Causal relationships are known to be interactions which involve complex and statistical variables that can be pervasive to provide the required evidence. The strength of a causal relationship will vary with the population, setting and time presented within a study and will solely depend on how the researcher decides to treat the choices made to affect the treatment and measurement of outcomes. For the scenarios below, the researcher will try to come up with an evaluation to provide evidence of a causal relationship. The elements to be evaluated will include association, time order and nonspuriousness.

Scenario 1 

You are interested in whether police officers with a college degree are more effective than those without a college degree. You have collected data on all NYPD police.

Scenario 2

You want to understand whether drug use as a teenager causes criminal activity later in life. You have data on a sample of 1,000 adults. You have two variables for each adult: (1) whether the individual used illicit drugs during their teenage years; and (2) whether the individual engaged in criminal activity as an adult. You find that 3% of adults in your sample who did not use drugs as teenagers have engaged in criminal activity, compared to 45% of adults who did use drugs as teenagers.

Scenario 3

You are interested in whether inmates who complete a vocational training program are more likely than other inmates to get a job when they are released from prison. You have data on a sample of 1,000 inmates in New York prisons. You have two variables for each: (1) whether the inmate completed a vocational training program while in prison; and (2) whether the inmate got a job when they were released. You find that 35% of inmates in both groups - those who did and did not complete vocational training - got a job when released.