New findings challenge assumptions about men’s reading habits

A longstanding belief in the publishing world suggests that men avoid reading fiction that centers on the lives of women. However, new research indicates that a protagonist’s gender has almost no impact on whether a man wants to continue reading a story. These findings appear in the Anthology of Computers and the Humanities.

The literary marketplace has historically skewed heavily toward men. For roughly two centuries, men wrote the majority of published novels. These books focused their narrative attention primarily on male characters.

That dynamic has shifted in recent years. Women now constitute the majority of published authors. In addition, women are now more likely to purchase and read books than men are.

This demographic change has sparked concern among some cultural commentators. There is an anxiety that literary fiction is becoming a pursuit exclusive to women. This worry often centers on the idea that boys and men are losing interest in reading as the representation of women increases.

Data from the industry shows a strong division between authors and readers based on gender. Men tend to read books written by men. Conversely, women tend to read books written by women.

Industry stakeholders often attribute this separation to a specific reader preference. They assume men are simply less willing to read books featuring women protagonists. This assumption suggests that publishers should release more stories centering on men to maximize their potential audience.

Federica Bologna, a doctoral student in information science at Cornell University, led a team to investigate this assumption. Co-authors included Ian Lundberg from the University of California, Los Angeles, and Matthew Wilkens from Cornell University. They noted that previous research on this topic was scarce.

Earlier studies on reader preferences often relied on small groups or interviews rather than large-scale data. Some of these smaller studies suggested that men prefer male protagonists. Others suggested that women were indifferent to character gender.

Bologna and her colleagues sought to determine if the gender of a character actually causes a reader to stop reading. They designed an experiment to isolate gender as a single variable. The team recruited approximately 3,000 participants living in the United States.

The participant pool was evenly split between men and women to ensure balanced data. The researchers excluded participants who identified as non-binary due to data limitations. The resulting sample size provided high statistical power for the analysis.

Participants read two short stories written specifically for the study. The researchers created original fiction to ensure no participant had seen the text before. One story focused on a character named Sam who goes hiking in the desert.

The second story depicted a character named Alex sketching in a coffee shop. The authors chose the names Sam and Alex because they are gender-neutral. This allowed the researchers to swap the genders of the characters without changing their names.

Crucially, the team randomized the pronouns used in each version of the stories. Half the participants read a version where Sam the hiker was a woman using “she/her” pronouns. In this version, Alex the artist was a man using “he/him” pronouns.

The other half of the participants read a version where the genders were swapped. For them, Sam was a man and Alex was a woman. This design ensured that the plot, setting, and dialogue remained identical for all readers.

Only the perceived gender of the main character changed between the groups. This approach is known as a vignette experiment. It allows researchers to attribute any difference in reader response directly to the specific variable they manipulated.

After reading the passages, participants had to answer comprehension questions. This step verified that they had actually read and understood the text. They were then asked to choose which of the two stories they would prefer to continue reading.

The researchers compared the probability of a reader selecting a story based on the protagonist’s gender. If the industry assumption were correct, men would be much less likely to choose the story when the protagonist was a woman. The results contradicted this prevailing wisdom.

When the protagonist was a woman, men chose the hiking story 76 percent of the time. When the protagonist was a man, men selected the hiking story 75 percent of the time. The statistical difference between these two numbers was effectively zero.

The presence of a female protagonist did not reduce the men’s desire to read the story. Being randomly assigned a female character increased the probability of a man choosing that story by only 0.8 percentage points. This result was not statistically distinguishable from having no effect at all.

Matthew Wilkens, an associate professor of information science, noted the clarity of the result. “This supposed preference among men for reading about men as characters just isn’t true. That doesn’t exist,” said Wilkens.

He emphasized that these findings challenge the anecdotes often cited in the publishing world. “That is contrary to the limited existing literature and contrary to widespread industry assumptions,” Wilkens added.

Women participants showed a different pattern than the men. They displayed a modest preference for stories featuring women. Women selected the hiking story 77 percent of the time when it featured a woman.

This probability dropped to 70 percent when the character was a man. The data suggests that while women leaned toward characters of their own gender, men remained indifferent. The gender of the character did not appear to be a deciding factor for male readers.

The authors acknowledged certain limitations in their experimental design. The study relied on just two specific short stories. It is possible that the genre of the story influences reader preferences in ways this experiment did not capture.

For instance, men might read more mysteries or thrillers. Those genres often feature male protagonists. If the study had used a different genre, the results might have differed.

Future research would need to randomize genre to see if that changes the outcome. Additionally, the use of unpublished fiction limits how well the study mimics real-world bookstores. In a bookstore, fame and marketing play a large role in what people choose.

However, using unpublished text provided strong internal validity. It prevented participants from recognizing the story or guessing the study’s intent. This ensures the responses were genuine reactions to the text itself.

Another limitation involved the demographics of the participants. The researchers excluded respondents with gender identities other than man or woman. This was necessary because they could not gather enough data on those groups to reach a statistical conclusion.

Bologna and her colleagues hope to include nonbinary readers in future work. Understanding how gender-nonconforming readers interact with character gender is a gap in the current science.

The study leaves open the question of why men predominantly read books by men. Since character gender is not the cause, other factors must be at play. The authors suggest that socialization or gendered expectations may influence reading habits.

Society may condition boys to view reading as a feminine activity. This could discourage them from reading at rates equal to girls. Alternatively, men may simply prefer the specific topics or writing styles found in books authored by men.

Despite these open questions, the study offers a clear message to publishers. The fear that writing about women will alienate male readers appears unfounded. Fiction editors need not reserve female protagonists for books marketed solely to women.

“Readers are pretty flexible,” Wilkens said. “Give them interesting stories, and they will want to read them.”

Bologna hopes this work will encourage the publishing industry to promote more books with a variety of girl and women characters. The team suggests that the industry creates a self-fulfilling prophecy by assuming men will not read about women. By breaking this cycle, publishers could offer a more diverse range of stories to all readers.

In future work, the researchers hope to explore whether these findings apply to other media. They question whether similar assumptions drive creators to avoid female protagonists in video games. If the same pattern holds, it would suggest that content creators across media are underestimating their male audience.

The study, “Causal Effect of Character Gender on Readers’ Preferences,” was authored by Federica Bologna, Ian Lundberg, and Matthew Wilkens.