Final

Question 1

Define a cohort study: A group of people _______________ is ____________ for a certain period of time to _____________________.

Answer 1

A group of people free of the disease of interest is identified and followed for a certain period of time to ascertain the occurrence of events.

The goal is to compare disease incidence between exposed and unexposed.

Question 2

What is the difference between RCT’s and cohort studies?

Answer 2

In an RCT, exposure is assigned by the investigator. In a cohort study, groups already exist in the source population and are selected into the study.

Question 3

What are three strengths of cohort studies?

Answer 3

Ability to study exposures that are difficult to randomize. Allows calculation of incidence. Multiple outcomes/exposures can be assessed.

Question 4

What is the main advantage of prospective studies, and the main disadvantage of retrospective?

Answer 4

We can see temporal ordering of exposure and disease.

Can only use available information which may be of lower quality or for different purposes.

Question 5

Studies will often include an __________________ that must occur before beginning to count person-time at risk of exposure effects (varies by subject matter).

Answer 5

induction/latency time

Question 6

Cohort studies most often deal with acute exposures (T/F).

Answer 6

False, chronic.

Question 7

____________ refers to how we will conceptualize exposure
based on our study hypotheses and what we currently understand
about the exposure/disease of interest

Answer 7

Dose representation

Question 8

What do we consider when defining exposure/dose?

Answer 8

Dose represention and causation; do we look for mean exposures, peak, cumulative?

What are the most important time periods of exposure?

Question 9

Cumulative dose is the sum of _____.

Answer 9

duration x frequency x intensity

Question 10

If the biological effects of a particular exposures are not well understood, it is a good idea to evaluate only one representations of exposure (T/F).

Answer 10

False, several!

Question 11

Selection bias is caused by conditioning on a __________________.

Answer 11

common effect of exposure and disease

Question 12

Exposure and Disease status should be independent of both _________ and _________ in order to avoid selection bias?

Answer 12

selection; retention

Question 13

As opposed to conditioning on a common effect, confounding is more about common _____.

Answer 13

causes

Question 14

What are the basic requirements for survival analysis definition?

Answer 14

Clear time scale and origin of time at risk; Precise definition of ‘event’ and how it will be measured; Accounting for censoring in the an analysis

Question 15

Why wouldn’t we do linear regression instead of survival analysis?

Answer 15

Time to event is usually quite skewed, and this would also ignore censoring.

Question 16

What are the two types of survival analysis models?

Answer 16

Parametric; assumes that the times follow a distribution.

Semi-parametric; assumes proportional hazards

Question 17

Right censoring causes an underestimation of the true time to event (T/F).

Answer 17

True

Question 18

S(t) describes the probability that ________.

Answer 18

an individual survives from the start of follow-up to a specified future time (t).

Question 19

The hazard function is the ____________ at time (t).

Answer 19

instantaneous event rate (incidence)

Question 20

The Cox Proportional Hazards Model estimates the effect of covariates on the hazard ratio. What is the hazard ratio?

Answer 20

The ratio of hazards betweenexposed/non-exposed, leaving the baseline hazard rate unspecified. Like an incidence rate raio.

h(t) = h0(t) * exp{b1x1 +b2x2+…+bpxp)

Question 21

The baseline hazard is the value of h(t) is all covariates __________.

Answer 21

equal zero

Question 22

What is the key (multiplicative) assumption of the Cox model?

Answer 22

The hazard in either group is a constant multiple of the hazard in any other; the baseline hazards are the same. Thus, the e^betas gives the hazard ratio for a change from x1 to x2. Curves don’t cross!

Question 23

Does hazard ratio depend on time?

Answer 23

No girl not at all

Question 24

How can violations of the proportional hazards assumption be addressed?

Answer 24

Including interaction terms between covariates and time, or by using strata variables. If the baseline isn’t the same for everyone, we can have different ones for different strata (stratified proportional hazards model).

+strata(variable)

Question 25

What is the disadvantage of stratifying baseline hazards?

Answer 25

That covariate will disappear as a predictor, so we won't see the effects of the stratification we'll just have controlled for the variable.

Question 26

In which ways are case-control studies more efficient than cohort studies?

Answer 26

Time (eliminate the need for follow-up) Personnel + lab analysis

Question 27

Ideally what proportion of cases in the source population will be included in the study?

Answer 27

All (minimizes standard error)

Question 28

Sampling must be independent of ______.

Answer 28

exposure

Question 29

Why should we select incident cases instead of prevalent cases?

Answer 29

Prevalence is related to disease duration / survival time. Sampling must be independent of exposure.

Question 30

What will an odds ratio approximate in each case?

Answer 30

Risk Ratio when controls are selected at the beginning of follow-up. Rate ratio when they're selected from members of the population at risk (density / risk-set sampling). Odds ratio when controls are selected from those who are non-cases at the end of follow-up, or risk ratio if the rare disease assumption is kept.

Question 31

What are the two basic rules of control selection in a case-control study?

Answer 31

They should be selected from the same population as the cases and independently if exposure.

Question 32

What are some alternatives to source sampling for controls?

Answer 32

Neighbourhood, friend, random digit dialing, hospital-based. Neighbourhood and friend may be related to exposure and not in the same source pop. All homes aren't likely to be recruited in random dialing. For hospital, make sure that controls are selected independently of exposure.

Question 33

What are the two types of matching in case-control studies and what is the effect?

Answer 33

Individual and frequency (distribution). We match on confounders but matching doesn't control for confounding, and those variables will have to be included as covariates to adjust for the bias. Matched analysis will have a slightly wider confidence interval.

Question 34

Which function do we use to model logistic regression?

Answer 34

logit, to ensure the range of pi is between 0 and 1.

Question 35

How do we interpret B1 in a logistic regression?

Answer 35

e^B1 is the odds ratio for a unit change in X. (e^B1)^2 is the OR for a two unit change

Question 36

What are the two opportunities to introduce selection bias into a case control study?

Answer 36

Control selection and case ascertainment

Question 37

Why is effect modification so important and informative?

Answer 37

Unlike biases which should be reduced, effect modification provides insight into the nature of the exposure-outcome relationship. It should be reported and understood. Helpful in identifying the groups of individuals that would benefit the most from an intervention.

Question 38

Effect modification examines whether the strength of an association related to the exposure ______________.

Answer 38

varies across strata of another variable

Question 39

Interaction refers to a situation where a risk factor and another factor ___________ each other.

Answer 39

strengthen or weaken

Question 40

What are the two main strategies for assessing and detecting Effect Mod?

Answer 40

Assessment of homogeneity: comparing effects of one exposure over the strata of a third. Comparing observed effects to expected joint effects under an additive or multiplicative disease model.

Question 41

Effect Modification could be caused by heterogeneity due to...?

Answer 41

random variability (stratify), confounding (we assume none), and other biases caused by error.

Question 42

Additive interactions are more relevant to public health because they reflect___________.

Answer 42

absolute differences in risk (e.g., numbers of additional cases)

Question 43

What is the difference between prediction and inference?

Answer 43

Prediction uses data to map inputs to outputs, requiring expert knowledge to specify the question, data sources, and variables. Inference estimates how a feature would change under different circumstances, requiring expert knowledge to define causal structure and identify biases.

Question 44

Good predictive models tell you about how to intervene to change the outcome (T/F).

Answer 44

False, they don't necessarily

Question 45

It is best if predictive models capture ___________________, otherwise predictions may have poor generalizability.

Answer 45

true causal relationships

Question 46

Health risk scores are useful in stratifying _______________ patients through prediction of health events.

Answer 46

low/high-risk

Question 47

Why are prediction model coefficients hard to interpret?

Answer 47

These models are typically quite complex so they can flexibly model the relationships of interest. Then, it's difficult to understand how any individual predictor influences the outcome.

Question 48

What is the most common metric used to evaluate prediction accuracy?

Answer 48

Mean squared error (MSE) which we want to be small for the test data. But it doesn't need to be too small (overfitting/noise inclusion).

Question 49

Cross-validation is often used to evaluate predictive models when ______________ are not available.

Answer 49

external test data

Question 50

What is a "hold-out" test sample?

Answer 50

A sample from the training set used only for model evaluation. We use the trained model to predict its values multiple times to estimate average performance.

Question 51

Describe leave-one-out cross validation

Answer 51

Testing on n-1 samples. Each data point is used as a test set once, repeated for every data point. MSE values are averaged and squared errors are calculated for each.

Question 52

Describe k-fold cross validation.

Answer 52

Fold into subsets by "k" (k=5/10) and use each, one at a time, as a test set. MSE values are averaged and squared errors are calculated for each.

Question 53

K-fold gives more accurate estimates than LOOCV because there is _______________ in the estimates of LOOCV.

Answer 53

more variance

Question 54

If the goal is prediction, we must think about confounding, bias, and model interpretation heavily (T/F).

Answer 54

False! This is true for inference. For prediction, interpretation is less important but the predictors should still capture the causal model in question.

Question 55

How do we approximate the target tissue dose?

Answer 55

We use proxies like urine and blood measures

Question 56

What are the steps in planning a study?

Answer 56

Define exposure and outcome, identify and recruit population, etc.

Question 57

What is an instrumental variable?

Answer 57

A factor which causes the exposure, not directly associated with the outcome or any confounders. A perfect proxy for exposure. We use it to extract variation in the exposure not influenced by confounders, and use this variation to estimate causal effect.

Question 58

Any paths between the instrumental variable and the outcome must pass through ______________.

Answer 58

exposure or be closed

Question 59

When does the IV method lead to incorrect inferences?

Answer 59

When the relationship between the instrumental variable and the outcome is affected by unmeasured confounding

Question 60

Propensity scores adjust for confounding by mimicking _____________ and making groups __________________.

Answer 60

randomization; as similar as possible with respect to known confounders. Identifies groups of people equally likely to receive treatment to minimize systematic differences

Question 61

What, technically, is a propensity score?

Answer 61

A balancing score. The predicted probability of exposure for every individual, using exposure as the dependent variable in a logistic regression model with potential confounders, interactions, and variables associated only with the outcome of the main relationship, regardless of significance.

Question 62

How are propensity scores used to mimic randomization?

Answer 62

Generate a new dataset of matched subjects based on similar propensity scores and then include the scores as a covariate and then to stratify. No guarantee of similarities between unmeasured confounders, though. It can also restrict sample size. 1:1 nearest-neighbour matching. Matched pair analysis.

Question 63

What does propensity score stratification do?

Answer 63

Creates strata based on individual score, then takes an average of the effect estimates across the different strata. Can reveal potential heterogeneity.

Question 64

Is it useful to use propensity score as a covariate?

Answer 64

Yes, if we include with other covariates it's a doubly robust method.

Question 65

PS can be used to adjust for confounding in observational studies, but unlike randomization, cannot account for __________________.

Answer 65

unknown or unmeasured confounding

Question 66

Name 3 advantages and 3 disadvantage of propensity scores.

Answer 66

Can adjust for confounding, adjust for more covariates than normal, and maximize comparability. Prediction is limited, can't handle time-varying exposures, can lower generalizability.

Question 67

Machine learning models train _______ to make predictions.

Answer 67

algorithms

Question 68

What is the main limitation of linear regression for prediction?

Answer 68

The assumed linear relationship between predictors and the outcome.

Question 69

Generalized additive models estimate the weight for each basis function to ___________________________________________.

Answer 69

minimize the residual square error of the model caputures non-linear relationships while still being interpretable.

Question 70

Name two assumptions of GAMs.

Answer 70

independent observations and normal distribution of residuals

Question 71

What are hyper-parameters in machine learning?

Answer 71

Parameters set before training that are tuned and influence the speed and accuracy of learning. Random forest (number of trees, node size, etc). XGBoost (number of sequential trees to build, depth of each tree). Dense neural net (learning rate, number of hidden layers).

Question 72

Name two methods for small samples in machine learning.

Answer 72

Transfer learning updates weights from a previously trained model with a new dataset. Data augmentation can increase the numbers of sample by randomly flipping images.

Question 73

Define training set, tuning set, and test set.

Answer 73

Primary dataset used to train the model. Subset used to tune the hyperparameters to optimize model performance. External dataset not used for model development.

Question 74

The quality of reference standard for model performance should be based on ______________________.

Answer 74

subject matter expertise

Question 75

What is overfitting in the context of machine learning?

Answer 75

Model performs really well on the training set but not so much on the test set; can happen if data in the two sets overlap. Pay attention to your data splits!

Question 76

Machine learning _____________ should be reproducible.

Answer 76

predictions

Question 77

The ___________ design is for answering causal questions at the group level.

Answer 77

difference-in-differences time and group are both confounders

Question 78

In occupational exposure studies, 8-hour ____________ are often used for dose representation.

Answer 78

Time-Weighted Average Exposures

Question 79

Describe what happens in a difference in differences study.

Answer 79

The first difference isolates the change in outcome over time within groups, controlling for group differences. The second difference compares the within group differences between the groups; how much change in the outcome we would expect in the treated group if the treatment had not occurred.

Question 80

What do we assume in a DiD study?

Answer 80

Exchangeability and the Parallel Trends Assumption. If no treatment had occurred, the difference between the treated group and the untreated group would have stayed the same post-treatment period as it was in the pre-treatment period. So the groups should have similar trajectories before treatment.

Question 81

How do we evaluate the parallel trends assumption?

Answer 81

Compare prior trends between groups, placebo test to estimate DiD at earlier time periods (there should be no difference).

Question 82

What is the structure of a DiD interpretation?

Answer 82

In [Year], compared with [Year], __________ experienced 7.7 additional ______________ for AMI per 1000 Medicare beneficiaries than would have been expected in the absence of ____________.

Question 83

What is a meta-analysis?

Answer 83

Quantitative approach for systematically combining results of previous research to arrive at conclusions about the body of research.

Question 84

Effect size is the measure of the ____________ and ________ of a relationship between variables

Answer 84

magnitude; direction

Question 85

What are the three types of heterogeneity?

Answer 85

Clinical, methodological, and statistical.

Question 86

The Cochrane Q statistic looks for heterogeneity by calculating ______.

Answer 86

the weighted sum of squared differences between individual study effects and the pooled effect across studies

Question 87

The I^2 statistic is the percentage of variation across studies that is due to _____________ rather than chance.

Answer 87

heterogeneity

Question 88

What is the difference between a fixed and random effects model in meta-analysis?

Answer 88

A fixed effects model assumes low heterogeneity (same effect for every study, variations due to random error) whereas a random effects model assumes the true estimate for each study varies and follow a normal distribution, with a higher CI. The random model allows for inter-study variability, although with wider confidence intervals, more equal weighting, and less stability.

Question 89

An ____________ funnel plot indicates clinical and methodological heterogeneity.

Answer 89

asymmetrical

Question 90

Name three ways to deal with publication bias?

Answer 90

Combine the results of the larger studies, file drawer method to estimate how many null studies it would take to make the results insignificant, trim and fill to duplicate and center the data and make new estimations.

Question 91

How do we find Simpson's paradox in meta-analysis.

Answer 91

Effect size is based on the comparison of a group with its own control group – avoids Simpson’s paradox. Could appear if we pool.