Investigating and explaining sex differences in episodic memory

Abstract: Sex differences can be seen in a lot of different areas, one of them being cognition where, among other things, men tend to perform better at more spatial tasks and women at more verbal. When it comes to episodic memory — the memory for past events in terms of the what, where, and when — there has been some circumstantial evidence of sex differences. In Study I of this thesis, we performed a meta-analysis on mean sex differences in episodic memory, using a dataset of 617 studies, totaling over one million subjects. This dataset consisted both of published and unpublished data, as well as several open databases. Here, the main finding was that, just like with other cognitive tasks, males tend to perform better on more spatial tasks while females tend to perform better on more verbal tasks. It could also be shown that females performed better when remembering faces, as well as tasks having to do with smell, touch, and different shades of colors. Further, it has been shown that males tend to be more variable than females for a lot of different traits, both biological and psychological. Even if men, on a group level, usually have larger variance when it comes to cognition in general, there is less support for saying anything about episodic memory. In Study II of this thesis, we performed a number of analyses on a somewhat reduced version of the dataset gathered in Study I, searching for variance differences between males and females. Here, there were some evidence showing that males were slightly more variable than females. However, through exploratory investigations we also found results suggesting that this pattern potentially could come about because of underlying methodological problems in the original research. Finally, there have been some large-scale studies suggesting that when it comes to improvements in cognition, women tend to benefit more than men from social progress and increased living conditions in a society. There has also been some evidence for this relationship when it comes to episodic memory. In Study III, we expanded upon these findings by examining the dataset gathered in Study I, which comprises more countries and a larger timeframe than any other investigation on this topic. Here, we could show that for verbal episodic memory, sex differences were related to several different indicators of living conditions tied to the year and country of each study. However, when pitted against each other, it was only the overall education and employment level that could be shown to be related to the outcome and not gender equality, which we expected would be the most important indicator. In this thesis, I first present a comprehensive background on the topics presented above, including some in-depth, possible explanations for why things are the way they are from evolutionary, biological, and social perspectives. I then go through the dataset that all three studies are based on, as well as present the result from them, including some of my own interpretations. Finally, I discuss some general themes that relate to the studies performed, including necessary choices when collecting data for meta-analyses, possible bias in the dataset, and statistical considerations when dealing with dependent data.