Cognitive Daily

What you remember about your life is almost certainly not accurate. Adults have very few memories before age five, and there is a systematic bias to the memories most people have for the rest of their lives. We are more likely to remember details about positive events like marriage and having children than we are to remember negative events like hospital stays or the death of a loved one.

Many studies have found that people appear to remember much more from their teens and 20s than the rest of their lives. A fifty-year-old might remember more about her 20s than her 30s, even though the events in her 30s were more recent. Is there something about the teens and 20s that make them more memorable? Or do our brains just lose the ability to form lasting memories as we age?

Systematically studying autobiographical memory is difficult because there isn’t an easy way to check and see if a reported memory is accurate. A researcher can identify a bias, but she can’t say whether the memories her subjects report are qualitatively different at different ages. I’ve written a memoir and I found that many of the things I “remembered” with utmost clarity turned out to be untrue. But it’s impractical for a researcher to fact-check the hundreds or thousands of memories she’s analyzing.

Still, studying the memories themselves — accurate or not — can tell us a lot about how memory works. A team led by Clare Rathbone asked 16 volunteers age 47-66 to report on the key memories about their lives, by coming up with 10 concepts that complete the phrase starting “I am …”. They then picked three that best defined themselves. Here are my top three: I am a writer, a father, and introspective. For each of these three statements, they were asked to generate ten specific memories of times when the statements were a significant part of their identity. So, for me, this would include when I published my first book, when I sold my business to become a full-time writer, and when I started this blog. Finally they were asked how old they were when each of those three defining statements became “a defining part of their identity.” I became a writer at age 36 when I finished my memoir, and a father at age 24. When I became introspective is a little harder to nail down. I’m going to say age 24 for that, too, because becoming a father also makes you think a lot about who you are.

What ages produced the most memories? Here are the results:

Other than finding very few memories below age 3, there’s no clear pattern to the results, and certainly no “bump” between the ages of 10 and 30, as earlier researchers had found. Indeed, there hardly appears to be any pattern to the results at all. But perhaps they were just looking at the data the wrong way. This graph takes a different view of the data, from the perspective of the identity-defining events:

So the year when each “self” (“father,” “writer,” “introspective”) is formed counts as 0 on this graph, which then charts how many memories from all respondents occurred on each year before and after the defining event. These memories are most likely to occur within a few years of the event itself. They’re also skewed to the right, meaning there are more memories of events after the identity-forming event than before. (Obviously this makes a lot of sense for events like having a child or getting married).

Even more interesting is what happens when we look at the years when identity-forming events occur:

The green columns show the years when respondents point to self-identity-defining moments (this is actually data from a second, larger online study). The most common years are between ages 11 and 30. The orange columns then compile data for only the first three memories respondents identified for each self. These memories peak in the 20s and 30s — a very similar pattern to the “bump” identified by earlier researchers.

Rathbone’s team believes this demonstrates that identity-forming events are what shape the biases in autobiographical memory: We remember more details about certain things because we now believe those things define who we are. The reason earlier researchers found general memory bump between 11 and 30 is because that happens to be when most of us experience the most significant events in our lives: the start of careers, marriage, and families.

This corresponds well to other researchers who have found that immigrants remember more details about the years surrounding their time of immigration than non-immigrants. So if you immigrate in your 30s, you’re more likely to have memories from your 30s than someone who immigrated in her 20s. Other studies have found a memory bump in people from Bangladesh corresponding to a period of political unrest in that country. So it seems that our memories are affected more by the events in our lives than just the physical development of our brains. We’re not all destined to remember more of our teens and 20s than other years; we’re just more likely to experience significant, life-changing events in those years than others.

C. J. Rathbone, C. J. A. Moulin, M. A. Conway (2008). Self-centered memories: The reminiscence bump and the self Memory & Cognition, 36 (8), 1403-1414 DOI: 10.3758/mc.36.8.1403

Comments

#1 Mr Paul
December 8, 2008

“We remember more details about certain things because we now believe those things define who we are.”

Seems the best Rathbone could claim is correlation, not causation. After all, isn’t it just as likely that these things define who we are because they are what we remember?
#2 Richard Simons
December 8, 2008

It seems to me that the way in which the data were collected is going to skew the results. If a defining moment, for example, was when I decided to emmigrate, that could not have taken place before I was reasonably adult. In my case, I was able to describe to my parents a place I had last seen when I was 3 years old (including details I had never been told), yet I would not consider these memories to be ‘defining moments’.

From the third graph, it looks like 1-2% of the sample have no memories of events before the age of 61, which surely is not correct. To me, this paper tells us more about people’s ages when they change their self-perception than about their memories.
#3 R E G
December 8, 2008

I seem to have a lot more childhood memories (whether valid or not) than my peer age group. I have always attributed it to my unsettled childhood. I can place my memories into specific years depending on the school/house/neighbourhood where they take place. That puts them into chronological order and gives them a place in my personal story, rather than just a highlight.

The late teens and early twenties are when most people move out of their family home and re-adjust to a new environment. They also spend a lot of time relating the stories of their new situation to friends and family left behind.

Anything out of the routine makes more of an impact. It just seems normal that the memories would be more vivid. As you get older new and different events also make an impact.
#4 Bjørn Østman
December 8, 2008

At a late age you have spent more time thinking about your early years than your later ones. Perhaps that is part of the reason why some have found that we recall our memories from early life better. Other factors could work on top of that.
#5 Fargo
December 8, 2008

A pet theory of mine, for a little while now, is that memories undergo compression of some sort, and that only unique pieces are really preserved. Smoking that first cigarette is unique, the ten thousand after are pretty much going to be the same. When you try to recall something that’s been highly compressed, perhaps the brain responds with a collection of things cribbed from the unique areas, padding out the time, as it were, and giving you something that’s more of a functional symbol based on expectation and the tiny bits of honestly recorded data.

Like lot’s of people don’t remember driving to work very clearly, I think it may be because you’ve derived nearly every unique bit you can from it by the nth commute. There simply isn’t any reason to track it, typically.

So, by the time you hit your mid-30′s, there’s a good chance you’ve settled into a routine of some kind, and have heavily mined the places you visit for unique data.

I’m just a computer guy though, so chances are good that I’m completely wrong.
#6 Joel Chan
December 9, 2008

Fargo: “perhaps the brain responds with a collection of things cribbed from the unique areas, padding out the time, as it were, and giving you something that’s more of a functional symbol based on expectation and the tiny bits of honestly recorded data”

Sounds Bayesian to me. I like it.
#7 marie holmes
December 20, 2008

I agree with Richard, that change in location or environment induces memory (in fact there is research in educational theory about this..at least in early childhood). I moved every two years until I was thirty, and I have many memories through these years. I have lived in the same spot for the last 8 years, and now I have a really hard time placing when or where something that has occurred during the past 8 years happened. Plus, I think what Fargo says is true: when you live someplace for too long, you have already ‘mined the area for unique data’ and are on autopilot. Did I mention I am ready to move!!
#8 Tony Jeremiah
December 30, 2008

What I find particularly interesting about the percent distribution pattern across the lifespan for the identity-forming events, is that it appears to be the reverse of the pattern observed for self-esteem across the lifespan.

The graph shows that the percentage of identity-forming events increases from 0-30 and decreases from 30 onward. OTOH, using the same online data collection method, Robins and Trzesniewski (2005) showed that self-esteem is highest in childhood (9-12); steadily decreases into adolescence and remains at a similar level into the 30s; it steadily rises again until the 60s before dropping again during the 70s onward.

So, if there is any connection between identity-forming events and self-esteem (and it seems reasonable to assume such a connection given theoretical links between self-concept formation and self-esteem) in the context of the data from the two studies, there does seem to be somewhat of a contradiction. That is, one would expect higher levels of identity-formation at particular ages to be associated with higher (and not lower) levels of reported self-esteem at the same ages.

What would be very interesting is a replication of the age of reported identity-forming events study, in conjunction with participants from each age category reporting both their current level of self-esteem, and, their (remembered) level of self-esteem during those events.
#9 acm
December 30, 2008

I dunno — their conclusion fits with the fact that a lot of people go through a lot of defining events in their teens (say, highschool and college) — at least times when they learned key life lessons and/or recognized certain things about themselves. That’s just developmentally when a lot of it happens. I’d say that the process continues through your 20′s and into your 30s (with or without marriage and kids), but probably after that, while you contine to learn and think, there’s less tangible epiphany (and/or less n-space in which you are totally ignorant and thus can learn really basic large lessons). certainly my density of awareness is tied to how differently I can now see many of the situations that felt intense then, while more recent events are seen through a pretty recognizable lens…
#10 Tony Jeremiah
December 30, 2008

@acm: Your comment is consistent with what Dave says; especially “The reason earlier researchers found general memory bump between 11 and 30 is because that happens to be when most of us experience the most significant events in our lives: the start of careers, marriage, and families. It’s also consistent with the concept of the social clock (i.e., what society expects a person to achieve at a particular age), which people may use as a means to make (social) judgments about how they are doing at a particular age.

What isn’t as obvious with this data is the (emotional) reactions to these identity-forming events when they actually happen, which is presumably what is tapped with questions concerning self-esteem; Are the reactions to these events positive or negative (e.g., the difference between a planned vs. unplanned pregnancy when one is 11-30 years of age).

Just going by two sources of data (one shown here and the other familiar to me), there’s some suggestion that self-esteem is lower at ages where the number of people having identity-formation experiences is highest (i.e., 21-30).

Since there’s no direct research evidence concerning this at the moment, some explanation would be required as to why the age of the highest identity-formation experiences, also corresponds with the age of lowest self-esteem. That is, why should the age when you’re achieving(?) particular identities correspond with lowest self-esteem?

One answer could be that around 11-30, this is the age people are experiencing some form of identity crisis (as James Marcia would suggest). When people begin to age, they may acquire a more solid identity (i.e., identity achievement). This is consistent with what you say in your last sentence. And it is possible that those who are still in an identity crisis after 30-something might also experience the so-called midlife crisis.

But more research would be needed to establish whether identity formation explains (the hypothesized) relationship between age of the highest identity forming events and levels of reported self-esteem at those ages.
#11 Electronic cigarette
July 19, 2009

Hi!Your website stands out from others I have visited so far, for it is more focused in instructional info on “how to” guides and I plan to learn from you.
The late teens and early twenties are when most people move out of their family home and re-adjust to a new environment. They also spend a lot of time relating the stories of their new situation to friends and family left behind.

Thank you!