Correlational studies on PTSD

Correlational studies are one of the most commonly used research methods in the study of Post-traumatic Stress Disorder (PTSD). In this post we’ll review WHAT correlational studies are, HOW they’re used to study PTSD and WHY they’re used. We’ll also review three key studies that can support this explanation. 

Research methods essays are the hardest to write, but that also means you have the potential to easily distinguish yourself from the rest. The following will help you do just that.

Read more: 

• Biological Etiology of PTSD: Abnormalities in the brain
• Exam Tips: How to write a research methods essay
• Mr. Dixon’s experiences with PTSD

What is a correlational study?

Correlational studies involve gathering data on two or more variables (called co-variables) and calculating a correlation coefficient to see how strongly they are related. A positive correlation is when one variable increases so do the other. A negative correlation happens when one variable decreases while the other increases.  The coefficient can range in value from -1.0 to + 1.0. A zero would be no correlation, a weak correlation is about 0.2-0.4, moderate is 0.4 to 0.6 and anything over 0.6 is considered a strong correlation.

How are they used to study PTSD?

Researchers gather data on the co-variables using questionnaires, interviews or other methods. In studies on PTSD, the Clinically Administered PTSD Scale (CAPS) is the tool most commonly used to gather data on PTSD symptoms. This is a structured interview that measures the severity of PTSD symptoms. Participants scores are then correlated with another variable that is hypothesized to have an effect on PTSD symptoms. Examples include volume and activity in particular areas of the brain (e.g. the hippocampus), cognitive appraisal styles and income (see example studies below).

Why are they used to study PTSD?

Correlational studies are commonly used when researchers are studying relationships between variables that cannot be manipulated in a laboratory setting. In studies on PTSD it is often the case that one or more of the covariables in the study would be difficult (if not impossible) to manipulate in a laboratory setting. It would be unethical, for example, to conduct an experiment to see if altering someone’s brain activity could cause symptoms of PTSD since this would cause harm (and might be irreversible). Animal experiments are a common alternative, but these are limited because animal models of PTSD might not generalize to humans. We can also only experiment on biological variables on animals, factors like income and appraisals cannot be manipulated in animal experimentation.

Another reason correlational studies are used is because PTSD is caused by experiencing a traumatic event. This makes conducting experiments difficult because it is unethical to traumatize someone for the purposes of research. Therefore, correlational studies gather data on participants who have already experienced some kind of trauma, like being in combat or a car accident.

Example Studies 

• Gilbertson et al. (2002) conducted a twin study on the hippocampus and PTSD. They found a negative correlation between hippocampal volume and PTSD symptoms in war veterans who had developed chronic PTSD (r = -0.64). This means those with smaller hippocampi tended to have worse symptoms.
• Hitchcock et al. (2015) conducted a study on kids and teenagers (7-17 years old). They used a CAPs scale adapted for children while the kids were in hospital being treated following a traumatic event (e.g., road traffic accident) and again six months later. They found that negative appraisals were positively correlated with PTSD symptoms (r = 0.31). This means more negative appraisals increased the risk for PTSD symptoms.
• Irish et al. (2011) conducted a study on gender differences in PTSD in a sample of people who had been in car accidents. They gathered their data on the participants while they were in hospital.  They found that women were more at risk for PTSD following a car accident. They also found that income was negatively correlated with initial PTSD symptoms (r = 0.20), after six weeks (-0.16) and after six months (-0.23). This supports other studies that show people with less income and from lower socioeconomic backgrounds (i.e. financially poor people) are more at risk for PTSD.

Exam Tip: The above three studies are excellent examples because they all used CAPs to measure PTSD symptoms and they also calculate correlation coefficients. They also represent each of the three approaches (biological, cognitive and sociocultural).

But….

No discussion of the research methods used in the study of PTSD is complete without considering the downsides. Here are some questions to get you thinking about the limitations of using correlational studies in the etiology of abnormal psychology.

1. Correlation does not mean causation: This is the fundamental limitation of correlational studies. How can you make this point using the above studies? (You can watch this video explaining how to evaluate correlational studies properly.)

2. Analyze the correlation coefficients: Review the summary of weak vs. strong correlations at the top of this page. What can you say about the conclusions from this study? What should researchers (and readers) bear in mind when discussing these findings?

---

References:

• Gilbertson MW, Shenton ME, Ciszewski A, Kasai K, Lasko NB, Orr SP, Pitman RK. Smaller hippocampal volume predicts pathologic vulnerability to psychological trauma. Nat Neurosci. 2002 Nov;5(11):1242-7. doi: 10.1038/nn958. PMID: 12379862; PMCID: PMC2819093. Link
• Hitchcock, C., Ellis, A. A., Williamson, P., & Nixon, R. (2015). The Prospective Role of Cognitive Appraisals and Social Support in Predicting Children’s Posttraumatic Stress. Journal of abnormal child psychology, 43(8), 1485–1492. https://doi.org/10.1007/s10802-015-0034-7. Link.
• Irish, L. A., Fischer, B., Fallon, W., Spoonster, E., Sledjeski, E. M., & Delahanty, D. L. (2011). Gender differences in PTSD symptoms: an exploration of peritraumatic mechanisms. Journal of anxiety disorders, 25(2), 209–216. https://doi.org/10.1016/j.janxdis.2010.09.004. Link.