An intriguing new study uses the 2016 US Presidential election as a tool to examine the organization of human memory.
The results show that events that occur around the same time are linked in memory. Remembering one past event tends to trigger the recall of other memories from that time. This chronological clustering makes intuitive sense, but it’s a theory that’s been debated in psychology for a while, under the name of the temporal-contiguity effect (TCE).
According to the authors of the new study, Mitchell G. Uitvlugt and M. Karl Healey, most previous studies of the temporal-contiguity effect have relied on memories created in the lab (e.g. by getting people to read an ordered list of items). Instead, Uitvlugt and Healey used real-world memories.
In the first experiment, n=1,051 MTurk participants were asked (shortly after election day) to recall events from the 2016 Presidential campaigns. Participants could describe the events in any order, and in their own words. On average, each participant recalled almost 9 distinct events. Here’s what people remembered.
Apart from the election results themselves, the most recalled event was Trump’s “Grab Them” tape being released, with ‘Peak of FBI Probe into Clinton’s emails’ a close runner-up.
The key test of the temporal-contiguity effect came when Uitvlugt and Healey looked at the ‘time lag’ between each participant’s consecutive recalled events. This revealed that short time-lags were more common than would be expected by chance, even correcting for the fact that news events that occured close in time also tended to have related content (semantic similarity).
So the temporal-contiguity effect was confirmed in the first dataset. A second experiment, in which participants were asked (on May 2nd) to recall news events from 2018, replicated the results. However, the magnitude of the effect was smaller in the second study, a difference that Uitvlugt and Healey say is difficult to interpret without further research.
Overall, the authors conclude that:
Our results suggest that mere temporal proximity can cause memories to be linked. This is true even when events are not deliberately studied and do not occur one after another in a chainlike sequence… the data suggest that the memory system naturally encodes information about temporal distance and uses that information during memory search.
This does seem common-sense to me. However, when I tried to find evidence of temporal-contiguity in my own memory, I confess that I drew a blank. I just can’t remember what else was in the news around, say, the 2016 Brexit referendum, even though I remember that event very clearly.
Perhaps my memory is organized differently? Or perhaps temporal-contiguity doesn’t work when you are deliberately trying to make it happen…?