Everyday Research Methods

Who can be "really, really smart?"

Which person is "really, really smart?" Which child is more likely to get good grades in school? Photo credit: wavebreakmedia/Shutterstock

Who thinks girls are as smart as boys? One study found that it depends on a kid's age and gender. In a set of empirical studies, a team of researchersinvestigated gender differences in gender stereotypes. CNN covered the research. 

The researchers asked young girls and boys (aged 5 to 7) to participate in several tasks.

In one test, researchers told the children a story about a person who was "really, really smart." "We were very careful to leave out any clues as to the person's gender," said researcher Lin Bian, a psychologist at the University of Illinois. The children were then asked to guess who the protagonist of the story was. [For this, kids were shown photos of four  unfamiliar adults, 2 males and 2 females, and asked to select one.] At age 5, both boys and girls picked out characters of their own gender. But within a year, that had changed. Boys still picked boys, she said -- but girls were picking boys, too.  "Suddenly, at age 6, girls became less likely than boys to do this," Bian said.

 

1. What are the variables in this study?

2. Sketch a small graph of the results they report. In your graph, how will you depict the result that age 6, "girls were picking boys, too"? Do you think this means that all the girls picked boys, or that they picked girls and boys equally?

3. This study uses a concurrent measures design. Why?

 

Here's a description of another task:

 

...in another task, the children were asked to choose who would get the best grades in school from four children, two girls and two boys. No difference was seen between the younger and older girls' likelihood of choosing their own gender.

 

4. What are the variables in this study? 

 

5. Sketch a small graph of the results this time.

 

6. One of the studies found gender stereotypes, and the other did not. How might you explain the difference in the two studies?

 

Correlates of stress in police officers

The death of a fellow officer on duty was rated as one of the most stressful events for police officers. Credit: W. Smith/Epa/REX/Shutterstock

What happens when people are exposed to very stressful events? A study has investigated this question using a sample of police officers. The study found a correlation between exposure to stressful job events and cortisol change over time. This is a journalist's take on the study. 

Here's an overview : 

For most people, cortisol, the vital hormone that controls stress, increases when they wake up. It's the body's way of preparing us for the day....[Now, a] study of more than 300 members of the Buffalo Police Department suggests that police events or conditions considered highly stressful by the officers may be associated with disturbances of the normal awakening cortisol pattern. That can leave the officers vulnerable to disease, particularly cardiovascular disease, which already affects a large number of officers.

The study's two main variables were the experience of major stress and cortisol patterns. First, read how they measured stress in the sample: 

For this study, participating officers assessed a variety of on-the-job stressors using a questionnaire that asks officers to rate 60 police-related events with a "stress rating." Events perceived as very stressful are assigned a higher rating.

Exposure to battered or dead children ranked as the most stressful event, followed by: killing someone in the line of duty; having a fellow officer killed on duty; a situation requiring the use of force; and being physically attacked.

Identifying the five most intense stressors police can face was significant, Violanti said. "When we talk about interventions to help prevent disease, it's tricky because these stressors are things that can't be prevented," he said. ... The survey showed that the officers experienced one of the five major stressors, on average, 2.4 times during the month before the survey was completed.

Second, read how they  measured cortisol patterns. Notice how in this case, the variable is operationalized not as a single outcome, but as a pattern over four time periods:

Cortisol was measured using saliva samples taken upon waking up, and 15, 30 and 45 minutes thereafter.

Here's how the journalist described and interpreted the result:

Officers who weren't as stressed showed a steep and steady, or regular, increase in cortisol from baseline. However, officers with a moderate and high major stress index had a blunted response over time.

That's because stress affects a system in the body known as the hypothalamic pituitary adrenal axis, or HPA Axis. When you're stressed, the HPA Axis elicits cortisol, a hormone that gets the body going and activates against the stressor, Violanti explained. Under normal circumstances, the body's cortisol pattern looks like a normal bell curve: It rises when we wake up, peaks around midday and comes back down at bed time.

"If you experience chronic stress or high stress situations, the cortisol can no longer adjust normally like this. So what happens with people under a lot of stress, the cortisol flattens out. For some people it goes down and others it goes up and stays up. That's called the dysregulation of the HPA axis," said Violanti, who served with the New York State Police for 23 years before shifting into academia.

Questions to answer:

1. Draw a well-labeled figure depicting the study's main result. (Will it be a bar graph, line graph, or a scatterplot? What are the best labels?)
   
   
2. What makes this a correlational study?
   
   
3. Evaluate the construct validity of a) the operationalization of stress and b) the operationalization of cortisol patterns. In your opinion, how well did they measure these two variables? 
   
   
4. Consider the external validity of this study. What are the characteristics of the Can we assume that the results will generalize to other cops? Do we know if the results generalize to other professions or people who have experienced stress? Why or why not? 
   
   
5. Can the study support the causal claim that "exposure to stress causes cortisol dysregulation in cops?" Consider temporal precedence (the directionality problem) as well as internal validity (the third variable problem).

 

Suggested Answers:

Question 1. You might wish to see how the authors graphed their results in the scientific paper, here. Look at Figure 1. 

Question 4 is about external validity. You might be interested to read the authors' take on this question: 

While the current study focused on Buffalo officers, the findings have implications for cops around the country, said paper co-author Michael Andrew, PhD, chief of the Biostatistics and Epidemiology Branch of the CDC/NIOSH Health Effects Laboratory Division in Morgantown, West Virginia.

"These findings show that exposure to major events inherent to police work may lead to a temporary reduction in the biological ability to respond to further stressful events. Since the major stressor events in this study were originally developed to reflect events that can apply to any police department, these results should generalize, more or less, to any police department in the U.S.," Andrew said, adding, "This points to the need for continued focus on supporting police officer health."

For Question 5, you already know that this study establishes covariance. However, temporal precedence is not very clear. It's possible that cops with poor cortisol regulation are more likely to be involved in future stressful events (for some reason). Internal validity is more of a problem, because, at least based on what's presented here, we don't know if they controlled for third variables such as what type of neighborhood the cops usually patrolled, or for personality characteristics such as impulsiveness, Type A personality, or other traits. For example, an impulsive personality might be associated with more stressors on the job, and might also be associated with cortisol patterns.