Deck 12: Causality

Question 1

What does the term ''All else equal" refer to, in the context of a causal relationship?

A) cases with similar characteristics on all variables but one
B) observational data for the same context at different times
C) cases with the same outcome value for a variable
D) observations with zero values for all independent variables

Answer 1

A

Question 2

Which factor must be assessed in order to declare an association between two variables as causal?

A) confounding bias
B) reverse causalty
C) internal validity
D) correlation

Answer 2

C

Question 3

If an association between two variables is determined to be causal, which two criteria could be used to strengthen the validity of that determination?

A) empirical association and mechanism
B) mechanism and doseresponse
C) nonspuriousness and time order
D) time order and doseresponse

Answer 3

B

Question 4

In a study, it was found that a 1-unit change in a variable X caused a 3.2 unit change in variable Y. What is this phenomenon termed as?

A) unary association
B) ternary association
C) noncausal association
D) empirical association

Question 5

What is the name given to the technique of combining results of several studies to arrive at a summary of findings?

A) confounding bias
B) meta-analysis
C) ceteris paribus
D) macro assessment

Question 6

To confirm causality, a researcher analyzes the data to verify that any change in an independent variable occurs prior to a corresponding change in the outcome variable. What is this type of verification called?

A) empirical analysis
B) longitudinal study
C) time order
D) reverse causation

Question 7

A research study found that what was hypothesized as the dependent variable, had value changes prior to corresponding changes in an associated variable, originally taken as an independent variable. What is the term given to such a finding?

A) empirical order
B) reverse causation
C) causal order
D) directional factor

Question 8

A cosmetics company observed that retail sales of a particular product fell in three consecutive years after a 5% price increase in 2016. The company did not study any other factors that could have affected sales, and, decided to reduce prices by 3%, going forward. What is this type of conclusion or deduction termed in research studies?

A) sample bias
B) post-hoc fallacy
C) inflated causality
D) temporal error

Question 9

Which type of study is appropriate for determining the time order between two variables?

A) cross-sectional
B) multivariate
C) repeated cross-sectional
D) longitudinal

Question 10

A study observed a positive association between number of insurance claims and income of the insured. On further analysis, it was found that the age of the insured had a strong positive association with income and helped explain the variation in the number of claims. What term is used to define the association between claims and income?

A) spurious
B) nonspurious
C) temporal order
D) causal order

Question 11

In the example given in question 10, what term is used to refer to the age variable?

A) response bias
B) spurious unit
C) confounding factor
D) interaction level

Question 12

What is a "mechanism", in the context of causal relationships?

A) an environmental condition that may affect association
B) a process that explains a cause-and-effect relationship
C) research steps taken to understand an association
D) empirical method for verifying cause-and-effect relationship

Question 13

What is a dose-response? Choose the best answer

A) a positive response to a specific dose in medical studies
B) a presence or absence of an outcome after a stimulus
C) a change in a predictor variable for the same outcome
D) a variation in size of the response with size of a predictor

Question 14

What is an alternative term often used for experimental studies?

A) randomized controlled trials
B) medical trials
C) treatment trials
D) panel controlled trials

Question 15

Why is there a control group in clinical trials?

A) in order to control for interaction effects between predictors
B) to assess how response varies between treatment and no treatment
C) to vary the size or magnitude of the independent variable
D) to distinguish between the type of treatment given to participants

Question 16

In an experimental study, the 36 participants of the study were arbitrarily placed into two equal groups of 18 participants each. What is this type of placement known as?

A) random sampling
B) cluster sampling
C) random assignment
D) stratification

Question 17

What is pretest measurement"?

A) measurement of the outcome prior to an intervention
B) value of the response at the time of the experiment
C) valuation of all variables before an experiment
D) verifying accuracy of the outcome value prior to analysis

Question 18

The participants of the study mentioned in question 16 are not informed whether they are being placed in the treatment or in the control group. What term is used to identify this type of study?

A) double-blinded
B) nonblinded
C) placebo blinded
D) single-blinded

Question 19

What does 'placebo effect' mean in the context of experimental studies?

A) no variation in the outcome with treatment
B) change in outcome without any treatment
C) no change in response after treatment
D) no change in response with a placebo

Question 20

What are the primary components of a research experiment? Choose the best answer.

A) Random assignment to treatment and control groups; posttest measurement
B) Medical test in a laboratory; repeated trials of the test
C) Test a new theory or experiment; repeated tests on the theory
D) Cross-sectional study; treatment and control conditions

Question 21

What is the primary difference between an observational and an experimental study?

A) the former does not do random assignment
B) the latter does not define a hypothesis
C) the former does not have groups in the study
D) the latter is done for clinical trials

Question 22

Why does an observational study have a greater risk of selection bias than an experimental study?

A) it does not do random sampling in every case
B) posttest measurement is not done by the study
C) There is no random assignment of an independent variable.
D) it may have missing or incomplete data

Question 23

How could a longitudinal observational study ensure 'time order'?

A) value the outcome before the predictor variable
B) measure the predictor and outcome at the same time
C) sort the data by ascending time for the outcome
D) measure the predictor before the outcome

Question 24

What type of a study is a 'Natural Experiment'?

A) cross-sectional
B) observational
C) randomized
D) experimental

Question 25

What does the term "nonequivalent groups" refer to?

A) treatment and control groups of unequal size
B) one group being weighted more than the other
C) difference in groups due to factors other than the predictor
D) comparison group containing more predictor variables

Question 26

Random assignment in experimental studies help reduce the effect of any confounders in the study.

Question 27

Designing a quasi-experiment in observational studies works as well as a randomized experimental study.

Question 28

Random sampling is often used in designing a study to ensure internal validity.

Question 29

A sample survey of voting trends in Florida's mid-term election results would have high external validity but may have low internal validity.

Question 30

A study showing a strong correlation between a dependent and independent variable, but with the presence of some bias, has good internal validity overall.

Question 31

A causal relationship is one where a change in the response variable causes a corresponding change in a predictor variable, keeping all other variables constant.

Question 32

An association which is subsequently shown to have reverse causation, implies changing the dependent to be an independent variable.

Question 33

A study on customer retention in the hospitality industry, found retention to be significantly associated with geographic location of the customer. This association was shown to be empirical by the study. Hence, the relationship between location and retention was causal in nature.

Question 34

Consistency of association refers to an association between two variables that has been observed in several different samples, and, in different study designs.

Question 35

The criteria of time order is satisfied when the independent variable occurs before the response variable.

Question 36

With the help of an example, explain "reverse causation" between two variables.

Question 37

A study found that drivers who resided in Trenton, New Jersey, were more prone to car accidents than other drivers in the state. The study did not look at several other characteristics of the driver, such as destination, location of the accident, and, age of driver, among others. What is this type of conclusion made by the study called, and why? Explain your answer.

Question 38

What is a confounding factor in the context of spurious relationships? Explain with an example.

Question 39

Once causality is established by assessing the three primary criteria for presence of the phenomenon, which two additional criteria help in strengthening the case for causality? Explain the process for applying these two criteria.

Question 40

Provide an example of an experimental study. Identify the treatment and control groups of the study.

Question 41

Distinguish between random assignment and random sampling. Illustrate with an example.

Question 42

What is "postmeasurement" in the context of an experimental study? Explain with an example.

Question 43

How is an observational study different from an experimental study? Illustrate with an example.

Question 44

Discuss how random assignment removes or greatly reduces confounding bias.

Question 45

What is a quasi-experiment? Explain this type of research design with an example.

Question 46

Design an observational study on a topic of your choice. Make sure that there are at least 2 independent variables in the study. Alternatively, you may use a study you designed for the essay questions in prior chapters, if appropriate. Discuss what steps you would take to assess the internal validity of a causal relationship between the outcome and one of the independent variables. Explain the reasoning for your assessment process.

Question 47

Design an experimental study on a topic of your choice, with at least two independent variables. Explain the process you would use to assess the internal validity of a causal relationship between the outcome and one of the independent variables. How does this process compare, in terms of complexity, with the process you used in question 1?

Question 48

Discuss the importance of ensuring internal validity in research studies. What are the implications if internal validity of a causal relationship cannot be confirmed in a research study?