Prejudice Shapes Interpretation
When I read psychology articles, whether in scientific journals or in the news, I always keep my ear tuned for bias. Although generally (probably) unintentional, biases often creep into how we interpret data. There are many books about the misuse of data (examples: How to Lie with Statistics, How to Use (and Misuse) Statistics, Statistics Done Wrong). It is shockingly easy to present accurate data in a way that totally misleads the reader about what it means. When it comes to psychology, there are multiple pervasive, fundamental, unscientific assumptions that reporters make when interpreting data to try to understand the human experience (See Psychology’s Misuse of Statistics and Persistent Dismissal of its Critics).
Science is Storytelling
Unless you are perhaps a robot, data by itself is boring. Reading the output of a statistical analysis, without understanding what it is saying, is sleep-inducing. Data only cohere (and become exciting!) in the presence of a story. The story brings a perspective on the data that binds it all together and turns a set of numbers into meanings. In the science world, these data-informed stories are first called hypotheses, and then as supporting evidence accumulates, they can become theories. You can see that storytelling in science is not only about effective communication. Storytelling is also the very essence of science itself. Who are we? Where do we come from? Why do we do the things we do? What is our place in the cosmos?
Storytelling is Subjective
One of the most important ways of approaching storytelling that I learned about in college was Derrida’s system of deconstructionism. He used it primarily for literary analysis, and really it can be applied to all stories and sets of ideas, including scientific ones. Essentially, Derrida recognized that all stories come with assumptions and biases. You have to start somewhere: to tell a story about heroism, for example, you have to start with assumptions about what heroism is and by whom, when, and how it can be demonstrated.
For example, I read, with horror, the depictions of heroism in The Song of the Nibelungs, the German epic. I don’t find mob-style violence heroic. At the same time, I did recognize heroism in The Epic of Gilgamesh. Ultimately, this is due in part to the assumptions I have around heroism. My assumptions align better with the those of the ancient Sumerian writer than those of the medieval German writer. When you pull at the threads of assumption that form what it means to be heroic, you start to recognize how contextually bound our morals can be. They are subjective.
If Science is storytelling, and storytelling is subjective, then… science is subjective?
Yes! Because theories are central to science, and theories are stories, science is inevitably subjective. Numbers may not lie, but everything that went into collecting and interpreting those numbers can. How did you make measurements? What did you decide to measure? What did you leave out? How did you fit those measurements together to come to a statistical finding? How did you incorporate those findings into other findings?
Scientists try to make their findings more objective by following standards. Assessment tools, statistical tools, diagnostic tools, treatment protocols. These are all standards psychologists use to try to be more objective. However, each of these tools comes with assumptions, based on other sets of standards. And those standards follow another set. This is reasonable, and in many ways, necessary. Really, there’s no other way to put together coherent stories in science.
And yet, this series of nested assumptions hits into a statistical validity issue: compounding error. Each assumption, by itself, may only bring a little error. But each layer is exponentially magnified by the one below. Until you have a scientific community that could not understand Copernicus displacing the Earth from the center of the universe.
Those scientists weren’t bad scientists. They had just failed to adequately question their assumptions, leading to a model of the universe that was wildly off what is now our assumed reality.
Assumptions in Psychology Reporting: Example Articles
What does this deconstruction of assumptions look like in psychology? I’ll briefly run through some articles of psychology reporting to show what I mean.
Neanderthal DNA and autism
Here’s a science news article that caught my eye, entitled, “Neanderthal DNA is more common in people with autism,” by Earth.com staff writer Rodielon Putol (posted 6/20/24). Our modern tools for splicing, modifying, and analyzing genetic nucleotides, such as CRISPR, are extraordinary. Studies like this enable us to zoom in on minute details, while still holding onto the broader picture, both now and across evolutionary history.
However, the reporting contains some worrying biases. They use the term “ancient” DNA. As if there were any other kind? The four DNA nucleotides have been around for billions of years. Certainly, how they get used and combined has changed throughout evolution, but just because a genetic unit was originally formed millions of years ago does not make it any less modern. Hands have been around for tens of millions of years, but don’t tell a pianist or basketball player that they are “ancient.” Context matters.
There is a reasonable desire to connect to our history. And yet this can create an unscientific distinction. The article implies that this non-homo-sapiens DNA was passed along, like a time bomb waiting to go off repeatedly in the present, as part of the constitution of some people, such as people with autism. The writer makes multiple assumptions here: us versus them; that more recent is somehow better; and that we can separate genetics from environment. None of these assumptions are valid.
Countering assumptions through other perspectives
In fact, from a genetic standpoint, we are contemporaries. Neanderthals appeared approximately half a million years ago, and then disappeared approximately 34000 years ago. In the low thousands. For genetic evolution, this isn’t even the blink of an eye – it’s the first hint of the blink of an eye. What makes us more modern than Neanderthals is not our genetics. It’s the externalized systems we have set up through civilization, which can evolve much faster than genetics. Libraries and schools. Bridges and tunnels. Tools and methods for agriculture. Weapons, airplanes. The internet. Try erasing all environmental knowledge of all of that, and we’ll see how “modern” we are then.
Consequences of this assumption are not minor. Suggestions of ‘older’ genetics fit neatly into Social Darwinist ideas. In Social Darwinism, evolution is ultimately not about fit but about improvement. Newer adaptations make better creatures. Humans aren’t just better adapted to the environment than, let’s say, chimpanzees: humans are simply better. This logic continues, Neanderthals may be better than chimpanzees, but we’re better than them. And then, if you have more of this lesser constitution in you, then you’re lesser. This is the basis for a term that is fortunately rarely heard anymore: the derogatory term, ‘throwback.’ People diagnosed with autism already have so many systemic barriers: these biases just reinforce the stigma that creates the barriers.
Article: Genetics and biological determinism
Sciencealert.com recently reviewed research on copy number variations as they pertain to people diagnosed with schizophrenia. The report, by Clare Watson (accessed 2/22/25), suggests in the title that “The Seeds of Schizophrenia May Be Planted in the Earliest Moments of Life.” The basis of the research is linking genetic variation to clinical outcomes. Just like in the Neanderthal study above, there have been a lot of these studies since we’ve had the technology to do them. But both the research itself, and the reporting, come with many assumptions.
One biased assumption is the use of terms like “error” to describe copy number variations. Descriptions use the word “error” a lot, but this is a misleading mistake. If we were to eliminate “error” from evolution, we wouldn’t be here. We’d still be in the primordial ooze. Likewise, if bacteria and viruses could not change, we would have eradicated most disease long ago. Mutations are a fundamental, necessary part of evolution. Change is not error. It may not always bring desirable outcomes. But desired outcomes do not come without it.
Here’s another biased assumption. The review of the paper states, “Research also suggests four out of five cases of schizophrenia can be traced back to genes inherited from the child’s parents, with more recent studies adding to the list of known genetic links. Yet that still leaves roughly 20 percent of cases without an obvious basis in inheritance.” What is the message here? This seems to correspond to the false idea that psychosocial challenges are “brain disease” (see my post on What is Mental Health for a discussion of this). In truth, 100% of everything we do has a basis in inheritance. You name it, researchers can find a genetic link. So why is “schizophrenia” getting a special call-out? The call-out implies that the experiences of schizophrenia are somehow ‘more’ biological, are carved into genetics, and inevitable. This is a eugenics idea, and it is untrue.
Biased components lead to biased conclusions
All of this leads back to the final bias, in what conclusion they reach. The research paper itself, reasonably enough, concludes that the two copy number variations “may contribute a small but significant part of the genetic architecture of schizophrenia.” I take issue with using a construct such as schizophrenia in the first place, but beyond that, this is indeed evidence that these CNVs, for some people, are part of the genetic architecture of their experiences. That’s all. This is a far cry from suggesting that “The Seeds of Schizophrenia May Be Planted in The Earliest Moments of Life.” Even seeds are not equivalent to the eventual plant, but the only inputs needed are sun, air, and water for the plant to germinate. No amount of sun, air, and water will turn a splice of DNA into schizophrenia.
Assumptions in Psychological Reporting: Trade journals
Unfortunately, general public writings are not the only ones to hold unscientific biases: they are found in trade publications, as well.
We will explore these next time…