Cadbury’s assorted biscuits were a real treat when I was growing up. We usually had them when someone gave us a box, which happened rarely enough to make the purple tin feel special.
One day during the Christmas holidays, after school had closed, I came home and saw the purple tin on the kitchen counter. It was already open, so I was hopeful. I crept into the kitchen, lifted the lid, and much to my disappointment, I found my mother’s paper recipes in the tin. She had kept the old biscuit tin to store them.
It takes hours each week to research, write, and produce at this level. Consider becoming a free or paid subscriber to support my work. Thank you.
That day, the purple tin seemed to tell me everything I needed to know about what was inside. Its purple lid still carried the promise of what it used to hold, and I had believed that promise before I looked inside.
René Magritte played with a similar gap in The Treachery of Images. He painted a pipe, then wrote beneath it: “This is not a pipe.” At first, we see the pipe. Then his words make us notice that we are not looking at a pipe but at an image of one. Magritte makes that slippage in our thinking visible.
Psychologists Daniel Kahneman and Shane Frederick described a related move in our thinking as attribute substitution: when a question is difficult, we may answer an easier, related one instead, without noticing the switch.
Kahneman illustrated the idea with an experiment by Fritz Strack, Leonard Martin and Norbert Schwarz.
Students were asked how satisfied they were with life overall and how many dates they had had in the past month. When the life-satisfaction question came first, the answers had little connection to dating.
But when the question about dating came first, the two answers became much more closely linked.
The students had just been prompted to think about one part of life, and that feeling became an easy way to judge life as a whole.
In the same way, a headline figure showing that women make up 30 or 40 per cent of a study can have a similar effect. The presence of women in the study is easy to see, so we let that answer stand in for a harder question: was the research designed to tell us what happens in women? and/or what does that figure actually tell us?
Rebecca Rechlin and her colleagues reviewed 3,193 papers from six neuroscience and psychiatry journals, comparing research published in 2009 and 2019. They looked at whether studies reported using both male and female subjects, and whether researchers tested if results differed by sex. The papers covered research with people, animals and cells.
By 2019, 68% of the papers reported using both sexes, up from 38% in 2009. Yet only 5% of papers in the 2019 sample tested whether the results differed by sex.
Imagine a study testing a new treatment in women and men. Researchers could calculate one average result for everyone, perhaps accounting for sex in their analysis.
Or they could ask whether the treatment works differently in women and men. Sex then becomes part of what the study is investigating. That is what it means to analyse sex as a discovery variable.
A study of childhood asthma shows why this kind of analysis can matter. Tesfaye Mersha and his colleagues looked at genetic markers (small variations in DNA) in girls and boys with asthma. When they analysed the children together, they found 33 markers associated with asthma.
When they looked at girls and boys separately, they found 47 markers.
Additionally, the two analyses overlapped on only 21 markers which appeared in both. Twenty-six markers were not identified in the combined analysis.
These markers were associations, not proof of what caused asthma. And an important disclaimer is that this was one study of selected genes in children. But it does shows how a pattern found in one sex can be harder to see when results for girls and boys are combined.
This does not mean every study should be designed to find a difference between women and men. Neuroscientist Liisa Galea and her colleagues distinguish between including both sexes in research and designing a study specifically to investigate sex differences.
The right approach, she says, depends on the question.
The aim is not to assume that women and men will respond differently. It is to make sure that, when sex could affect the result, the study can test for it. Sometimes it will find a difference; sometimes it will not. The evidence needs to let us know which it is.
Research builds the evidence base on which science-backed product development depends. It helps establish what a condition looks like, whom a treatment or diagnostic might serve, and what remains uncertain.
When sex could affect the result but researchers have not examined it, that uncertainty travels forward into product development.
That is why I treat “science-backed” as a question to investigate, not a conclusion. What evidence supports the product’s central claim? How was that evidence generated, and what remains to be tested?
Watch the documentary Why Better Science Isn’t Enough to Improve Women’s Health. If you’d like more conversations with the people shaping women’s health, subscribe to our YouTube channel.
My new book, The Billion Dollar Blindspot is now available on Amazon and in bookstores.
About Maryann Selfe
Maryann Selfe is an institutional investor and the Founder and Chief Investment Officer of Calabar Capital, an investment platform backing exceptional managers and founders in women’s health. She is the author of The Billion Dollar Blindspot and writes about women’s-health investing and capital allocation.
Connect on LinkedIn · Discover the Book · Learn more about Calabar Capital
The Billion Dollar Blindspot is an independent publication. Its content is provided for informational and educational purposes only and does not constitute financial advice, an investment recommendation or an offer to buy or sell any security.
If you are interested, I'd love to have you as a guest on my YouTube channel. DivergentBrainworks.com
NeuroawesomeLife.com
https://www.youtube.com/@neuroawesomelife
So true. Not every study needs to be sex-differentiated. That is overkill. But when it matters, it matters.