Real world memory for order


For centuries, researchers have been intrigued by the psychology of episodic memory and in particular with the way personal events, situations and experiences from our past are organised in the brain. One influential theory to emerge out of this literature suggests that episodic memories are linked according to time (Howard et al., 2015). Evidence for this idea comes primarily – if not exclusively – from laboratory list learning experiments. In the laboratory when individuals are presented with and then tested on a list of words, they tend to recall successive words that were learned nearby in time (Clark & Bruno, 2021). This is known as the temporal contiguity effect, and it occurs with a forward bias. That is, once an individual recalls a word, transitions in the forward direction (i.e., transitions to words that were presented after the current word in the list) are more likely than transitions in the backward direction (Healey et al., 2019). This is known as the asymmetry effect, and it manifests despite individuals being free to recall the words in any order. Temporal contiguity and asymmetry effects are so robust in the laboratory that many have argued they are universal principles of episodic memory (Healey & Kahana, 2014; Kahana et al., 2008). However, laboratory studies of how people recall lists do not match up with studies of real-world behaviour. For example, when participants in Fonti's (2021) study were asked to recall the list of locations they visited on a particular day, they did not follow the same order. To better understand the factors driving this difference between laboratory and real-world memory, it is important to determine whether people can remember the order of real-world items. That is, if people are explicitly asked to recall the locations they have visited on a particular day in order, are they able to? Answering this question is the aim of the current study.

Research Questions / Hypotheses

We intend to replicate Fonti’s (2021) study, but with a serial recall paradigm, to determine whether people have order information available when recalling their visited locations. We also aim to compare serial recall as it appears in the laboratory with serial recall as it appears in this experiment, to see whether critical benchmark findings from the laboratory hold in the real-world (i.e., primacy and recency effects, retention interval effects and protrusion effects). Firstly, we hypothesise that people will be more likely to output successive locations that were visited nearby in time, with a preference for forwards order, mapping onto contiguity and asymmetry effects. Secondly, we hypothesise that people will recall locations visited at the start and end of a day with greater accuracy than locations visited in the middle, mapping onto primacy and recency effects. Thirdly, we hypothesise that as the retention interval (i.e., the distance between the day of encoding and the memory test) increase, recall accuracy will decrease. Finally, we hypothesise that people will be more likely to intrude a location from another day that occurred in the same serial position as the location they are attempting to recall, mapping onto protrusion effects. We also expect that people will be more likely to intrude a location from another day that was visited at the same time as the location they are attempting to recall.


24 people signed up to the study. Only 3 people attended Part 1 of the study. Of these participants no one met the eligibility requirement of having their Google Map location services on. Therefore, no one could complete Part 2 of the study (i.e., no one completed the memory test).


Participants were required to sign up for an account. They were asked to collect 30-days of Google Map location data and upload this to their account. After their location data was uploaded, a memory test would appear in their account on the site. For each day of the month, this test asked them to select which locations they visited in the order they visited them. It was expected to take approximately 30 minutes to complete.


This is an ongoing study. The results have not been analysed, but are expected to be in accordance with the above hypotheses.


Further data collection will occur over Semester 2 and will form the analysis for a Fourth Year Research Project.