Is there any evidence to support the oft quoted belief that women who live together have synchronised menstrual cycles? Rachel Gilby investigates the possible role of pheromones and the research that might (but only might) offer some explanation for this curious phenomenon.
The human menstrual cycle is an infradian rhythm because it lasts for more then 24 hours. Menstruation itself occurs at the end of the cycle when the lining of the womb is shed as no eggs have been fertilised since the last menstruation. As part of the human menstrual cycle, oestrogen and progesterone, as well as other hormones, control the release of eggs from the ovary, the thickening of the lining of the uterus and then the menstruation. This cycle happens repetitively throughout a woman’s life from when she hits puberty until she reaches the menopause at around the age 50 and prepares the body for the possibility of carrying a child.
Much research has been done into how, in some cases of females all living together, there can be synchronisation of their menstrual cycle. The first notable piece of work is from Martha McClintock in 1971. Whilst at university she noticed how her and her roommates seemed to have a synchronised menstrual cycle and that this could not simply be a coincidence. She then looked at the idea that pheromones, biochemical substances that are released into the air, acting like external hormones, may play a role as they are chemical messengers that pass on messages from one member of a species to another. She took a number of female friends and dorm sharers and took pheromones from them and wiped them onto the upper lip of each other. The upper lip as it was assumed this would be the best place for them to be taken in through the nostrils and go into the body in a way that mimicked how it would do naturally in real life. She then repeated this process and found that eventually, a large proportion of women in the experiment now had synchronised menstrual cycles.
In a follow up ten year longitudinal study, in 1988, McClintock and Stern then looked at how pheromones from other women effected the menstrual cycles of women with infrequent, random, ovulation. They took 29 females between the ages of 20 and 35 with a history of irregular ovulation and gathered pheromones from 9 of them at different points in their cycles. They did this to see if they could lengthen or shorten the women’s cycles by placing pheromones from different stages of other women’s cycles on their upper lip. Their study showed that 68% responded to the pheromones with their cycles being lengthened or shortened depending on where in the cycles the pheromones they were given were from.
McClintock’s studies were both in natural settings making them high in ecological validity. The fact that the women’s lives were not altered in anyway other than by the addition of the independent variable (pheromones) and the dependent variable (menstrual cycle times) is a huge benefit as if they were put in a synthetic environment/situation this could have lead to synthetic results which are not true in real life situations. Another strength of the studies is that they could be easily replicated and that her second study continued to support and give evidence for what was found in her first study: that pheromones played a role.
As with many studies, there were also some weaknesses in McClintock’s studies. Firstly no control group was used such as looking at how much of her results were affected by the pheromones and how much was effected by the placebo effect and knowledge of the other women’s cycles. Being a quite naturalistic study that was performed over a long period of time, it was also hard for other factors to be controlled such as diet and stress levels, known to have an effect on the cycles too. The samples are also something McClintock’s studies can be criticised for, only taking into account a small amount of females with a small age range. Her original sample was hugely unrepresentative of the population, with only people in that area, of that class, with that type of lifestyle and that age range were used as they were all university students living together. The small sample size also meant that each synchronised person had a huge effect on results, how can we be sure that this wasn’t due to other causes or when their cycles were originally? McClintock’s studies also did not show significant enough results to suggest that the pheromones are the complete cause of the synchronisation of the human menstrual cycle.
Strassmann suggested that the results of the work done by McClintock may have been influenced by bias. McClintock and Stern admitted that they had made adjustments slightly to the data in order to take into account the effect of nasal congestion on the results. With the small sample size, this could have resulted in a big change in results and Strassmann responded to this by saying that ‘it would be useful to know what a priori criteria were employed in making such adjustments, and whether the data analysis part of the project was done blind’, suggesting that there may have been a lack in logical facts used in making the adjustments and that it was not done objectively, instead, the adjustments were done to influence the results and help prove the hypothesis.
The study has also been criticised for not showing cause and effect, only a correlation between the pheromones and menstrual cycles. There has been little investigation into other factors and if they have more or less of an effect than the pheromones. As with things such as heart disease, could there be a number of influencing factors?
To conclude, the research into the synchronisation of the human menstrual cycle may not be extensive enough to give a definite indication as to what causes the synchronisation. Although there are many strengths to McClintock’s work, there are also many weaknesses and the reliability of her data is also questionable.
The experiment took place in order to investigate the effects of previous events on perception, in this case it was seeing a picture which was either a group of animals or a group of people and then later when shown a picture participants had to describe what they had perceived. The null hypothesis for this experiment is that there will be no correlation between the pictures that the participants were originally given and the way in which they perceived the second image. The alternative hypothesis however is that there will be a difference depending on which images were seen by the participants. The experiment was used to investigate various theories about perception such as that from Vernon 1955 who described the ‘Perceptual Set’ and said that it worked in two ways. The first is where the perceiver has certain expectations, in this case due to the image already seen and therefore will focus their attention on particular aspects of sensory data. He calls this the selector. The second part is where the perceiver knows how to classify name and interpret certain data and therefore know what to draw from it, he calls this the perceiver. The main aim of this experiment however is to replicate the study of perception conducted by Bugelski and Alampay who investigated the importance of expectation in the perceptual set, they found that those who had previously been shown images of animals were more likely to see the stimulus as being a rat because they had preconceived expectations. The other aim is to understand the conventions for writing psychological investigations using a simple experiment in order to practice this.
In order to test this a sample involving the whole of the psychology class was taken making it an opportunity sample, some of the class were given stimulus cards depicting animals where the rest had cards with people on. This therefore used the experimental design of independent groups because there were two groups with different stimuli. The actual picture was then put upon the board and each participant was asked to write down what they had seen. In this case it was likely to be either a man or a rat depending on the stimulus they have previously been given. The results were then interpreted using chi square which is a statistical test that allows accuracy of results to be seen.
This is the formula for chi squared:
Here is an example of the stimuli given:
Participants were able to give consent because it was made clear that they didn’t have to participate and could remain anonymous if necessary. On the other hand informed consent could not be given because this would affect the results of the experiment as demand characteristics could be displayed making them less reliable. A debrief was carried out however which explained why the experiment had taken place meaning the experiment was fairly ethical due to these measures.
Here is a table of the raw results seen in the experiment:
Here is the table of the raw results inputted into the chi squared equation:
The results could be described using Gregory’s theory of indirect perception which states that people base their perceptions on prior knowledge and past experiences hence why the participants used their previous knowledge of the stimulus to base their perception upon. Gregory says that perception occurs as a result of hypothesis testing where the brain attempts to guess and process the image based on information previously stored in long-term memory. Here in these results however there appears to have been a fault in the perception which he would explain to be due to a faulty hypothesis hence the differing perceptions although there is still a weak correlation. So to conclude, although using chi squared the results appear to be insignificant, we can see results which begin to prove the alternative hypothesis that perception is dependent upon the stimulus seen but they are not consistent enough to provide a reliable conclusion.