Midterm

Question 1

What is the goal of the counterfactual model?

Answer 1

To estimate a true causal contrast?

Question 2

Define confounding, selection bias, and information bias.

Answer 2

Confounding is a non-causal association
that is observed between a given exposure and an
outcome owing to a third variable. Lowers exchageability.

Selection bias is

Information bias is

Question 3

DAGs can be used to communicate hypothesized relationships and ________.

Answer 3

identify and understand potential sources of measured/unmeasured selection bias or confounding.

Question 4

How is the line of best fit obtained?

Answer 4

By minimizing the sum of squared errors (i.e. the vertical distance between each point
and line)

Question 5

The _______ is the
proportion of the total
variability explained by the model that includes the covariate x.

Answer 5

coefficient of determination (R^2)

1-SSy-SSxy

Question 6

Name the four assumptions of linearity.

Answer 6

Constant variance, normality of errors, X-Y relationship is linear, and errors are all independent.

Question 7

What do we call the difference between the measured population mean of X (μx) and the true population mean of T (μT)? How is this different from precision?

Answer 7

Bias. Precision is about low variance of the measurement error itself.

Question 8

When is measurement error non-differential?

Answer 8

When the sensitivity and
specificity of exposure assessment is equal for both groups

Question 9

Which type of measurement error moves estimates closer to the null?

Answer 9

Non differential.

Differential moves estimates toward or away or change directions.

Question 10

If exposure measurement error is non-differential, the bias
in the effect estimate is a function of ________.

Answer 10

precision (random error)

Question 11

How does measurement error on exposure differ from outcome?

Answer 11

Outcome measurement, because it’s the dependent variable, won’t change the slope estimate but will increase its standard error and widen the confidence interval. Exposure will bias toward the null.

Question 12

The sample size needed under non-differential measurement error is proportional to ______________, which is the assumed common exposure variance among cases and controls.

Answer 12

standard deviation

Question 13

When everyone is assigned the same exposure, true exposures vary normally around ____.

Answer 13

Group values

Question 14

Berkson-type random exposure measurement error is not expected to bias effect estimates (i.e. slopes in regression models) but there is still a loss of ______ (i.e. wider confidence intervals) and __________.

Answer 14

precision; reduced power

Question 15

Describe the Berkson model of bias.

Answer 15

Random error is attached to the true exposure value, independent of the observed; lowering precision.

Question 16

Increasing sample size can minimize the impact of measurement
error in continuous outcome variables (T/F).

Answer 16

True

Question 17

How does error in outcome measurement change categorical/continuous variables?

Answer 17

Categorical outcome measurement error will produce a bias toward the null. Continuous outcome error has no effect, if you recall, on slope and just reduces precision.

Question 18

Even when the
expected direction of bias is toward the null (because of non-differential exposure misclassification) bias away from the null can occur because ______.

Answer 18

Any study is just one realization, one sample which could be affected by random error in large/small ways each time and only evens out with multiple trials. So what we expect doesn’t always happen.

Question 19

Misclassification and mixing levels leads to bias toward the null (T/F)

Answer 19

False, it can happen away!

Question 20

Confounders are typically only CAUSALLY associated with the exposure (T/F).

Answer 20

False, the outcome!

Question 21

Subject matter expertise is the best way to identify potential confounders (T/F).

Answer 21

True

Question 22

How do we adjust for confounding through stratification?

Answer 22

Separate confounder into strata and calculate stratum-specific associations, then pool if homogenous.

Question 23

Name two limits of stratification as a method for controlling confounding.

Answer 23

Leaves room for residual confounding in continuous variables and adjusting for multiple variable requires specific estimates for every different combination.

Question 24

The adjusted regression coefficient is the expected change in the mean value of ___ per unit change in X keeping _______.

Answer 24

Y; all other variables constant

Question 25

Residual confounding occurs when categories are _______, or when confounders are measured with _____ or ______.

Answer 25

too broad; error; unmeasured / left out

Question 26

The distribution of a confounder has to be different across groups (e.g. case/controls) to cause confounding (T/F).

Answer 26

False, especially when it's associated with the outcome.

Question 27

The overall impact of including multiple parameters in the model depends on their _____ with X.

Answer 27

correlation Variance of B1 is based on variance in X, error in Y, sample size, and the correlation between parameters.

Question 28

What are 3 methods for modelling nonlinear relationships?

Answer 28

Dummy variables, quadratic terms, and splines.

Question 29

Selection bias occurs when exposure and ______ both affect inclusion in the analysis.

Answer 29

outcome

Question 30

What does conditioning on a collider do?

Answer 30

Induces a spurious relationship between the variables leading into it.

Question 31

Confounding is the presence of common effects while selection bias is conditioning on common causes (T/F).

Answer 31

False, the opposite!

Question 32

Effect modification refers to the situation where the strength of association between exposure and outcome differs across ____________.

Answer 32

levels of a third variable

Question 33

Explain the difference between additive and multiplicative interaction.

Answer 33

Additive interaction is when the absolute risk changes across levels of a third variable, multiplicative interaction is present when the RR varies. You can have additive without multiplicative if the ratios are all the same.

Question 34

The interaction term is the excess change in the outcome not explained by the _____.

Answer 34

Sum of the individual effects of two independent predictor variables.

Question 35

What are two things to be aware of when interpreting interaction terms?

Answer 35

Heterogeneity due to small sample size and confounding/error across strata of the effect modifier.

Question 36

Case-crossover studies are used to examine the acute health effects of ____________.

Answer 36

intermittent exposures with short induction times

Question 37

What kind of confounding is not present in case-crossover studies?

Answer 37

Confounding due to variables which do not change within single individuals across the reference/case periods. Transient co-exposures may cause confounding, though.

Question 38

What are the three ways you can have selection bias in a case crossover study?

Answer 38

nonrepresentative case selection, differential case survival (exposure influences survival), control time not independent of exposure.

Question 39

Exposures during case and reference periods must be _______, to avoid ______________.

Answer 39

independent of each other; carryover effects

Question 40

We use a time-stratified design if experiencing the event doesn’t impact the likelihood of ___________, so control periods can be selected before and after the event.

Answer 40

subsequent exposures.

Question 41

What analysis do we perform for case-crossover studies?

Answer 41

Logistic regression to get Odds Ratios

Question 42

Case-crossovers look at the ____ in exposure between two periods, rather than the absolute level.

Answer 42

change

Question 43

What's the major con of case-crossover studies?

Answer 43

They only use information from a single point in time, so they capture prevalence rather than incidence. Also cannot determine relationships.

Question 44

Ecological studies are good for studying variability __________ and for prevention at a _____________.

Answer 44

between homogenous populations; population level

Question 45

Name the three types of ecological measures.

Answer 45

Aggregate measures (characteristics within the group) Environmental measures (physical characteristics of a location) Global measures (group characteristics not reducible to characteristics at the individual level)

Question 46

What do you call it when within-group correlations are different from between-group and you try to conflate the two?

Answer 46

Cross-level inference (ecological bias)

Question 47

What three steps have to be at the group level in order to make an ecological study?

Answer 47

measurement, analysis, and inference

Question 48

What are the pros and cons of ecological studies?

Answer 48

Less expensive (good starting point + existing measurements) and increased generalizability potential. Hard to control confounding, identify temporal relationships, and threat of ecological fallacy.

Question 49

The Poisson distribution is used for counts—if events happen at a constant rate over time, the Poisson distribution gives the probability of ___________.

Answer 49

X number of events occurring in time T

Question 50

Count frequencies are often positively skewed with ________________.

Answer 50

most values being low and relatively few high values

Question 51

In Poisson distribution, the mean event rate is supposed to be equal to the ______.

Answer 51

variance

Question 52

In Poisson Regression the dependent variable is the log of ________.

Answer 52

the expected count

Question 53

How does logE[Yi] respond with each parameter in the model?

Answer 53

linear increase!

Question 54

exp(B1) in a Poisson model approximates to?

Answer 54

Rate Ratio for a one unit increase

Question 55

What is the offset parameter?

Answer 55

logTi (the time that unit “i/p” was at risk) The offset parameter is used to account for differences in exposure time or population size when modeling count data.

Question 56

The effect of over-dispersion is that the precision of effect estimates is not correct, leading to ______.

Answer 56

too-narrow confidence intervals

Question 57

θ is the _____ over-dispersion parameter

Answer 57

Quasi-Poisson (more flexible variable function)

Question 58

Splines fit a number of different ______ (usually cubic) over the range of the data and are joined smoothly at _______ to examine nonlinear forms.

Answer 58

polynomial curves; knots

Question 59

Splines can be used to flexibly adjust for ______.

Answer 59

confounders

Question 60

What are two ways to optimize time-series models?

Answer 60

lag models to correct for delayed effects and sensitivity analyses related to spline knots. Autocorrelation (nonindependent days) is unclear though.

Question 61

_____ analysis is useful when a report makes policy recommendations/actions and draws inferences about causality

Answer 61

Bias

Question 62

Non-participation always leads to selection biased results (T/F).

Answer 62

False, only if participation depends on exposure AND outcome (confounding) We can know this by asking nonparticipants to fill out a short questionnaire

Question 63

Where do we get information on an unmeasured confounder?

Answer 63

A sub-sample of the study population OR educated guesses based on other studies

Question 64

______ can be conducted to collect measurements using both the “gold standard” method and the less accurate/reliable method

Answer 64

Internal validation studies

Question 65

What does a simple bias analysis leave out of the picture, other than the other types of biases?

Answer 65

uncertainty in the bias parameter itself

Question 66

What are the parameters for simple bias analysis for selection bias vs. confounding vs. measurement error?

Answer 66

Contingency table --> multiply OR by selection OR. Specify association between confounder and the outcome for the unexposed, as well as the prevalence in both groups Reshuffle contingency table based on sensitivity and specificity values for each cell.

Question 67

By repeatedly sampling from bias parameter distributions, the output out probabilistic analysis is ________ which can be interpreted as a point estimate and frequentist confidence interval.

Answer 67

a distribution of adjusted estimates of association

Question 68

Multidimensional bias analysis is an extension of simple bias analysis that assigns probability distributions to ________ at the same time

Answer 68

multiple bias analysis