This article originally appeared on VICE US.
You and your partner signed up to participate in a study about relationships—a cakewalk for a strong couple like the two of you, really. The prompt feels simple enough: All you have to do is list aspects of your partner that you don’t like. You take your pencil and write: Cranky when tired. You pause, chew your lip, tapping the pencil against the side of the desk until it comes to you. Can be self-righteous during arguments. Doesn’t do the dishes unless I ask.
You sit back in your chair and sneak a look at the clock. Plenty of time left, but you’re basically done. Except… your partner, behind you, is still writing. At a steady pace. A little fast, even. At the top of the experiment, the administrator told you you’d both received the same prompt. Which means…
You lean forward again, tightening your grip on the pencil, heat rising behind your ears. You write: Snores. Unenthusiastic about oral (giving). Hates reading. Performatively journals in public. Bad cook. Flaky. Unbeknownst to you, your partner has actually been told to list items in their home, so they are actually dashing off an inventory of rugs, lamps, and pillows, while you’re seething about a fight you had months ago about blanket hogging. And, depending on where you scored on an earlier questionnaire, researchers would likely have concluded this response stemmed from your low self-esteem, and that bringing this frailty into your relationship meant it was prone to fracturing over minor events like this, and thus, doomed.
At least, that’s what a new book pushing a “negativity is bad, positivity is good” thesis would like you to think. But at least one attempt to replicate the results of this study found that whether one partner has a meltdown in this scenario had no ties to self-esteem at all; it’s just a genuinely fucked up thing to do to a set of two people.
This study’s findings have been used to bolster the connections between poor self-image and poor relationship health, most recently in The Power of Bad: How the Negativity Effect Rules Us and How We Can Rule It, by John Tierney and Roy F. Baumeister, released on the last day of 2019 and excerpted in The Atlantic on Thursday. This experiment was first conducted in 2002 by psychologists John Holmes and Sandra Murray to support the theory that people with low self-esteem project their poor self-image onto their partners.
In the experiment, people whose survey responses indicated low self-esteem reacted negatively to the conditions above and added to their lists, while people whose survey answers indicated higher self-esteem were unbothered by their partner’s apparent novella of complaints and stayed firm in their initial attitudes.
“The insecure people were reacting needlessly, because in reality they were valued by their partners just as much as the secure people were,” the Power of Bad authors wrote. “But they projected their own self‑doubts into their partners’ minds. They assumed their partners would judge them as harshly as they judged themselves.”
It is extremely difficult to imagine anyone remaining cool in the scenario above, no matter how self-confident they may be, especially given the fact that it relies on outright deception. This suspicion (that nobody is immune to this level of psychological warfare) was confirmed thanks to an independent replication of the experiment in 2017. The subjects in the replicated experiment reacted negatively to the aspect listing ploy, but this time, researchers found almost no difference between “secure” and “insecure” participants, findings that were inconsistent with the initial research.
These results are not totally surprising given the fact that social psychological research is notoriously tough to replicate. When reached for comment, Baumeister, one of The Power of Bad’s co-authors, said he had not seen the study before citing Holmes’ and Murray’s original experiment.
“There are many such failures with various phenomena,” Baumeister told VICE via email. “But this one is more disturbing than most because they did get a significant difference on the manipulation check. (Most failures to replicate various things have no significant manipulation check, which means they were not really testing the hypothesis.)” That difficulty of replication renders findings like the one from Holmes’ and Murray’s 2002 study shaky ground to base anyone’s understanding of relationships—let alone a portion of a book—on.
Relationships aren’t all about splitting a burrata before dinner or posting pictures together at someone else’s wedding; but this guilt-inducing research, cherry-picked to strengthen an argument, is no harbinger of romantic failure. It would be perfectly reasonable for any of us to feel undermined in this scenario that’s specifically designed to make people feel undermined!
Sometimes, insecurity or negativity stems from a genuinely messed up situation rather than some internal failure of confidence—and whether they’re a social psychologist or the love of your life, reacting to that treatment accordingly doesn’t make you a bad person. It makes you a perceptive one. Any book that bends so far over backward to lay personal blame for a systemic problem bears further scrutiny.
Follow Katie Way on Twitter.