Everyday Research Methods

Has the pandemic changed people's personalities?

The study reported that average levels of extroversion, agreeableness,
and conscientiousness all went down from before the pandemic
when compared to a year after lockdowns. 
Credit: TheVisualsYouNeed/Deposit Photos

During the lockdowns of the pandemic, many researchers developed and tested interesting research questions. They collected data to test COVID-related hypotheses, and now many of those studies have completed the peer-review process and are coming out in publication. 

One such study was interested in whether the prolonged social isolation of the pandemic might have resulted in shifts in people's personality traits, on average. The study was covered by multiple media outlets, including the online magazine Fortune. The original empirical research is available open-acesss, in PLOS ONE. 

Researchers have long been interested in the question of how, and how much, people's personalities change. They've asked the extent to which people become more conscientious when they transition into a full-time job, or whether people become more agreeable as they get older. The study summarized by this journalist is in the same spirit. 

The Fortune journalist summarized the study this way: 

Past studies have found personalities are generally immune to environmental factors like hurricanes and earthquakes, but a group of researchers found that the global COVID-19 pandemic may be an exception. 

The researchers used data from the Understanding America Study of the University of Southern California to examine results from over 7,000 U.S. adults ages 18 to 109 before and during the pandemic. Personality was measured using a scale to assess five traits: neuroticism (is moody), openness (has an active imagination), conscientiousness (is a reliable worker), agreeableness (is generally trusting), and extraversion (is talkative).

The study echoed previous findings that neuroticism actually decreased from the period before the pandemic started to the early spring months of 2020, which may in part be because the beginning of the pandemic caused major outpourings of support with people rallying together and could have improved individual emotional stability, Sutin says. It may also be because people were able to point to a stressor as the reason for their overwhelm or anxiety versus blaming themselves, she speculates. 

However, as the pandemic wore on, personality traits changed more drastically. Measures of extroversion, openness, agreeableness, and conscientiousness all declined compared to participants’ pre-pandemic personalities. Experts say shifts like this usually take a decade.

Questions

a) This is a type of quasi-experiment. What is the quasi-experimental variable? What are the dependent variables? 

b) Which quasi-experimental design (from Chapter 13) best fits this study? Here are your options: Non-equivalent control group posttest-only design, Non-equivalent control group pretest-posttest design, Interrupted time series design, or Non-equivalent control group interrupted time series design?

c) Pick the dependent variable of extroversion. Sketch a graph of the results described by the journalist. If you want to follow the style of the graphs in the textbook, you would put time on the x-axis and extroversion on the y-axis. 

d) Now you can look at how the empirical study presented the results by looking at Figure 1 of the empirical article. The published graph probably looks different from yours (in c, above) in several ways. First, they don't show absolute levels of extroversion; instead, they graphed changes in extroversion--in 2020 and 2021, compared to a pre-panedemic baseline. Specifically, you will see a bar graph in which rates of each trait were either higher (positive bars) or lower (negative bars) than they were before the pandemic. You'll also see that they graphed these changes separately for different age groups. Study the bar graphs in Figure 1 and identify the traits and subgroups that had the largest changes. 

e) Think about the external validity of this study. The sample came from the Understanding America Survey. Visit the survey's website and see if you can figure out if the sample used in the study was collected using probability sampling techniques or not (hint: use their "methodology" tab).

 

Stress and food: experimental and quasi-experimental evidence

In one study, department store workers ate more unhealthy foods
when they were working more demanding shifts. Is such a study likely to be
an experiment, a quasi-experiment, or a correlational study?
Credit: 22January/Shutterstock

A Washington Post journalist integrated several health psychology studies in an article about the relationships between stress and eating. For many people, stress and anxiety is associated with weight gain. The journalist summarized several examples of studies that support, and help explain, this link. Read them and think about whether they are  experimental or quasi-experimental. 

Here's the first study the journalist described (we'll call it Study A):

In laboratory studies, scientists have found that administering synthetic versions of cortisol to people causes them to eat substantially more calories than people who are given a placebo. That’s in part because cortisol reduces your brain’s sensitivity to leptin, also known as the satiety hormone, which regulates your appetite and makes you feel full.

a) There are two variables in this study. What are they? What are their levels? For each variable, decide if it is measured or manipulated. 

b) Do you think this was an experiment or a quasi-experiment? Explain your reasoning.

Here's the next one, which we'll call Study B:

In one study of department store workers, people ate more sugar, saturated fat and overall calories when they had to work long, demanding shifts compared with when they worked less stressful shifts with lighter workloads.

c) There are two variables in this study. What are they? What are their levels? For each variable, decide if it is measured or manipulated. 

d) Do you think this was an experiment or a quasi-experiment? Explain your reasoning.

Here's another, which we'll call Study C:

Even stress from activities we enjoy can lead to overeating. In one study, researchers followed ardent football fans in different cities. They found that fans whose NFL teams lost on Sunday consumed more calories and saturated fat the next day. Fans whose teams won ate less food and saturated fat the following day. The scientists found similar results when they looked at the dietary patterns of French soccer fans.

e) There are two variables in this study. What are they? What are their levels? For each variable, decide if it is measured or manipulated. 

f) Do you think this was an experiment or a quasi-experiment? Explain your reasoning.

So far, the evidence suggests that stress in one's life might lead people to eat more of types of foods that are not good for us. Here's a final study that was designed to test a way to prevent stress from affacting our eating habits. We'll call it Study D:

In a study published last year, [researcher Janet] Tomiyama and her colleagues recruited 100 adults with elevated stress levels and split them into two groups. Everyone was trained to do a daily, six-minute stress reduction exercise called progressive muscle relaxation, which involves tensing and relaxing your muscles from toes to head. You can find an example of it here. This deep relaxation technique has been shown in studies to reduce stress and anxiety.

But members of one group were assigned to eat a serving of fresh fruit such as sliced pineapple, honeydew and pears, about five minutes into each of their daily progressive muscle relaxation sessions. After one week of this, the researchers found that eating just the fruit alone made the participants feel less stressed and put them in a better mood. By pairing the fruit with a relaxation exercise, their brains began to view the fruit as something that reduced their stress levels — essentially turning the fruit into comfort food.

g) There are two variables in this study. What are they? What are their levels? For each variable, decide if it is measured or manipulated. 

h) Do you think this was an experiment or a quasi-experiment? Explain your reasoning.

i) Bonus question: Study D included an element of classical conditioning, which you probably learned about in Introductory Psychology. Can you identify the conditioned stimulus and conditioned response in this example? 

j) Rank the four studies (A, B, C, and D, above) in terms of which ones you believe have the strongest evidence (based on their methodology).

Can a cuddly object or blanket reduce adult anxiety?

Does holding a comfort object actually reduce anxiety in adults? That's an empirical question!  Photo: NDanko/Shutterstock

Apparently, about half of American adults still have a childhood teddy bear, and 40% still sleep with it (we can look into the methodology of this survey in a later post!). Do such comfort objects provide anything beyond sentimental feelings?  What about the products marketed to help calm people or help them feel less lonely--products like weighted blankets, or the robotic, stuffed PARO seal, which was created to engage older adults with dementia? 

Time magazine journalist Jamie Friedlander Serrrano covered a variety of studies on these types of comfort objects. The story described several studies that have experimentally tested the effects of weighted blankets, teddy bears, and the PARO seal. One study featured by the journalist was published in the empirical journal, PLOS One. This study focused on a robotic cushion that was plush and soft, and which also mimicked breathing by inflating and deflating every few seconds.

The journalist explained that the researchers wanted to see if holding this object would help students feel less anxious before a stressful oral math exam. 

Students in the test group held the object—a plush, baby-blue cushion the size of a throw pillow that automatically inflates and deflates, mimicking inhaling and exhaling—for eight minutes before their exam. One control group did a guided breathing meditation instead, while another control group didn’t do anything special. Haynes and her team found that clutching the stuffed breathing cushion reduced anxiety as much as doing the guided meditation.

“This indicated that the cushion could be similarly effective as a breathing meditation for anxiety,” says [scientist] Haynes, who is now completing her postdoctoral fellowship at Saarland University in Germany. “We didn’t give the students in the experiment any guidance about using the cushion. We didn’t tell them to follow it with their breathing or anything—it was just purely the act of holding it as it slowly breathed that eased their anxiety. I think we thought it would help with anxiety, but we were pleasantly surprised that it was as effective as the breathing meditation.”

Lywood—who is currently working on commercializing the breathing cushion through her company Sooothe—believes the findings highlight our innate need for touch, even if the source isn’t human or alive, for that matter. “We take touch for granted,” she says. But because many people have been deprived of it during the pandemic, she points out, “we are sort of rediscovering how valuable it is.”

Questions

1. Using the table below, classify the two main variables mentioned in this study (hint: One variable was manipulated and one was measured)

Variable name (conceptual variable)	How was this variable operationalized?	What are this variable's levels?	Was this a manipulated or measured variable?	Was this an IV or a DV?

2. Was the independent variable manipulated as independent groups or within-groups? 

3. Which of the four basic experimental designs was used: posttest only? pretest-posttest? repeated measures? concurrent measures?  Explain your answer. 

4. Sketch a little bar graph of the results of the anxiety result.

5. The journalist doesn't specify whether people were randomly assigned to the three conditions, and when you check the PLOS One article, it also doesn't mention that people were randomly assigned. I don't think we should assume that researchers did NOT take this step--sometimes they just forget to mention it. In any case, why would you need to randomly assign people to the three conditions? That is, what kinds of alternative explanations does random assignment help rule out?

6. One of the authors of the study also created, and hopes to market, the "breathing" cushion. Is this an ethical problem for the study? Is this a scientific problem for the study? 

"Fetuses smile for carrots but grimace over kale"

This fetus's facial expression would be judged as more of a cry or grimace. Photo courtesy of FETAP (Fetal Taste Preferences) Study, Fetal and Neonatal Research Lab, Durham University

This fetus's facial expression would be judged as smiling (or lip-pulling). Photo courtesy of FETAP (Fetal Taste Preferences) Study, Fetal and Neonatal Research Lab, Durham University

Even though kale is having a moment as a vegetable (and kale salad is certainly one of my favorites), most folks would agree that carrots taste better than kale.  A new study, covered by multiple news outlets, has demonstrated that even in utero, fetuses agree. They seem to be able to tell the difference between carrots and kale, and they like carrots better!

As you can imagine, this study was a media darling. Here are some excerpts of a report on this study as covered by journalists at CNN. 

The scientists used in-utero facial expressions to infer how babies felt about what their parent had eaten: 

Fetuses create more of a “laughter-face” in the womb when exposed to the flavor of carrots consumed by their mother and create more of a “cry-face” response when exposed to kale [...]

The study looked at the healthy fetuses of 100 women between the ages of 18 and 40 years who were between 32 and 36 weeks pregnant in northeast England.

From this, 35 women were [assigned at random] into an experimental group that consumed an organic kale capsule, 35 were put into a group that took a carrot capsule, and 30 were put into a control group that was not exposed to either flavor.

After a waiting period of 20 minutes post-consumption, the women underwent 4D ultrasound scans, which were compared to 2D images of the fetuses.

Lip corner pulling, suggestive of a smile or laughter, was significantly higher in the carrot group compared to the kale and control group. Whereas ... casting down of the lower lip...– suggestive of a crying face – [was] much more common in the kale group than the other groups.

Questions

1. Using the table below, classify the three main variables mentioned in this study (hint: One variable was manipulated and two were measured)

Variable name (conceptual variable)	How was this variable operationalized?	What are this variable's levels?	Was this a manipulated or measured variable?	Was this an IV or a DV?

2. Was the independent variable manipulated as independent groups or within-groups? 

3. Which of the four basic experimental designs was used: posttest only? pretest-posttest? repeated measures? concurrent measures?  Explain your answer. 

4. Sketch a little bar graph of the results of the "lip pulling" result.

5. Can this study support the causal claim that "eating a carrot pill causes fetuses to make a smiling face"? Apply the three causal criteria. 

6.  The CNN study mentioned the following in their reporting. Which of the four big validities is this addressing? 

Participants were asked not to consume any food or flavored drinks one hour before their scans. The mothers also did not eat or drink anything containing carrot or kale on the day of their scans to make sure it would not influence the results.

7. The 100 fetuses in this study were not a random sample of any population--they were a convenience sample. Why do you think these researchers did not prioritize external validity in this study? Do you think that's a major problem, or not? 

8. Use the Internet to find other media coverage of this study--note any similarities and differences in how the journalists discuss the findings. 

Here's a link to the original empirical article (paywalled). 

Time alone with your thoughts?

Credit: Ground Picture/Shutterstock

Imagine sitting alone with your thoughts for 20 minutes. You wouldn't be allowed to distract yourself with a book, phone, or even walking around. How much do you think you would you enjoy it?  

When this question was asked of participants in a series of studies, most reported that they'd not enjoy time alone very much.  But this study also tested how well their predictions matched their actual experience. The empirical research was published in the Journal of Experimental Psychology: General, and  summarized in a trade magazine for the American Psychological Association. According to the journalist, 

...study lead author Aya Hatano, PhD, of Kyoto University in Japan[, said] “Our research suggests that individuals have difficulty appreciating just how engaging thinking can be. That could explain why people prefer keeping themselves busy with devices and other distractions, rather than taking a moment for reflection and imagination in daily life.”

The research team conducted six studies. Here is an overview of some of them: 

In a series of six experiments with a total of 259 participants, the researchers compared people’s predictions of how much they would enjoy simply sitting and thinking with their actual experience of doing so. In the first experiment, they asked people to predict how much they would enjoy sitting alone with their thoughts for 20 minutes, without being allowed to do anything distracting such as reading, walking around or looking at a smartphone. Afterward, participants reported how much they had enjoyed it.

The researchers found that people enjoyed spending time with their thoughts significantly more than they had predicted. This held true across variations of the experiment in which participants sat in a bare conference room or in a small, dark tented area with no visual stimulation; variations in which the thinking period lasted for three minutes or for 20 minutes; and one variation in which the researchers asked people to report on their enjoyment midway through the task instead of after it was over. In every case, participants enjoyed thinking more than they had expected to.

a) The researchers conducted six different studies. Do these sound like direct replications? Replication-plus-extensions? or Conceptual replications? (you can skip this question if you haven't done Chapter 14 yet)

b) The journalist identifies these as "experiments," so let's assume that each study contains a manipulated variable. Take this study, as described by the journalist:

The researchers found that people enjoyed spending time with their thoughts significantly more than they had predicted [in a variation of] the experiment in which participants sat in a bare conference room or in a small, dark tented area with no visual stimulation. 

In this study, one of the independent variables was "predicted vs. actual".  What is the other main independent variable, or IV? What is the dependent variable, or DV? 

c) Do the same for this study:

variations in which the thinking period lasted for three minutes or for 20 minutes;

In this study, one of the independent variables was "predicted vs. actual". What is the other IV? What is the DV?

d) And for this one: 

In another experiment, the researchers compared one group of participants’ predictions of how much they would enjoy thinking with another group’s predictions of how much they would enjoy checking the news on the internet. Again, the researchers found that people underestimated their enjoyment of thinking. The thinking group expected to enjoy the task significantly less than the news-checking group, but afterward, the two groups reported similar enjoyment levels.

In this study, one of the independent variables was "predicted vs. actual". What is the other IV? What is the DV?

e) Challenge question: For the study above (comparing time alone to looking at news on the internet)  and graph the result. You can compare your graph to Figure 6 in the empirical article (available next to the APA journalist's story, here).

Here's a final caveat about the results. I appreciated this statement because it puts the statistical outcome in concrete and realistic terms: 

It is important to note that participants did not rate thinking as an extremely enjoyable task, but simply as more enjoyable than they thought it would be, according to Murayama. On average, participants’ enjoyment level was around 3 to 4 on a 7-point scale. Future research should delve into which types of thinking are most enjoyable and motivating, according to Murayama. “Not all thinking is intrinsically rewarding, and in fact some people are prone to vicious cycles of negative thinking,” he said.

 

 

What your taste in music says about your personality

According to the study, which personality trait is associated with upbeat, danceable music? Credit: Prostock-studio/Shutterstock

This splashy heading from Inc.com reads:

Here's what your taste in music says about your personality, according to a huge new study. whether you like Mozart or metal really does reveal something about your personality, new research finds

 

This was a correlational study, and provides a good chance to practice concepts from Chapter 3 and 8. Here is how the journalist summarized the methods and results:

...researchers from University of Cambridge and Israel's Bar-Ilan University used an online quiz to extract information on the musical preferences and personalities of more than 285,000 study subjects from 53 countries (you can try the quiz yourself here). They also set up a website where 71,000 visitors from 34 countries rated music clips and took a personality quiz. 

The team used well-known, scientifically validated frameworks for categorizing both music and personality. The aptly named "Music" framework categorizes music using five dimensions: mellow, unpretentious, sophisticated, intense, and contemporary. The familiar Big 5 personality framework categorizes people according to openness, conscientiousness, extroversion, neuroticism, and agreeableness. 

Did the research team find any correlations between the type of music someone enjoyed and their personality type? Yup, reported the researchers recently in the Journal of Personality and Social Psychology. 

a) You could think of this study as having two big variables or ten smaller variables. What would be the two "big" variables? What would be the ten smaller variables? 

b) What makes this a correlational study? 

c) Here is a description of one of their results:

Extroverts the world over tend to like upbeat, danceable contemporary music. 

Sketch a well-labelled scatterplot of this correlation. 

d) Here is another result: 

The study found fans of the Nirvana classic "Smells Like Teen Spirit" and other intense music are higher in neuroticism on average.

Sketch a well-labelled scatterplot of this correlation. 

e) Here is a third result:

Those high in conscientiousness tended to avoid more intense music...

Sketch a well-labelled scatterplot of this correlation. 

f) And here's one more:

Looking for an agreeable type? Then put on some Marvin Gaye and see how they react. The researchers found a strong correlation between liking mellow music and being high in agreeableness.

 

g) Now, here are two quotes from the journalist's article. For each quote decide which of the four big validities this point is addressing.

 

1. The team used well-known, scientifically validated frameworks for categorizing both music and personality. 

2. They ... set up a website where 71,000 visitors from 34 countries rated music clips and took a personality quiz. 

 

h) What is a question you could ask about this study's statistical validity?

 

i) Inc.com is a business magazine, so they framed the study in terms of job hiring:

 

Despite its sometimes dubious scientific basis, the personality testing industry rakes in around $2 billion annually telling companies their new hire is an INFJ or revealing her Hexaco type. [This] study suggests your business might be able to save some money and just ask candidates about their taste in music instead. 

What do you think? Does the research study support the idea that we could use music preferences in job interviews to select certain personality traits? Why or why not?  (to read the Inc.com take on this question, read the full article). 

 

 

Diet + Exercise: A factorial example

The study found that both exercise and diet are important. Is the effect additive or interactive? Photo credit:Dragana Gordic/Shutterstock

You might have heard someone claim that because they exercise regularly, they can eat whatever they want and still be healthy. When people say this, they seem to mean that either exercise OR diet can be sufficient to foster a healthy lifestyle. However, a recent study counters this "either/or" view. Like many other studies, this one found evidence that both diet and exercise contribute to health outcomes in adults.

In this blog post, we'll explore how the study can be viewed as a factorial design (Chapter 12), and we will discuss how its main finding can map on to different factorial outcomes. 

The empirical study appeared in the British Journal of Sports Medicine, which may be paywalled. I learned about the study from number of journalists who covered it, including this accessible summary in Healthline, which we'll use in this post.  

Here are the opening bullet points from the Healthline story: 

• A new study finds that in order to live longer, people need to eat healthily and exercise regularly.
• While a balanced diet and regular exercise independently have plenty of health benefits, the study notes the greatest reduction in mortality risk comes from combining diet and exercise together.

a) These bullet points mention the study's three main variables. One of the variables is mortality risk, or longevity (lifespan). What are the other two variables?

b) Chapter 12 concerns factorial designs, in which there are two (or more) independent variables (IVs) and one dependent variable (DV). Reflect back on the three variables you listed in a), above, and identify which are the IVs and which is the DV.  In addition, identify the levels of each IV. 

Here are some details about the study's method and results, as summarized by the journalist:

For the study, the researchers evaluated the health and exercise data of 346,627 individuals over the course of 11 years. They tracked how much exercise each participant completed in an average week along with how intense their physical activity was. The team also followed the participants’ eating habits. During the study window, 13,869 participants died — 2,650 from heart disease and 4,522 from adiposity-related cancers.

The research team found that any type of regular exercise was associated with a lower risk of mortality, and those who both ate well and exercised had the lowest risk of dying.

In addition, while exercise and diet are independently linked to a lower risk of mortality, high levels of exercise cannot entirely mitigate the harms of a poor diet.

I made up three patterns of data to illustrate possible outcomes of this study. You can see the graphs below, labelled Outcomes A, B, and C. Take a moment to orient yourself to the graphs, noticing how the DV is on the y-axis, one IV is on the x-axis, and the other IV is depicted using line color:

 

c) Which of these outcomes (A, B, or C, above) is consistent with the idea that EITHER exercise OR diet alone is sufficient to maintain a longer lifespan? Explain your reasoning.

d) The journalist wrote that "The research team found that any type of regular exercise was associated with a lower risk of mortality, and those who both ate well and exercised had the lowest risk of dying [i.e., the longest lifespan].”  Two of these outcomes (A, B, or C) are consistent with this statement. Which ones? 

e) In your own words, describe the difference between Outcome A and Outcome B.

f) Based on what the journalist wrote, can you tell which of these two (Outcome A or Outcome B) the study found? Specifically, which one could go with how the journalist described the study's outcome:

“In our study, those who ate a poor-quality diet and were active still had substantially reduced mortality risk than those who ate a poor-quality diet and were inactive,” Melody Ding, MPH, PhD, the lead author of the study and an associate professor of health and medicine in the Sydney School of Public Health at the University of Sydney, told Healthline. “It’s just that you had the maximum risk reduction when doing both things right.”

g) Just for more practice: Two of the outcomes above depict an interaction effect (an interaction between diet and exercise). Which 2 outcomes are they? Explain your reasoning--how do you know?

Suggested answers:

a) Besides longevity, a second variable is diet quality and a third variable is exercise habits.

b) Longevity is the DV. Diet quality is an IV (levels could be good vs. poor diet). Exercise habits are also an IV (levels good be regular exercise vs. no exercise)..

c) Outcome C. Outcome C shows that you have the same lifespan if you exercise or eat a good diet, and that doing both does not increase your lifespan any more than doing one or the other (all three conditions--exercise alone, diet alone, and both--show the same lifespan). Note that this is not actually the result that the study found.

d) Both Outcome A and Outcome B are consistent with the journalist's statement. 

e) The difference between Outcomes A and B can be summarized with the terms "additive" versus "interactive."
Outcome A shows an additive effect--A good diet adds (in this example) 3 years to your life; exercising adds 3 years, and if you do both, it adds 6 years. One is good, both is better.  
Outcome B shows an interactive, or multiplicative effect. It shows that if you do both diet AND exercise, you have an especially long lifespan. That is, in this example, a good diet adds 3 years to your life; exercising adds 3 years, and if you do both, it adds 11 years. One is good, both is amazing (kaboom!). In Chapter 12 you'll learn that this is an interaction--the combination of the two IVs results in something even bigger than simply adding the two together.

f) In my view, we can't tell which outcome the study found based on what the journalist wrote. However, since they used the terms "both" rather than "especially", it seems like it might be Outcome A. My reading of the empirical study backs this up--it seems to be an additive effect (not interactive). 

g) Outcome B and C both show interactions, but of different types. First, you can see the interactions in both because the lines are not parallel--there are differences of differences. That is, the effect of diet is different, depending on if we're talking about regular exercise or no exercise. 

In Outcome B, there is an interactive, or multiplicative effect--I sometimes call this an "especially" interaction. It shows that if you do both diet AND exercise, you have an especially long lifespan (compared to doing either one alone).

In Outcome C, there is a different kind of interaction--I might call it an "only if" or an "it depends" interaction. You could explain it by saying that lifespan is reduced only if you don't exercise AND don't eat well. If you exercise, or diet, or both, you'll live longer. 

Tuskegee Syphilis Study: An anniversary

Four months after Heller's story broke, the Tuskegee study was terminated with this letter.
Credit: U.S. National Archives

In this photo from the 1950s, a man participating in the study had his blood drawn.
Credit: U.S. National Archives

The year 2022 marks the 5oth anniversary of a major news event--journalist Jean Heller's story about the Tuskegee syphilis study. In Chapter 4 you can read about the study, in which poor Black men in Alabama were recruited and purposefully not treated for syphilis. The news story first appeared in the Associated Press (the AP). To honor its 50-year anniversary, the AP re-ran their original news report from July, 1972. 

Here is the lede (the opening passage) from Heller's original story:

For 40 years the U.S. Public Health Service has conducted a study in which human guinea pigs, denied proper medical treatment, have died of syphilis and its side effects.

The study was conducted to determine from autopsies what the disease does to the human body.

PHS officials responsible for initiating the experiment have long since retired. Current PHS officials, who say they have serious doubts about the morality of the study, also say that it is too late to treat syphilis in any of the study’s surviving participants.

In Chapter 4 (Ethics), you can read how the Tuskegee syphilis study led to the Belmont Report, which outlined major changes in how research is evaluated and conducted in the United States. The Tuskegee study violated three distinct ethical principles: Respect for Persons, Beneficence, and Justice. 

a) In your own words, define each of these principles. 

b) Read each of the quoted passages below, from Heller's original story. Decide what ethical problems each one seems to illustrate. Can you map each passage to one of the core ethical principles of the Belmont report (Respect for Persons, Beneficence, or Justice)?

b1) The experiment, called the Tuskegee Study began in 1932 with about 600 black men mostly poor and uneducated, from Tuskegee, Ala., an area that had the highest syphilis rate in the nation at the time.

b2) As incentives to enter the program, the men were promised free transportation to and from hospitals, free hot lunches, free medicine for any disease other than syphilis and free burial after autopsies were performed.

b3) The Tuskegee Study began 10 years before penicillin was discovered to be a cure for syphilis and 15 years before the drug became widely available. Yet, even after penicillin became common, and while its use probably could have helped or saved a number of the experiment subjects, the drug was denied them, Dr. J.D. Millar says.

Don Prince, another official in the venereal disease branch of CDC, said [...] “I don’t know why the decision was made in 1946 not to stop the program,” Prince said. “I was unpleasantly surprised when I first came here and found out about it. It really puzzles me.”

As part of the anniversary, the AP also ran a story-behind-the-story about how Heller researched and broke the story.  Here is Heller reflecting on how she felt as she was breaking the story: 

Normally, reporters celebrate these “Eureka” moments. But Heller felt no such elation.

“I knew that people had died, and I was about to tell the world who they were and what they had,” she says, her voice dropping. “And finding any joy in that ... would have been unseemly.”

Heller also said,

“As much injustice as there was for Black Americans back in 1932, when the study began, I could not BELIEVE that an agency of the federal government, as much of a mistake as it was initially, could let this continue for 40 years,” says Heller. “It just made me furious.”

What were some outcomes of Heller's reporting? 

Nearly four months after the story ran, the study was halted.

The government established the Tuskegee Health Benefit Program to begin treating the men, eventually expanding it to the participants’ wives, widows and children. A class-action lawsuit filed in 1973 resulted in a $10 million settlement.

c) As you reflect on this story, what are you most concerned about? What questions do you still have? 

 

Having more kids linked to lower cognitive function

The study found that parents with more than two children experienced greater cognitive decline as they aged, compared to parents with only two children--at least in Northern Europe.  Photo credit: fotoliza/Shutterstock

The website Study Finds produces  summaries of science research, publishing multiple summaries per day. This example concerns a study showing that how many kids people have had (in their younger years) predicted their later-life cognitive function. Somewhat confusingly, the print headline reads, "having more kids linked to weaker brain function," but the accompanying video segment includes the causal claim, "Kids can weaken your brain". 

Here's the opening segment from the journalist:

Everything in moderation — even kids? Researchers from Columbia University and Université Paris-Dauphine report having more than two kids may have a negative impact on late-life cognition. The study shows that older parents with just two children appeared sharper cognitively than those with three.

a) There are two main variables in this statement so far. What are they? Do you think these variables were manipulated or measured? 

b) Identify the verbs in the bolded statements above, and explain which ones are association claim verbs and which are causal claim verbs. 

c) On the face of it, the study found a simple correlation between having more than two kids (variable A) and cognitive function (variable B). You know that correlation is not causation, of course. What are some possibilities that explain this relationship? Consider temporal precedence first: Come up with a causal story in which A leads to B, as well as an alternative story in which B leads to A. Then come up with a C variable (a third variable, or internal validity problem) that might plausibly be related to both A and B. 

Now that you've assessed the basics of this bivariate correlational study,  here's more of what the journalist said about the study:

...study authors analyzed data provided by the Survey of Health, Aging and Retirement in Europe (SHARE). Featuring representative samples of the older populations across 20 European countries and Israel, SHARE was an ideal match for this project. Each individual was at least 65 years old and had at least two biological children.

Via a series of complex econometric methods, the research team was able to successfully disentangle various causal relationships from simple associations. In simpler terms, the analysis found that having three or more kids (as opposed to just two) is related to worse late-life cognition. This held up among both moms and dads.

[...]

“The negative effect of having three or more children on cognitive functioning is not negligible, it is equivalent to 6.2 years of aging,” Dr. Bonsang notes.

The journalist referred to "complex econometric methods" because the study itself used a creative strategy to try to test the relationship of having more than two kids on cognitive functioning, while ruling out potential alternative explanations. The researchers started with the assumption that parents who choose to have more than one kid often wish to have at least one boy and one girl. Indeed, in their large European data set, they found evidence that when people start out with two kids of the same sex (such as two boys or two girls), they are more likely to go for a third child, compared to people whose first two kids were a boy and a girl. Then, when the researchers statistical controlled for this variable (the sex composition of the first two kids), they observed a strong decline in cognitive function in later years. 

The upshot is that the researchers tried to take advantage of a natural experiment that goes something like this: When people's first two kids are the same sex, they're more likely to have a third child, which is likely to be followed by cognitive decline in later years. In contrast, when people's first two kids are different sexes, they're less likely to have a third child, which is followed by less cognitive decline. Because sex of one's first two children is more or less randomly assigned, it serves as an experimental variable. 

Importantly, they found the effect of children on cognitive decline only in the Northern European countries (not other regions such as Western, Eastern, or Southern). 

Here are some brief questions for you. 

Which of the four big validities (internal, external, construct, or statistical) is each of these statements addressing? Each quote comes from the journalist's coverage. 

d)  "...study authors analyzed data provided by the Survey of Health, Aging and Retirement in Europe (SHARE). Featuring representative samples of the older populations across 20 European countries and Israel" (Which validity is this?)

e) “'The negative effect of having three or more children on cognitive functioning is not negligible, it is equivalent to 6.2 years of aging,' Dr. Bonsang notes." (Which validity is this?)

f) Which two validities are left? Ask a question that would provide information about the remaining two validities.

g) Is it a problem that this effect was found only in the Northern European sample, and not the other three samples? Why or why not?

 

Life expectancy gap in different U.S. counties

An example of politically polarized health outcomes occurred during the pandemic.  Democratic-led areas were more likely to pass masking and lockdown mandates, and these areas later had lower COVID death rates compared to Republican-led areas.
Photo credit: MikeEdwards/Depositphotos

The American political system is polarized--increasingly so in the last few decades. Such polarization has gridlocked national politics in the past few years. In contrast, U.S. states and counties are passing legislation, and increasingly, this legislation is reflecting polarized local politics.

A new study decided to look at whether these local political differences were linked to actual health outcomes. The researchers showed that county by county, the average life expectancy in a particular U.S. county is associated with that county's presidential voting patterns. Specifically, the study found that 

how a state votes in presidential elections helps predict life expectancy for people living in that [county].

The study was summarized by journalists at NPR, and the story takes the form of a conversation between two journalists. It reported:

...political affiliation is one factor shaping the life span of people around the country according to this new study. Researchers went back about 20 years. They analyzed a CDC database that collects death data on, essentially, every individual in the U.S. They linked this data with federal election data, looking specifically at how counties voted in presidential elections and also in state governor's races.

a) So far, the journalist has mentioned two major variables in this study. What are they?

b) Were these two variables measured or manipulated? 

Now here is some more detail: 

Now, overall, mortality rates in the U.S. had been declining in the early 2000s. People were generally living longer. But the researchers found some of those gains have faded away. And when they looked at areas losing ground, they found Republican counties were losing more. Bottom line, people in Republican-leaning counties appear to be more likely to die prematurely. 

In many correlational studies in psychology, the person is the unit of analysis (the unit of focus). But in this study, the county is the unit of analysis. So for each county, the researchers recorded the death rates and the voting patterns. 

c) Take a moment to sketch a scatterplot of the relationship described above. Put "average age at death" on one axis and "percentage of votes for Republicans" on the other. Should your scatterplot show a positive or a negative relationship? 

d) What does one dot in your scatterplot represent? 

The researchers used regression to rule out several possible alternative explanations for this relationship, as you'll read about in this next passage:

Now, it's not something as simple as Republican counties being older. The researchers adjusted for age. Nor is it just an urban-rural divide in the country. They found the mortality gap held up in suburban and urban counties that vote Republican, too.

Note that you can see the results of these analyses in the empirical journal article. 

e) What were the variables they controlled for in their analysis?

The rest of the journalist's discussion is about mediators--they discuss WHY there might be a link between Republican voting and death rates. Here's one example: 

[researcher Dr. Steven Woolf argues that] if you look at policies such as the expansion of Medicaid, access to health care ... tobacco control, gun legislation, drug addiction - a whole range of policies have an impact on health and mortality rates. Democratic states have supported more of these. But Republican states have gone the other direction.

Researcher Woolf mentioned the pandemic, as well:

States with Republican leaders were pushing back on COVID-19 vaccination and enforcement of public health policies. And what we basically had was a controlled experiment where, you know, some states adopted this proactive approach, other states didn't. And we had an outcome that could be measured within weeks. And we saw massive differences in death rates.

f) After reading the passage above, you should be able to sketch the different mediators they are discussing, using the following template:

   Republican voting pattern. ---> __________ --> Higher death rate

g) What other potential mediators could you propose for this link?

h) Are there potential third variables that you think the researchers might wish to control for in their regression analysis?