Verified by 
Key points
- Most psychological research findings are subtle, not headline-grabbing.
- Research findings in themselves have no meaning or wisdom.
- It is the interpretation and generalizability of findings which creates meaning.
- Media interpretations of research findings can do more harm than good.
Findings of psychological research now regularly escape the ivory towers of academe to permeate popular media. This results in misunderstandings, misinterpretations, and misquoting of research findings. It’s a common occurrence these days to hear arguments in which both sides cite the same data to support different, if not opposite conclusions.
The Meaning of Research Findings
Valid research confirms or dis-confirms probable relationships among variables. Research findings themselves have no other meaning and no wisdom. It is the interpretation and generalizability of the findings, applied to real-life contexts, which creates meaning.
While researchers work hard to reduce bias in design and methods, lay interpretations of findings are embedded in bias. Generalizability depends on delineation or definition of the variables under study (how the researchers operationalize them), as well as methods of measurement and the sample under observation. Yet methods, operational definitions, and samples almost never appear as context for citations in the media. It's quite usual for authors to use complex terms like "intimacy" in ways that go well beyond the operational definition in the research cited as authority.
Complexity is the enemy of statistically significant findings. The enormous pressure on researchers to produce findings of statistical significance leads some to oversimplify the delineation of variables, diluting real-life relevance.
Most psychological findings concern average outcomes in the group under study, which do not necessarily tell us anything about individuals. For example, consider the empirically valid, if awkward statement:
“Men are taller than women.”
It means that were you to select, at random, around 50 men and 50 women, the average height of the men would exceed the average height of the women, although there would be some tall women and short men in the respective groups. My colleague, Pat Love, and I use this example to start presentations. She’s 5-foot-10 and I’m barely 5-foot-6. She pats me on the head as we make the empirically-supported statement that men are taller than women.
Because random selection of representative samples is difficult, researchers try to reduce within-group variance to accentuate between-group differences. In our example, a hidden bias would guide the selection of average height men and women.
Beware of Media Claims
Consumer attention is the fickle god of media in the era of micro attention spans. Attention-seizing headlines find little or no support in the research cited within the article. Citations rarely go beyond the logical fallacy of appeal to authority, much less offer proof of claims.
When the popular media cites research, look for statements like:
“Research suggests that blank (findings) might apply to some people.”
Unless you're willing to look up cited studies and evaluate the methods employed and the evidence presented, and consider the acknowledged limitations of the research and calls for further research on the topic listed at the end of the study, turn a skeptical ear to statements like:
“Research shows (or proves) that…”
“Because this happened to you as a child, you will experience (blank) as an adult.”
“You feel this way because….”
“Your partner acts that way because….”
These offenses occur frequently on websites that try to outdo competitors by saying something new:
"A new study shows (proves).…”
Some interpretations become self-fulfilling prophecy for unsuspecting consumers. Common examples:
“Adverse childhood experience causes intimacy problems in adult relationships.”
“Your partner has this personality disorder if these symptoms are present.…”
See Explanations as Self-fulling Prophecy for a fuller discussion.
The only credibility a psychological explanation has for you is its ability to predict in your life and relationships. Form hypotheses from any explanation that appeals to you and test them.
Popular media misrepresents research—and does more harm than good—by oversimplifying complex psychological phenomena and conflating concepts:
“Love is about getting your needs met. Relationships are happier when partners get their needs met.”
I’m most sensitive to misinterpretations of research in my clinical specialty of resentment, anger, and abuse in relationships. For example, most studies of anger management interventions use self-report questionnaires instead of behavioral outcomes, such as recidivism rates. The fact that blame, denial, and avoidance are integral to problem anger calls into question the reliability of self-report. Partners who live with study subjects might slightly overestimate the frequency and intensity of anger incidents but not to the degree that the subjects will underestimate their own after enduring an anger-management class. Yet partners are almost never queried.
Similarly, studies that show benefits of fighting and expressions of anger, also measured with questionnaires, never consider the human loathing of pain without gain. Couples who hurt each other are predisposed to find some gain in their pain and report benefits they don't experience, at least not for very long. Such outcomes are reliable only if measured at different intervals over time, which, to my knowledge, has never occurred.
Rigorous peer-review in professional journals ameliorates some problems with findings. Authors and editors list limitations of the research and call for more research on the topics. Those caveats and limitations are rarely mentioned in media citations of studies. In fact, peer-review cannot temper the misinterpretations rampant in popular media. Many psychological studies either have non-representative samples, are designed inadvertently to attain specific findings, or achieve statistical significance through elaborate statistical manipulation, with little or no real-world importance. Some are tautological. If A and B are essentially the same, A does not predict B. Common examples are the robust correlation of love and affectionate behavior or racism and discriminating behavior. Distressingly few studies are (or will be) replicated. Replication of findings is key to scientific knowledge.
It's extremely difficult to know something and extremely easy to think we already do.
Research citations in media discussions of complex psychological phenomena support, at best, informed opinion, not scientific knowledge. While informed opinion can be valuable in negotiating life’s challenges, scientific pretensions are misleading and sometimes harmful.
Wisdom is never contained in psychological studies. It emerges from knowing which research findings help describe human nature or guide beneficial behavior or reliably predict outcomes.
