So, if you work on identity, belonging, or equity—or if you’re just nosy and want the tea—I encourage you to read what Mary has to say (and check out the Science Advances paper!).
marycmurphy.substack.com?utm_source=n...
It made clear how much was riding on how we measure these experiences. Seven years of measurement work later, the Science Advances paper is, in a lot of ways, my response to that conversation. www.science.org/doi/full/10....
I remember ending that conversation in a good place—a greater understanding of each other. But it stayed with me.
...which, as Mary points out in her Substack, only emerges under a very specific set of conditions. I, however, was talking about the underlying psychological experience.
After taking a beat, I asked what they meant and it became clear we were talking past each other. They were talking narrowly about a consequence of threat—specifically, the link between stereotype threat and test performance...
So, to hear someone flat out say it wasn’t real was...something else.
It’s that sense, when you walk into a space, look around and take in the various cues of the setting, and wonder: Because of who I am, will others stereotype me here? Can I come to belong here? Can I be my full, authentic self here? Will I be harassed or physically unsafe here?
I remember pausing. Because what I meant by social identity threat was something I had experienced myself, and that I’d heard described firsthand by so many others.
For them it was probably one of a hundred conversations that two weeks. But I will remember this particular conversation forever, because when I mentioned studying social identity threat, they looked me dead in the eye and said: “Well, stereotype threat isn’t real.”
For those who haven’t been, it’s basically summer camp for us social psychology nerds. I was at a mixer chatting with another attendee—as you do—about research. [What do you study? Very interesting. What do you study? On and on.]
It’s hard to know exactly when a particular research idea starts. Things can percolate, semi-formed for a while before any sort of aha! moment. But one of the more clarifying moments for me was at SISPP 2019.
Full disclosure: in it, she mentions our recent social identity threat measurement paper in Science Advances (which I hope folks will take a look at!), so Mary’s post felt like a good opportunity to share the story of why I became so obsessed with measuring social identity threat in the first place.
Whether you know a lot or a little about stereotype threat research, it’s a fascinating read.
In it, she takes on some recent claims that stereotype threat is dead, drawing on the current state of the evidence and her own experience in the field, and does so in a way that feels like you’re having a one-on-one conversation with her.
My former PhD advisor, Mary Murphy, published a Substack this week that’s well worth your time: The Report of Stereotype Threat’s Demise Has Been Greatly Exaggerated.
Many thanks to my wonderful collaborators: @arianahdzcolm.bsky.social, @drdorainnegreen.bsky.social, Heidi Williams, @austinkuzdal.bsky.social, Juan Ospina, Courtney Moore, Gunjan Agarwal, Andrew Heckler, Jenny Crocker, Ken Fujita, @marycmurphy.bsky.social, and Steve Spencer!
We hope you'll give the paper a read and, if the SITC Inventory is useful to you, we'd love for you to use it!
Paper: doi.org/10.1126/scia...
Why does this matter? Knowing where threat concentrates—which groups, which settings, what forms—helps make the case that specific settings need to change, and points toward what targeted, context-level responses might look like.
And although no single study can establish universality, our findings collectively make a compelling case for it—documenting threat across gender, race, sexuality, age, weight, political orientation, religion, mental health status, and citizenship status.
In the field, women taking college physics reported greater gender-related threat than men overall, but especially in a male-dominated physics course.
In experiments, conservatives anticipated more threat in liberal-leaning group discussions and liberals more in conservative-leaning ones; and both Black and white participants anticipated more race-related threat when they were the sole representative of their racial group.
For example, in surveys, LGBQ+ folx reported greater sexuality-related threat than heterosexuals, peaking in unwelcoming settings (e.g., religious spaces) and lowest in welcoming ones (e.g., friend groups).
Along the way, we found compelling support for core tenets of social identity threat theory. Threat proved situational.
The result is the Social Identity Threat Concerns (SITC) Inventory—a 23-item measure validated across eight studies (N = 5,002).
So, we asked, how do you measure “a threat in the air”? We started by combing through open-ended descriptions of threat experiences (N = 1,286), identifying themes and building a taxonomy of the many forms threat takes.
Direct self-report measures existed too, but most were built for a single identity, context, or form of threat. That made them ill-suited for testing core theoretical claims—that threat is situational, multifaceted, and nearly universal.
But this approach led many to conflate the experience of threat with its downstream consequences—and to make the dubious assumption that if the consequences aren't there, the threat isn't either (or worse, that it 'isn't real').
Early research mainly measured outcomes theoretically linked to threat: diminished performance, physiological anxiety, domain avoidance. The logic was that if we see these effects, we can infer that threat is present.
Theory tells us it's (1) a vigilance state, (2) multifaceted, (3) situational, and (4) nearly universal. So how do you capture threat's many forms flexibly enough to apply across a nearly infinite set of identity-by-context combinations? It's been a longstanding challenge.
This paper is all about social identity threat—what Claude Steele famously called 'a threat in the air': ever-present in certain settings, but notoriously difficult to pin down.