You are here

Causality

19 October, 2015 - 17:38

When designing a research project, how issues of causality are attended to will in part be determined by whether the researcher plans to collect qualitative or quantitative data. Causality refers to the idea that one event, behavior, or belief will result in the occurrence of another, subsequent event, behavior, or belief. In other words, it is about cause and effect.

In a qualitative study, it is likely that you will aim to acquire an idiographic understanding of the phenomenon that you are investigating. Using our example of students’ addictions to electronic gadgets, a qualitative researcher might aim to understand the multitude of reasons that two roommates exhibit addictive tendencies when it comes to their various electronic devices. The researcher might spend time in the dorm room with them, watching how they use their devices, follow them to class and watch them there, observe them at the cafeteria, and perhaps even observe them during their free time. At the end of this very intensive, and probably exhausting, set of observations, the researcher should be able to identify some of the specific causes of each student’s addiction. Perhaps one of the two roommates is majoring in media studies, and all her classes require her to have familiarity with and to regularly use a variety of electronic gadgets. Perhaps the other roommate has friends or family who live overseas, and she relies on a variety of electronic devices to communicate with them. Perhaps both students have a special interest in playing and listening to music, and their electronic gadgets help facilitate this hobby. Whatever the case, in a qualitative study that seeks idiographic understanding, a researcher would be looking to understand the plethora of reasons (or causes) that account for the behavior he or she is investigating.

In a quantitative study, on the other hand, a researcher is more likely to aim for a nomothetic understanding of the phenomenon that he or she is investigating. In this case, the researcher may be unable to identify the specific idiosyncrasies of individual people’s particular addictions. However, by analyzing data from a much larger and more representative group of students, the researcher will be able to identify the most likely, and more general, factors that account for students’ addictions to electronic gadgets. The researcher might choose to collect survey data from a wide swath of college students from around the country. He might find that students who report addictive tendencies when it comes to their gadgets also tend to be people who can identify which of Steven Seagal’s movies he directed, are more likely to be men, and tend to engage in rude or disrespectful behaviors more often than non-addicted students. It is possible, then, that these associations can be said to have some causal relationship to electronic gadget addiction. However, items that seem to be related are not necessarily causal. To be considered causally related in a nomothetic study, such as the survey research in this example, there are a few criteria that must be met.

The main criteria for causality have to do with plausibility, temporality, and spuriousness. Plausibility means that in order to make the claim that one event, behavior, or belief causes another, the claim has to make sense. For example, if we attend a series of lectures during which a student’s incessant midclass texting or web surfing gets in the way of our ability to focus on the lecture, we might begin to wonder whether people who have a propensity to be rude are more likely to have a propensity to be addicted to their electronic gadgets (and therefore use them during class). However, the fact that there might be a relationship between general rudeness and gadget addiction does not mean that a student’s rudeness could cause him to be addicted to his gadgets. In other words, just because there might be some correlation between two variables does not mean that a causal relationship between the two is really plausible.

The criterion of temporality means that whatever cause you identify must precede its effect in time. As noted earlier, a survey researcher examining the causes of students’ electronic gadget addictions might find that more men than women exhibit addictive tendencies when it comes to their electronic gadgets. Thus the researcher has found a correlation between gender and addiction. So does this mean that a person’s gadget addiction determines his or her gender? Probably not, not only because this doesn’t make any sense but also because a person’s gender identity is most typically formed long before he or she is likely to own any electronic gadgets. Thus gender precedes electronic gadget ownership (and subsequent addiction) in time.

Finally, a spurious relationship is one in which an association between two variables appears to be causal but can in fact be explained by some third variable. In the example of a survey assessing students’ addictions to electronic gadgets, the researcher might have found that those who can identify which of Steven Seagal’s films the actor himself directed also exhibit addiction to their electronic gadgets. 1 This relationship is exemplified in (Missing in original).

So does knowledge about Seagal’s directorial prowess cause gadget addiction? Probably not. A more likely explanation is that being a man makes a person both more likely to know about Seagal’s films and more likely to be addicted to electronic gadgets. In other words, there is a third variable that explains the relationship between Seagal movie knowledge and electronic gadget addiction. This relationship is exemplified in (Missing in original).

Let’s consider a few additional, real-world examples of spuriousness. Did you know, for example, that high rates of ice cream sales have been shown to cause drowning? Of course that’s not really true, but there is a positive relationship between the two. In this case, the third variable that causes both high ice cream sales and increased deaths by drowning is time of year, as the summer season sees increases in both (Babbie, 2010). 2 Here’s another good one: it is true that as the salaries of Presbyterian ministers in Massachusetts rise, so, too, does the price of rum in Havana, Cuba. Well, duh, you might be saying to yourself. Everyone knows how much ministers in Massachusetts love their rum, right? Not so fast. Both salaries and rum prices have increased, true, but so has the price of just about everything else (Huff & Geis, 1993). 3 Finally, research shows that the more firefighters present at a fire, the more damage is done at the scene. What this statement leaves out, of course, is that as the size of a fire increases so, too, does the amount of damage caused as does the number of firefighters called on to help (Frankfort-Nachmias & Leon-Guerro, 2011). 4 In each of these examples, it is the presence of a third variable that explains the apparent relationship between the two original variables.

In sum, the following criteria must be met in order for a correlation to be considered causal:

  1. The relationship must be plausible.
  2. The cause must precede the effect in time.
  3. The relationship must be nonspurious.

What we’ve been talking about here is relationships between variables. When one variable causes another, we have what researchers call independent and dependent variables. In the example where gender was found to be causally linked to electronic gadget addiction, gender would be the independent variable and electronic gadget addiction would be the dependent variable. An independent variable is one that causes another. A dependent variable is one that is caused by another. Dependent variables depend on independent variables.

Relationship strength is another important factor to take into consideration when attempting to make causal claims if your research approach is nomothetic. I’m not talking strength of your friendships or marriage (though of course that sort of strength might affect your likelihood to keep your friends or stay married). In this context, relationship strength refers to statistical significance. The more statistically significant a relationship between two variables is shown to be, the greater confidence we can have in the strength of that relationship. We’ll discuss statistical significance in greater detail in “Reading and Understanding Social Research”. For now, keep in mind that for a relationship to be considered causal, it cannot exist simply because of the chance selection of participants in a study.

Some research methods, such as those used in qualitative and idiographic research, are not conducive to making predictions about when events or behaviors will occur. In these cases, what we are instead able to do is gain some understanding of the circumstances under which those causal relationships occur: to understand the how of causality. Qualitative research sometimes relies on quantitative work to point toward a relationship that may be interesting to investigate further. For example, if a quantitative researcher learns that men are statistically more likely than women to become addicted to their electronic gadgets, a qualitative researcher may decide to conduct some in-depth interviews and observations of men and women to learn more about how the different contexts and circumstances of men’s and women’s lives might shape their respective chances of becoming addicted. In other words, the qualitative researcher works to understand the contexts in which various causes and effects occur.