Social Media “Science” as Digital Confirmation Bias

Social Media has lots of potential to engage us and connect us to information, people and ideas. How do we manage the information that is presented to us as “scientific”? How do we ensure that we are properly critical of how information, positioned as knowledge, contributes to how we think the world already functions? How do we know we are not participating in “Digital Confirmation Bias”? Social Media is meant for fast consumption. Fast response. Quick integration. It can push us to do additional research. Ask more questions. Seek out more knowledge. But algorithms are unconcerned with whether we do or not. Algorithms are concerned with financial bottom-lines. Let’s use an example to demonstrate how this works. This is not meant to be a direct criticism of the posting entity nor of the original research project itself (those are other conversations). More importantly it is a critique of ourselves as we participate in consuming digital “knowledge”.

A posting by “@goodneuroscience” on Instagram about a University of Wisconsin research project claiming “a mother’s voice doesn’t just comfort, it lowers her daughter’s stress hormones within minutes. Studies show hearing mom speak cuts cortisol and boosts oxytocin, just like a hug.” @neuroscience claims it is about “Brain Health | Neuroplasticity | Neurobiology.  Helping you understand your brain (without the jargon). Learn what your brain wishes you knew. The Overthinking Toolkit”. Not in itself problematic though Paolo Freire would argue about the omission of “jargon” (yet another discussion).

Let us look first at only the post itself and its potential impact on the casual social media consumer. What does this post “say”? What does this post “claim”? What does this post “teach”? What does this post “reaffirm”? What does this post “assume”? What does this post “omit”?

What does this post “say”?

It says, mothers have voices that impact their daughters’ stress. And that impact is positive. Who doesn’t want to be less stressed? And that this impact is relatively immediate. Who doesn’t want fast and quick relief from stress? This is confirmed by “studies”, an implicit connection that learned people, often named scientists, confirm this information. If it was studied, then research was done, the scientific method was employed, it was vetted by peer review and deemed valuable information that the world would benefit from. And finally, it is confirmed with a “hug”. Hugs are great. Who doesn’t like a hug? (Actually, lots of people.) Mother’s hugging their daughter’s is a good thing right?

What does this post “claim”?

This post “claims” that all of this information from this particular study is important. Or else why post it? It claims that mothers are an important category of person. It claims that the relationship between mothers and daughters are critical to life experiences of the daughters. It claims that mothers hold an important “gift” to bestow on their female children, just through their voice.

What does this post “teach”?

This post teaches us that stress is a negative force in our lives created by elevated levels of ‘cortisol’ that must be combatted with ‘oxytocin’. These “jargon” terms seem important because we probably don’t really know what they are or what they mean. They sound important and technical (unlike hug), so their very names alert us to their importance. Especially after claims that jargon was meant to be avoided so if it is included then it is critical that we pay attention (even without learning more about what these terms allude too). But it is just enough jargon to confirm that this must be “scientific” and therefore we presume we have learned something just from their presence in the text.

What does this post “reaffirm”?

This post reaffirms that “mothers” are very important. Very important to their daughters (we will deal with sons later) especially. Commonsense understandings of motherhood tells us that mothers are the most important individual connected to child development. “Mothers” implies biological mothers and their biological offspring. So, it reaffirms that biology is more vital over care. Nature not nurture. We see these patterns reinforced in decisions about childcare, and legal claims to child custody when parents divorce. It is reinforced in popular culture. It is reinforced in literature on parenting. Often with opening statements claiming the mother is the most important in the child’s development. It affects social roles in relation to responsibilities to parenting. This became apparent in COVID lockdown in situations where parenting responsibilities fell more harshly on wives versus husbands even when the wives were making more money in the household, were in high level positions as faculty or leadership in corporations.

What does this post “assume”?

This post assumes that only mothers can provide this relief. This post assumes it only impacts female children (we will deal with trans/intersex/nonbinary later). This post assumes sons do not matter or at the very least are not impacted by their mother’s voices. More importantly it assumes that all mothers are “good” at being mothers. That all mothers are invested in helping their daughters feel better. It assumes that being a mother is so natural they have a direct impact on chemical processes in their daughter’s bodies without physically coming in contact with them (i.e. not a hug). That’s almost magical in terms of powers. If this is true, then maybe we might employ mothers to speak in order to diffuse political and social tensions to avoid conflict globally. Are all mother’s caring and nurturing? It assumes that all women are the same fundamentally at a biological level. It assumes that daughters will call their mothers when they are stressed above anyone else - another parent, sibling, friend, partner, spouse, or any other caring adult they might have in their life. It also assumes that mothers don’t have to put any work into being a “good” parent. It must be natural i.e., what they are meant to do.

Finally, what does this post “omit”?

This post omits fathers, sons, trans, intersex, and non-binary children. It omits adoptive parents. At minimum it relegates those types of families as less important by implicitly asserting that biology is at the core of this experience. It omits queer parents (what happens if you have two dads? SOL?) It omits all of the science that shows that nurture is vitally important to the development and life experiences of children. It omits all of the other potential adults, peers in the life of this person who may be more comforting. I am not even denying that a comforting voice could have an impact on chemical processes in a person’s body. After a quarter century of advising and teaching college students on how to deal with their parents about getting an unsatisfactory grade or that they are choosing anthropology instead of accounting as their major, I could visibly watch students with whom I had no biological connection (and I am not a mom) become less stressed. They were often most stressed because they had to tell their mom they got a B. This post omits science fundamentally. In many research spaces this would be labeled “pseudoscience”. For me that means it bears the trappings of science. Familiar enough to trick us into thinking this is “real” science” but in fact is far from what would count as science because it only reasserts biased assumptions about gender and sex.

The post also omits, most significantly, the demographics of the actual people these assertions are based on. Any good quantitative researcher will tell you that one of the biggest burdens of quantitative research is confirmation bias. The results of the research reflect more about the researchers than those being researched. It is well understood that in the recruitment of participants for research one must be extremely intentional in ensuring that the participants are not a mere reflection of the researchers in terms of various demographics (i.e. race, class, gender, sex, religion, education, nationality, language etc.). In recruitment of participants researchers are often biased (implicitly and explicitly) in who the people they approach (familiarity and comfortability being major factors) and the strategies of recruitment (where is the research advertised? where is the research held?) which then result in a population that can be more similar to the researchers. For this particular example, who were these mothers and daughters? When and where was the research conducted? (middle of the day during the work week? weekend? evening? was it only advertised at the university where the P.I.’s worked?). It would be safe to assume that under minimal restrictions (location and time of day) working class mothers would not be able to participate. Much less low-income mothers who might be working 50-60hrs a week needing transportation to and from work to find additional time in the day to travel to a university campus to participate in a research project. Were the participants compensated? Paid? Given meals? Reimbursement for travel expenses? Would mothers who are abusive choose to participate in a study like this, giving their daughter an opportunity to share what their lives were actually like to outsiders? Highly unlikely. And highly unlikely that these mothers would be a source of comfort after a poor performance on a test. All of these may have been accounted for in the actual study (again that would be an entirely different review) but for the casual social media surfer who glances at the first slide and “hearts” it, what influence does such a “headline” have on them as they move on to the next post?

This response is anything but a social media post. It was not quick. It was not quick to consume. It was lengthy. It was not specifically reaffirming and hopefully left the reader with more questions. It challenges confirmation bias. This is not meant to be a statement to avoid social media or social media that focuses on scientific, or others forms of knowledge building. Rather, like in any situation connected to knowledge production, we must be intentional in our absorption of knowledge. This is more a sample of how one in all scenarios of knowledge production should and can determine what one is meant to learn and understand. If you are thinking about your own processes of knowledge production, or now realizing maybe you have not done a meta-analysis of your research processes then hopefully this has been helpful.