Causation & Third Variables

Question 1

Third Variables - extraneous variable

Answer 1

• Any variable outside of the experimental set up that EITHER influences the Independent or Dependent variable
• So any variable that could influence the experiment but is not included can be considered extraneous (Lurking, Confounder, Mediator, Moderator)

Question 2

Confounder - third variables

Answer 2

• Any variable that influences BOTH the Independent and Dependent variable of your experiment
• Reason why correlational research cannot establish causation (there could always be a unknown confounder)

Question 3

What is a moderator?

Answer 3

• A third-variable that influences the relationship between two variables
• Can be the source of interactions

Question 4

What is a mediator?

Answer 4

• Mediator: a third variable that explains the indirect relationship of the independent and dependent variable.
• (“man in the middle”)

Question 5

What is common response?

Answer 5

A correlation of two variables can be because of causal relationship or a lurking variable (Third- Variable).
The lurking variable can be the actual reason for changes in the explanatory and dependent variable and there is no actual relationship between explanatory and dependent (Common response)
Example: There is an association, but no causal relationship between use of sunscreen and skin cancer. Both factors are influenced by amount of strong sunlight.

Question 6

confounder - associations and correlations

Answer 6

The lurking variable can also be a confounder. Both the explanatory and dependent variable do have a connection, but they are both influenced by a third variable.
Example: People who do sport have a lower risk of heart attack. So there is a negative association between doing sport and having a heart attack. A confounding variable could be age, because more young people tend to do more sport because they have more energy. Being young also reduces the risk of heart attacks, because fatty deposits that clog up the blood vessels surrounding the heart take decades to develop

Causation: sport - heart attack
Confounder sport - age - heart attack

Question 7

Exam questions:
Traumatic Brain injuries in children often lead to lower school performances. However it was observed that this effect is reduced in children out of families with high socio economic status, possible because families can afford better rehabilitation programs. On the flipside, the effect is accentuated in children out of families with a lower socio economic status.

A. Socio economical status could be a Moderator
B. Socio economical status could be a Mediator
C. Socio economical status and school performance show Common Response to traumatic brain injury
D. Socio economical status could be a Confounder

Answer 7

Answer: A

Question 8

Exam question:
There is a positive association between degree of education and money spent per month. What is the most likely reason for this?

A. Socio Economic Status could be a Moderator
B. Job Salary could be a Mediator
C. Closeness to a shopping center could be a Confounder
D. Money spend and Degree of Education show a common response to Job Salary

Answer 8

Answer: B