**Principal Investigator(s):**
**Mahzarin Banaji**
Harvard University
Email: [mahzarin_banaji@harvard.edu][1]
Home page: [http://www.people.fas.harvard.edu/~banaji/][2]
**Sample size**: 3010
**Field period**: 03/12/2019-10/25/2019
**Abstract**
“A father and his son are in a car accident. The father dies on the spot. The son, badly injured, is taken to the hospital. In the ER, the attending surgeon looks at the boy and says, 'I can't operate on this boy. He's my son!' How could this be?”
This now famous riddle was first brought into the living rooms of Americans in an episode of the 1970s comedic TV series, All in the Family. Yet, even fifty years later, few reach the obvious yet illusive answer: the surgeon is the boy’s mother. Indeed, despite making strides towards equality over the past forty years, gender parity within medicine remains an elusive goal. Women constitute half of medical school graduates, yet 35 percent of practicing physicians and only 19.2 percent of surgeons. One possible cause is societal stereotypes, which like “pictures in our heads” connect men more than women to medicine. We presented the classic “son” version of the surgeon riddle alongside two variations (“daughter”, “child”) to investigate the presence, magnitude, and psychological and demographic correlates of the surgeon=male stereotype in the United States.
**Hypotheses**
Research Question: What proportion of Americans will realize that the surgeon could be the boy’s mother?
Research Question: Will changing the gender from “son” (classic riddle) to “daughter” or “child” increase the incidence of mother responses?
Research Question: Are certain demographics (e.g., women) more likely to provide mother responses?
Research Question: Are mother responses predicted by statistical information (e.g., men constitute 80% of surgeons in the United States; frequency-based hypothesis), or on a belief about the essence of the social category gender (i.e., “men are naturally more suited to be surgeons”; essentialist hypothesis)?
Research Question (Exploratory): Does having a close family member (e.g., mother) or relative (e.g., cousin) who is a female surgeon or doctor increase the likelihood a person will provide a mother response?
**Experimental Manipulations**
Participants were assigned one of three versions of the riddle in which the gender (“son”, “daughter”, “child”) of the child was manipulated. Additionally, the order of the self-report questions, as well as the question anchors (e.g., “MEN are more…” vs. “WOMEN are more…”) were counterbalanced.
**Outcomes**
Participants’ answer to the surgeon riddle (variable: B1) was treated as the main outcome variable. Participants’ answers were dichotomized such that responses that included “mother” as a possibility were assigned a value of 1, and all other responses (e.g., “Step-Father”, “Priest”) were assigned a value of 0. This dichotomized value is hereafter referred to as “mother response” or “mother inference.” Importantly, as we were interested in participants’ spontaneous reactions to the riddle, only participants who reported having no prior exposure to the riddle (variable: B2) were included in the results reported below.
Additionally, we captured participants responses to the following three questions:
a) What do you think the gender breakdown of practicing surgeons in the United States is? (100% Male/0% Female to 0% Male/100% Female, in increments of ten percent);
b) Which of the following statements do you most agree with? (“Men are naturally more suited to be surgeons than women”; “Men and women are naturally equally suited to be surgeons.”; “Women are naturally more suited to be surgeons than men”);
c) Are you or any of your close family members doctors or surgeons? [Select all that apply]
**Summary of Results**
We presented the classic “son” version alongside two variations (“daughter”, “child”) to investigate the presence, magnitude, and psychological and demographic correlates of the surgeon=male stereotype in the United States. Strikingly, only 18% of Americans naïve to the riddle reported that the surgeon could be the mother. However, changing gender from “son” to “daughter” or “child” increased mother responses to 23% and 40%, respectively, suggesting that gender-fair language can alleviate the expression of bias. Further, we found evidence to suggest that women, younger individuals, and participants self-identifying as White were more likely to provide mother responses that their counterparts. Finally, we examined the competing frequency-based and essentialist hypotheses, i.e., whether mother responses were predicted by assumptions about the gender distributions within surgery or beliefs about whether men or women are more “naturally suited” to be surgeons. Importantly, using Bayesian posterior models we find conclusive evidence that essentialist beliefs but not participants’ estimated gender breakdown of surgeons in the United States predicted mother responses.
[1]: mailto:mahzarin_banaji@harvard.edu
[2]: http://www.people.fas.harvard.edu/~banaji/