable’s effect is not linear
b. If the errors are correlated
c. If the errors have a non-normal distribution
d. If there is an interaction between two variables

Answer 20

a. If a variable’s effect is not linear

d. If there is an interaction between two variables

Question 21

In the logistic regression models
Select one or more:
a. different types of variables can appear among the predictors.
b. the number of the predictors must be the same as those of dependent variables,
c. the number of the predictors can be arbitrary,
d. the predictors must be binary,

Answer 21

a. different types of variables can appear among the predictors.
c. the number of the predictors can be arbitrary,

Question 22

The odds ratio …
Select one:
a. is a number between 0 and 1,
b. is an integer,
c. can be an arbitrary number.
d. is a non-negative number,

Answer 22

d. is a non-negative number,

Question 23

A categorical variable, which can take 4 different values, can be substituted by ..... indicator variables.
Select one:
a. 3 
b. 4
c. 5
d. 1

Answer 23

a. 3

Question 24

The following data are from a screening test result. Calculate the positive predictive value!
Screening result
Confirmed
Not confirmed
Positive
2100
800
Negative
145
15400

Select one:

a. 99,1%
b. 72,4%
c. 93,5%
d. 87,6%
e. 57,9%

Answer 24

b. 72,4%

Question 25

Please select the correct answer!
Select one or more:
a. Biomarkers are very important tools for “translational medicine”.
b. The disadvantage of biomarkers is that provide information about the course of the disease only
following the standard clinical diagnostics.
c. Predictive biomarker estimates how well a given treatment will be successful for a given patient.
d. The microarray technology cannot be used in clinical settings as biomarker.

Answer 25

a. Biomarkers are very important tools for “translational medicine”.
c. Predictive biomarker estimates how well a given treatment will be successful for a given patient.

Question 26

Biomarkers are called laboratory, visual, electrical (e.g. EEG) etc. the parameters obtained from
measurements, and they are:
Select one or more:
a. correlate with the presence or severity of the disease.
b. can be used to characterize the time course of a therapeutic intervention.
c. can predictive for success of a therapeutic intervention.
d. previously they were called as soft clinical endpoint.

Answer 26

a. correlate with the presence or severity of the disease.
b. can be used to characterize the time course of a therapeutic intervention.
c. can predictive for success of a therapeutic intervention.

Question 27

What type would you choose to examine a rare exposition?
Select one:
a. cohort study
b. case control study

Answer 27

a. cohort study

Question 28

Blood pressure of the cases is measured by an experienced nurse, while blood pressure of
controls is measured by a beginner.
Options:
Select one:
a. selection bias
b. recall bias
c. random error
d. information bias

Answer 28

d. information bias

Question 29

We examine the methods of catching up of disadvantageous children with a cohort study. The
best pupils are leaving the cohort.
Options:
Select one:
a. response bias
b. selection bias 
c. instrumential bias
d. random error

Answer 29

b. selection bias

Question 30

What are the properties of a confounder?
Select one or more:
a. associated with the exposure 
b. independent risk factor of the disease 
c. an effect of the exposure 
d. not an effect of the exposure

Answer 30

a. associated with the exposure
b. independent risk factor of the disease
d. not an effect of the exposure

Question 31

Select one or more:

a. detailed description of the symptoms, diagnosis and treatment of one/more patients
b. Observation of effect the average level of an exposure to a population level end point
c. strong evidence
d. has no control group

Answer 31

a. detailed description of the symptoms, diagnosis and treatment of one/more patients
d. has no control group

Question 32

Can a casual relationship determined from ecological studies?
Select one:
a. No
b. Yes
c. Only in special cases

Answer 32

c. Only in special cases

Question 33

In which type of study can we expect to the highest number of uncontrolled confounders and biases?
Select one:
a. cross sectional study
b. cohort study
c. case control study
d. case report

Answer 33

d. case report

Question 34

Which factor determines a study to be observational or interventional?
Select one:
a. decision of the researcher
b. who determines the group for a subject (control or exposed): the researcher or the subject
c. number of enrolled subjects
d. number of groups in the study

Answer 34

b. who determines the group for a subject (control or exposed): the researcher or the subject

Question 35

Please select the correct answer!
Select one or more:
a. Statistics is a branch of science.
b. The word “statistics” has more than one meaning. It means a computed (derived) data. For example,
the average is a „statistic”.
c. The objective of statistical inference is to estimate the uncertainties due to sampling.
d. Statistics means collection of data published in yearbooks by health authorities like WHO.

Answer 35

a. Statistics is a branch of science.
b. The word “statistics” has more than one meaning. It means a computed (derived) data. For example,
the average is a „statistic”.
c. The objective of statistical inference is to estimate the uncertainties due to sampling.
d. Statistics means collection of data published in yearbooks by health authorities like WHO

Question 36

When recording data in a table,
Select one or more:
a. Qualitative properties (healthy or sick) are often encoded by numbers. For example, being in the
control group is typically denoted with 0.
b. Statistical programs denote missing data with special characters.
c. We recommend using Microsoft Excel because it treats foreign (Hungarian) characters well.
d. Statistical variables have what is called “type” like binary, continuous and so on.

Answer 36

a. Qualitative properties (healthy or sick) are often encoded by numbers. For example, being in the
control group is typically denoted with 0.
b. Statistical programs denote missing data with special characters.
d. Statistical variables have what is called “type” like binary, continuous and so on.

Question 37

By type of variables:
Select one or more:
a. Age is a binary variable.
b. The patient’s weight is a continuous variable.
c. Drugs that a patient is taking is a nominal variable.
d. Severity of the disease (like Stage 1, 2, 3 and 4) is a continuous variable.

Answer 37

b. The patient’s weight is a continuous variable.

c. Drugs that a patient is taking is a nominal variable.

Question 38

Please select the correct answer!
Select one or more:
a. Pain score (none, mild, severe) is an example for an ordinal variable.
b. Binary data is often transformed logarithmically to have a normal distribution.
c. Binary transformation of continuous data (e.g. blood pressure) (high, low) leads to information loss.
d. Typing of variables is based on the possible values a variable can take.

Answer 38

a. Pain score (none, mild, severe) is an example for an ordinal variable.
c. Binary transformation of continuous data (e.g. blood pressure) (high, low) leads to information loss.
d. Typing of variables is based on the possible values a variable can take.

Question 39

Please select the correct answer!
Select one or more:
a. The geometric mean should be used when the distribution is right skewed.
b. The median of data is sensitive to outlier data.
c. The average is sensitive to the symmetry of the distribution.
d. The mean or the median should be published, giving both at same time is unnecessary.

Answer 39

a. The geometric mean should be used when the distribution is right skewed.
c. The average is sensitive to the symmetry of the distribution.

Question 40

Please select the correct answer!
Select one or more:
a. Twenty-five percent of the sample elements are smaller than Q1.
b. The Quartiles (Q1, Q2, Q3) give three-point summaries of the data.
c. Q2 is known otherwise as a median.
d. Twenty-five percent of the sample elements are higher than Q3.

Answer 40

a. Twenty-five percent of the sample elements are smaller than Q1.
b. The Quartiles (Q1, Q2, Q3) give three-point summaries of the data.
c. Q2 is known otherwise as a median.
d. Twenty-five percent of the sample elements are higher than Q3.

Question 41

Please select the correct answer!
Select one:
a. The standard deviation is a scale independent parameter.
b. The standard deviation is also called as variance.
c. The mean can computed only for normally distributed data.
d. The coefficient of variation is a scale independent parameter.

Answer 41

d. The coefficient of variation is a scale independent parameter.

Question 42

Please select the correct answer!

Select one or more:

a. If we collect enough data then the distribution of our data will be bell-shape like (“normal”)
distribution.
b. The calculated confidence interval is a “random” interval which varies from between samples.
c. The width of the confidence interval of mean depends on the sample size.
d. The 90% confidence interval of the mean covers the population with a probability of 90%.

Answer 42

b. The calculated confidence interval is a “random” interval which varies from between samples.
c. The width of the confidence interval of mean depends on the sample size.

Question 43

The good estimation is:
Select one or more:
a. Relevant 
b. Unbiased 
c. Consistent 
d. Significant 
e. Effective 
f. Biased

Answer 43

b. Unbiased
c. Consistent
e. Effective

Question 44

Which statement(s) are true regarding to the power of a statistical test?
Select one or more:
a. The power of a statistical test is the probability of having a corresponding (to the sample) or more extreme value if the null hypothesis is true.
b. It depends on the p-value calculated on a sample.
c. It depends on the effect size and sample size.
d. The power of a statistical test is the probability of rejecting the null hypothesis if the alternative hypothesis is true.

Answer 44

c. It depends on the effect size and sample size.
d. The power of a statistical test is the probability of rejecting the null hypothesis if the alternative hypothesis is true.

Question 45

Give the meaning of unbiased estimation.
Select one:
a. The standard error of the estimation (the variability of the estimates) is small.
b. The expected value of the estimates in different samples is equal with the estimand.
c. The estimation contains the same amount of information as thewhole sample
d. The variability of estimates decreases with increased sample size.

Answer 45

b. The expected value of the estimates in different samples is equal with the estimand.

Question 46

Select the correct statements about ’incidence rate’.

Select one or more:

a. Its application in epidemiology is limited, because its interpretation is highly influenced by the loss to follow-up.
b. Incidence rate is the instantaneous risk for developing the disease at any point in time.
c. Incidence rate can be calculated as the number of new diseases developing during the follow-up period divided by the total person-time.
d. Cumulative incidence is a synonym for incidence rate.

Answer 46

b. Incidence rate is the instantaneous risk for developing the disease at any point in time.
c. Incidence rate can be calculated as the number of new diseases developing during the follow-up period divided by the total person-time.

Question 47

The Year of Nuremberg Code
Select one:
a. 1947 
b. 1936
c. 2013
d. 1960

Answer 47

a. 1947

Question 48

Role of the Ethical Committee
Select one:
a. implementation
b. licensing
c. reviewing 
d. design

Answer 48

c. reviewing

Question 49

Which is not necessary in a good inform consent?
Select one:
a. the end day of the study 
b. the experimental character
c. insurance
d. the number of enrolled patients

Answer 49

a. the end day of the study

Question 50

The Milgram experiment:
Select one:
a. there is no such experiment, I haven’t heard about it
b. is a famous experiment on the withdrawal of drinks containing vitamin C
c. is the social psychology’s obedience test
d. is an experiment on psychological burden of medical students

Answer 50

c. is the social psychology’s obedience test

Question 51

Which is not the researcher’s task?
Select one:
a. adherence to and compliance with scientific principles
b. best knowledge of the scientific field
c. protection of personal data
d. representation of the interests of individual patients

Answer 51

d. representation of the interests of individual patients

Question 52

The main role of ethical comittee regarding the trial protocoll is
Select one:
a. licensing
b. to inspect
c. the ethical review 
d. to define

Answer 52

c. the ethical review

Question 53

Which is not true for observational clinical trials?
Select one:
a. there is no active intervention
b. can be defined well legally
c. data protection is important
d. no need to notificate or licence

Answer 53

d. no need to notificate or licence

Question 54

The retrospective use of medical data
Select one:
a. is retrospective data collection, so it is not an observational study
b. can be done at any time without licensing
c. is not a real research, the Declaration of Helsinki does not name it
d. is research, the Declaration of Helsinki does name it

Answer 54

d. is research, the Declaration of Helsinki does name it

Question 55

The Declaration of Helsinki
Select one:
a. is the legal regulation of clinical trials
b. is the legal regulation of medical staff
c. is the revelation of the principles of GCP
d. is the international standard of clinical trials

Answer 55

d. is the international standard of clinical trials

Question 56

Oviedo Convention in 1998 involves
Select one:
a. the ban on cloning 
b. the regulation of researchers
c. data protection
d. the costs of the research

Answer 56

a. the ban on cloning

Question 57

Which is not part of Good Clinical Practice?
Select one:
a. ensuring that data are accurate
b. ensuring that results are believable
c. ensuring that results are accurate
d. guiding for clinical trials of medical devices

Answer 57

d. guiding for clinical trials of medical devices

Question 58

What has to be true at the same time for confounding to occur? The confounder…
Select one or more:
a. Has an effect on the endpoint 
b. Has an effect on the exposure 
c. Is associated with the endpoint 
d. Is associated with the exposure

Answer 58

a. Has an effect on the endpoint

d. Is associated with the exposure

Question 59

What does model estimation mean?
Select one:
a. To determine the uncertain value of a parameter of the sample based on the population
b. The determine the certain value of a population parameter based on a sample
c. To determine the certain value of a parameter of the sample based on the population
d. To determine the uncertain value of a population parameter based on a sample

Answer 59

d. To determine the uncertain value of a population parameter based on a sample

Question 60

What is the advantage of stratification?
Select one:
a. Universal (does not require any further assumptions)
b. Gives easily interpretable results
c. Continuous confounders can be easily handled
d. Provides one single result

Answer 60

a. Universal (does not require any further assumptions)

Question 61

What does additivity mean?
Select one:
a. The impact of a variable on the endpoint does not depend on the impact of another variable
on the endpoint
b. The value of a variable doesn’t depend on another variable’s impact on the endpoint
c. The impact of a variable on the endpoint does not depend on another variable’s value
d. The value of one variable doesn’t depend on the value of another variable

Answer 61

c. The impact of a variable on the endpoint does not depend on another variable’s value

Question 62

If two variables’ effect is not additive, we can also say that…
Select one:
a. There is no interaction between them
b. There is an interaction between them
c. They are independent from each other
d. They are not independent from each other

Answer 62

b. There is an interaction between them

Question 63

A variable has 4 possible outcomes. How many dummy variables do we need to enter it into a
model?
Select one:
a. 3
b. 1
c. 4
d. 2

Answer 63

A variable has 4 possible outcomes. How many dummy variables do we need to enter it into a
model?
Select one:
a. 3

Question 64

If we enter a continuous variable into the linear regression model as-is, what do we assume of its impact on the
endpoint? It is…
Select one:
a. Linear
b. Additive 
c. Quadratic
d. Non-linear

Answer 64

a. Linear

Question 65

What is multicollinearity?
Select one:
a. If explanatory variables explain each other
b. If explanatory variables explain the response variable
c. If the response variable explains itself
d. If the response variable explains the explanatory variables

Answer 65

a. If explanatory variables explain each other

Question 66

Which of the following is a property of an estimator?
Select one or more:
a. Monotone 
b. Unbiased 
c. Consistent 
d. Efficient

Answer 66

b. Unbiased
c. Consistent
d. Efficient

Question 67

Which of the following is the type of the test that tests the significance of a group (possibly
including all) of variables?
Select one:
a. F-test
b. χ2 -test 
c. z-test
d. t-test

Answer 67

d. t-test

Question 68

Confounding violates which of the following model assumptions?
Select one:
a. Error normality
b. Linearity 
c. Lack of autocorrelation
d. Exogeneity

Answer 68

d. Exogeneity

Question 69

Which is the most widely used metric to measure the goodness-of-fit of a model?
Select one:
a. K2
b. H2
c. A2
d. R2

Answer 69

d. R2

Question 70

What are the possible applications of a statistical model?
Select one or more:
a. Interpretation (with numerical results)
b. Interpretation (with visualization)
c. Prediction
d. Clairvoyance

Answer 70

a. Interpretation (with numerical results)
b. Interpretation (with visualization)
c. Prediction

Question 71

What can cause overfitting?
Select one or more:
a. Apparently non-complex model, but one that was reached through many fine-tuning steps
b. A model assumption is violated
c. Too large sample size
d. Too complex model, e.g. too many variables are included

Answer 71

a. Apparently non-complex model, but one that was reached through many fine-tuning steps
d. Too complex model, e.g. too many variables are included

Question 72

What should we do if we have to select from candidate explanatory variables?
Select one or more:
a. Filter them based on redundancy
b. Filter them based on medical considerations
c. Use automatic model selection methods without validation
d. Do not filter them, but use special estimation methods

Answer 72

a. Filter them based on redundancy
b. Filter them based on medical considerations
d. Do not filter them, but use special estimation methods

Question 73

The odds …
Select one:
a. can be an arbitrary number. 
b. is an integer,
c. is a number between 0 and 1,
d. is a non-negative number,

Answer 73

d. is a non-negative number,

Question 74

Which statement is true?
Select one or more:
a. Let us suppose that for comparison of an exposed group with another one we calculate an
odds ratio concerning the occurrence of a disease. The estimation of the odds ratio
generally becomes different if we take into consideration also the sex and the age of the
subjects.
b. Interaction between predictors are not allowable in the logistic regression models.
c. We can choose any value of a categorical variable for reference value (i.e. any of the
groups determined by the variable can be the reference category).
d. If there are several predictors in a logistic regression model then also a negative value can
appear among the odds ratios estimated from this model.

Answer 74

a. Let us suppose that for comparison of an exposed group with another one we calculate an
odds ratio concerning the occurrence of a disease. The estimation of the odds ratio
generally becomes different if we take into consideration also the sex and the age of the
subjects.
c. We can choose any value of a categorical variable for reference value (i.e. any of the
groups determined by the variable can be the reference category).

Question 75

If we investigate the occurrence of a disease among people with high blood pressure and people with
low blood pressure and we get an odds ratio equal to 2, then
Select one or more:
a. the odds of the occurrence of the disease among people with low blood pressure is half as
big than in the group of the people with high blood pressure.
b. the probability of the occurrence of the disease among people with low blood pressure is
half as big than in the group of the people with high blood pressure,
c. the odds of the occurrence of the disease among people with high blood pressure is two
times greater than in the other group,
d. the probability of the occurrence of the disease among people with high blood pressure is
two times greater than in the other group,

Answer 75

a. the odds of the occurrence of the disease among people with low blood pressure is half as
big than in the group of the people with high blood pressure.
c. the odds of the occurrence of the disease among people with high blood pressure is two
times greater than in the other group,

Question 76

By a statistical test it is worth investigating whether an odds ratio
Select one:
a. is significantly different from 1,
b. is positive,
c. correlates significantly with the probabilities.
d. is significantly different from 0,

Answer 76

a. is significantly different from 1,

Question 77

In the logistic regression models
Select one:
a. the estimated coefficients can be interpreted as odds,
b. the estimated coefficients can be interpreted as odds ratios,
c. the estimated coefficients can be interpreted as probabilities.
d. odds ratios can be calculated from the estimated coefficients,

Answer 77

d. odds ratios can be calculated from the estimated coefficients,

Question 78

In the logistic regression models if we interpret the odds ratio belonging to a value of the categorical
predictor, the point of reference is
Select one:
a. the reference category, 
b. the previous category,
c. the dependent variable.
d. the previous predictor,

Answer 78

a. the reference category,

Question 79

In the logistic regression models the odds ratio belonging to a continuous predictor
Select one or more:
a. cannot be less than 1.
b. can be interpreted so that the odds of the occurrence of the examined outcome will be
multiplied by this number if the value of the predictor increases with one unit,
c. depends on the measure unit of the predictor,
d. can be interpreted so that the odds of the occurrence of the examined outcome will be
multiplied by this number if the value of the dependent variable increases with one unit,

Answer 79

b. can be interpreted so that the odds of the occurrence of the examined outcome will be
multiplied by this number if the value of the predictor increases with one unit,
c. depends on the measure unit of the predictor,

Question 80

Please select the correct answer!
Select one:
a. The clustering algorithms can be used to specify the group’s FDR value.
b. The hierarchical clustering algorithm combines groups that are not significantly different from each
other.
c. The result of hierarchical clustering is presented with the so-called heatmap.
d. In a microarray experiment clustering helps to find genes which expression is increased due to
treatment.

Answer 80

c. The result of hierarchical clustering is presented with the so-called heatmap.

Question 81

The following data are from a screening test result. Calculate the sensitivity!
Screening result
Confirmed
Not confirmed
Positive
2100
800
Negative
145
15400

Select one:

a. 93,5%
b. 72,4%
c. 57,9%
d. 87,6%
e. 99,1%

Answer 81

a. 93,5%

Question 82

The researcher decides to use values under 20 µg/dl as normal limits for a screening test and the test becomes commercially available. One of Your patients has a test result of 27 µg/dl . You conclude that…
Select one:
a. A confirmation test will be needed as she may or may not have cancer
b. The patient does not have cancer of the ovary
c. This is a false-negative test
d. This test is not sensitive enough to detect cancer
e. The patient has cancer of the ovary

Answer 82

a. A confirmation test will be needed as she may or may not have cancer

Question 83

Please select the correct answer!
Select one or more:
a. Bonferroni correction is a statistical procedure that reduces the chance of making false statements
when more than one statistical hypothesis tests are carried out.
b. Bonferroni correction means that the FDR value cannot be more than 50%.
c. The “omics” technology is characterized by measuring several even 1000000 parameters
simultaneously.
d. The Bonferroni correction is recommended to evaluate „. OMICS” data.

Answer 83

a. Bonferroni correction is a statistical procedure that reduces the chance of making false statements
when more than one statistical hypothesis tests are carried out.
c. The “omics” technology is characterized by measuring several even 1000000 parameters
simultaneously.

Question 84

Please select the correct answer!
Select one or more:
a. Blood pressure is a surrogate marker for antihypertensive drugs.
b. Stroke risk is a commonly used biomarker.
c. Every biomarker is a surrogate endpoint.
d. The “surrogate endpoint” is an endpoint which is accepted by the authorities instead of the usual
clinical endpoint.

Answer 84

a. Blood pressure is a surrogate marker for antihypertensive drugs.
d. The “surrogate endpoint” is an endpoint which is accepted by the authorities instead of the usual
clinical endpoint.

Question 85

If the researcher chooses values under 25 µg/dl as normal limits for the test, which of the following statements is true?
Select one:
a. The test will be 100% specific
b. The test will be 100% sensitive
c. All persons with cancer will have a positive test
d. No false-negative tests will occur
e. There will be some false positive tests

Answer 85

e. There will be some false positive tests

Question 86

Please select the correct answer!
Select one or more:
a. The microarray technology cannot be used in clinical settings as biomarker.
b. The disadvantage of biomarkers is that provide information about the course of the disease only
following the standard clinical diagnostics.
c. Biomarkers are very important tools for “translational medicine”.
d. Predictive biomarker estimates how well a given treatment will be successful for a given patient.

Answer 86

c. Biomarkers are very important tools for “translational medicine”.
d. Predictive biomarker estimates how well a given treatment will be successful for a given patient.

Question 87

If the researcher chooses values under 30 µg/dl as normal limits for the test, which of the following statement is true?
Select one:
a. The test will be 100% sensitive
b. Some persons without cancer will test positive
c. The test will be 100% specific
d. All persons with cancer will have a positive test
e. There will be some false positive tests

Answer 87

c. The test will be 100% specific