stud
bud
Sign up
Log in
Home
Library
Explore
SAT
More
Epidemiology: Measures of Association & Causal Reasoning I Flashcards | StudBud
Explore
/
Epidemiology: Measures of Association & Causal Reasoning I
Epidemiology: Measures of Association & Causal Reasoning I
200 terms
·
StudBud Administration
·
1 play
Imported from external platform
Flashcards
200 cards
F
Practice quiz
P
Host live
L
Questions
200
Answers
1
Which study design is best for proving time sequence?
Case-control
Cross-sectional
Ecological
Prospective cohort
2
Why might cause "A" be considered strongest in a study?
Only factor contributing to disease
Has the highest OR in stats
Appears in every causal pie
Most commonly studied risk factor
3
What is an "exact match" in matching?
Non-diseased subject must be same age as diseased subject
Non-diseased subject must be within a meaningful age range
Non-diseased subject must be randomly assigned
Non-diseased subject must have different exposure status
4
What is the purpose of statistical testing?
To find the cause of disease
To count cases in a study
To infer population values
To replace biology with math
5
Why are ratio measures transformed before CIs?
No transformation is needed
They aren’t symmetrical around 1
They are normally distributed
They range from -1 to 1
6
Which is an example of "experimental evidence" supporting causation?
Lab studies confirming a biological mechanism
Anecdotal reports from doctors
A public opinion survey on risk factors
A case report linking exposure to disease
7
What does a higher OR or RR indicate?
Lower disease prevalence
More important cause of disease
Proof of causation
Stronger exposure-disease link
8
What does a 95% confidence interval mean?
The true value is always in the interval
It applies to individuals, not populations
A 95% CI is more accurate than a 99% CI
The interval has the true value 95% of the time
9
Why is matching used in epidemiological studies?
To estimate the effect of a variable
To control for confounding variables
To randomly assign exposure groups
To introduce selection bias
10
What does a large chi-square value indicate?
Sample size is too small
Observed and expected values are nearly identical
Strong association between exposure and outcome
Null hypothesis should be accepted
11
Who first formulated principles leading to Koch’s Postulates?
Edward Jenner
F.G. Jacob Henle
Louis Pasteur
John Snow
12
Which method compares the amount of disease between an exposed and unexposed group?
Prevalence difference
Case fatality rate
Cumulative incidence
Ratio measures
13
Which is an example of a causal relationship?
People who exercise eat healthier
Neutering an animal causes sterility
Owning a pet is linked to more exercise
Students who study together score higher
14
What describes a causal association?
One variable directly affects another
A correlation that needs no evidence
A random relationship between two variables
Any statistical link, regardless of biology
15
What is causal inference?
The process of proving all hypotheses
A scientific method to infer causation
Direct measurement of causation
A statistical test to prove causation
16
What do relative measures assess in epidemiological studies?
The total population at risk
The difference in absolute risk between two groups
The ratio of disease occurrence between two groups
The number of individuals exposed to a risk factor
17
What is the core idea behind Karl Popper’s refutationism?
Observations always confirm theories
Hypotheses must always be inductively derived
Hypotheses can never be proven, only falsified
A single experiment proves scientific truth
18
What does inductive reasoning rely on?
Rejecting all hypotheses
Deductive proofs
Observing patterns to infer general laws
Assuming correlation equals causation
19
Which is NOT one of Hill’s criteria for causation?
Biologic plausibility
Time sequence
Statistical significance
Dose-response relationship
20
What is the most common ratio used in matching?
10:1
2:1
1:1
5:1
21
Epidemiological study designs can be used to assess relationships between:
Medical treatments and patient satisfaction
Diseases and potential causative factors
Environmental changes and government policies
Population growth and economic trends
22
In which type of study is the odds ratio most commonly used?
Case-control studies
Cohort studies
Cross-sectional studies
Randomized controlled trials
23
What does a dose-response relationship indicate?
A single exposure always causes disease
Causal relationships don’t need exposure variation
Increasing exposure increases disease risk
Exposure and disease occur simultaneously
24
What concerns cumulative evidence rather than a single study?
Criteria for causation
Study error (bias)
Intrinsic relationships
Chance (random variation)
25
What does specificity suggest about a cause?
Should have no health effects
Causes multiple diseases
Linked to one disease
Must be genetic
26
Which of the following is NOT typically calculated from a 2 × 2 table?
Absolute risk reduction
Rate ratio
Risk difference (attributable risk)
Odds ratio
27
Why use P-values in analysis?
Measures CI size
Objective association test
Removes bias
Confirms causation
28
Which measure compares probability of disease in exposed vs. unexposed groups?
Rate ratio
Attributable risk
Case fatality rate
Risk ratio
29
Which measure compares incidence rate in exposed vs. unexposed groups?
Risk ratio
Prevalence ratio
Odds ratio
Rate ratio
30
What is true about analogy in causation?
Strongest evidence
Does not prove causation
Replaces experiments
Required for causation
31
Why are confidence intervals important in epidemiology?
Eliminate random error in studies
Always include exact population parameter
Replace need for statistical tests
Quantify expected variability in estimates
32
Who developed the concept of induction in scientific reasoning?
John Stuart Mill
Francis Bacon
Karl Popper
Isaac Newton
33
What does the reliability coefficient (RC) show?
Type I error probability
Standard deviation
Sample size correction factor
Multiplier for standard error
34
What does number needed to treat (NNT) represent?
Cases needed to calculate risk difference
Fraction of disease in a population due to exposure
Animals treated to prevent one negative outcome
Probability an animal will develop disease after treatment
35
Why are concordant pairs ignored when calculating the matched OR?
Increase study’s statistical power
Reflect effect of matching variables, not exposure effect
Provide no useful statistical information
Bias results toward the null
36
Why is specificity the least useful criterion?
Cannot be tested
Only applies to infections
Causes often have multiple effects
Does not apply to cancer
37
How does correlation differ from association in epidemiological studies?
Correlation measures statistical relationship association is broader
Correlation always implies causation
Association is only used for categorical data
Association is a subset of correlation
38
Which example challenges specificity?
A gene linked to one disorder
Smoking and lung cancer
Smoking and heart disease
TB infection and tuberculosis
39
Which measure has a null value of 1?
Probability / Risk
Risk Difference (Attributable Risk)
Rate
Risk Ratio
40
What does coherence mean in causal inference?
Only lab tests are valid
Findings align with current knowledge
All studies must match exactly
Requires experimental data
41
What is an example of a two-sided test?
H₀: μ₁
H₀: μ₁ ≤ μ₂, Hₐ: μ₁ > μ₂
H₀: μ₁ > μ₂, Hₐ: μ₁ ≤ μ₂
H₀: μ₁ = μ₂, Hₐ: μ₁ ≠ μ₂
42
How can dose-response trends be distorted?
Excluding statistical significance
Using different category groupings
Ignoring the threshold effect
Including too many participants
43
How are odds mathematically defined?
The difference between two probabilities
The probability that something will not occur
ratio of prob something will occur to the prob it will not occur
The probability that something will occur
44
What does it mean if P(A, B) ≠ P(A) * P(B)?
The null hypothesis is true
The two factors are not independent
No association exists between the factors
The two factors are independent
45
How many pieces of data are required to calculate measures of association using a 2 × 2 table?
Five
Two
Four
Three
46
How can preconceived beliefs affect epidemiological conclusions?
They prevent statistical associations from being analyzed
They always confirm the null hypothesis
They can influence how people interpret data
They eliminate bias in data collection
47
What does the null hypothesis state in epidemiology?
There is no association between the factors being studied+C31:C54
Every correlation implies causation
All risk factors directly cause disease
Disease prevalence is the same in all populations
48
What is a limitation of statistical significance?
Replaces experiments
Always due to sampling error
No confidence intervals
May not be important
49
Why are Koch’s Postulates important?
They classify all microbes as pathogens
They apply to non-infectious diseases
They prove all diseases have a microbial cause
They help determine if a microorganism causes disease
50
What is the range of possible values for risk difference?
0 to 1
-∞ to +∞
Any positive value
-1 to 1
51
What does the causal pie model suggest about timing?
Some causes happen long before disease appears
Only one factor contributes at a time
All causes act at the same time
Time is not a factor in disease
52
What is TRUE about the argument: "Pigs have wings. All creatures with wings can fly. Pigs can fly."?
Valid but unsound
Neither valid nor sound
Sound and valid
Sound but invalid
53
What is the primary focus of analytical epidemiology?
Describing disease distribution in populations
Recording disease prevalence without analysis
Testing hypotheses related to disease causation
Providing medical treatment for diseases
54
Why doesn’t correlation imply causation in epidemiology?
All associations are assumed to be random
Only a small subset of correlated associations are truly causal
Epidemiologists avoid studying causal mechanisms
Every exposure leads to disease
55
When is the odds ratio a good approximation of the risk ratio (RR)?
When prevalence is high
When sample size is large
When case-control studies use 2:1 matching
When outcome is rare (
56
What does "risk" refer to in epidemiology?
The number of cases recorded over time
The total number of exposures in a study
The severity of a disease in a population
The probability of an event occurring within a certain time period
57
What is a key limitation of ratio measures?
They only apply to chronic diseases
They cannot determine how strongly a factor is associated with an outcome
They do not provide information on absolute risk or impact
They require direct biological testing
58
Which is an example of a dose-response relationship?
One virus exposure always causes infection
A genetic mutation always leads to disease
More smoking is linked to higher mortality
Diet has no effect on cholesterol
59
Why is consistency important for causation?
Every study must report the same results
Similar results across multiple studies
A single study is enough for causation
Consistency eliminates statistical analysis
60
If chi-square test results in P-value < 0.05, what does this suggest?
Exposure directly causes the outcome
Study needs a larger sample size
Null hypothesis must be accepted
Statistical evidence of association
61
What does analogy suggest in causal inference?
New causes must be unique
A cause must be genetic
A similar known cause supports causation
Only direct experiments matter
62
Which of the following is TRUE about correlation and causation?
Causal relationships are easier to prove than associations
Every correlated association is also causal
Correlation does not necessarily mean causation
Epidemiologists always assume correlation equals causation
63
Why is experimental evidence strong for causation?
Can be done for all exposures
Allows high control over variables
Removes need for statistics
Only valid method
64
Why is logic valuable in causal reasoning?
Dissects arguments and analyzes structure
Replaces statistical methods in epidemiology
Eliminates errors in scientific observations
Confirms causation without empirical evidence
65
Why is consistency not helpful in only one study?
All studies must use the same methods
One study does not establish a pattern
A single study lacks statistical significance
Single studies are always inaccurate
66
What is TRUE about strength of an association?
Guarantees causation
Represents strength of evidence
Represents strength of the cause itself
Weak association means no relationship
67
Which is an example of inductive reasoning?
All planets orbit stars, so Earth orbits the Sun
Sun rises every day, so it will rise tomorrow
Newton’s law says thrown objects fall
Birds have feathers, so a feathered creature must be a bird
68
What does absolute impact quantify in epidemiological studies?
The strength of association between an exposure and disease
The proportion of individuals who recover from disease
The likelihood of disease occurring in the unexposed group
The consequences in the population after exposure to a risk factor
69
What is confounding bias?
A type of selection bias
Mixing of effects between variables
Random error in measurements
Errors in data entry
70
How does study design affect time sequence?
Some designs are better at proving sequence
All designs establish time order equally
Study design does not affect causation
Only experiments determine time sequence
71
What is the usual threshold for significance?
P > 0.05
P
P
P
72
What makes an argument valid?
Argument is widely accepted by experts
Argument is based on empirical data
Conclusion logically follows from premises
Premises are all factually true
73
When does confounding occur?
When data is collected inconsistent
When another variable distorts the true effect
When sample sizes are too small
When groups are not randomly selected
74
What is the usual confidence interval level in studies?
99%
90%
50%
95%
75
What transformation is used for ratio measures?
Exponential
Natural logarithm (log)
Standard deviation
Square root
76
What measure is similar to standard error (SE)?
P-value
Incidence rate
Risk ratio
Standard deviation (SD)
77
Which type of measure calculates the difference in absolute risk between two groups?
Prevalence rates
Measures of impact
Incidence ratios
Ratio measures
78
Why is the statistical null hypothesis often opposite of a biological hypothesis?
Null hypothesis always proves causation
Biological hypotheses cannot be tested directly
Statistical tests always support alternative hypothesis
Statistical methods require a "no effect" assumption
79
Which criterion refers to whether a causal relationship is biologically reasonable?
Strength of association
Experimental evidence
Coherence
Biologic plausibility
80
If an outcome is common (>10% prevalence), what is a potential issue with using the odds ratio?
Underestimates the true risk ratio
Cannot be calculated
Same as the prevalence ratio
May overestimate the true relative risk
81
How do epidemiological studies identify risk factors?
By testing treatments on affected individuals
By analyzing disease symptoms only
By comparing incidence or prevalence between exposed and unexposed groups
By recording individual case studies
82
What is an association in epidemiology?
Proof that exposure leads to disease
A dependence between two variables
A direct cause-effect relationship
A conclusion based only on biology
83
Why is experimental evidence not always possible?
Cannot test chemicals
Always takes decades
Some exposures cannot be controlled
Too expensive
84
Which measure has a range of [0, ∞] and a null value of 1?
Probability / Risk
Risk Difference (Attributable Risk)
Odds Ratio
Rate
85
Which of Hill’s criteria states that an exposure leads to one disease?
Specificity
Analogy
Biologic plausibility
Coherence
86
What describes deductive reasoning?
Uses personal experience to generalize
Forms hypotheses based on probability
Moves from specific observations to general conclusion
Moves from general principle to specific case
87
What determines which component has a stronger association?
Prevalence of component in the population
Level of statistical significance
Number of studies on each factor
Number of diseases linked to each factor
88
Which measure has a range of [0,1] and a null value of 0?
Odds
Rate Ratio
Probability / Risk
Odds Ratio
89
Which measure requires disease-free time rather than simple counts?
Risk ratio (RR)
Odds ratio (OR)
Rate ratio
Prevalence ratio (PR)
90
How is a log-transformed CI converted back?
Add SE to mean
Exponentiate values
Multiply by 1.96
Square root
91
How can confounding be prevented?
By ignoring missing data
By using different tools for each group
By adjusting for confounders in analysis
By randomly selecting participants
92
Who formalized Koch’s Postulates in 1877?
Alexander Fleming
Robert Koch
Rudolf Virchow
Joseph Lister
93
Matching can be done at which level?
Only at the group level
Only at the individual level
Both group and individual level
Only in randomized trials
94
Which measure has units expressed as "cases per time"?
Probability
Odds Ratio
Risk Ratio
Rate
95
Which measure has a null value of 0 and can extend to infinity?
Rate
Rate Ratio
Risk Ratio
Odds
96
What is a threshold effect in a dose-response relationship?
All exposures impact disease equally
Association is always linear
No effect until exposure reaches a certain level
Effect suddenly decreases at high doses
97
Which type of hypothesis investigates biological mechanisms of disease?
Statistical hypothesis
Biological hypothesis
Epidemiological hypothesis
Mathematical hypothesis
98
Which of the following describes a risk factor?
A factor that guarantees disease development
A symptom of an existing disease
A factor associated with an increased risk of an event occurring
Any environmental exposure
99
What does the rate ratio compare?
Probability of an event occurring in a population
Difference in absolute risk between two groups
Ratio of two rates
Proportion of disease cases in a population
100
How is strength of association estimated?
Observing case reports
Measuring statistical significance alone
Using OR and RR
Using absolute risk differences
101
Why does smoking challenge specificity?
Not statistically proven
Causes more than one disease
Only linked to lung issues
Does not always cause cancer
102
How do regular and matched OR differ in case-control studies?
Regular OR analyzes individuals, matched OR analyzes pairs
Regular OR only valid for rare diseases, matched OR is not
Regular OR is used only in cohort studies
Matched OR includes concordant pairs in calculation
103
Which falls within the scope of logic?
Conducting laboratory experiments
Studying reasoning about probability and causality
Measuring disease incidence in populations
Collecting raw epidemiological data
104
What is the purpose of Koch’s fourth postulate?
Prove all microorganisms cause disease
Ensure microorganism survives in different environments
Determine antibiotic effectiveness
Confirm it can be recovered from an infected host
105
What does vaccine efficacy measure?
Total number of vaccinated individuals in a study
Fraction of disease in unvaccinated group prevented by vaccine
Percentage of vaccinated individuals who develop disease
Probability vaccinated individual will get sick
106
Which measure is used in case-control studies to estimate disease likelihood?
Rate ratio
Risk ratio
Prevalence ratio
Odds ratio
107
What is the null value for a rate ratio?
∞
0.5
1
0
108
Which statistical test is commonly used to compare RR and OR?
Linear regression
Chi-square test
ANOVA
T-test
109
Why do odds not have meaning outside of their approximation to probabilities?
Odds are only meaningful when converted into probabilities
Odds are always greater than probabilities
Odds cannot be used to compare groups
Odds are just another way of expressing risk, not probability
110
What is the primary interest of epidemiologists studying disease?
Identifying statistical correlations
Describing random variations in health outcomes
Proving that all associations are causal
Determining what causes disease in populations
111
What is the purpose of a one-sided hypothesis test?
Determine if one group is greater/smaller
Replace confidence intervals
Test for any possible difference
Ensure results are always significant
112
What Z statistic is used for a 95% confidence interval?
1.96
1.64
3.14
2.58
113
What does a large P-value suggest?
Strong evidence of an effect
No association or effect
Hₐ is correct
Proof H₀ is false
114
What happens after all necessary causes occur but before disease appears?
Disease must appear immediately
Other coincidental factors may occur
No additional factors influence disease
Disease outcome becomes uncertain
115
What is the purpose of a P-value?
Measures chance in results
Replaces confidence intervals
Finds disease causes
Determines needed sample size
116
When should risk difference be used?
Only in case-control studies
When comparing multiple exposure groups
When study design allows calculation of probabilities
Only when analyzing rare diseases
117
What determines "strength" of association in a population?
Strength of causal factor itself
Statistical significance level
Number of cases diagnosed
Prevalence of component causes
118
How is the chi-square value interpreted for statistical significance?
Use it to calculate mean risk ratio
Divide by the sample size
Compare to a critical value in a chi-square table
Transform into a confidence interval
119
What does biologic plausibility refer to?
Statistical power of a study
Findings align with biological knowledge
Proof of causation in one study
Strength of association between variables
120
What describes inductive reasoning?
Applies general law to a specific situation
Makes assumptions based on contradictions
Moves from specific observations to general conclusion
Uses math formulas to prove theories
121
Which of the following best defines an association in epidemiology?
Proof that an exposure directly causes disease
A random occurrence with no statistical significance
A mathematical relationship between an exposure and disease
A laboratory-confirmed mechanism of disease development
122
What is NOT a use of statistical testing?
Making decisions on hypotheses
Giving biological evidence
Proving absolute causation
Evaluating study results
123
What does it mean when an individual is "exposed" to a risk factor?
The individual has a genetic predisposition to the disease
The individual is immune to the disease
The individual is guaranteed to develop the disease
The individual has been in contact with the risk factor
124
Which measure quantifies absolute impact of exposure by risk difference?
Odds ratio
Rate ratio
Prevalence ratio
Risk difference (Attributable Risk)
125
Why is a sound argument not necessarily proof of truth in empirical sciences?
Empirical observations can contain errors
Valid arguments cannot be sound in science
Scientific conclusions do not require logic
Logic does not apply to scientific reasoning
126
Why is biologic plausibility subjective?
Based only on statistics
Requires experimental human data
Science evolves over time
Only applies to infectious diseases
127
Which criterion is most important for causation?
Strength of association
Biologic plausibility
Time sequence
Specificity
128
Which is an example of an association, not causation?
Virus exposure leads to infection
Pet adopted from shelter is more likely to be sterilized
Drug inhibits bacterial growth in lab
Smoking damages lung tissue leading to cancer
129
What is Koch’s third postulate?
Disease must be treatable with a vaccine
Absent in healthy individuals
Isolated microorganism causes disease when inoculated
Must be destroyed by antibiotics
130
Which 1964 report helped establish causal criteria in epidemiology?
Bradford Hill Criteria
US Surgeon General’s report on smoking & health
Koch’s Postulates Report
WHO Global Disease Report
131
Why is Koch’s second postulate difficult to satisfy?
Some pathogens exist in healthy carriers
All microbes exist in pure culture in nature
Some diseases come from non-living agents
Pathogens cannot be isolated
132
Which criterion considers whether findings align with existing knowledge?
Coherence
Specificity
Analogy
Experimental evidence
133
What do ratio measures indicate in epidemiological studies?
The difference in risk between two populations
The total number of exposed individuals
How strongly a factor is associated with an outcome
The absolute risk of disease development
134
What is the null hypothesis (H₀)?
Strong association exists
Must always be accepted
No effect or difference
Always biological
135
Which equation correctly converts odds to probability?
Probability = 1 – odds
Probability = (1 – odds) / odds
Probability = odds / (odds + 1)
Probability = odds * (1 – odds)
136
In a matched case-control study, what is the unit of statistical analysis?
Individual subjects
Exposure groups
Population prevalence
Case-control pairs
137
Which measure compares disease cases present in a population at one time?
Prevalence ratio
Odds ratio
Rate ratio
Risk ratio
138
What does "strength of association" refer to?
How strongly exposure links to outcome
Number of studies on the relationship
Statistical confidence level
Proportion of the population exposed
139
Why is cumulative incidence useful in risk assessment?
It replaces the need for statistical analysis
It measures disease severity rather than occurrence
It eliminates the need for control groups
It provides a quantitative estimate of risk over time
140
Which philosopher proposed causal reasoning criteria in 1862?
Karl Popper
John Stuart Mill
Francis Bacon
René Descartes
141
What is TRUE about inductive reasoning?
Always guarantees conclusions are true
Only valid when based on mathematical proofs
Often based on repeated personal observations
Uses general principles to predict specific cases
142
Why are epidemiological study designs important?
They help determine causal relationships between variables
They replace clinical trials for treatment testing
They ensure all correlations indicate causation
They eliminate the need for data collection
143
Which of the following best describes an epidemiological association?
A mathematical relationship between two variables
A proven causal link between two variables
A guaranteed effect of exposure on disease
A factor that directly increases disease severity
144
What describes "conjecture and refutation"?
Hypothesis testing for falsification
Using beliefs for conclusions
Accepting only undisproven hypotheses
Proving theories with repeated experiments
145
What is a major limitation of Koch’s Postulates?
Require human experimentation
Not all pathogens can be cultured
Only apply to bacterial infections
Assume all diseases have a single cause
146
Which is an example of deductive reasoning?
I have seen white swans, so all swans are white
A falling apple means all apples fall the same way
Illness after eating shellfish means all shellfish cause illness
All mammals have lungs, so a dog has lungs
147
What is a key principle of matching in case-control studies?
Do not match on variables whose effects you want to estimate
Matching replaces need for statistical analysis
Matching eliminates all bias in a study
Always match on every possible variable
148
How should individual component causes be considered in the pie model?
Some causes are always stronger
Most common cause is the most important
Rare causes should be ignored
No cause is more important than another
149
What is TRUE about causal inference in epidemiology?
Statistical significance ensures causation
A single study proves causation
Requires both evidence and reasoning
Causal relationships can always be measured
150
Why is time sequence important for causation?
Proves exposure is the only cause
Exposure must occur before the outcome
Exposure & outcome must be simultaneous
Exposure can happen after outcome
151
Which of the following is true about range-based matching?
Range should be random
Range should be both convenient and biologically meaningful
Range should not consider biological relevance
Range should be as wide as possible
152
Why can't ratio measures assess absolute risk?
They only compare disease occurrence between groups
They focus solely on environmental exposures
They do not involve statistical calculations
They only apply to infectious diseases
153
Why is appropriate study design essential in analytical epidemiology?
To avoid the need for statistical analysis
To prevent disease transmission
To confirm all hypotheses are correct
To establish valid associations
154
What is the purpose of epidemiological study designs?
To develop vaccines for infectious diseases
To cure diseases using statistical analysis
To describe disease symptoms only
To measure associations between exposures and outcomes
155
What is TRUE about the causal pie model?
All causes happen simultaneously
A single cause is always sufficient
Final cause occurs just before disease starts
Causes follow strict order without overlap
156
What is a problem with overusing biologic plausibility?
Makes lab experiments unnecessary
New theories can explain any association
It removes need for statistical tests
Guarantees causation
157
How does refutationism relate to statistical hypothesis testing?
Assumes all null hypotheses are true
Proves alternative hypotheses with certainty
Uses observations to reject hypotheses
Focuses on consistency with data
158
How is risk difference interpreted in epidemiology?
Shows odds of disease given exposure
Measures strength of association between exposure and disease
Predicts future disease prevalence
Estimates how much risk in exposed group is due to exposure
159
Which criterion suggests that similar exposures may also be causal?
Analogy
Consistency
Dose-response relationship
Strength of association
160
How is NNT calculated?
1 / Relative risk
1 / Attributable risk
1 / Odds ratio
1 / Prevalence ratio
161
Which criterion refers to increasing disease risk with higher exposure?
Specificity
Biologic plausibility
Strength of association
Dose-response relationship
162
How does coherence differ from biologic plausibility?
Proves causation alone
Ignores lab findings
Considers all scientific knowledge
Only about statistics
163
Which equation correctly converts probability (p) to odds?
Odds = p + (1 – p)
Odds = p / (1 – p)
Odds = (1 – p) / p
Odds = p * (1 – p)
164
What does the etiologic fraction estimate?
Fraction of disease in exposed group due to exposure
Total number of disease cases in a study
Relative risk of exposure
Probability of disease occurring in a population
165
Which is an example of analogy in causation?
TB and bacteria exposure
Crohn’s disease and M. avium
All infections cause disease
Smoking and lung cancer
166
What does risk difference (attributable risk) measure?
Probability of exposure in diseased individuals
Difference in risk between exposed and unexposed groups
Proportion of a population affected by a disease at one time
Strength of association between exposure and disease
167
Which of the following is a common variable used for matching in studies?
Breed
Age
Sex
All of the above
168
What does the causal pie model assume about disease causation?
Disease occurs once all component causes are present
Only infectious diseases follow this model
Random chance plays a major role
All variation is due to unknown causes
169
What does logic primarily investigate?
Prevalence of disease in populations
Statistical significance of findings
Structure of statements and arguments
Validity of scientific experiments
170
What distinguishes a causal relationship from an association?
Causal relationships only apply to chronic diseases
Causal relationship meanschange in one variable directly affects another
Correlations never suggest causation
Associations always prove causation
171
Why is time sequence essential for causation?
Time sequence alone proves causation
Cause must occur before effect
Effect must happen immediately
Cause and effect must happen at the same time
172
What is the null value for risk difference?
Any value above 0
1
0
-1
173
Why are errors in observation problematic in epidemiology?
Always result in false associations
Can destroy validity of causal conclusions
Make statistical tests unnecessary
Eliminate need for hypothesis testing
174
What do epidemiological studies primarily measure?
The genetic basis of diseases
Associations between variables
Direct causation between variables
The effectiveness of medical treatments
175
What do epidemiologists actually calculate in studies?
Strength of an association between variables
Direct causation between exposure and outcome
Absolute proof one factor causes another
Exact biological mechanism of disease
176
Which of the following does NOT have a unit of measurement?
Probability
Odds Ratio
All of the above
Rate Ratio
177
Which measure has a range of [-1,1]?
Risk Ratio
Risk Difference (Attributable Risk)
Probability
Odds Ratio
178
What does analytical epidemiology estimate to determine potential disease causes?
Random statistical variations
Strength of epidemiological associations
Disease prevalence
Environmental risk factors
179
Which statement about causation is TRUE?
A single observational study determines it
All statistical associations are causal
It requires strong associations & logic
If two variables are correlated, one causes the other
180
What does the strength (magnitude) of an association indicate?
Likelihood exposure causes disease
Absolute risk in all populations
Total number of cases in a population
Number of people exposed to the risk factor
181
Who developed widely used criteria for causation in epidemiology?
Louis Pasteur
Francis Bacon
John Stuart Mill
A.B. Hill
182
What may explain exceptions to consistency?
Consistency applies only to lab studies
Random errors in study design
All risk factors change over time
Statistical significance ensures consistency
183
What is TRUE about deductive reasoning?
It always leads to factually correct conclusions
It relies on personal experience
It starts with observations and builds to general rule
If premises are true, conclusion must be true
184
Which statement about risk factors is TRUE?
Some risk factors may actually be protective
Risk factors are only relevant to infectious diseases
All risk factors directly cause disease
Risk factors are based only on genetic predisposition
185
What does "consistency of findings" mean in Hill’s criteria?
Association is usually insignificant
Observed across multiple studies
Found in only one study
Exposure always leads to disease
186
What is the primary purpose of the chi-square test in a 2 × 2 table analysis?
Measure individual-level risk factors
Estimate effect size of an exposure
Determine significant difference between proportions
Calculate confidence intervals for the risk ratio
187
What does a P-value represent?
Effect size
Probability of data if H₀ is true
Probability Hₐ is true
Probability H₀ is false
188
How are expected values in a chi-square test calculated?
Take square root of sample size
Sum the observed frequencies
Multiply marginal totals and divide by total sample size
Calculate mean of all observed values
189
Which formula is used to calculate risk difference?
[a / (a + c)] ÷ [b / (b + d)]
[a × d] / [b × c]
[a / (c + d)] - [b / (a + b)]
[a / (a + b)] - [c / (c + d)]
190
What must be true about a microorganism per Koch’s Postulates?
Must infect humans first
Must be in all disease cases in pure culture
Only exists in labs
Causes multiple unrelated diseases
191
Who published early work on causation in 1862?
John Snow
John Stuart Mill
Robert Koch
A.B. Hill
192
Why is it difficult to determine causation in epidemiology?
Correlations always imply causation
Population-level studies rarely reveal direct biological mechanisms
Causal effects only occur in clinical trials
All data collected in epidemiology is flawed
193
What is the study of reasoning called?
Statistics
Logic
Philosophy
Epidemiology
194
Which field commonly uses NNT in evidence-based medicine?
Clinical medicine
Environmental science
Epidemiology
Genetics
195
Which type of matching is required for calculating a matched OR?
Group matching
2:1 matching
1:1 matching
Frequency matching
196
Why is analogy a weak causal criterion?
Does not use evidence
Can justify almost anything
Only applies to infections
Never statistically valid
197
Which best describes a sound argument?
Argument based on statistical significance
Argument widely accepted in philosophy
Valid argument with true premises
Argument proven scientifically
198
How can probabilities of disease occurrence be compared?
By comparing case studies of individual patients
Using only absolute risk measures
By measuring disease symptoms only
Using relative and absolute measures
199
What does a confidence interval (CI) represent?
Range of values reasonably consistent with data
Exact true value of association in population
Probability an individual will develop disease
Measure of statistical significance of a test
200
If two factors are independent, what is true about their probabilities?
P(A, B)
P(A, B) = P(A) * P(B)
P(A, B) is always zero
P(A, B) > P(A) * P(B)
