Would babies benefit if their families received extra cash every month? Photo: Cavan Images/Alamy Stock Photo

When psychologists and sociologists take a big-picture view of early childhood development, they almost always notice a correlation between income and child behavior. Specifically, children being raised in poverty tend to perform worse than higher-income children on language development, executive function, socio-emotional development, and even brain activity. These facts are based on correlational research (where both income and child outcomes are measured variables). What would happen if we could experimentally test the effect of extra income? We could then test whether extra income causes child outcomes to improve later on.  That’s the goal of a major study called Baby’s First Years (BFY). Here’s a summary of the project:

NBER (May 2025) posted new findings from the Baby’s First Years RCT, studying the effects of providing unconditional cash transfers to low-income, new mothers in the US. This high-quality RCT found no impacts on any of the study’s primary child development outcomes at the four-year follow-up.

And:

Developmental differences between children growing up in poverty and their higher-income peers are frequently reported. However, the extent to which such differences are caused by differences in family income is unclear. To study the causal role of income on children’s development, the Baby’s First Years randomized control trial provided families with monthly unconditional cash transfers. One thousand racially and ethnically diverse mothers with incomes below the U.S. federal poverty line were recruited from postpartum wards in 2018-19, and randomized to receive either $333/month or $20/month for the first several years of their children’s lives. After the first four years of the intervention (n=891), we find no statistically significant impacts of the cash transfers on four preregistered primary outcomes (language, executive function, social-emotional problems, and high-frequency brain activity) nor on three secondary outcomes (visual processing/spatial perception, pre-literacy, maternal reports of developmental diagnoses).  Possible explanations for these results are discussed.

Questions

1. This was an experiment. What was the manipulated variable, and what were its levels?
2. Why did the researchers randomly assign families to the two conditions? What does this allow them to do? (Use the terms “third variable” or “internal validity” in your response)
3. There were four major dependent variables. What were they?
4. Sketch a bar graph of one of the results of the study.
5. Here’s the full report from the Baby’s First Years Scroll to Figure 2, on page 41, and study the interval plot you see there. What does one of the squares mean? What does the vertical dashed line indicate? What does one of the horizontal bars mean? All of these interval plots include the dashed line….what does this mean, and how is it related to the summary of the study?
6. Does this study support the claim that “providing low-income families with an extra $4000 per year causes their children have better developmental outcomes? Apply the three causal criteria of covariance, temporal precedence, and internal validity.
7. Chapter 11 concerns null effects. First, why does this study show a null effect? Second, I would argue that this study’s design helps us trust the null finding as “real” and not due to methodological weaknesses. Review the concepts in Table 11.3 and apply at least one of them to this study.

Here’s an important detail: An early set of data from the BFY study did show an improvement; specifically, at one year of age, babies in the extra cash group were showing improved brain activity (here’s a  description of that result). But in this 4-year follow-up, there was no difference between the two cash groups on brain activity.

This is a joint blog post by two W.W. Norton & Co. authors (and pals): Beth Morling and Jess Hartnett.  We first “met” through our blogs and social media years ago (Beth's blog on teaching RM, which you are reading now, and Jess's blog on teaching stats, notawfulandboring, here). Our friendship love language is DMing each other with ideas for the other’s blog posts. Recently, we thought of each other when we came across a story out of New Orleans. It illustrates how it is never too late for science to self-correct; in this case, to atone for human rights violations.

New Orleans mayor LaToya Cantrell speaks at the memorial service honoring the people whose remains were returned. Photo Courtesy Jacob Cochran/Dillard University

In the 1880s, the skulls of 19 African Americans, all of whom had died in a New Orleans hospital, were sent to a researcher in Germany. Recently, they were returned to New Orleans, LA. This story was covered by New Orleans NPR affiliate WWNO:

“Last weekend, Dillard University held a memorial service and jazz funeral to honor 19 Black New Orleanians whose skulls were wrongfully taken from Charity Hospital and sent to Germany in the 1880s.

The ceremony shed light on the legacy of racist pseudoscience. Dr. Henry Schmidt, a New Orleans physician, is believed to have given the skulls to a German researcher studying phrenology, a discredited pseudoscience that falsely claimed skull shape could determine racial characteristics.”

Presumably, Schmidt’s study was being conducted in a tradition of race “science.” During the 19th and early 20th centuries, many psychologists attempted to argue that brain mass, skull shapes, and cognitive ability are superior in Whites (for psychological reviews see Thomas et al, 2023;  Winston, 2020). 

Here is a link to the NPR article, as well as a link to Dillard University’s information about the repatriation of the remains.

Beth and Jess came up with three ideas for using this example in statistics and research methods classes. 

1. Use the story to teach ethics (Chapters 4 and 14 in Beth’s book).
   This use of Black Americans’ skulls in research would be a strong introduction to research ethics. A good place to start would be to help students review and apply the Three Core Principles (in the U.S, these are known as the Belmont Principles). Specifically, after reading the synopsis, students could tackle the following questions:
   
   a) The principle of Respect for Persons refers to the intrinsic worth and dignity of all people. 
   i) Come up with two questions you could ask about the study to decide the extent to which the 1870 Leipzig study (which collected skulls from around the world, not just New Orleans) might have met or violated this principle.
   ii) The principle of Respect for Persons specifically obligates researchers to obtain informed consent from research participants. Based on your reading of this story, do we have enough information to know if the Black Americans whose skulls were part of this study gave their consent? What would you want to know?
   
   b) The principle of Concern for Welfare, or Beneficence, directs researchers to maximize benefits to people and protect them from harm. It addresses people’s quality of life, including health, privacy, and community membership. This principle also asks people to consider the social importance of the research question.  The Canadian ethics code specifically includes the protection of human “biological materials” here. What aspects of the Leipzig study, if any, met this principle? Which did not?
   
   c) The principle of Justice is concerned with societal fairness, including whether one social group bears a disproportionate burden of research participation. What aspects of the Leipzig study, if any, met this principle? Which did not? (Hint: You might consider signs that the researchers took advantage of low-income Black Americans specifically.)
   
   d) The Three Core Principles are a valuable framework for analyzing whether a study was conducted ethically. But they might not address one of the most egregious issues. Specifically, the Principles don’t seem to capture the racist intent of the original research question. Or do they? Where in the Three Core Principles might we address whether a study is being conducted in good faith?
   (In Beth’s book (Chapter 14) you can read about additional examples of racist research questions.) 
   
   
2. Use the story to teach the scientific method (Chapters 1 and 2 in Beth’s book, Chapter 5 in Jess's book).
   Consider this quote from an interview with Dillard University history professor Dr. Eva Baham: “I often reference Thomas Jefferson, who also looked at skulls to determine personality, to determine intellect. There was no hypothesis. There was a conclusion first. And then the research was done to prove themselves right.”
   
   f) Reflect on the quote from Dr. Baham and reread about Merton’s scientific norms in Chapter 1 (Morling, Table 1.1). Which scientific norms is Baham's quote addressing?
   
   g) Look up a couple of definitions of the term pseudoscience. Is Dr. Baham suggesting that the Leipzig study was pseudoscience? Make some connections between the two. 
   
   h) In this quote, Baham seems to be accusing Thomas Jefferson of confirmation bias (Morling, Chapter 2). Why?
   
   i) How does this quote apply to the logic behind Null Hypothesis Significance Testing (NHST, covered in Chapter 5 of Jess’s book)? How does Thomas Jefferson’s logic (described by Baham) run counter to NHST’s emphasis on providing evidence to reject the null hypothesis? Specifically, does it seem like Jefferson was concerned with rejecting the null, or proving the alternative?   (Bonus: How does the NHST process help prevent confirmation bias?)
   
   
3. Use the story to address self-correction–at multiple levels. How does science correct itself? (Chapter 14 in Beth's book, Chapter 6 in Jess's book)
   When we think about self-correction in science, we typically think of changes in the standards surrounding research methodology and statistics. Self-correction has led to new practices such as research pre-registration, transparent data sharing, and reporting effect sizes alongside p-values (Hartnett, Chapter 6).
   
   The story of Dillard University is a unique approach to self-correcting science because it extends it to ethics. Certainly, the scientific community has long repudiated the phrenology/eugenics research for which the skulls were used. But Dillard University took correction a step further by respectfully handling the remains used in this “research”. Read the NPR piece and visit the Dillard website to learn more about all of the work that went into the repatriation. How did Dillard try to restore dignity to the people whose remains were used without their consent?  

Are you planning to try one of these teaching ideas in your class? Let us know how it goes!

Were all the variables in this study measured? Were any manipulated?
Photo: Caftor/Shutterstock

Psychology Today columnist Sebastian Ocklenburg summarized a recent study under the headline, The Dark Side of Emojis: A Surprising Link to Narcissism. Let's take a look.

It's a recent correlational study that tested the link between various personality traits and emoji use. (The empirical article can be found here.) 

The columnist provides a straightforward summary of the study's method here:

To clarify this question, a new study, just published in the scientific journal Current Psychology, scientifically investigated the link between gender differences in emoji use and different personality variables (Kennison and co-workers, 2025). In the study, the research team, led by Shelia M. Kennison of Oklahoma State University, gathered data from 285 undergraduate students (135 male students, 145 female students, and 5 students with other genders). First, the scientists gathered information on the volunteers’ emoji use. The volunteers had to rate the frequency at which they used 40 different commonly used emojis when communicating. They were also asked how often they used emojis when writing text messages, social media posts, and replies to social media posts of other people.

Moreover, the scientists assessed different aspects of personality in the volunteers using questionnaires. These included the so-called Big Five (openness, extraversion, agreeableness, conscientiousness, and neuroticism), the Dark Triad (narcissism, Machiavellianism, and psychopathy), and sensation-seeking.

a) The headline of the column was "The Dark Side of Emojis: A Surprising Link to Narcissism.". Is this a frequency, association, or causal claim?

b) Based on what you've read so far, classify the variables in this study (hint: I've counted 10 variables). 

c) Was this study correlational or experimental? 

Now let's read about the study's results. I'll present three findings: 

...First, the results indicated that women used both positive and negative emojis more often than men did based on their responses about their usage of the 40 common emojis. Women also indicated that they used emojis more often than men when writing text messages, social media posts, and replies to social media posts of other people. 

...higher emoji use was linked to higher narcissism in both women and men, with highly narcissistic women showing the most emoji use....

In terms of the Big Five, both men and women with higher extraversion reported more emoji use than introverted people. 

d) Take a look at the three results above. Select two you'd like to sketch a graph of. Then set up the axes for each graph. Remember to label your axes mindfully, and think about whether a bar graph or a scatterplot is the best graphing approach. 

e) Some of the results from this study showed moderators, and the moderator was gender. Here is one example of a moderator patterns in the results:

For men only, higher Machiavellianism was linked to more emoji use. 

Sketch a graph of this moderator. Hint: Make a scatterplot (what will go on the x and y axes?), using one color or shape for men dots and another color/shape for women dots. 

f) Here is another example of a moderator pattern:

For men only, it was also found that people with high neuroticism (indicating mood instability and difficulties dealing with negative emotions) specifically used more negative emojis than other people. 

Sketch a graph of this moderator. Hint: Make a scatterplot (what will go on the x and y axes?), using one color or shape for men dots and another color/shape for women dots.