owning the personal past

21
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/271772940 “Owning” the personal past: Adolescents’ and adults’ autobiographical narratives and ratings of memories of recent and distant events Article in Memory · February 2015 Impact Factor: 2.09 · DOI: 10.1080/09658211.2014.995673 · Source: PubMed READS 38 3 authors, including: Nicole Hättenschwiler University of Applied Sciences and Arts N… 6 PUBLICATIONS 1 CITATION SEE PROFILE Marina Larkina Emory University 18 PUBLICATIONS 140 CITATIONS SEE PROFILE All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately. Available from: Nicole Hättenschwiler Retrieved on: 24 May 2016

Upload: nicole-haettenschwiler

Post on 20-Mar-2017

89 views

Category:

Documents


2 download

TRANSCRIPT

Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/271772940

“Owning”thepersonalpast:Adolescents’andadults’autobiographicalnarrativesandratingsofmemoriesofrecentanddistantevents

ArticleinMemory·February2015

ImpactFactor:2.09·DOI:10.1080/09658211.2014.995673·Source:PubMed

READS

38

3authors,including:

NicoleHättenschwiler

UniversityofAppliedSciencesandArtsN…

6PUBLICATIONS1CITATION

SEEPROFILE

MarinaLarkina

EmoryUniversity

18PUBLICATIONS140CITATIONS

SEEPROFILE

Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,

lettingyouaccessandreadthemimmediately.

Availablefrom:NicoleHättenschwiler

Retrievedon:24May2016

This article was downloaded by: [Nicole Hättenschwiler]On: 04 February 2015, At: 12:27Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

MemoryPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/pmem20

“Owning” the personal past: Adolescents’ andadults’ autobiographical narratives and ratings ofmemories of recent and distant eventsPatricia J. Bauera, Nicole Hättenschwilerb & Marina Larkinaa

a Department of Psychology, Emory University, Atlanta, GA, USAb Institute Humans in Complex Systems, University of Applied Sciences,Northwestern Switzerland, Olten, SwitzerlandPublished online: 02 Feb 2015.

To cite this article: Patricia J. Bauer, Nicole Hättenschwiler & Marina Larkina (2015): “Owning” the personal past:Adolescents’ and adults’ autobiographical narratives and ratings of memories of recent and distant events, Memory,DOI: 10.1080/09658211.2014.995673

To link to this article: http://dx.doi.org/10.1080/09658211.2014.995673

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensorsmake no representations or warranties whatsoever as to the accuracy, completeness, or suitabilityfor any purpose of the Content. Any opinions and views expressed in this publication are the opinionsand views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy ofthe Content should not be relied upon and should be independently verified with primary sources ofinformation. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly orindirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial orsystematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distributionin any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found athttp://www.tandfonline.com/page/terms-and-conditions

“Owning” the personal past: Adolescents’ and adults’autobiographical narratives and ratings of memories

of recent and distant events

Patricia J. Bauer1, Nicole Hättenschwiler2, and Marina Larkina1

1Department of Psychology, Emory University, Atlanta, GA, USA2Institute Humans in Complex Systems, University of Applied Sciences, NorthwesternSwitzerland, Olten, Switzerland

(Received 28 August 2014; accepted 1 December 2014)

Adults and adolescents are characterised as having different perspectives on their personal orautobiographical memories. Adults are recognised as having vivid recollections of past events and asappreciating the meaning and significance of their autobiographical memories. In development, thesequalities are noted as absent as late as adolescence. To evaluate the assumption of developmentaldifferences, we directly compared autobiographical memories of adults and adolescents drawn from eachof several periods in the past, using measures of narrative quality (coded independently) andparticipants’ own subjective ratings of their memories. Adults’ narratives of events from the previousyear and for the “most significant” event of their lives were coded as more thematically coherent relativeto those of adolescents’; the groups did not differ on thematic coherence of narratives of early-life events(ages 1–5 and 6–10 years). The ratings that adults and adolescents provided of their autobiographicalmemories were similar overall; differences were more apparent for early-life events than for more recentevents and indicated stronger mnemonic experiences among adolescents than adults. The pattern offindings suggests that whereas adults have more sophisticated narrative tools for describing thesignificance of events and their relation to the corpus of autobiographical memories, adolescents aswell as adults have vivid recollective experiences as well as personal and subjective perspective on theevents of their lives and their memories thereof.

Keywords: Adolescents; Autobiographical memory; Narrative; Self; Subjective ratings.

Autobiographical or personal memories aredefined as vivid recollections of specific past eventsthat have meaning or significance to one’s self.Autobiographical memories are formed early inlife (e.g., Peterson, Grant, & Boland, 2005; Reese,Jack, & White, 2010; Tustin & Hayne, 2010),yet there are pronounced developmental changesin the robustness of autobiographical memory

throughout childhood and beyond (Bauer &Larkina, 2013; see Bauer, in press, for discussion).One aspect that is presumed to change overdevelopmental time is the perspective that indivi-duals take on their memories. Adults are recog-nised as appreciating themeaning or significance ofpast events to their lives, thus putting the “auto” inautobiographical memory (e.g., Bluck & Alea,

Address correspondence to: Patricia J. Bauer, Department of Psychology, Emory University, 36 Eagle Row, Atlanta, GA 30322,USA. E-mail: [email protected]

Portions of this research were submitted to the faculty of the University of Berne, Switzerland, as partial fulfillment of therequirements for a master's degree for Nicole Hättenschwiler.

© 2015 Taylor & Francis

Memory, 2015http://dx.doi.org/10.1080/09658211.2014.995673

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

2008; Bluck & Habermas, 2001; Fivush, 2011;Fivush & Zaman, 2014; McAdams, 2001). Thisstance is noted as absent from autobiographicalnarratives as late as adolescence (Fivush, 2012;Habermas & de Silveira, 2008). Moreover, prior toadulthood, memories of past events are charac-terised as lacking important features of episodicand thus autobiographical memory, such as vividrecollection or phenomenological reliving (e.g.,Wheeler, 2000). The primary source of evidencefor these perspectives are the narratives thatchildren and adolescents produce about pastevents. Their narratives tend to feature relativelyfew attributes that establish the temporal andspatial context of events (Reese et al., 2011) andwhich convey a subjective sense of the significanceof events to the narrators’ lives (Bohn & Berntsen,2008; Fivush & Zaman, 2014; Habermas & Bluck,2000). Importantly, direct comparisons of thequalities of adults’ autobiographical memoriesand those of children and adolescents are rare inthe literature (though see Bauer & Larkina, 2013),making it difficult to evaluate claims of featurespresent in adults’ autobiographical recollectionsbut absent in those younger. In the presentresearch, we compared autobiographical memoriesof adolescents and adults drawn from each ofseveral periods in the past, using measures ofnarrative quality and participants’ own subjectiveratings of their memories. The data were used totest suggestions of differences in the perspectiveadults and adolescents have on memories of theirpersonal pasts.

An element crucial to definition of an episodicmemory as autobiographical is evidence of aunique, personal perspective on the event that isthe subject of the memory (Bauer, 2007, in press;Bluck & Alea, 2008; Conway, 2005; Habermas &Köber, 2014; Nelson & Fivush, 2004). Amongadults, evidence of the meaning or significanceplaced on memories of past events comes fromthree primary sources. First, the fact that adultsremember some events over long periods of timeis taken as prima facie evidence of their import-ance to the individual. Second, in their narrativesabout past events, adults include informationthat indicates vivid recollection of the experi-ences, such as details that locate the events intime and place. They also feature subjectiveevaluations and interpretations of events, andthey explicitly create linkages between autobio-graphical events and experiences, future plansand self-understanding (Bluck, Alea, Habermas,& Rubin, 2005; Pillemer & Kuwabara, 2012;

Reese et al., 2011). Third, when adults describetheir memories by rating them, they indicate thattheir memories are vivid, as well as personallyrelevant and significant (e.g., Bauer, Stennes, &Haight, 2003; Howes, Siegel, & Brown, 1993;Waters, Bauer, & Fivush, 2014; Weigle & Bauer,2000; West & Bauer, 1999). These sources con-verge to suggest that adults have a unique,individual and personal perspective on theirmemories—in effect, they “own” their autobio-graphical memories.

In studies with children and even adolescentsas subjects, there is less evidence of appreciationof the personal relevance or significance ofmemories of past events. Young children, inparticular, do not remember the events of theirlives over periods as long as adults do. Indeed,there is growing evidence that throughout thefirst decade of life, children’s memories are morevulnerable to forgetting, relative to those ofadults (e.g., Bauer & Larkina, 2014; Bauer,Burch, Scholin, & Guler, 2007; Peterson, Warren,& Short, 2011); the younger the child at the timeof the event, the more susceptible the memory isto forgetting (Bauer & Larkina, 2014; Morris,Baker-Ward, & Bauer, 2009; see Bauer, in press,for discussion).

The narratives that children provide about pastevents are also noted as lacking in some of thefeatures that characterise adults’ narrativereports. There are pronounced changes in narrat-ive length and quality over the first decade of life(e.g., Habermas, Negele, & Mayer, 2010; VanAbbema & Bauer, 2005). It is not until 10–12years of age that children effectively orient thelistener to the time and place of events andmaintain and elaborate on the topics of pastevents (e.g., O’Kearney, Speyer, & Kenardy,2007; Reese et al., 2011). Especially salient omis-sions are the causal connections between andamong actions and events (e.g., because, so that)that aid in understanding of why events unfoldedas they did and which characterise adults’ narra-tives (e.g., Bauer et al., 2005; Habermas et al.,2010). It is only in adolescence that individualsconstruct an extended life narrative and convey asense of narrative self-identity (e.g., Bohn &Berntsen, 2008; Fivush & Zaman, 2014; Habermas& Bluck, 2000; see Bohn & Berntsen, 2014).

The third source of evidence regarding themeaning of memories of past events to adults,namely, subjective ratings, is virtually absent fromthe developmental literature. With children, useof subjective ratings seemingly is confined to

2 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

those of confidence in old/new recognition ofwords or pictures from a study set (e.g., Ghetti,Qin, & Goodman, 2002), and valence and arousalof emotional picture stimuli (e.g., Leventon,Stevens, & Bauer, 2014). It does not appear thatsubjective ratings of the qualities of personal orautobiographical memories have been elicitedfrom children or adolescents. As a result, it isnot possible to determine how pre-adults them-selves evaluate their memories of the naturallyoccurring events of their lives.

The major purpose of the present research wasto test suggestions of differences in the perspect-ive adults and adolescents have on memories oftheir personal pasts. We addressed this questionby obtaining narratives and subjective ratings ofpast events from adolescents (ages 12–14 years)and adults (ages 18–23 years); the materials wereobtained in the context of an online survey. Wechose these age groups because they are on eitherside of a transition in identity development(Erikson, 1968). Further, based on analyses ofautobiographical narratives, they are presumed tohave different relationships with their memoriesof past events. As noted, adults’ narratives reflecta personal or subjective perspective on the eventsof their lives. In contrast, even as late as adoles-cence, children’s narratives are lacking in thisperspective (e.g., Bluck & Habermas, 2001;Fivush, 2011; Habermas & Bluck, 2000; McA-dams, 2001). Direct comparison of participantsbefore and after the hypothesised transitionstands to inform the question of whether adultsand adolescents place different meaning or signi-ficance on their recollections of past events.Although it would have been desirable to extendthe developmental sample further into childhood(i.e., below age 12 years), we did not do sobecause of concerns that younger children wouldhave difficulty completing the protocol, which, asdescribed below, required written production ofseveral narratives and completion of multi-item,multiple-point rating scales. Moreover, we hadconcerns that parents of children younger thanage 12 years would be uncomfortable with theirchildren independently completing an online sur-vey describing personal events and experiences.

We coded the participants’ narratives aboutpast events along three dimensions of coherencethat are theoretically and empirically distinct:context, chronology and theme (Reese et al.,2011). We focused on the attribute of narrativecoherence because of the central role that well-developed narratives are presumed to play in

building a structure of personal experiences in away that facilitates subjective reflection and helpsto understand life (Bruner, 1987). Specifically,coherent narratives provide a chronological se-quence of events and aid the narrator and listenerin placing events on a timeline. They also featureintentions, motivations, thoughts and emotionsthat give structure and meaning to the events(Linde, 1993). In short, the coherence of anarrative is thought to reflect the sense ofpersonal meaning of experienced events (Reeseet al., 2011). The context dimension of narrativecoherence reveals the extent to which a reader(or listener) observes evidence that the event thatis the subject of the memory report is located intime and place, and the extent to which thetemporal and spatial context of the event iselaborated and specified. The chronology dimen-sion reveals the extent to which the narration istemporally organised. The theme dimension cap-tures the extent to which the coder perceivesmaintenance of a distinct topic in the narrativeand importantly for present purposes, the extentto which the narrative features interpretations ofthe meaning of the event, and connections to andintegration with other personal events and experi-ences, future plans and the significance of theevent to the self. In prior research, the dimensionof theme has been shown to be especially sensit-ive to developmental change (Reese et al., 2011).

Critically, coding of narratives for any attribute—including coherence—is an act performed not bythe person recollecting the past event, but by anindependent coder. As such, it does not providea direct window onto the perception that theprovider of the narrative has on the memory orthe event. Moreover, it is inevitably confoundedwith developmental changes in narrative produc-tion. This further distances the coding from theprovider’s perception. In this regard, use ofsubjective ratings can be expected to be espe-cially revealing of the perspective an individualhas on her or his recollections. In spite of thispotential, ratings of subjective qualities such asthe importance of the event at the time ofthe experience and in retrospect, the vividnessof the recollection, and the completeness ofthe memory, for example, do not tend to beincluded in developmental studies; we know ofno comparison of ratings of these attributes byadolescents and adults. In the present research,we used participants’ subjective ratings of theirmemories as a source of evidence regarding how

“OWNING” THE PERSONAL PAST 3

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

they think about their memories of the naturallyoccurring events of their lives.

We asked the participants to narrate and ratememories from the recent and distant past.Specifically, we asked both age groups to providememories and ratings for each of three eventsfrom each of three periods: (a) ages 1–5 years, (b)ages 6–10 years and (c) the previous year. Adultsalso provided memories and ratings of the periodof 11–15 years. Both age groups also narrated andrated their memory of the “most significant”event of their lives. We used the data in plannedcontrasts of adolescents’ and adults’ responseswithin each period.

A strong version of the suggestion that adultsand adolescents have different perspectives onthe events of their lives and their memories ofthem would lead to the prediction of between-group differences in memory quality, regardlessof the period of time from which the memorieswere drawn. Moreover, the differences would beexpected to be apparent in both narratives andsubjective ratings; the narrative dimension oftheme would be expected to be especially sensit-ive to group differences because it reflects nar-rative interpretations of the meaning andsignificance of the event to the self. Thesepredictions follow from the argument that asubjective perspective on the corpus of memoriesof past events is developed in adulthood, asindividuals come to evaluate their past experi-ences in terms of their significance and meaningto the self. Thus, regardless of when the eventstook place, adults’ narrative reports and subject-ive ratings of their memories should feature asense of vivid recollection with personal or indi-vidual perspective. In contrast, it is argued thatadolescents have not yet gained autobiographicalperspective on the events and memories of theirlives and thus should not convey it either innarrative descriptions or in subjective ratings.

An alternate version of the suggestion thatadults and adolescents have different perspectiveson their memories of past events would allow thatpersonal meaning or significance might be appar-ent only for memories formed at (or near) thetime that a subjective perspective was adopted,namely, in adulthood. In this case, memoriesformed earlier in life (1-5 years and 6-10 years)might be recalled with less “autobiographical”flavour, by both adults and adolescents. In con-trast, age group differences should be apparent innarratives and subjective ratings of memories ofthe “most significant” event and memories of

events from the previous year of life, which foradults was ages 17–22 years and for adolescentswas ages 11–13 years. For adults, memories ofthese events should be characterised by vividrecollection and personal reflection, whereasthese qualities should be less apparent in adoles-cents’ memories. Moreover, under the assump-tion that achievement of a subjective perspectiveoccurs gradually, we would expect to see moreevidence of it in narratives and ratings of eventsand memories from the period of early adoles-cence (11–15 years), relative to childhood (6–10years). We tested this by comparing adolescents’narratives and ratings of events and memoriesfrom the period of 6–10 years and adults’ narra-tives and subjective ratings from the period of 11–15 years. This contrast features memories fromthe same temporal distance in the past (i.e., onaverage, 5-7 years in the past), yet the memorieswere formed in different developmental periods(childhood vs. early adolescence).

As just outlined, both the strong and alternateversions of the suggestion that adults and adoles-cents have different perspectives on their mem-ories lead to predictions that the pattern of agegroup differences would be similar whether basedon narratives that the individuals produce or ontheir subjective ratings of their memories. Yetthese two windows on autobiographical memorydo not by necessity, afford the same view. Asnoted earlier, when the source of evidence abouta memory is a narrative description of it, thevividness, personal relevance, meaning, or signi-ficance of the experience or the memory is in theeye of the beholder—an independent coder. Thatindividual’s perception of the memory is inevit-ably confounded with the quality of the narrativedescription of it. Because narrative competencedevelops slowly, over the first decade of life andbeyond, it is possible that independent coding ofnarratives underestimate the extent to whichnarrators have adopted a subjective perspectiveon their memories (see Bauer, in press, fordiscussion). If this is the case, then fewer differ-ences between adults’ and adolescents’ may beobserved in their subjective ratings of events andmemories, relative to their narrative descriptionsof them.

In summary, in the present research, we usednarratives and subjective ratings to directly com-pare the qualities of autobiographical memoriesof adolescents and adults. We compared thequalities in both recent and distant events. Evid-ence of differences in both the narrative

4 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

description and subjective ratings of events andmemories, regardless of the time in life fromwhich they were drawn, would be consistent witha strong version of the suggestion of a differentperspective on autobiographical memories byadults and adolescents. Evidence of differencesin more recent and the “most significant” event—but not in more remote events—would be con-sistent with an alternate suggestion that a sub-jective perspective infuses events and memoriesformed later, but not earlier, in life. In light of thepossibility that independent coding of narrativesmay underestimate the personal meaning orsignificance of events, especially among youngernarrators, we expected fewer differences betweenadults’ and adolescents’ in their subjective ratingsof events and memories, relative to their narrat-ive descriptions of them.

METHODS

Participants

A total of 110 participants took part in the study.Of the participants, 54 were adolescents (26females, 28 males; M age = 13.01 years, SD =0.80, range 12–14 years) and 56 were adults (29females, 27 males; M age = 20.51 years, SD =1.30, range 18–23 years). Adolescents wererecruited from an existing pool of families whohad volunteered to take part in child develop-ment research. Though no specific information onSES was obtained, the pool comprises largelymiddle- to upper middle-class families in whichone or both parents are college educated. Theadults were undergraduate and graduate studentsrecruited through flyers and online advertise-ments on the Emory University campus. Basedon self-report, the racial composition of thesample was 19% African-American, 17% Asian,56% European-American and 3% mixed race;the remaining 5% of participants did not reporttheir race. Four-percent of the sample self-identi-fied as Hispanic. The adults and the parents ofthe adolescents provided online written informedconsent for their own or their children’s parti-cipation (respectively). Adolescents provided on-line written assent to take part. After completingthe survey, participants received a gift certificate.All procedures were reviewed and approved bythe Emory University Institutional ReviewBoard.

Procedure

Data were collected via an online survey usingSurvey Monkey (www.surveymonkey.com). Adultparticipants and the parents of adolescent partici-pants were sent a direct link to the survey usingelectronic mail. Upon accessing the survey, adultparticipants were asked to complete the consentprocedure. Parents of adolescent participants wereencouraged to review the entire survey beforeconsenting to their adolescents’ participation.Upon receiving parental consent, the adolescentsthemselves were asked for their assent to particip-ate. They were instructed to complete the surveyindependently (i.e., without parental assistance).All participants then were asked to provide thedemographic information of birth date, surveydate, gender, race and ethnicity, and for adults,the last year completed in college. All participantsfulfilled the inclusion criterion of having English astheir first language (the requirement for English asthe first language was provided in the recruitmentmaterials).

In the survey proper, participants were askedto think about specific events that occurred in oneplace at one time in their lives. A sample surveyfor one memory report is provided in theAppendix. They were asked to provide a writtennarrative of each event and to include “as muchdetail as you can” about the event. Adolescentswere prompted to provide a total of 10 memoryreports. They described three memories fromeach of three periods: 1–5 years, 6–10 years andthe previous year of life. At the end, they wereasked to describe the “most significant” event oftheir lives. Adults were prompted to provide atotal of 13 memory reports. They described threememories from each of four periods: 1–5 years,6–10 years, 11–15 years and the previous year oflife. At the end, they were asked to describe the“most significant” event of their lives.

After describing each memory, participantswere asked to provide their age at the time ofthe event, in years and months. They also wereasked to provide a descriptive title for the event.Participants then used 9-point Likert-type scalesto provide ratings of their subjective experienceof the event that gave rise to the memory, of thememory itself and of the extent to which theyshared their memory with others. To captureparticipants’ ratings of the subjective experienceof the event that gave rise to the memory, weasked them to rate the event on valence (very

“OWNING” THE PERSONAL PAST 5

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

negative to very positive), arousal (very bored/calm to very excited/nervous) and uniqueness(occurs often to occurs very rarely). We alsoasked about the significance of the event at thetime of the experience (not at all important tovery important), and the significance of the eventnow (not at all important to very important). Tocapture participants’ ratings of the qualities of thememory of the event, we asked them to rate thememory on vividness (very unclear to very clear)and completeness (referred to as “coherence” inthe survey; very few details to many details). Wealso asked them to rate their confidence that theirmemory accurately captured the details of theevent (very unsure to very sure). Participants alsoindicated whether their memory of the event was(a) as if they experienced it (first person), (b) as ifthey were watching it happen (third person) or(c) both. Finally, to assess the extent to whichparticipants shared their memories with others,we asked them to rate the frequency of retellingof the narrative with others (never to very often).

The prompts for memories from the differentperiods were presented in a pseudo-randomorder, such that memories from the same periodwere not prompted in immediate succession, andthe prompt for the “most significant” eventalways was provided last. Each of the fourpseudo-random orders was used approximatelyequally often across participants and age groups.Questions that prompted ratings of the events,the memories themselves and the frequency ofretelling of the narratives were interspersed and

presented in the same order for all memories, forall participants.

Coding and reduction of data

The length of each narrative was calculated inwords, using Microsoft Word. The coherence ofthe memory narratives was coded on the threedimensions of the Narrative Coherence CodingScheme (NaCCs) developed by Reese et al.(2011). The three dimensions are context (orient-ing in time and space), chronology (relating eventcomponents along a timeline) and theme (main-taining and elaborating on topic). Each dimensionwas coded on a 4-point scale, from 0–3. The levelsof the dimensions are described in Table 1.

All narratives were coded by a single indi-vidual (the second author). For purposes ofestimating reliability of coding, a second inde-pendent researcher (the third author) coded 20%of the narratives. Average reliability for eachdimension was calculated using Cohen’s kappa.Estimates of reliability ranged from .82 to .85 foradults and .82 to .86 for adolescents (ps < .001),indicating “outstanding” reliability (Landis &Koch, 1977).

For purposes of analysis, for each period (1–5,6–10, 11–15 [adults only], previous year) wecalculated the mean length of the narratives (inwords) and the mean score for each coherencedimension (context, chronology, theme; as arguedand demonstrated by Reese et al. [2011], thedimensions are theoretically and empirically

TABLE 1The narrative coherence dimensions of context, chronology, and theme, and the levels of each dimension (derived from the NaCCs

categories in Reese et al. [2011])

Score Context Chronology Theme

0 Neither time norlocation information isprovided in the story.

Narrative contains no information abouttemporal order.

Narrative is off-topic or described with severaldistracters that make identification of thetopic difficult.

1 Time point or location atany level of specificity.

Some events on the timeline and fewer thanhalf of the temporally relevant actions can beordered on a timeline with confidence.

Topic is identifiable yet the narrative includesnegligible development of causal linkages,personal evaluations and reactions orelaborations of actions.

2 Time and location of theevent and one of thedimensions is specific.

Between 50% and 75% of the relevant actionscan be placed on a timeline but the entire storycannot with confidence be ordered from startto finish.

The narrative includes interpretations and/orelaborations of previously reported actions.

3 Time and locationinformation is specific.

Almost all the temporally relevant actions canbe ordered.

In addition to the requirements for score 2memories are connected to otherautobiographical experiences, future plansor self.

6 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

distinct and thus were treated separately in ana-lyses). We also calculated the mean subjectiverating for each dimension (valence, arousal,uniqueness, significance then, significance now,vividness, completeness, confidence in details andfrequency of retelling). The mean values wereused in all analyses. Because there was only asingle “most significant” event, analyses of themost significant event were based on the rawscores and ratings. Finally, for each period, wetallied the percentage of memories that partici-pants indicated were in first person, third person,and both.

RESULTS

Events reported

As is frequently observed in studies in whichparticipants are free to report on any type of pastevent (e.g., Morris et al., 2009), the majority ofthe events upon with the participants reportedwere moderately positive experiences, includingschool events, family outings and celebrations andvacations. Few negative experiences were refer-enced, with those included consisting primarily ofminor injuries or accidents. For the “most signi-ficant” event, adults’ and adolescents alikereported on family and leisure or vacation activ-ities; experiences with friends and family; andbirthdays, gifts and holidays. Additionally, adultsreported on their reception of the news that theyhad been accepted to college.

Missing data and preliminary analyses

One adolescent indicated that he “did not under-stand” the instruction to report on events fromthe “previous year.” This participant’s data wereexcluded from analysis of this period. Addition-ally, due to an error in survey design, one adultwas not prompted for her or his age at the time ofthe events from the previous year, and two adultswere not prompted for their ages at the time ofthe most significant event of their lives. Addition-ally, data for three adolescents were excludedfrom analyses of the “most significant” event: oneadolescent did not respond to the prompt and twoindicated that the most significant event in theirlives was their own birth, but that they did notremember it. In each case, degrees of freedomwere adjusted accordingly.

To ensure that participants responded appro-priately to the prompts to provide memories fromspecific periods of time, we conducted analyses ofthe average age of participants at the time of theevents reported from each of the time periods. Asreflected in Table 2, Panel a, participants per-formed as directed and provided reports of eventsfrom within the specified time frames. For theperiods of 1–5 and 6–10 years, the average ages ofparticipants at the time of events did not differfor adolescents and adults. For the previousyear and most significant event, the average agesof the adolescents at the time of the events wereyounger, relative to adults. In addition, forthe previous year and most significant events,we examined the length of delay between thetime of the event and the session date (see Table2, Panel b). The average delays did not differsignificantly between the adolescents and adults.

Because of reports in the literature of genderdifferences in adolescents’ and adults’ autobio-graphical memories (see Fivush & Zaman, 2014,for a review), we conducted analyses for possiblegender differences. We conducted a total of65 statistical tests (4 narrative variables plus 9ratings × 5 period comparisons); only 5 of thetests reached the conventional level of statisticalsignificance. Across age groups, females producedlonger narratives than males for events from theperiod of 6–10 years (Ms = 121.30 and 93.30,SDs = 62.29 and 61.83) and for the previous year(Ms = 141.60 and 109.10, SDs = 75.00 and 86.58;ts(108) = 2.36 and 2.10, ps < .05, respectively). Formemories from the period 1–5 years, females hadhigher scores on the context dimension of narrat-ive coherence, relative to males (Ms = 2.05 and1.67, SDs = 0.48 and 0.48; t(108) = 4.24, p < .001,respectively). In terms of their ratings of theirmemories, across age groups, females indicatedthat the events on which they reported were moreunique than males, both from the period 6–10years (Ms = 7.70 and 6.93, SDs = 1.34 and 1.65)and from the previous year (Ms = 7.35 and 6.73,SDs = 1.58 and 1.59; ts(108) = 2.93 and 2.04,ps < .05, respectively). There were no othergender differences. Given the relative paucity ofgender-related effects, gender was not consideredin subsequent analyses.

Narrative dimensions

Statistics describing the narratives provided bythe participants are provided in Table 3, Panel a.

“OWNING” THE PERSONAL PAST 7

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

As reflected in the table, the average lengths ofthe narratives (in words) provided by adolescentsand adults did not differ for the period of 1–5years. For all other periods, adults providedlonger narratives, relative to adolescents. In con-sideration of the differences in narrative length,the number of words per narrative was controlledin subsequent analyses.

To test for possible age-related differences innarrative coherence, we conducted one-way ana-lyses of covariance (ANCOVA) for each of thenarrative dimensions (n = 3), for each periodcomparison (n = 5), controlling for narrativelength.1 Because for the period 1–5 years, thedifference in narrative length was not statisticallysignificant, we also conducted analyses of thisperiod without the statistical control. The resultsof analyses did not differ when narrative length

was and was not controlled. In light of the largenumber of planned comparisons conducted, inaddition to conventional levels of statistical signi-ficance, we also evaluated the results of theANCOVAs against an alpha adjusted for thenumber of comparisons (p < .0033: p = .05/15 =.0033). Table 3 reflects the results of both theuncorrected and corrected comparisons.

On the dimension of contextual coherence,adolescents and adults differed in their narrativesfrom the previous year. However, the effect fellbelow the adjusted level of statistical significance(p < .0033). There were no other significantdifferences in contextual coherence. On thedimension of chronology, none of the analyseswas statistically significant.

On the dimension of thematic coherence, ado-lescents and adults did not differ for the periods of1–5 and 6–10 years. The narratives that adultsproduced about events from ages 11 to 15 yearswere more thematically coherent than those pro-duced by adolescents about events from the samedistance in the past, namely, the period of 6–10years. Additionally, adults’ narratives of eventsfrom the previous year and about the most signi-ficant event of their lives were more thematicallycoherent than those produced by adolescents.Though the effects for events from childhoodversus adolescence (6–10 years vs. 11–15 years)

TABLE 2Average age at the time of events from each time period for adolescents and adults (Panel a) and average delay between events

and memory reports for the previous year and most significant events (Panel b)

GroupTest of difference

Adolescents Adults T-test valueTime period M (SD) Range M (SD) Range (df = 108)*

Panel a: Average age at the time of events1–5 years 4.29 (0.77) 4.35 (0.67) 0.47

1.64–5.67 2.81–5.926–10 years 8.55 (0.83) 8.55 (0.89) 0.02

7.03–10.53 6.50–10.6111–15 years NA 13.47 (0.86) NA

11.64–15.44Previous year 12.10 (1.83) 19.80 (1.35) 6.91***

11.00–14.61 17.20–22.94Most significant 10.85 (2.88) 17.34 (3.72) 7.64***

1.08–14.17 4.78–23.08Panel b: Average delay between events and memory reports for the previous year and most significant events (in years)Previous year 0.89 (0.37) 0.81 (0.38) 1.10

0.27–1.86 0.19–1.84Most significant 2.16 (2.32) 3.17 (3.40) 1.76

0.03–11.18 0.06–13.75

Degrees of freedom for the previous year = 106 and for the most significant event = 103. ***p < .001.

1The data also could have been subjected to a group(adults, adolescents) × period (1–5, 6–10, previous year, mostsignificant event) multiple analysis of variance (MANOVA)with the three coherence dimensions entered as separatedependent variables. Such an analysis would have the potentialto yield uninterpretable main effects of period. Moreover, weexplicitly predicted the possibility of interactions of Group ×Period, as well as different patterns for the different dimensionsof coherence. As such, the MANOVA approach likely wouldhave resulted in a larger number of statistical tests. For thisreason, we did not adopt it.

8 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

TABLE 3Means (and standard deviations) for narrative length and coherence (by dimension) (Panel a), subjective ratings of events (Panel b), subjective ratings of memories (Panel c) and ratings

of frequency of retelling (Panel d), for adolescence and adults, from each time period

Time period

Dimension or rating by age group 1–5 years 6–10 years 11–15 yearsa Previous year Most significant

Panel a: Length (in words; no maximum) and narrative coherence dimension (max = 3.0)Length Adolescence 71.31 (48.81) 92.72 (64.96) NA 96.11 (61.15) 119.50 (172.40)

Adults 81.76 (41.92) 121.30 (58.98) 140.72 (92.65) 153.50 (90.33) 187.70 (176.60)t-test value 1.21 2.42* 2.03* 3.89*** 2.05*

Context Adolescence 1.85 (0.46) 1.99 (0.52) NA 1.83 (0.66) 1.85 (0.78)Adults 1.88 (0.57) 2.26 (0.59) 2.21 (0.48) 2.27 (0.51) 2.18 (0.83)

(ANCOVA length controlled) 0.01 F(1, 107) 3.43 F(1, 107) 3.26 F(1, 107) 7.78**/ns F(1, 106) 1.74 F(1, 104)Chronology Adolescence 0.67 (0.57) 0.76 (0.61) NA 0.69 (0.64) 0.88 (1.01)

Adults 0.82 (0.70) 0.89 (0.51) 0.83 (0.51) 0.95 (0.52) 1.18 (0.92)(ANCOVA length controlled) 1.43 F(1, 107) 1.12 F(1, 107) 0.32 F(1, 107) 2.76 F(1, 106) 2.22 F(1, 104)Theme Adolescence 1.13 (0.30) 1.19 (0.35) NA 1.22 (0.38) 1.41 (0.53)

Adults 1.23 (0.34) 1.26 (0.37) 1.56 (0.52) 1.59 (0.47) 2.04 (0.85)(ANCOVA length controlled) 1.55 F(1, 107) 0.02 F(1, 107) 9.62** F(1, 107) 8.18**/ns F(1, 106) 16.08*** F(1, 104)Panel b: Subject ratings of the events (max = 9)Valence Adolescence 5.74 (1.79) 6.21 (1.81) NA 6.70 (1.61) 7.65 (2.47)

Adults 4.98 (1.44) 4.97 (1.58) 5.61 (1.44) 6.15 (1.69) 5.84 (3.56)t-test value −2.44*/ns −3.84*** −1.93 −1.71 −3.02***/ns

Arousal Adolescence 6.71 (1.35) 7.23 (1.59) NA 7.52 (1.30) 8.53 (1.24)Adults 5.09 (1.90) 5.60 (1.85) 6.22 (1.66) 6.62 (1.60) 7.46 (2.21)t-test value −5.14*** −4.03*** −3.24***/ns −3.23**/ns −3.04**/ns

Uniqueness Adolescence 7.23 (1.65) 6.86 (1.60) NA 6.20 (1.65) 7.22 (2.15)Adults 7.80 (1.30) 7.82 (1.37) 8.21 (0.86) 7.83 (1.08) 8.59 (1.17)t-test value 2.02*/ns 3.37*** 5.54*** 6.13*** 4.16***

Significance Adolescence 6.33 (1.95) 6.83 (1.61) NA 7.16 (1.51) 8.71 (1.10)Then Adults 5.68 (1.84) 6.44 (1.71) 6.48 (1.83) 6.75 (1.40) 8.18 (1.71)

t-test value −1.82 −1.25 −1.07 −1.46 −1.88Significance Adolescence 4.57 (2.08) 5.24 (1.83) NA 5.86 (2.02) 7.96 (1.82)Now Adults 3.85 (1.93) 4.45 (1.89) 4.55 (1.65) 6.01 (1.53) 7.95 (1.71)

t-test value −1.88 −2.21*/ns −2.04*/ns 0.43 −0.04Panel c: Subject ratings of the memories (max = 9)Vividness Adolescence 5.17 (1.76) 6.67 (1.34) NA 7.79 (1.02) 8.25 (1.25)

Adults 4.36 (1.79) 5.61 (1.47) 6.02 (1.25) 7.62 (1.04) 7.79 (2.05)t-test value −2.40*/ns −3.97*** −2.66***/ns −0.86 −1.41

Complete Adolescence 4.73 (1.71) 5.89 (1.41) NA 6.75 (1.39) 7.59 (1.68)Adults 3.57 (1.81) 4.79 (1.75) 5.40 (1.60) 7.20 (1.43) 7.55 (1.90)

“OW

NIN

G”THE

PERSO

NALPAST

9

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

TABLE 3 (Continued)

Time period

Dimension or rating by age group 1–5 years 6–10 years 11–15 yearsa Previous year Most significant

t-test value −3.43*** −3.61*** −1.70 1.65 0.10Confidence Adolescence 5.80 (1.85) 7.02 (1.36) NA 8.02 (0.89) 8.20 (1.25)

Adults 4.79 (1.81) 5.94 (1.54) 6.52 (1.26) 7.77 (0.98) 8.02 (1.33)t-test value −2.29**/ns −3.90*** −2.00*/ns −1.35 −0.71

Perspective Adol 1st 65% 73% NA 76% 72%Adult 1st 59% 67% 71% 77% 71%Adol 3rd 9% 7% NA 5% 6%Adult 3rd 25% 17% 11% 7% 11%Adol both 26% 20% NA 19% 22%Adult both 16% 16% 18% 16% 18%

Panel d: Subjective rating of frequency of retelling (max = 9)Retelling Adolescence 3.78 (1.88) 4.28 (1.41) NA 4.95 (1.99) 6.35 (2.09)

Adults 2.94 (1.77) 3.16 (1.77) 3.15 (1.77) 4.17 (2.00) 4.77 (2.91)t-test value −2.43*/ns −3.65*** −3.68*** `−2.04*/ns −3.21**/ns

aThe test for difference between adolescents and adults for the period 11–15 years is a comparison between adolescents for period 6–10 and adults for period 11–15 years. T-test (df) =108 (for the previous year df = 107; for the most significant df = 105).

For the dimensions of coherence in Panel a, alpha corrected for multiple comparisons = p < .0033. For Panels b, c and d, alpha corrected for multiple comparisons = p < .0011. In allpanels, effects that met conventional levels of statistical significance, but failed to reach the more stringent alpha, are indicated by “ns” next to F or t value.

*p < .05; **p < .01; ***p < .001.

10BAUER,H

ÄTTENSC

HW

ILER,L

ARKIN

A

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

and the most significant event remained statistic-ally significant with the adjustment for multipletests (p < .0033), the effect for the previous year fellbelow the corrected level. Thus, the age of theparticipant at the time of the report did not haveconsequences for narrative characteristics aboutevents from early in life (ages 1–5 and 6–10 years).In contrast, group differences in thematic coher-ence emerged in reports of events from adoles-cence versus childhood (i.e., 11–15 years vs. 6–10years) and from adulthood versus adolescence (i.e.,for the most significant event with a less robusttrend for events from the previous year).

Subjective ratings

To test for possible age-related differences inparticipants’ subjective ratings of the events thatgave rise to their memories, the qualities of thememories themselves, and the frequency withwhich they shared the memories with others, weconducted one-way analyses of covariance(ANCOVA) for each of the rating scales (n =9), for each period comparison (n = 5). We didnot conduct formal tests of possible differences inthe perspective that participants had on theirmemories (first vs. third person, or both). In lightof the large number of planned comparisonsconducted, in addition to conventional levels ofstatistical significance, we also evaluated theresults of the ANCOVAs against an alphaadjusted for the number of comparisons (p <.0011: p = .05/45 = .0011). Table 3 reflects theresults of both the uncorrected and correctedcomparisons.

To assess potential age group differences inparticipants’ ratings of the subjective experience ofthe events that gave rise to their memories, weasked them to rate the events on valence, arousal,uniqueness, significance then and significance now.As reflected in Table 3, Panel b, on the dimensionsof valence, arousal, significance then and signific-ance now, adolescents provided higher ratings thanadults. In the cases of valence and arousal, withoutthe correction for multiple comparisons, adoles-cents’ ratings tended to be statistically significantlyhigher than those of adults (8 of 10 comparisons).With the correction (p < .0011), only three signi-ficant differences remained; they were confinedto the periods 1–5 and 6–10 years. The groupsdid not differ significantly in their ratings ofthe significance of the events at the time of theexperience. Ratings of the significance of the

events at the time of the report (i.e., significancenow) differed significantly only in the cases ofcomparison of adolescents’ and adults’ ratings ofevents from ages 6–10 years, and in comparison ofadolescents’ ratings of events from ages 6–10 yearsand adults’ ratings of events from ages 11–15 years.Yet with the correction for multiple comparisons(p < .0011), these differences no longer weresignificant. The one exception to higher ratings byadolescents than adults was on the dimension ofsignificance now, for events from the previousyear; the difference was not statistically significant,even before correction. On the dimension ofuniqueness, adults had significantly higher ratingsthan adolescents for all periods; for every periodexcept 1–5 years, the differences remained signi-ficant after correction. Thus, in general, relative toadults, adolescents rated the events that gave riseto their memories as less unique. Importantly, theydid not rate the events that gave rise to theirmemories as less significant, either at the time ofthe experience or to the present day.

To assess potential age group differences inparticipants’ ratings of the qualities of theirmemories of the events, we asked them to ratetheir memories on vividness, completeness andconfidence in memory accuracy. Participants alsoindicated whether their memory of the event wasin first-person perspective, third-person perspect-ive or both perspectives. As reflected in Table 3,Panel c, for the dimensions of vividness, comple-teness and confidence, adolescents tended toprovide higher ratings than adults. For the dimen-sions of vividness and confidence, for the periodsof 1–5 years, 6–10 years, and 6–10 years foradolescents and 11–15 years for adults, thedifferences were statistically significant beforecorrection for multiple comparisons. After thecorrection (p < .0011), the groups differed sig-nificantly only for the period of 6–10 years. Forthe dimension of completeness, the differences inratings were significant for the periods of 1–5 and6–10 years, both before and after correction;the comparison of 6–10 years for adolescentsand 11–15 years for adults was not statisticallyreliable. For all three dimensions, ratings ofmemories of events from the previous year andthe most significant event did not differ betweenthe age groups. The ratings thus indicate that formemories of events more distant in time, adoles-cents’ phenomenological experience of theirrecollections was that they were as or more vividand complete than the recollections of adults.Adolescents also had greater confidence in their

“OWNING” THE PERSONAL PAST 11

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

distant memories, relative to adults. The patternis not surprising given that, on average, theamount of time between the occurrence of theevents and the report was shorter for adolescentsthan adults. Consistent with this observation, thedifferences were less apparent in participants’ratings of memories of more recent events (pre-vious year and most significant), for which thelength of delay between the time of the event andthe session date did not differ between the agegroups.

The perspective that adolescents and adultstended to provide on their memories (first-per-son, third-person, both) was nominally similar.Because there were multiple observations perperiod (except for the most significant event),there is not an appropriate statistical test todetermine whether the patterns differed statistic-ally. Visual inspection of Table 3, Panel c,indicates a predominance of first-person perspect-ive for both age groups, for all periods. Ingeneral, both adolescents and adults indicatedthat they had both first- and third-person per-spective more frequently than third-person alone.The one exception was that adolescents tended tohave both first- and third-person perspective ontheir memories from the period 1–5 years (relat-ive to third-person alone), whereas for thisperiod, adults had a nominally higher percentageof third-person alone relative to both first- andthird-person perspective.

Finally, to assess the extent to which partici-pants shared their memories with others, weasked them to rate the frequency of retelling ofthe narrative (never to very often). As reflectedin Table 3, Panel d, for all periods, adolescentsindicated more frequent retellings of the eventsrelative to adults. All the comparisons werestatistically significant before the correction.After correction for multiple comparisons (p <.0011), only the differences for 6–10 years and6–10 years for adolescents and 11–15 years foradults remained significant.

DISCUSSION

Theories of the development of autobiographicalor personal memory posit that over develop-mental time, there are important changes in theperspective that individuals take on past eventsand their memories of them. Based on thequalities of their narrative reports and on theirsubjective ratings of their memories, adults are

recognised as appreciating the meaning or signi-ficance of past events to their lives. Based ontheir narrative reports, this stance is noted asabsent from memories as late as adolescence(e.g., Fivush, 2012; Habermas & de Silveira,2008). Again based largely on their narrativereports, adolescents’ memories also are charac-terised as lacking important features of episodicand thus autobiographical memory, such as loca-tion in time and place and a sense of vivid,veridical recollection (e.g., Reese et al., 2011;Wheeler, 2000). In the present research, weprovided a direct comparison of the qualities ofadults’ autobiographical memories and thoseof adolescents, thereby permitting evaluation ofclaims of features unique to adults’ autobiograph-ical recollections. We compared autobiographicalmemories of adolescents and adults from each ofseveral periods encompassing distant to recentevents. We used measures of narrative quality aswell as participants’ own subjective ratings of theevents and their memories of them. Subjectiveratings can be expected to be especially revealingof the perspective an individual has on her or hisrecollections, both because they are uncon-founded by developments in narrative productionability and because they are provided by theindividuals themselves, versus by an independentcoder.

Analysis of the narrative qualities of adoles-cents’ and adults’ autobiographical memoryreports yielded no significant differences on thedimension of chronological coherence. Contex-tual coherence differed only for narratives ofevents from the previous year, with the advantagegoing to adults. The effect did not survive statist-ical comparison for multiple tests, however. Asexpected, the narrative dimension of thematiccoherence proved most sensitive to age-relateddifferences. This dimension reflects the narrator’sinterpretations and elaborations of the reportedactions and thus a sense of the significance of theevent to the self. The highest score on thedimension indicates that the event beingdescribed is connected to and integrated withother autobiographical experiences or futureplans (Reese et al., 2011). It is thus suggestivethat the narrator has created a life story orautobiography of which the described event is apart (e.g., Bohn & Berntsen, 2008; Habermas &Bluck, 2000).

On the dimension of thematic coherence, weobserved between-group differences on the “mostsignificant” event, on events from the previous

12 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

year, and on events from the period 11–15 yearsfor adults versus 6–10 years for adolescents; theeffect for the previous year did not survive thecorrection for multiple comparisons. The patternof differences is consistent with both the strongand alternate versions of the suggestion thatadults and adolescents have different perspectiveson the events of their lives and their memories ofthem (at least as conveyed through narrative).For adults, based both on their average age at thetime of the report (20.5 years), and on theiraverage ages at the times of the most significantevent and events from the previous year (17.3 and19.8 years, respectively), a personal or subjectiveperspective on the events would be expected(e.g., McAdams, 2001; Reese et al., 2011). Foradolescents, the reports were obtained (13.0years) and the events were experienced (10.9and 12.1 years, for most significant and previousyear, respectively), at ages before which theywould be expected to convey personal perspect-ive on the events of their lives through narrative.The logic extends to events reported by adultsfrom ages 11–15 years (average 13.5 years) andby adolescents from ages 6–10 years (average8.6 years). This contrast features narratives aboutevents from the same temporal distance in thepast (i.e., on average, 5–7 years in the past), yetthe memories were both formed in differentdevelopmental periods (adolescence vs. child-hood) and reported in different developmentalperiods (adulthood vs. adolescence).

The between-group differences on the “mostsignificant” event, on events from the previousyear, and on events from the period 11–15 yearsfor adults versus 6–10 years for adolescents, areconsistent with both the strong and alternateversions of the suggestion that adults and adoles-cents have different degrees of ownership of theremembered experiences of their lives. In con-trast, the pattern of findings for the periods 1–5and 6–10 years are consistent with the alternate,but not with the strong version of the suggestion.We observed minimal, and nonsignificant, differ-ences in the thematic coherence of reports ofevents from the periods of 1–5 and 6–10 years.Memories from these periods were formed whenboth age groups presumably lacked a personalperspective on events. Nevertheless, by strongversions of the suggestion (e.g., Bluck & Alea,2008; Bluck & Habermas, 2001; Fivush, 2011,2012; Fivush & Zaman, 2014; Habermas & deSilveira, 2008; McAdams, 2001), adults would beexpected to have adopted a more personal or

subjective perspective on the events and to reflectthat perspective in their narrative retellings of theexperiences. The fact that adults had notobviously infused their memories of these earlyevents with a more subjective perspective (relat-ive to adolescents), even though they haddeveloped the narrative skills to do so, impliesthat at least as tested in the present research,adults had no greater ownership of these early-life events, relative to adolescents. Overall, thepattern is more consistent with the alternate thanthe strong version of the suggestion of differencesin perspective on memories of past events byadults and adolescents.

Analysis of the coherence dimension of adults’and adolescents’ narratives provided evidenceconsistent with characterisation of the groups ashaving different perspectives on their autobio-graphical memories. Yet going beyond the stat-istical significance of mean differences in thematiccoherence provides reason for caution in thisconclusion. Moreover, consideration of partici-pants’ subjective ratings of their memories pro-vided yet more reason to be circumspect.

As reflected in Table 3, Panel a, adults earnednarrative thematic coherence scores of 1.56, 1.59and 2.04, for the periods 11–15 years, previousyear and most significant event, respectively.Their scores were statistically significantly differ-ent from those of adolescents, with correspondingscores of 1.19, 1.22 and 1.41, for the periods of 6–10 years, previous year and most significantevent, respectively (though the difference forprevious year failed to reach the corrected levelof statistical significance). Critically, for both the6–10 versus 11–15 comparison and for the previ-ous year, for both age groups, the modal rating ofthematic coherence was 1 (76% and 73% foradolescents for 6–10 and previous year, and 54%and 52% for adults for 11–15 and previous year).This means that most participants—in both agegroups—provided “negligible development of thetopic” in terms of causal linkages, personalevaluations and reactions or elaborations ofactions (Table 1). It was only on the “mostsignificant” event that a substantial percentageof adults earned scores of 2 (29%) or 3 (37%),and thus provided evidence of the narrator’sinterpretations and elaborations of the events(score = 2) and their integration with otherautobiographical experiences, future plans andself-identity (score = 3). In contrast, for adoles-cents, even on the most significant event, themodal rating of thematic coherence was 1; 37%

“OWNING” THE PERSONAL PAST 13

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

earned a score of 2 and only 2% earned a score of3 (no narrative earned a score of 0). There aretwo points to be made about this pattern. First,modal differences in scores were confined to themost significant event. On the balance of thenarratives, at least under the conditions of thepresent research, adolescents and adults weremore similar than they were different. Second,the narratives about the most significant eventwere different. When describing this definingevent of their lives, adults featured evidence ofinterpretation, elaboration and integration oftheir autobiographical experiences into a lifenarrative. Although the task called for it, adoles-cents did not make the autobiographical featuresof this most significant event apparent in theirnarratives.

Another observation about the thematiccoherence scores earned by adolescents andadults in the present research is that they arenominally lower than those observed for 11- to14-year olds and adults in Reese et al. (2011).Across the separate samples of adolescents andadults included in the analysis by Reese andcolleagues, the average thematic coherence scoreswere around 2 and 2.5, respectively. In contrast,in the present samples, the average thematiccoherence scores for events from the previousyear and the most significant event were 1.22 and1.41 for adolescents and 1.59 and 2.04 for adults,respectively. We attribute the apparent differ-ences to methodological factors. First, in thepresent research, the narratives were obtained inan online survey, without any direct interactionwith an experimenter. For the samples includedin Reese et al., participants reported on theirmemories either verbally in one-on-one inter-views or in hand-written reports collected in aclassroom or laboratory in the presence of anexperimenter. Second, in the present research,participants reported on the relatively ordinaryand mundane events that make up everyday life.In much of the research included in Reese et al.,participants were asked to report on highly emo-tional events and experiences. Finally, in thepresent research, participants had a relatively highburden, in that they were asked to report on 10(adolescents) or 13 (adults) events. They also ratedeach memory on 10 dimensions. In contrast, inmuch of the research included in Reese et al.,participants reported on four to six events; noratings were collected. Alone or in combination,these methodological factors are a likely explana-tion for the nominally lower thematic coherence of

the narratives in the present research, relative toReese et al. (2011). Importantly, the methodolo-gical factors cannot explain why, in the presentresearch, only modest differences were observedbetween adolescents and adults: both age groupsexperienced them; increased participant burdenwould be expected to have a larger negative impacton adolescents relative to adults and thus shouldhave exaggerated—rather than diminished—differences between the groups.

Though the levels of thematic coherence in thepresent research were nominally lower for bothadolescents and adults, relative to Reese et al.(2011), there was ample evidence that even theadolescents had autobiographical perspective onthe events on which they reported. The evidencecame from their subjective ratings. As arguedabove, subjective ratings have an advantage overnarrative descriptions in that they are directreports of the provider’s perception of the eventor the memory, whereas narrative descriptionsare evaluated through the eyes of an independentobserver. Moreover, ratings are not confoundedwith developmental changes in narrative ability,per se. In spite of these advantages, subjectiveratings are not a common tool in research onautobiographical memory in development. In thepresent research, direct comparison of subjectiveratings by adults and adolescents yielded littleevidence that adults have greater personal per-spective on the events of their lives, relative toadolescents. Indeed, in general, adolescentstended to give higher ratings, relative to adults.In terms of the events that were the subject oftheir memories, adolescents rated the experiencesas more positive and resulting in greater arousal,relative to adults. Few of the differences reachedthe corrected level of statistical significance, how-ever. At the same time, adolescents indicated thatthe events were less unique than those describedby adults. Importantly, the adolescents and adultsdid not differ in their ratings of the importance ofthe events at the time they were experienced.Once corrected for multiple comparisons, adoles-cents’ and adults’ ratings of the current signific-ance of the events also did not differ significantly.Overall, with the exception of the quality ofuniqueness, adolescents’ and adults’ ratings ofthe events that gave rise to their memories weremore similar than different.

The subjective ratings that the participantsprovided also lacked evidence of differentialperspective on the memories themselves amongadolescents relative to adults. For memories of

14 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

events from the previous year and the mostsignificant event, the ratings of vividness, com-pleteness and confidence in the details of thememory did not differ between adolescence andadults. For events from the period 6–10 years,adolescents and adults differed on all three ofthese dimensions, even after correction for mul-tiple comparisons. The only other difference thatsurvived the correction for multiple comparisonswas on completeness, for memories from theperiod 1–5 years. In these cases, adolescents ratedthe features of their memories higher than adults.The (mostly nonsignificant) differences betweenthe groups could be interpreted as evidence ofdifferential perspective on early-life events byadolescents and adults (albeit opposite theexpected direction). However, a more parsimoni-ous interpretation is in terms of the length of thedelay between the event and the report: the delaywas shorter for adolescents than adults, whichcould lend itself to higher ratings. As well,children reported more frequent retellings of theevents, relative to adults. This could contribute togreater vividness, completeness and confidence(though why it would not also apply in the casesof the previous year and most significant event isnot immediately clear). Finally, for both adoles-cents and adults, the majority of events weredescribed as being seen from first-person per-spective. Overall, ratings of events were moresimilar than different for adolescents and adults.The most pronounced differences were on distantevents, not the more recent events for whichadults may have been assumed to have adopted amore personal or subjective perspective orownership.

Taken together, the results of the presentresearch indicate greater similarity than differ-ence in adolescents’ and adults’ perspective ontheir autobiographical memories. Even in theirnarratives, the groups differed only on the dimen-sion of thematic coherence, and then only on themost significant event and on events from 6–10years for adolescents and 11–15 years for adults.Adolescents and adults also differed little interms of their rating of the events that gave riseto their memories and of the memories them-selves. The exceptions tended to favour adoles-cents (i.e., they provided higher ratings thanadults) and were largely confined to events fromearlier in life. Indeed, with the exception of thedimension of uniqueness, for events from theprevious year and the most significant event,

none of the statistical tests for differencesbetween adolescents’ and adults’ ratings reachedthe corrected level of significance. For the com-parison of 6–10 years for adolescents and 11–15years for adults, only the tests for uniqueness andfrequency of retelling reached the corrected levelof statistical significance. The fact that the eventsthat were the subjects of the memory reportswere rated as more unique by adults than byadolescents makes the absence of differences onthe other dimensions even more striking.

Another factor that should be considered inevaluation of the degree of difference betweenthe adult and adolescent samples in the presentresearch is the reminiscence bump. A robustfinding in the literature is that older adults tendto have a larger number of memories from theperiod of 10 or 15 to 30 years, relative to the yearsboth before and after (e.g., Glück & Bluck, 2007;Jansari & Parkin, 1996). Adults also tend to ratememories from this period as especially vivid andas high in personal relevance and significance(e.g., Conway, Singer, & Tagini, 2004). In the pres-ent research, the memories upon which adultsreported for the previous year and the mostsignificant event were from within the bumpperiod. Fewer of the memories reported by theadolescents came from this period. Based on thedifferential degrees of sampling from the period ofthe reminiscence bump, prominent differencesbetween the groups could reasonably have beenexpected. Yet for the most part, they were notobserved.

In conclusion, the results of the presentresearch provide qualified support for the sugges-tion that adults have greater personal or subject-ive perspective on the events of their lives,relative to adolescents. Yet the support wasconfined to observation of higher scores on thedimension of thematic coherence for adults’narratives of (a) events from the periods of 11–15 years compared to the narratives of adoles-cents from the same temporal distance in time,namely, the period 6–10 years and (b) the mostsignificant event, relative to these same periodsfor adolescents; the same pattern was observedfor events from the previous year, though theeffect was less robust. Thus, as evaluated by anindependent coder, adults’ memories from theseperiods featured more evidence of a subjectiveperspective than adolescents’ memories from thecorresponding periods.

There also were qualifications to the support;they came in three forms. First, whereas evidence

“OWNING” THE PERSONAL PAST 15

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

of the personal or subjective perspective wasmore apparent in adults’ narratives, relative toadolescents’, it was not especially pronouncedeven in the older age group. In most cases, themodal score on the thematic coherence dimen-sion was 1 for both adults and adolescents. It wasonly on the most significant event of their livesthat the majority of adults’ narratives earned ascore of 2 or 3 on this dimension. Second, therewere no differences in the thematic coherence ofnarratives about events from the periods 1–5 and6–10 years. The absence of differences on early-life events is inconsistent with the strong form ofthe suggestion that adults have adopted anevaluative perspective on past events and mem-ories thereof, a perspective lacking among ado-lescents. Third, as evidenced by their ratings ofthe events that gave rise to their memories and ofthe memories themselves, a personal or subjectiveperspective is not missing from adolescents’autobiographical memories (even though it wasless apparent in their narrative descriptions).Adolescents rated the events of their lives asemotional, arousing and significant, both whenthe events occurred and in the present day. Theyalso rated their memories as vivid, complete andaccurate. These are not the ratings of a popula-tion that lacks personal or autobiographical per-spective on past events or ownership of theirmemories of them. Instead, the data suggest thatadolescents as well as adults have auto in theirautobiographical memories.

ACKNOWLEDGEMENTS

We thank Claudia M. Roebers, University of Berne,Switzerland, for her comments on the work, andmembers of the Memory at Emory Laboratory fortheir assistance with various aspects of the research.Special thanks to the participants who devoted theirtime to this effort.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by theauthors.

FUNDING

Support for this research was provided by EmoryCollege of Arts and Sciences.

REFERENCES

Bauer, P. J. (in press). A complementary processesaccount of the development of childhood amnesiaand a personal past. Psychological Review.

Bauer, P. J. (2007). Remembering the times of our lives:Memory in infancy and beyond. Mahwah, NJ:Lawrence Erlbaum.

Bauer, P. J., Burch, M. M., Scholin, S. E., & Guler, O.E. (2007). Using cue words to investigate thedistribution of autobiographical memories in child-hood. Psychological Science, 18, 910–916.doi:10.1111/j.1467-9280.2007.01999.x

Bauer, P. J., & Larkina, M. (2013). The onset ofchildhood amnesia in childhood: A prospectiveinvestigation of the course and determinants offorgetting of early-life events. Memory. doi:10.1080/09658211.2013.854806

Bauer, P. J., & Larkina, M. (2014). Childhood amnesiain the making: Different distributions of autobio-graphical memories in children and adults. Journalof Experimental Psychology: General, 143, 597–611.doi:10.1037/a0033307

Bauer, P. J., Stark, E. N., Lukowski, A. F., Rademacher,J., Van Abbema, D. L., & Ackil, J. K. (2005).Working together to make sense of the past: Mothers’and children’s use of internal states language inconversations about traumatic and nontraumaticevents. Journal of Cognition and Development, 6,463–488. doi:10.1016/j.cogdev.2010.08.007

Bauer, P. J., Stennes, L., & Haight, J. (2003). Repres-entation of the inner self in autobiography:Women’s and men’s use of internal states languagein personal narratives. Memory, 11(1), 27–42.doi:10.1080/741938176

Bluck, S., & Alea, N. (2008). Remembering being me:The self-continuity function of autobiographicalmemory in younger and older adults. In F. Sani(Ed.), Self-continuity: Individual and collective per-spectives (pp. 55–70). New York: Erlbaum.

Bluck, S., Alea, N., Habermas, T., & Rubin, D. C.(2005). A tale of three functions: The self-reporteduses of autobiographical memory. Social Cognition,23(1), 91–117. doi:10.1521/soco.23.1.91.59198

Bluck, S., & Habermas, T. (2001). Extending the studyof autobiographical memory: Thinking back aboutlife across the life span. Review of General Psycho-logy, 5(2), 135–147. doi:10.1037//1089-2680.5.2.135

Bohn, A., & Berntsen, D. (2008). Life story develop-ment in childhood: The development of life storyabilities and the acquisition of cultural life scriptsfrom late middle childhood and adolescence. Devel-opmental Psychology, 44, 1135–1147. doi:10.1037/0012-1649.44.4.1135

Bohn, A., & Berntsen, D. (2014). Cultural life scriptsand the development of personal memories. In P. J.Bauer & R. Fivush (Eds.), The Wiley-Blackwellhandbook on the development of children’s memory(pp. 626–644). West Sussex: Wiley-Blackwell.

Bruner, J. (1987). Life as narrative. Social Research, 54(2), 11–32.

Conway, M. A. (2005). Memory and the self. Journal ofMemory and Language, 53, 594–628. doi:10.1016/j.jml.2005.08.005

16 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

Conway, M. A., Singer, J. A., & Tagini, A. (2004). Theself and autobiographical memory: Correspondenceand coherence. Social Cognition, 22, 491–529.doi:10.1521/soco.22.5.491.50768

Erikson, E. H. (1968) Identity youth and crisis. London:W. W. Norton.

Fivush, R. (2011). The development of autobiograph-ical memory. Annual Review of Psychology, 62, 559–582. doi:10.1146/annurev.psych.121208.131702

Fivush, R. C. (2012). Subjective perspective and per-sonal timeline in the development of autobiograph-ical memory. In D. Berntsen and D. C. Rubin(Eds.), Understanding autobiographical memory:Theories and approaches (pp. 226–245). New York,NY: Cambridge University Press.

Fivush, R., & Zaman, W. (2014). Gender, subjectiveperspective, and autobiographical consciousness. InP. J. Bauer & R. Fivush (Eds.), The Wiley-Blackwellhandbook on the development of children’s memory(pp. 586–604). West Sussex: Wiley-Blackwell.

Ghetti, S., Qin, J., & Goodman, G. S. (2002). Falsememories in children and adults: Age, distinctive-ness, and subjective experience. DevelopmentalPsychology, 38, 705–718. doi:10.1037/0012-1649.38.5.705

Glück, J., &Bluck, S. (2007). Looking back across the lifespan: A life story account of the reminiscence bump.Memory & Cognition, 35, 1928–1939. doi:10.3758/BF03192926

Habermas, T., & Bluck, S. (2000). Getting a life: Theemergence of the life story in adolescence. Psycho-logical Bulletin, 126, 748–769. doi:10.1037/0033-2909.126.5.748.

Habermas, T., & Köber, C. (2014). Autobiographicalreasoning is constitutive for narrative identity: Therole of the life story for personal continuity. In K. C.McLean & M. Syed (Eds.), The Oxford handbookof identity development. Oxford: Oxford UniversityPress.

Habermas, T., Negele, A., & Mayer, F. B. (2010).““Honey, you’re jumping about”—Mothers’ scaf-folding of their children’s and adolescents’ lifenarration. Cognitive Development, 25, 339–351.doi:10.1016/j.cogdev.2010.08.004

Habermas, T., & de Silveira, C. (2008). The developmentof global coherence in life narratives across adoles-cence: Temporal, causal and thematic aspects. Devel-opmental Psychology, 44, 707–721. doi:10.1037/0012-1649.44.3.707

Howes, M., Siegel, M., & Brown, F. (1993). Earlychildhood memories: Accuracy and affect. Cognition,47(2), 95–119. doi:10.1016/0010-0277(93)90001-C

Jansari, A., & Parkin, A. J. (1996). Things that gobump in your life: Explaining the reminiscencebump in autobiographical memory. Psychology andAging, 11(1), 85–91. doi:10.1037/0882-7974.11.1.85

Landis, J. R., & Koch, G. G. (1977). The measurementof observer agreement for categorical data. Biomet-rics, 33, 159–174.

Leventon, J. S., Stevens, J. S., & Bauer, P. J. (2014).Development in the neurophysiology of emotionprocessing and memory in school-age children.Developmental Cognitive Neuroscience, 10, 21–33.doi:10.1016/j.dcn.2014.07.007

Linde C. (1993). Life stories: The creation of coherence.New York: Oxford University Press.

McAdams, D. P. (2001). The psychology of life stories.Review of General Psychology, 5(2), 100–122.doi:10.1037//I089-2680.5.2.100

Morris, G., Baker-Ward, L., & Bauer, P. J. (2009). Whatremains of that day: The survival of children’s auto-biographical memories across time. Applied CognitivePsychology, 24, 527–544. doi:10.1002/acp.1567

Nelson, K., & Fivush, R. (2004). The emergence ofautobiographical memory: A social cultural devel-opmental theory. Psychological Review, 111, 486–511. doi:10.1037/0033-295X.111.2.486

O’Kearney, R., Speyer, J., & Kenardy, J. (2007).Children’s narrative memory for accidents and theirpost-traumatic distress. Applied Cognitive Psycho-logy, 21, 821–838. doi:10.1002/acp.1294

Peterson, C., Grant, V., & Boland, L. (2005). Childhoodamnesia in children and adolescents: Their earliestmemories. Memory, 13, 622–637. doi:10.1080/09658210444000278

Peterson, C., Warren, K. L., & Short, M. M. (2011).Infantile amnesia across the years: A 2-year follow-upof children’s earliest memories. Child Development,82, 1092–1105. doi:10.1111/j.1467-8624.2011.01597.x

Pillemer, D. B., & Kuwabara, K. J. (2012). Directivefunctions of autobiographical memory: Theory andmethod. In D. Berntsen & D. C. Rubin (Eds.),Understanding autobiographical memory: Theoriesand approaches (pp. 181–201). New York, NY:Cambridge University Press.

Reese, E., Haden, C. A., Baker-Ward, L., Bauer, P.,Fivush, R., & Ornstein, P. A. (2011). A multi-dimensional model of narrative coherence acrossthe lifespan. Journal of Cognition and Development,12, 424–462. doi:10.1080/15248372.2011.587854

Reese, E., Jack, F., & White, N. (2010). Origins ofadolescents’ autobiographical memories. CognitiveDevelopment, 25, 352–367. doi:10.1016/j.cogdev.2010.08.006

Tustin, K., & Hayne, H. (2010). Defining the boundary:Age-related changes in childhood amnesia. Devel-opmental Psychology, 46, 1049–1061. doi:10.1037/a0020105

Van Abbema, D., & Bauer, P. J. (2005). Autobio-graphical memory in middle childhood: Recollec-tions of the recent and distant past. Memory, 13,829–845. doi:10.1080/09658210444000430

Waters, T. E. A., Bauer, P. J., & Fivush, R. (2014).Autobiographical memory functions served by mul-tiple event types autobiographical memory func-tions. Applied Cognitive Psychology, 28(2), 185–195. doi:10.1002/acp.2976

Weigle, T. W., & Bauer, P. J. (2000). Deaf and hearingadults’ recollections of childhood and beyond. Mem-ory, 8(5), 293–309. doi:10.1080/09658210050117726

West, T. A., & Bauer, P. J. (1999). Assumptions ofinfantile amnesia: Are there differences betweenearly and later memories? Memory, 7(3), 257–278.doi:10.1080/096582199387913

Wheeler, M. A. (2000). Episodic memory and auton-oetic awareness. In E. Tulving and F. I. M. Craik(Eds.), The Oxford handbook of memory (pp. 597–608). New York: Oxford University Press.

“OWNING” THE PERSONAL PAST 17

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

APPENDIX

Example survey for one memory report to which participants responded online.

18 BAUER, HÄTTENSCHWILER, LARKINA

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5

“OWNING” THE PERSONAL PAST 19

Dow

nloa

ded

by [

Nic

ole

Hät

tens

chw

iler]

at 1

2:27

04

Febr

uary

201

5