Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

Winning with Data Science > 4 - Making Decisions with Data > Flashcards

4 - Making Decisions with Data Flashcards

1
Q

What was the conclusion of the researchers’ cost-effectiveness analysis regarding the new drug ClaroMax?

A

The new drug was both more expensive and less effective than the old drug.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What does causality refer to in data analysis?

A

The underlying causal relationships between phenomena: X causes Y.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is a common example of causality in daily life?

A

An apple a day keeps the doctor away.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the main goal of data analysis?

A

Determining causality.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is a confounder?

A

A factor that has a causal relationship with both the outcome and the intervention.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Fill in the blank: Randomization helps to ensure a roughly even distribution across _______.

A

[confounders]

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the purpose of a placebo in a drug trial?

A

To serve as a fake drug that has no therapeutic effect.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the significance of randomization in clinical trials?

A

It accounts for both measurable and unmeasurable confounders.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

In the context of the ClaroMax study, what are important attributes to account for?

A
  • Baseline age
  • Prior medical conditions
  • Stage of cancer
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

True or False: In a clinical trial, it is essential for the intervention groups to be 100 percent identical across all variables.

A

False.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What could be a potential issue when trying to recruit comparable groups for a clinical trial?

A

It may be difficult to measure all important confounders.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is an example of a variable that is not a confounder in the ClaroMax study?

A

The length of their toenails.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How can researchers determine which food caused stomach issues in a hypothetical scenario?

A

By cloning themselves to isolate the impacts of different foods.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What does the process of cloning illustrate in the study of causality?

A

The importance of controlling for all possible variables.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is the expected outcome if randomization is properly applied in a study?

A

The distribution of characteristics should replicate the overall population.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the challenge with ensuring even distribution of confounders in clinical trials?

A

Some confounders may be unmeasurable.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What is the role of a control group in a clinical trial?

A

To provide a baseline for comparison against the treatment group.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Fill in the blank: The key insight in establishing causality is that groups need to be similar enough across important _______.

A

[confounders]

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

How does randomization help in a study with unknown confounders?

A

It helps ensure that both measurable and unmeasurable confounders are evenly distributed.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

What might happen if a study finds a difference in survival between two groups that are not comparable?

A

It may lead to incorrect conclusions about causality.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What is the relationship between randomization and sample size in ensuring equal distribution?

A

With a large enough sample size, randomization leads to expected equal distribution.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What is a confounder?

A

A variable that influences both the independent and dependent variables, potentially leading to a false association.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

How does randomization help in causal inference?

A

It preserves the distribution of attributes among intervention groups, allowing for causal conclusions.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

What is A/B testing?

A

A framework for designing and executing mini randomized experiments to answer causal questions.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
What key question format is typically used in causal studies?
'What is the effect of [intervention] on [outcome]?'
26
What is a common pitfall in defining a causal question?
The use of vague terms in the outcome definition.
27
What should outcomes in a causal study be?
Clearly defined and meaningful.
28
What are inclusion and exclusion criteria?
Criteria that describe who will be included or excluded from the study population.
29
What is generalizability in the context of research?
The extent to which study conclusions can be applied to populations beyond the studied group.
30
What is an important consideration when recruiting participants?
How the recruitment methodology affects the generalizability of the study results.
31
How can randomization be operationalized in a study?
By randomly allocating interventions to groups.
32
What is one method for collecting data in a clinical trial?
Nurses recording the date and time of drug administration.
33
What is the effect-size estimate?
A quantitative measure of the magnitude of the experimental effect.
34
What does a p-value indicate?
The statistical significance of the results.
35
What happens when randomization is not possible?
Causal conclusions can be drawn from observational data using statistical techniques.
36
What are natural experiments?
Situations where interventions are allocated to groups somewhat randomly due to chance.
37
What is an example of a natural experiment in economics?
The effect of minimum wage increases on unemployment using neighboring states as comparison groups.
38
What is the Difference-in-Differences estimation?
A statistical technique used to determine the causal effects of a treatment by comparing the changes in outcomes over time between a treatment group and a control group.
39
What is the assumption made when analyzing policy changes as natural experiments?
The timing of the policy enactment is essentially random.
40
Fill in the blank: A/B testing is used to find the most effective _______.
[marketing and advertising approaches]
41
True or False: Perfect concordance of confounders is necessary for causal inference.
False
42
What must be ensured during data collection in clinical trials?
That data on the intervention and outcome is captured accurately and timely.
43
What is the main purpose of defining intervention and outcome variables clearly?
To allow for reproducibility in the study.
44
What is a natural experiment?
A situation where a causal conclusion can be drawn due to an external change affecting different groups differently ## Footnote Examples include policy changes or minimum wage increases.
45
What is the interrupted time series methodology?
A method that analyzes data before and after a policy change to evaluate its impact ## Footnote It treats the period before the policy as a control group and the period after as an intervention group.
46
Define regression discontinuity design.
A design that analyzes the impact of a treatment based on a cutoff point along a continuum ## Footnote For instance, eligibility for Medicare at age 65.
47
What is the significance of the age cutoff of 65 in the context of health insurance?
It serves as a discontinuity for comparing uninsured patients under 65 and Medicare patients over 65 ## Footnote This comparison helps attribute differences in hospitalization rates to health insurance.
48
What is matching in observational data?
A statistical technique that emulates a randomized experiment by creating comparable groups ## Footnote It matches individuals on attributes except the one of interest.
49
What is selection bias?
A distortion that occurs when the selected sample differs systematically from the population of interest ## Footnote This can affect the generalizability of study results.
50
What is measurement error?
A discrepancy between what is measured and the actual value ## Footnote It can be systematic (consistent error) or random (variability in error).
51
Define nonresponse bias.
Bias that occurs when certain groups do not respond to surveys, skewing the results ## Footnote Typically, only extreme opinions are reflected in the responses.
52
What is interrater reliability?
The degree to which different raters give consistent estimates of the same phenomenon ## Footnote High interrater reliability indicates good agreement among raters.
53
What is intrarater reliability?
The consistency of a single rater's assessments over time ## Footnote Variability can indicate fatigue or inconsistency in judgment.
54
What is reporting bias?
A bias that occurs when the reported data does not accurately reflect the true situation ## Footnote Examples include media coverage that overrepresents certain events.
55
Define publication bias.
A bias where only certain types of findings are published, often favoring positive results ## Footnote This can lead to a misleading understanding of the evidence base.
56
What is p-hacking?
The practice of manipulating data analysis to find statistically significant results ## Footnote It often involves testing multiple hypotheses until a significant result is found.
57
What is the difference between correlation and causation?
Correlation indicates a relationship between two variables, while causation implies that one variable directly affects the other ## Footnote Misinterpreting correlation as causation can lead to incorrect conclusions.
58
What is a key pitfall in data analysis?
Biases and errors can skew interpretations and mislead conclusions ## Footnote Awareness of these biases is crucial for accurate data science.
59
What is the main mistake made by the committee regarding client retention in Michigan?
Confusing correlation with causation.
60
What explains the high retention of clients in Michigan?
Most clients were from a single employer with high employee-retention rates.
61
What is the Conditional Probability Fallacy?
Confusion of the inverse.
62
True or False: Just because everyone with Ebola has a high fever means that everyone with a high fever has Ebola.
False.
63
What societal issue arose from confusion of the inverse after 9/11?
Discrimination against Muslim Americans.
64
What does a positive COVID-19 test result imply about the individual?
Not everyone who tests positive actually has COVID-19.
65
What example illustrates the difference between absolute and percentage change?
Enrollment increased from 10 to 13 people per 10,000.
66
What is needed to contextualize changes in advertising budget?
The total advertising budget.
67
What led to an increase in breast cancer diagnoses without an increase in death rates?
New screening measures for mammograms.
68
What is the myth of the single study?
Placing too much weight on the results of a single study.
69
What type of evidence is considered the highest quality?
Multiple randomized experiments.
70
What analysis technique combines results from multiple studies?
Meta-analysis.
71
What differentiates a systematic review from a meta-analysis?
Systematic reviews typically do not include quantitative synthesis of results.
72
What should consumers of studies focus on instead of individual studies?
Overall synthesis of evidence across all available studies.
73
What key question should be asked to determine if a study is high-quality?
How different is this body of evidence from several rigorously conducted, bias-free randomized experiments?
74
What should be considered when evaluating the applicability of evidence?
The specific context or problem.
75
What is a key factor in determining the sample size needed for a study?
Statistical power.
76
How can confounding factors be addressed in study design?
By accounting for them.
77
What type of bias might affect the study's conclusions?
Selection bias.
78
What is the importance of data infrastructure in experiments?
To capture the results accurately.
79
What should be done if randomization is not possible?
Approximate randomization using observational data.
80
What does the balance of evidence refer to?
Looking at all relevant evidence, not just a selection.
81
What is essential for the reproducibility of results?
Transparent and well-documented methods.
82
How can conflicts of interest affect evidence interpretation?
They can bias the conclusions drawn from the evidence.

Winning with Data Science (11 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions