RESS Flashcards
Epidemiology is the “study of the distribution & determinants of health related states or events in specified populations”. Which of these is a health related state?
a) Smoking
b) Death
c) Age
B) Death
Other health related states (or outcome measures): incidence/prevalence, hospital admissions, birth weight, life expectancy.
Age and smoking are determinants (risk factors/exposures/causes)
Following the scientific method, what do you do after proposing a hypothesis?
a) Modify it
b) Reject it
c) Test it
c) Test it
- Observe
- Propose hypothesis
- Test hypothesis
- Reject and modify hypothesis
- Do not reject and continue to test hypothesis
Which type of population study design typically follows patients up over time to see if they develop a disease of interest?
a) Case series
b) Cross-sectional study
c) Cohort study
d) Case-control study
e) Ecological study
c) Cohort study
Which type of population study design typically selects patients with/without a disease and looks back over time to determine their exposure?
a) Case series
b) Cross-sectional study
c) Cohort study
d) Case-control study
e) Ecological study
d) Case-control study
Which study design describes a sample of cases with the same disease?
a) Case series
b) Cross-sectional study
c) Cohort study
d) Case-control study
e) Ecological study
a) Case series
Which study design examines variations between geographical areas?
a) Case series
b) Cross-sectional study
c) Cohort study
d) Case-control study
e) Ecological study
e) Ecological study
Which study design studies a group of people at a single point in time?
a) Case series
b) Cross-sectional study
c) Cohort study
d) Case-control study
e) Ecological study
b) Cross-sectional study
What type of health outcome is ‘mortality’?
a) Record-based
b) Biological/clinical
c) Clinician/patient-reported
a) Record-based
What type of health outcome is ‘quality of life’?
a) Record-based
b) Biological/clinical
c) Clinician/patient-reported
c) Clinician/patient-reported
What term means that an outcome measures accurately what it meant to measure?
a) Validity
b) Reliability
c) Responsiveness
a) Validity
What term means that repeating the test will bring about the same results?
a) Validity
b) Reliability
c) Responsiveness
b) Reliability
Which term means that you can detect real changes in outcomes when they occur?
a) Validity
b) Reliability
c) Responsiveness
c) Responsiveness
What measures how many people have a disease at a specific point in time?
a) Incidence
b) Prevalence
c) Person-time
d) Mortality
e) Case fatality rate
b) Prevalence
What measures the number of new cases of a disease over a specific time period?
a) Incidence
b) Prevalence
c) Person-time
d) Mortality
e) Case fatality rate
a) Incidence
How do you calculate incidence rate?
New cases/population at risk
How do you calculate prevalence?
People with disease/Population
How do you calculate mortality?
Number dying from disease/population
How do you calculate case fatality?
Number dying from disease/number with disease
Considering a common, short duration disease such as influenza, which of the following would you expect to see?
a) Low incidence, low prevalence
b) Low incidence, high prevalence
c) High incidence, low prevalence
d) High incidence, high prevalence
c) High incidence, low prevalence
Lots of people get it but they are cured quickly, therefore only few people have it at a single point in time
How would you interpret a relative risk (risk ratio) of 1? The exposure:
a) Is protective
b) Is harmful
c) Has no impact
c) Has no impact
We are comparing 2 groups – exposed and unexposed.
Relative risk is a ratio, so a value of 1 means a ratio of 1:1 which means that the risk for the exposed is the same as the unexposed and therefore the exposure has no more (or less) impact than the control
A relative risk (risk ratio) of >1 means what?
Exposure is harmful
> 1 means the risk in the exposed group and greater than the risk in the control group
A relative risk (odds ratio) of <1 means what?
Exposure is protective
<1 means the risk in exposed group is less than the risk in the control group
Which of these is not a valid keyword search technique?
a) Truncation
b) Adjacency searching
c) Wildcard
d) Wild key
d) Wild key
? and # are examples of ____ that are used when searching for literature
wildcards
E.g. behavio?ral to check for US and UK spelling. Searches for one character that may or may not be present.
E.g. Wom#n to search for both woman and women. Searches for one character, but there is definitely a character there.
What search term could you use to find both these titles:
- Eating disorder
- Eating related disorder
adj2
adjacency searching
What symbol is used for truncation when searching for literature?
Asterisk *
Which of the following would be returned if you search for anorexi* in Medline? (select one or more)
a) Anorex
b) Anorexia
c) Anorexic
d) Anorexia nervosa
b, c and d
The asterisk is for truncation
Which Boolean operator would you use to only get results that include both concepts?
a) Anorexia AND CBT
b) Anorexia OR CBT
c) Anorexia NOT CBT
d) Anorexia NOR CBT
a) Anorexia AND CBT
OR will find anorexia on its own, and CBT on its own but not necessarily both
NOT will find anorexia where CBT isn’t mentioned
Which Boolean operator would you use to search for synonyms?
a) Anorexia AND CBT
b) Anorexia OR CBT
c) Anorexia NOT CBT
d) Anorexia ADJ CBT
b) Anorexia OR CBT
What type of data is weight?
a) Continuous
b) Discrete
c) Ordinal
d) Nominal
a) Continuous
Can be any number, no restriction. Even if you measure to a certain number of decimal places.
What data type do you get if you use the Likert scale?
a) Continuous
b) Discrete
c) Ordinal
d) Nominal
c) Ordinal
Which is what makes it difficult to analyse – you used correlation, t-tests etc. to look at continuous variables. You shouldn’t use those methods for this variable
What type of data is being described:
Numerical value, no limitation on values e.g. weight. Even if recorded in whole units.
Continuous data
What type of data is being described:
Numerical value, limitation on values e.g. number of hospital appointments
Discrete data
What type of data is being described:
Ordered category e.g. Likert scale, stage of disease
Ordinal data
What type of data is being described:
Non-ordered category. E.g. sex, blood group
Nominal
How would you summarise the average of a non-normally distributed variable?
a) Mean
b) Median
c) Standard deviation
d) Inter-quartile range
b) Median
Median best for skewed data. Mean for normally distributed data. The mean is affected by abnormally high or low values, the median is not.
SD and IQR are measures of spread. Use SD with the mean and IQR with median
How would you check on the distribution of a continuous variable? Plot a:
a) Boxplot
b) Bar chart
c) Histogram
d) Scatterplot
c) Histogram
The area of a bar in a histogram is proportional to the frequency (frequency histogram), so height is proportional to class width. Histograms are plotted for continuous variables.
What method may be used to assess associations between 2 categorical variables? (select one or more)
a) Risk/odds
b) T-test
c) Correlation
d) Linear regression
a) Risk/odds
What method may be used to assess associations between 2 categorical variables?
(select one or more)
a) Risk/odds
b) T-test
c) Correlation
d) Linear regression
c and d
How would you assess association between a single categorical exposure and continuous outcome?
T-test
How would you assess association between a single continuous exposure and continuous outcome?
Correlation
How would you assess association between multiple exposures and continuous outcome?
Linear regression
The p-values measure the:
a) Effect of a treatment
b) Accuracy of a statistical test
c) Probability of a result
c) Probability of a result
So p=0.05 means a 1 in 20 (5%) chance of an event happening.
Combined with a hypothesis , it’s a 5% chance of seeing the results you saw if the null hypothesis were true.
Small chance => reject null. Large change = cannot reject null.
P-value does not provide “proof” but “evidence”
The effect size is also important, but different. The confidence interval is also important.
P-value can’t tell you whether the statistical test was accurate, or the correct one to use, or whether you used it correctly
A set of weighing scales varies according to climate. This is an example of:
a) Error
b) Bias
a) Error
Random error
- Error due to random factors
- Measurement error
- Error from chance fluctuations in the profile of our sample
Systematic error
- Error due to non-random factors
- Measurement bias, confounding bias etc.
A distortion of an association between 2 variables due to a common shared cause is called:
a) Confounding bias
b) Experimenter bias
c) Information bias
d) Selection bias
a) Confounding bias
A systematic difference between patients chosen vs. not chosen for a study is called:
a) Confounding bias
b) Experimenter bias
c) Information bias
d) Selection bias
d) Selection bias
Bias due to Distortion of association due to a common shared cause is called ____
Confounding bias
Bias due to the behaviours and actions of the experimenter is called ____
Experimenter bias
Bias due to systematic error in reporting, measurement or recording of information is called ____
Information bias
Bias due to systematic difference between those selected into a study sample and those not selected is called ____
Selection bias
How do you calculate standard error?
stnd. deviation/sqrt(n)
What is standard error?
A measure of precision
What is standard deviation?
A measure of spread