(Inside Science) -- A long-held theory in evolutionary psychology suggests that heterosexual women's attraction to certain types of men ebbs and flows with their menstrual cycles. But a recent study from the University of Southern California in Los Angeles casts doubt on that claim, causing friction between different schools of thought in psychology.
In the last 20 years, various evolutionary psychologists published studies demonstrating that women are more likely to find highly masculine, dominant men more attractive when they are ovulating. In these studies, masculinity is defined by various cues like vocal quality, facial structure, scent and musculature. These traits are thought to indicate genetic fitness, or a high volume of "good genes."
Additionally, researchers suggest that when women are ovulating and fertile, they are more tuned in to these cues for genetic fitness. In an evolutionary sense, this subconscious ability would help a woman choose a good mate with whom to produce a child.
Wendy Wood, a professor of psychology at the University of Southern California, aggregated data from 45 published studies and 13 unpublished studies on this topic. She and her colleagues then sifted through the data, looking for patterns in a meta-analysis -- a type of statistical analysis that looks for patterns across multiple studies. They published their findings in the journal Emotion Review.
"Overall we didn't find any effect of cycle phase on women's preferences," Wood said. Instead, they found that women are attracted to highly masculine men throughout their cycles.
A typical 28-day menstrual cycle is composed of various phases. About halfway through each cycle, surges in certain hormones cause one of the ovaries to release an egg and the cervix to open. These conditions, which last approximately six days, are just right for an egg to be fertilized if it comes into contact with sperm. Outside of that window, a woman is unlikely to become pregnant.
Wood said that among the studies she analyzed, those that found shifts in women's preferences used imprecise measures of fertility. Instead of testing for hormones that would indicate a woman's fertile phase, the researchers used a "counting" method. They would ask a woman when the first day of her last period was, and then count the number of days it had been since then to determine approximately what phase of her cycle she was in.
"Women don't remember necessarily what phase they're at, but also there's an unreliability in the cycles themselves," Wood said. Every woman's cycle is different. Few women consistently have a 28-day cycle, so some women might ovulate on day 10 whereas others ovulate on day 20.
The studies that found an effect usually had a fertile phase of nine to 10 days. The phase's length was determined by counting the number of days from the first day of each woman's period, determining each woman's fertile phase, and then averaging across the participants.
"The message is that women's biology doesn't necessarily drive particular social judgment, but we're all biological beings, that's a given," Wood said. "The idea that menstrual cycles will affect who women find attractive just doesn't hold up in the research literature."
Not everyone agrees with Wood's findings. Steven Gangestad, an evolutionary psychologist at the University of New Mexico in Albuquerque and author of many of the studies in Wood's report, took issue with her conclusions.
"They did find certain effects, but they dismissed those and explained them away," Gangestad said. He felt that Wood and her colleagues' analysis was too broad and that the patterns of attraction were drowned out in the noise of all the other data.
Gangestad pointed to another meta-analysis published recently in the journal Psychological Bulletin that used many of the same studies as Wood's but found a completely different result. Wood said that the other meta-analysis's lead author, Kelly Gildersleeve, was her student in a class on meta-analysis.
"This is not an uncommon phenomenon to find two meta-analyses that find different conclusions," said Christopher Schmid, a professor of biostatistics at Brown University in Providence, Rhode Island who specializes in meta-analysis.
Schmid only looked at Wood's study, but said that he is unsure how much it can or can't say about this theory in evolutionary psychology. Part of the problem is the nature of individual studies. Most of the time, the studies test just one or a few factors by having participants look at photos or listen to voice recordings.
"You're only using one of your five senses if you're looking at a picture," he said. "I think [Wood has] done as much as you can with these data, but there's a lot of conversion from one scale to another."
The existence of multiple conclusions though, highlights the importance of making sure that psychological research takes factors of both biology and culture into account.
"My view is that neither culture nor biology could win out here, instead I'm hoping that more adequate theoretical models that are based on both will be developed in the future," Wood said.
Cynthia McKelvey is a science writer based in Santa Cruz, California. She tweets @NotesofRanvier