Esta obra está bajo una licencia de Creative Commons Reconocimiento-NoComercial-SinObraDerivada 3.0 Unported.

Parent-child conversations in three urban middle-class contexts:

Mothers and fathers reminisce with their daughters and sons in

Costa Rica, Mexico, and Germany

Lisa Schröder

University Duisburg-Essen, Germany

Abstract. The present study investigated culture- and gender-differences in mother- and father-child reminiscing

with 3-year old daughters and sons in urban middle-class families from Costa Rica, Mexico, and Germany. Families of

the three contexts were overall similarly elaborative and children contributed a similar amount of memory elaborations.

However, context speciﬁ c use of different elaborative elements related to speciﬁ c elaborative styles. Compared to the

Latin American families, conversations in German families were least socially oriented. Across contexts, parents talked

more about social aspects with daughters than with sons. Costa Rican mothers and fathers were equally elaborative,

whereas German and Mexican mothers were more elaborative than fathers. We found similarities but also speciﬁ cities in

parent-child conversations about the past across these contexts with similar educational backgrounds.

Key words: Parent-child conversations, reminiscing, autobiographical memory, gender, culture, elaborations.

Resumen. El presente estudio investigó diferencias según cultura y género en las conversaciones sobre el pasado de

niños y niñas de 3 años de edad con sus progenitores en contextos urbanos en Costa Rica, México y Alemania. En los tres

contextos, el nivel de elaboración general de las conversaciones fue similar y los niños produjeron una cantidad similar

de recuerdos. Sin embargo, se encontraron diferencias culturales en algunos aspectos de elaboración. En comparación

con las familias latinoamericanas, las conversaciones de las familias alemanas estaban orientadas socialmente en menor

medida. En todos los contextos, los progenitores hablaron más acerca de aspectos sociales con las hijas que con los

hijos. Padres y madres costarricenses evidenciaron niveles de elaboración similares, mientras que las madres alemanas y

mexicanas mostraron niveles de elaboración mayores que los padres. Nuestros hallazgos indican la presencia de similitudes

y diferencias en las conversaciones sobre el pasado en familias con niveles educativos similares provenientes de distintos

contextos culturales.

Palabras clave: conversaciones padre-hijo, reminiscencia, memoria autobiográﬁ ca, género, cultura, elaboraciones.

Actualidades en Psicología, 27(115), 2013, 49-73

ISSN 2215-3535

Lisa Schröder, Faculty of Medicine, University Duisburg-Essen. E-mail: lisa.schroeder@uni-due.de

Heidi Keller, University of Osnabrück. E-mail:

Astrid Kleis, University of Osnabrück. E-mail:

Postal address: Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy, LVR-Klinikum Essen,

Wickenburgstr. 21, 45147 Essen, Germany.

Heidi Keller

Astrid Kleis

University of Osnabrück, Germany

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

Introduction

Cultural values and practices can be regarded

as adaptive strategies that have evolved due to

speciﬁ c contextual demands (Rogoff, 2003).

Contextual demands vary as a function of the

ecosocial characteristics of a context. Depending

on the respective ecosocial characteristics,

speciﬁ c cultural models can be deﬁ ned (Keller,

2011; Keller & Kärtner, 2013). In the present

study we investigated three cultural contexts

representing two cultural models: Middle-class

families from Costa Rica and Mexico on the one

hand and middle class families from Germany

on the other.

All families had high formal education;

however, varied in their household-types

(extended in the Latin American and nuclear in

the German context), the number of children,

and the maternal age at the ﬁ rst child’s birth.

According to the ecosocial conception of

cultural models (Keller, 2011; Keller & Kärtner,

2013) the Western, German middle class families

can be regarded as representing the model of

psychological autonomy, whereas the Latin

American middle class families from Costa Rica

and Mexico can be regarded as representing the

model of communal psychological autonomy and

relatedness. This means, that in both models the

child’s psychological autonomy

is emphasized;

however, in the Latin American contexts, it is

also important that children develop strong

family bonds and obligations. As previous studies

demonstrated, these orientations are reﬂ ected in

parental socialization goals, ethnotheories, and

early mother-child interactions of Costa Rican

and Mexican middle-class families compared to

German middle-class families (Kärtner, Borke,

Maasmeier, Keller, & Kleis, 2011; Keller, Borke,

Yovsi, Lohaus, & Jensen, 2005; Keller, Lamm,

Abels et al., 2006). In the present study we

looked at a social practice becoming particularly

relevant for child development with the onset

of language: Parent-child reminiscing during the

preschool years. We investigated the intersection

of culture and the individual characteristic

of gender in mother-child and father-child

past event conversations with daughters and

sons. Revealing parent, gender, and culture

speciﬁ c mechanisms of reminiscing style and

child development, is important to understand

developmental pathways; in addition, it can be

relevant to conceptualize sensitive intervention

programs for families in order to advance

children’s development of language-, and

literacy- skills (e.g., Reese, 1995; Reese, Leyva,

Sparks, & Grolnick, 2010).

Mother-child reminiscing across cultures

Talking about past events with children is a

common practice in many different cultures

(e.g., Eisenberg, 1985; Farrant & Reese, 2000;

Melzi, 2000; Mullen & Yi, 1995; Wang, 2001)

and inﬂ uences children’s understanding of

themselves as a person in time, embedded in

a cultural context (Conway & Pleydell-Pearce,

2000). Early reminiscing is a social, language-

based practice that is particularly revealing

to understand culture speciﬁ c socialization

practices and children’s developing self-concepts.

Cultural emphases are reﬂ ected in the way how

parents involve children in these conversations

on a structural level as well as in what they focus

on of an event.

So far, cross-cultural research on early

reminiscing has mainly focused on mother-

child conversations. These studies revealed that

In contrast to “action autonomy” in non-Western, rural farming

contexts, see Keller & Otto, 2011.

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

mothers vary in their degree of structuring

conversations more or less elaboratively (see

Fivush & Haden, 2003 for an overview). The

elaborative style is characterized by questions

(particularly open-ended Wh-questions) that

embellish the conversation with information

as well as regular positive feedback to child

contributions (e.g., Fivush & Fromhoff, 1988;

Reese & Fivush, 1993). This reminiscing style

provides a scaffold for children to jointly

reconstruct past events. In contrast, a low

elaborative style is characterized by a high degree

of repetitions (mothers repeating their own

utterances) and more closed questions (yes/no

questions). Children have fewer opportunities to

contribute to the conversation.

The high elaborative style is common in

Western middle-class families and has been

related to the socialization goal of autonomy:

children are encouraged to contribute their

own opinions and memories to conversations

and are thus treated as quasi-equal interlocutors

(e.g., Wang, 2007). In contrast, mothers of

Asian (mostly Chinese), African (Cameroonian),

and Oceanian (New Zealand Māori) middle-

class families have been reported as being less

elaborative (e.g., Reese, Hayne, & MacDonald,

2008; Schröder, Keller, Kärtner et al., 2013; Wang,

2001). The low elaborative style has been linked

to their socialization goals relating to compliance

with social roles: the child is following the lead

of the mother and learns cultural messages in

a teacher-novice format. This style is also and

particularly pertinent in rural farming families

with low formal education where children have

to grow into their social roles as fast as possible

(Schröder et al., 2013). Across and within

cultures, maternal elaborative style is related to

children’s memory contributions concurrently

and predictively (e.g., Farrant & Reese, 2000;

Reese, Haden, & Fivush, 1993; Schröder et al.,

2013; Wang, 2006; 2007).

Also when looking at the content level, i.e. what

mothers reminisce about, cultural differences

emerge in congruence with the underlying

socialization goals: mothers in Western middle-

class families talk about the child, as the central

character of an event and focus on the child’s

personal preferences and opinions, whereas

in non-Western contexts, mothers talk more

about other persons (i.e. social context) than

the child and teach moral values and rules (Han,

Leichtman, & Wang, 1998; Schröder, Keller,

Tõugu et al., 2011; Wang 2001; Wang, 2004;

Wang, Leichtman, & Davies, 2000).

Reminiscing in Latin American families

Latin American middle-class mothers

(Peruvian: Melzi, Schick, & Kennedy, 2004;

Costa Rican: Schröder et al., 2013) and working-

class families from Central American countries

living in the US (Melzi, 2000), are found to be

similarly elaborative when reminiscing with their

children compared to Western middle-class

families (see different results for low-income

Latino US families, Sparks, 2008). However,

mothers have been found to vary in their way to

use different elaborative elements(for a review of

reminiscing in Anglo middle-class families and

Latino families see Carmiol & Sparks, in press).

Melzi and colleagues (Melzi, 2000; Melzi et al.,

2004) have revealed that mothers with Latin

American background, structure conversations

with many open questions and let the child take

the narrator role, with mothers functioning

as participators or elicitors. In contrast, Euro-

American mothers ask less open -but more

closed- questions, and in that way contribute

also information of their own memories, thus,

serving as conarrators or constructors. On the

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

content level, conversations of Latin American

families -similar to Asian families- focus on

social aspects of events (i.e. are family-centered;

Cristofaro & Tamis-Lemonda, 2008) and talk

about emotions more often, compared to Euro-

American mothers (Eisenberg, 1999).

Differences between mothers and fathers

Mothers are typically studied in socialization

research since they are considered to be

the primary and most important caregivers.

However, fathers in middle-class families are

regarded as more and more important for child

development in general (e.g., Lamb & Tamis-

Lemonda 2004; Lamm & Keller, 2012) and child

language development speciﬁ cally (e.g., Pancsofar

& Vernon-Feagans, 2006; Tamis-LeMonda,

Shannon, Cabrera, & Lamb, 2004). Thus, in

the present study we looked at both mother-

child and father-child reminiscing. Studies with

fathers in the context of reminiscing are rare,

especially from a cross-cultural perspective. So

far, the ﬁ ndings comparing mothers and fathers

are heterogeneous for both dimensions: how

parents structure conversations and what they

talk about.

The structure of mother- and father- child reminiscing

For the conversational structure of parents,

some studies with Euro-American middle-

class families ﬁ nd fathers and mothers to be

equally elaborative (Reese & Fivush, 1993;

Reese, Haden, & Fivush, 1996). Other studies

with preschoolers but also older children, ﬁ nd

mothers to be more elaborative than fathers

(e.g., Zaman & Fivush, 2013). With respect to

children’s gender, some studies ﬁ nd mothers

equally elaborative with boys and girls within

different cultures (e.g., Fivush & Wang, 2005;

Melzi, 2000; Melzi et al., 2011; Schröder et al.,

2013). Yet in other studies with Euro-American

middle-class families, mothers and fathers were

found to be more elaborative with daughters

than with sons (Fivush, Berlin, McDermott-

Sales, Mennuti-Washburn, & Cassidy, 2003;

Reese & Fivush, 1993; Reese et al., 1996). Thus,

ﬁ ndings vary with respect to similarities and

differences of mothers and fathers in structuring

conversations with daughters or sons.

The content of mother- and father- child reminiscing

On the content level, parental differences

have especially been revealed with respect

to talk about emotions and social aspects of

events. Mothers in Euro-American middle-class

families have been found to refer to emotions

more often than fathers do (Fivush, Brotman,

Buckner, & Goodman, 2000), other studies

did not ﬁ nd such differences (Adams, Kuebli,

Boyle, & Fivush, 1995; Kuebli & Fivsuh, 1992).

Regarding children’s gender, studies revealed

mothers referring more often, to emotional as

well social aspects of events with their daughters

than with their sons (e.g., Buckner & Fivush,

2000; Fivush et al., 2003; see Fivush & Buckner,

2003 for a review). Also Mexican mothers

were found to talk more about emotions with

daughters than sons (Eisenberg, 1999) and Latin

American parents living in the US talked more

about socially oriented activities with daughters

and more about action-based events with sons

(Cristofaro & Tamis-Lemonda, 2008). Peruvian

mothers, however, talked more about emotional

aspects with sons than with daughters (Melzi &

Fernandez, 2004). Thus, the most robust result

may be a more social oriented focus during

reminiscing with daughters than with sons.

The present study and hypotheses

In the present study we aim to investigate the

structure and the content of parent-child past-

event conversations in three cultural middle-

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

class contexts representing two cultural models.

Based on previous studies with similar middle-

class contexts, we formulated several hypotheses:

How: Elaborative structure

H1. Overall elaborativeness: We expected that

contexts would not differ with respect to their

overall elaborativeness.

H2. Use of different elaborative elements: We

expected that Latin American parents would ask

more open questions (elicitation/participator

style) than the German parents, who in turn

were expected to ask more closed questions or

to make statements similar to Euro-American

middle-class families (conarrator/constructor

style).

H3. Child memory contributions: We expected

that children would contribute a similar amount

of memory elaborations across contexts.

What: Content

H4. Social orientation: We expected that

conversations of the Latin American parent-

child dyads would have a greater focus on social-

than nonsocial aspects of an event compared

to the German context. Furthermore, we

hypothesized that parents across contexts would

talk more about social aspects of events with

daughters than with sons.

With respect to mothers’ and fathers’

elaborativeness with their daughters or sons,

we had no explicit hypotheses due to the mixed

ﬁ ndings of previous studies.

Method

Participants

Families from three different urban cultural

contexts participated in this study: n = 19 families

from San José, the capital of Costa Rica; n =

14 families from Guanajuato, a city in Central

Mexico; and n = 22 families from Osnabrück,

a city in Northern Germany. In some families

it was not possible to assess conversations for

both parents. Because of that, the ﬁ nal samples

consisted of 18 mother- and father- child

conversations from San José, Costa Rica (28%

girls); 12 mother- and father- child conversations

from Guanajuato, Mexico (42% girls); and18

mother- and father- child conversations from

Osnabrück, Germany (56% girls). Due to

withdrawing consent, unavailability, or technical

problems, conversations were not available

of one father of the Costa Rican context,

one father of the Mexican context, and three

fathers and one mother of the German context.

Children were 36 months on average (ranging

from 34 to 37 months). The distribution of

children’s gender did not differ signiﬁ cantly

across contexts, χ²(2, N = 48) = 2.86, p = .24.

However, within contexts there were more

boys in the Costa Rican (13 of 18 children) and

Mexican (7 of 12 children) contexts, and more

girls in the German context (10 of 18 children).

Families from all three contexts held a high

formal education and accordingly, did not differ

in their years of formal schooling for mothers or

fathers. All parents had at least 10 years of formal

education and thus, at least secondary school

level (“secundaria”, “mittlere Reife”). In the

Costa Rican and Mexican contexts, household

sizes were signiﬁ cantly bigger compared to

the German families. Mothers of the German

context were signiﬁ cantly older when they had

their ﬁ rst child compared to mothers of the

Costa Rican and the Mexican context. In the

Costa Rican context 56% of children were

ﬁ rstborns, 36% of children were ﬁ rstborns in

the Mexican context, and all children of the

German context were ﬁ rstborns, χ

(2, N =

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

49) = 14.79, p < .01. For means and statistical

indices of the sociodemographic information

see Table 1.

Procedure

Families were visited at home by female trained

research assistants, either being native speakers

or bilingual Spanish - German speakers. Mothers

and fathers were asked to discuss two shared

past events with their daughter or son each.

In line with the procedure proposed by Reese

and Fivush (1993), the research assistants asked

parents to select events that had occurred within

the last four weeks, and that parent and child

had experienced together; these events should

be speciﬁ c and not routine events. Furthermore,

events were supposed not to contain a story line,

for example talking about the content of a book

or ﬁ lm. Parents were asked to choose the place

for this conversation and no time limit was set.

Mother- or father- and child pairs were left alone

in their place of choice during the conversations.

Conversations with each parent took place within

one week. The order of mothers or fathers

having conversations was ﬁ rst counterbalanced.

Conversations were audiotaped and subsequently

transcribed verbatim.

Coding

The coding procedure for parental structuring

of conversations during reminiscing was adopted

from Reese and Fivush (1993).

Type of talk. Four different types of talk were

differentiated: a) Event talk: all talk directly

referring to the speciﬁ c shared past-event in

past tense; e.g., “What happened at the lake?”;

b) Associative talk: all talk related to the event;

e.g., general knowledge like “how is your cousin’s

name?”, talk related to another past event or the

future, e.g., “I want to go there again.”, or the

present; e.g.; “Where is the present you got now?”;

c) Meta-talk: all talk about the situation itself; e.g.,

“I want to talk about our visit to grandma now.”

or about memory performance; e.g., “You can’t

remember?”; and d) Off-topic talk: all talk about

aspects not related to the past event: e.g., “Sit

down”, “I want to go out now”, “I am hungry”.

Table 1

Mean differences of sociodemographic information among samples

Osnabrück

M (SD)

San José

M (SD)

Guanajuato

M (SD)

F-value

(η

)

(error, sample)

Maternal years of formal education 14.50

(3.75)

12.61

(2.43)

15.17

(4.13)

2.40, n.s. (45, 2)

Paternal years of formal education 14.39

(3.54)

15.35

(3.10)

15.64

(1.86)

.71, n.s. (45, 2)

Maternal age when having ﬁ rst child 30.67

(3.60)

25.40

(4.24)

26.33

(6.29)

6.54** (.23) (46, 2)

Number of persons living in household 3.56

(0.92)

4.44

(1.29)

4.75

(1.14)

4.79* (.18) (47, 2)

Percentage of ﬁ rst borns 100 % 56 % 36 %

Note. Means of samples not sharing indexed letters differ signiﬁ cantly from each other.

* p < .05, ** p < .01

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

All types of talk were coded based on

propositions (i.e. independent sentences) with

the exception of off-topic talk that was coded per

turn for parent and child. For each conversation

the starting point was set when the parent asked

the ﬁ rst speciﬁ c question (or made a speciﬁ c

statement) about the event under discussion;

thus, coding for a conversation always started

with event talk. A conversation was deemed as

ﬁ nished when only meta-talk or off-topic talk

continued. Thus, conversations always ended

with event talk or associative talk. Only within

event talk, speciﬁ c codes were differentiated

(e.g., different types of questions). Associative

talk was not differentiated for different types

(thus, associative talk was applied to questions,

conﬁ rmations etc.).

Elaborative structure of event talk. Coding unit

for the elaborative structure of event talk were

independent sentences (i.e., propositions). This

means that a structure code was applied for

each independent clause (potentially including

dependent clauses). For example both of the

following sentences received one structure code:

“Do you remember where we were yesterday?”

“Do you remember what we did yesterday when

we visited grandma and grandpa?”

The coding scheme within event talk (adopted

from Reese & Fivsuh, 1993) was exclusive

and exhaustive. For the present study we only

consider different elaborative elements as well

as repetitions. Following, the here analyzed

structure codes on the child’s and parent’s part

are outlined:

A) Child memory elaborations: Child

utterances that contained new information

about the speciﬁ c past event that had not been

mentioned before by mother/father or child,

were coded for child memory elaboration.

This code could also apply to just one word

said by the child in case he/she did not build

a whole sentence; e.g., M: “Who was there?”

C: “Lena.” In this case the child received one

memory elaboration (if this person had not

been mentioned before).

B) Parental elaborations: Elaborations -in

contrast to repetitions- contain new information

about the event under discussion that had not

been mentioned before neither by the parent

nor the child. Elaborations thus embellish the

conversation with new aspects. Elaborations

were coded for different elements:

a. Open questions: All open question

elaborations, meaning Wh-questions, fell

into this category; e.g.: M: “What did we

do yesterday?”; F: “Where did you play

yesterday?”. Thus, children had to answer

with a word or sentence instead of just “yes”

or “no”.

b. Closed questions: All embellished questions

only demanding a “yes” or “no” answer of the

child; e.g., F: ”Did you swim yesterday?”; M:

“Do you remember where we have been at the

weekend?”. The last example demonstrates

that also questions containing a Wh-question-

word, could be built as a closed question.

c. Statements: All embellished utterances

not being a question but just a statement

fell into this category; e.g., F: “Yesterday we

went to the lake.” Tag-questions also fell into

this category, e.g., M: “You were swimming,

weren’t you?”.

d. Conﬁ rmations: This code was applied

to general positive feedback (evaluation),

like “Yes”, “Right”, “That’s correct” etc.

Furthermore, repetitions of the child’s

answer also fell into this category; e.g.: C: “We

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

played in the sand.” M: “You were playing in

the sand?”. Mostly, repetitions of the child

answers occur as closed questions. However,

they have a different function compared to

closed elaborative questions. Repeating the

answer of an interlocutor, functions as a

conﬁ rmation in that it demonstrates interest

and active listening.

C) Parental repetitions: Parental utterances

(statements or questions) that repeated

information the mother or father had mentioned

before; e.g.: F: “What did we do yesterday?”

C: “Don’t know” F: “Tell me what we did

yesterday?”. Thus, only repetitions of their own

utterances –not the child’s utterances- fell into

this category. Also prompts not leading to a new

aspect and containing no new information were

coded for this category; e.g.: “Tell me about it.”,

“Say.”, “Tell.”.

For ﬁ nal analyses we collapsed all elaborative

elements (a.-d.) and calculated a difference

score by subtracting repetitions from the

total elaboration score. Thus, we achieved

an elaboration-repetition difference score as

an overall indicator for elaborativeness. In

addition, the different elaborative elements were

considered separately in the following analyses.

Content. The content was coded independently

of the structure for each subject-verb unit

within event talk as well as associative talk.

This means, in contrast to structure coding,

more than one content code could be applied

within an independent sentence, namely also for

dependent sentences: “What did we do when the

sun was starting to shine?” In this example two

content codes were applied, one for “What did

we do…” and one for “when the sun…”. For

the present study we only consider two content

codes relevant to our study: social context and

nonsocial context (similar to Buckner & Fivush,

2000). We thought of these two codes as a strong

indicator of social orientation.

A) Social context references: This code was

applied for parental and child utterances referring

to other persons but not the mother or the child

as the subject of a sentence; e.g., M: “What was

your cousin doing?” C: “Climbing.” In this case,

both mother and child would have received

one code for social context each because the

utterances are referring to another person as the

subject (implied by the verb in the child’s case)

of the sentence.

B) Nonsocial context references: This code

was applied for parental and child utterances

not referring to persons but referring to objects,

food, animals, numbers, physical appearances

or attributes; e.g., F: “What color did the car

have?” C: “Red.” In this case also both, father

and child would have received a code for

nonsocial context.

Also here, we calculated a difference score

for social context references minus nonsocial

context references as the ﬁ nal measure. Thus,

the higher the score, the more socially oriented

-relative to nonsocial aspects- conversations

were, with a score close to zero indicating an

equal talk about social and nonsocial aspects.

Reliability. Two research assistants coded

independently 20% of the transcripts from each

context and inter-rater reliability was calculated.

Transcripts of the Osnabrück context were

coded in German; the Costa Rican and Mexican

conversations were transcribed in Spanish ﬁ rst

and then translated into English by bilingual

research assistants. English transcripts were

coded for these contexts as trained coders were

ﬂ uent in German and English. Cohen’s kappa

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

was k = .89 on average for structure coding

and k = .87 for content coding across contexts.

Discrepancies were resolved by discussion.

The same two research assistants coded the

remaining transcripts.

Results

Plan of Analyses

In preliminary analyses we compared the

number of events talked about by mothers

and fathers across contexts and explored all

variables for extreme outliers. Furthermore,

we explored differences in the length of

conversations and the type of talk referred to

during reminiscing (event-, associative-, meta-,

and off-topic- talk) in order to see whether the

task of reminiscing was performed similarly

across contexts.

For main analyses we conducted several

repeated measure ANOVAs and MANOVAs

with cultural context (3 levels) and children’s

gender (2 levels) as the between subject-factors,

and parent (mother or father) as the within-

subject factor for all analyses. According to

our hypotheses, we analyzed 1) the overall

elaborativeness of parents (relative to

repetitiveness) with the elaboration-repetition

difference score as the dependent variable;

2) investigated differences in the parental use

of different elements of the elaborative style

(open questions, closed questions, statements,

and conﬁ rmations); 3) differences in the

amount of child memory elaborations and 4)

the social orientation of conversations based

on the social- nonsocial- context difference

score. Finally, we calculated simple Pearson’s

correlations (two-tailed) across and among

children’s memory elaborations and the parental

elaborativeness within each context.

Since violating sphericity is critical when

conducting repeated measure analyses of

variance with small sample sizes, degrees of

freedom were corrected with the Greenhouse-

Geisser method in analyses with more than one

within subject factor. Uncorrected F-values

with corrected p-values and effect-sizes (partial-

eta squared: η

) are reported. Signiﬁ cant

main and interaction effects were followed by

post-hoc t-tests with LSD (lowest signiﬁ cant

difference), with a signiﬁ cance level of p < .05.

Marginally signiﬁ cant effects with p < .10 were

also followed up in case the effect size was at

least medium (>.06; Cohen 1988).

Preliminary Analyses

Number of events. Preliminary results

demonstrated that the number of events

discussed by mothers varied from one to three

events. The distribution of number of events

differed signiﬁ cantly among contexts for

mothers, χ

(4, N = 49) = 10.58, p < .05. In the

Costa Rican context, 78% (n = 14) of mothers

discussed two events, 17% (n = 3) one event,

and one mother discussed three events. In the

Mexican context, half of the mothers discussed

one event with their children and the other half,

two events. In the German context all mothers

discussed two events with their children, except

one mother talking about one event. The number

of events discussed by fathers varied from one

to four events; however, the distribution did not

vary signiﬁ cantly among contexts.

In order to keep means comparable, we

divided all codes by the number of discussed

events, for mothers and fathers. Thus, all means

analyzed and illustrated refer to means per event

(see Reese & Fivush, 1993).

Outliers. Next, we explored variables for

extreme outliers; meaning for upper or lower

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

outer fence values (deﬁ ned by a coefﬁ cient =3).

For all child and parental variables we found

three extreme outliers in the German context

(for maternal and paternal open questions

elaborations, and paternal conﬁ rmations, all

from different participants) and four extreme

outliers in the Mexican context: three from the

same father (his open question elaborations,

statement elaborations and repetitions) and one

for open question elaborations of a mother. All

extreme outliers were upper fence values. Due

to the small context sizes we did not want to

exclude these participants. Instead, we decided

to replace outliers by the next closest value as

suggested by Tukey (1962) cited in Sachs (2003).

In this way the direction of the outlier is still

taken into account but not as an extreme.

Conversational length and type of talk. In

order to compare the total length of coded

conversations, and the different types of talk

(event-, associative-, meta-, and off-topic- talk),

we computed a repeated 3 (cultural context) x

2 (children’s gender) measure MANOVA with

parent and type of talk (including parental as well

as child variables) as the within subject factors. A

signiﬁ cant main effect was revealed for type of

talk, F(72, 2)

= 253.40, p < .001, η

= .86. Post-

hoc tests revealed across contexts, that parent-

child event talk (M = 42.40, SD = 2.10) was

signiﬁ cantly higher than all other types of talk,

whereas meta-talk referred to by parents and

children, was signiﬁ cantly lower compared to all

other types of talk. The amount of associative

talk and off-topic talk did not differ signiﬁ cantly

(associative: M = 8.09, SD = 6.89; meta-talk: M =

2.62, SD = 2.56; off-topic: M = 7.40, SD = 7.72).

A marginally signiﬁ cant main effect for context

revealed that conversations of the Mexican

parent-child dyads were signiﬁ cantly longer

than conversations of the German and Costa

Rican dyads, which did not differ from each

other, F(2, 42) = 3.07, p = .06, η

= .13. Longer

conversations of the Mexican conversations were

due to including signiﬁ cantly more associative-,

meta-, and off- topic talk compared to the other

two contexts but not event talk, which did not

differ signiﬁ cantly among contexts (see Figures

1 and 2 for proportions and means).

A marginally signiﬁ cant main effect per

parent demonstrated that there was a trend for

mother-child conversations being overall longer

than father-child conversations across contexts,

F(1, 42) = 3.24, p = .08, η

= .07. A signiﬁ cant

interaction effect for Type of Talk x Parent

revealed that this effect was based on mother-

child dyads talking signiﬁ cantly more about the

speciﬁ c past event (event talk: M = 47.87, SD =

17.69) than father-child dyads (M = 36.47, SD =

19.54), F(78, 8) = 2.14, p < .05, η

= .18. Parent-

child conversations did not differ signiﬁ cantly

from each other with respect to associative talk

(mother-child: M = 7.58, SD = 7.61; father-child:

M = 8.61, SD = 9.39), meta-talk (mother-child:

M = 3.04, SD = 3.69; father-child: M = 2.20,

SD = 2.25), or off-topic talk (mother-child: M

= 7.46, SD = 10.78; father-child: M = 7.35, SD

= 9.52).

Taken together, across the three contexts,

mothers and fathers, and daughters and sons,

the proportion of event talk accounted for the

biggest part of conversations, indicating that

families conformed to the task to reminisce

about past events (see Figures 1 and 2).

Main Analyses

Conversational Structure. As outlined in the

method part, all codes analyzed in the following

were only applied within event talk, i.e. when

mother- or father- child dyads were talking

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

51,44

42,05

51,21

5,19

7,06

11,92

2,42

2,06

5,46

6,78

2,89

15,33

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Osnabrück

San José

Guanajuato

Off -topic talk

Meta -memory talk

Associative talk

Eventtalk

Figure 1.

Percentages and frequencies (indicated on bars) of different type of talk during

mother-child conversations within each context..

Note. Number on bars indicate mean frequencies per event.

Figure 2.

Percentages and frequencies (indicated on bars) of different type of talk during

father-child conversations within each context.

Note. Number on bars indicate mean frequencies per event.

30,33

41,64

37,9

7,29

7,48

12,28

0,94

2,76

3,25

6,7

13,35

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Osnabrück

San José

Guanajuato

Off -topic talk

Metamemory talk

Associative talk

Eventtalk

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

about the specific past event in past tense.

The coding unit was proposition (i.e.

independent sentences).

Overall elaborativeness. First, we investigated

the overall elaborative structure by analyzing

the difference score of total elaborations minus

repetitions (see method part) by conducting

a repeated 3 (cultural context) x 2 (children’s

gender) measure ANOVA with parent as the

within subject factor and the elaboration-

repetition difference score as the dependent

variable. Across parents, the difference score

did not differ among contexts. In line with our

hypothesis, parents were thus, overall having

similarly elaborative conversations with their

children in the three cultural contexts. However,

across contexts, mothers were overall more

elaborative than fathers; with a main effect of

parent, F(1, 42) = 6.45, p < .05, η

² = .13. A

marginally signiﬁ cant interaction effect for

Context x Parent, F(2, 42) = 2.57, p = .09, η

= .11 revealed that mothers’ difference score in

the German and Mexican contexts was higher

than fathers’ difference score, whereas in the

Costa Rican context mothers and fathers were

equally elaborative. Mothers of the German

and Mexican context were adopting the most

elaborative style; fathers of these contexts, the

least elaborative style, and mothers and fathers

of the Costa Rican context, both lay in-between

(see Table 2 for all means). There was no effect

for children’s gender.

Elaborative elements. Next, we explored the use of

different elaborative elements. For this purpose

we calculated a repeated 3 (cultural context) x

2 (children’s gender) measure ANOVA with

parent (2) and structure (4) as the within subject

factors. There was a main effect for parent with

F(1, 42) = 4.77, p < .05, η

² = .10 demonstrating

that the use of elaborative elements was higher

in mothers than in fathers across contexts. This

indicates that the higher elaboration-repetition

difference score in mothers was not due to

fathers being more repetitive, but rather using

less elaborative elements.

Furthermore, there was as expected a

signiﬁ cant interaction effect for Elaboration-

Element x Context, F(6, 80) = 2.62, p < .05,

²= .11. This effect was due to context speciﬁ c

use of different elaborative elements as follow-

up t-tests demonstrated. Parents of both Latin

American contexts used more open questions

compared to the German context (Costa Rica:

t(34) = 2.00, p = .05, d = .67; Mexico: t(28)

= 2.14, p = .04, d = .80). Costa Rican and

German parents conﬁ rmed child contributions

to a similar degree and both more often than

Mexican parents (Costa Rica: t(28) = 1.70, p =

.10, d = .63; Germany: t(28) = 2.02, p = .05, d

= .75). Furthermore, parents of the German

context made more elaborative statements than

Costa Rican parents t(34) = 1.89, p = .07, d =

.63, whereas Mexican parents’ use of statements

did not differ from any of the two contexts and

lay thus, in-between. Despite Mexican parent’s

use of closed questions was quite high (see

Figure 3), contexts did not differ signiﬁ cantly in

asking closed questions during conversations (ps

> .10). Thus, compared to the other contexts,

German parents made many elaborative

statements and gave positive feedback, (i.e.

conﬁ rmations) and asked less open elaborative

questions. Costa Rican parents asked many open

elaborative questions and regularly conﬁ rmed

child contributions, but made few elaborative

statements. Mexican parents also asked many

open elaborative questions, made elaborative

statements to a medium degree and were low in

conﬁ rming child contributions (see Figure 3).

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

In line with our hypothesis, these results

indicate culture speciﬁ c reminiscing styles that

can be related to the constructor and elicitor

styles (Melzi et al., 2011). There was no signiﬁ cant

effect for gender, nor an interaction effect for

parent, indicating that parents were similar in

adopting these styles within contexts and did so

similarly with girls and boys.

In order to substantiate our hypotheses,

we also tested for signiﬁ cant differences

within contexts. In line with previous research

(Melzi et al., 2011), we tested for intra-cultural

differences between parents’ request of

information (through open or closed elaborative

questions) and parents provision of information

(elaborative statements). In both the Costa Rican,

t(17) = 7.71, p < .001, d = 2.64, and Mexican,

t(11) = 5.70, p < .001, d = 2.43, contexts, parents

asked signiﬁ cantly more elaborative questions

compared to providing information through

statements. In the German context, parents

requests and provision of information did not

differ signiﬁ cantly from each other, t(17) = 1.43,

p > .10.

Child Memory Elaborations. In order to

investigate children’s contributions of memory

elaborations during conversations, we calculated

a repeated 3 (cultural context) x 2 (children’s

gender) measure ANOVA with parent as

the within subject factor and child memory

elaborations as the dependent variables.

As expected, there was no effect for cultural

context. There was a marginally signiﬁ cant

interaction effect for Context x Parent with F(2,

42) = 2.59, p = .09, η

² = .11. This interaction

effect was due to children of the German

context contributing signiﬁ cantly more memory

elaborations when reminiscing with their

mothers than with their fathers, t(17) = 2.89, p

< .05, d = 1.37; in the Costa Rican and Mexican

Osnabrück

San José

Guanajuato

Mean frequencies per event

pos. feedback

open elab.

closed elab.

statement elab.

Figure 3.

Mean values and standard-error for different elaborative elements as a

function of cultural context.

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

Table 2

Means (Standard deviations) for conversational structure codes

Osnabrück, Germany San José, Costa Rica Guanajuato, Mexico

Girls Boys Total Girls Boys Total Girls Boys Total

Mother-child

Elaboration-repetition

difference score

21.25

(9.04)

21.31

(9.82)

21.28

(9.12)

14.90

(6.20)

15.12

(7.86)

15.06

(7.26)

19.13

(14.48)

16.44

(9.74)

17.33

(10.92)

Total Repetitions

3.30

(1.95)

4.25

(2.41)

3.72

(2.15)

6.90

(5.16)

4.45

(4.41)

5.13

(4.61)

4.38

(3.94)

8.31

(5.06)

7.00

(4.93)

Conﬁ rmations

8.05

(3.17)

6.94

(2.23)

7.56

(2.77)

5.20

(3.40)

6.37

(3.43)

6.05

(3.36)

4.88

(5.72)

5.06

(2.67)

5.00

(3.67)

Elab. Open Questions

5.50

(1.87)

5.31

(2.27)

5.42

(1.99)

7.00

(2.15)

6.19

(3.79)

6.42

(3.37)

5.63

(4.23)

7.56

(1.82)

6.92

(2.81)

Elab. Closed Questions

3.70

(3.28)

4.50

(3.02)

4.06

(3.10)

5.40

(3.73)

3.71

(2.25)

4.18

(2.73)

4.63

(3.09)

7.31

(3.28)

6.42

(3.35)

Elab. Statements

7.30

(5.77)

8.81

(7.55)

7.97

(6.46)

4.20

(1.92)

3.29

(2.89)

3.55

(2.63)

8.38

(5.41)

4.81

(2.84)

6.00

(4.02)

Child memory

elaborations

8.80

(4.72)

5.63

(3.04)

7.39

(4.27)

6.50

(2.96)

5.10

(2.76)

5.49

(2.80)

3.38

(3.64)

4.56

(3.34)

4.17

(3.33)

Father-child

Elaboration-repetition

difference score

14.12

(8.09)

12.56

(11.72)

13.43

(9.58)

18.00

(2.76)

13.37

(8.48)

14.66

(7.56)

15.21

(11.65)

8.90

(5.74)

11.00

(8.22)

Total Repetitions

2.37

(2.85)

3.83

(3.66)

3.02

(3.22)

1.92

(0.96)

6.63

(5.17)

5.32

(4.88)

8.67

(4.97)

6.15

(4.19)

6.99

(4.41)

Conﬁ rmations

4.15

(3.05)

3.85

(3.02)

4.02

(2.95)

5.67

(2.26)

5.42

(4.67)

5.49

(4.08)

4.00

(4.97)

3.07

(1.88)

3.38

(3.93)

Elab. Open Questions

4.30

(4.00)

2.69

(1.46)

3.58

(3.17)

4.63

(0.42)

5.91

(3.94)

5.56

(3.37)

5.38

(2.69)

5.07

(2.29)

5.17

(2.31)

Elab. Closed Questions

3.53

(2.74)

5.00

(7.57)

4.19

(5.31)

4.33

(1.03)

4.54

(2.43)

4.48

(2.10)

7.88

(7.64)

3.36

(2.61)

4.87

(5.02)

Elab. Statements

4.50

(3.65)

4.85

(5.06)

4.66

(4.20)

5.28

(2.53)

4.13

(3.53)

4.45

(3.25)

6.63

(1.60)

3.53

(1.82)

4.56

(2.27)

Child memory

elaborations

4.17

(3.30)

3.56

(3.58)

3.90

(3.34)

4.87

(2.14)

5.82

(5.62)

5.56

(4.85)

5.38

(6.37)

3.92

(2.76)

4.40

(4.05)

Note. Some mean values include replaced outliers (see p. 58)

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

contexts, children were contributing as many

memory elaborations when reminiscing with

their mothers as when reminiscing with their

fathers. These results conform to the parental

style differences across contexts and parents:

mothers in the German context being overall

more elaborative than fathers and mothers, and

fathers in the Costa Rican context, being equally

elaborative. However, results for the Mexican

context fall out of this correspondence to the

parental style. Also for this analysis, no gender

effects were revealed.

Content of Conversations. As outlined in the

method part, all codes analyzed in the following

were applied within event talk, and associative

talk. The coding unit was subject-verb

constructions (i.e. independent and dependent

sentences separately).

Social orientation. In order to investigate

the social orientation of conversations, we

calculated two repeated 3 (cultural context) x

2 (children’s gender) measure ANOVAs with

parent as the within subject factor and social-

nonsocial- context difference score as the

dependent variable for parents and children

respectively. There were signiﬁ cant main effects

for the analysis of parents’ conversational social

orientation, however, no signiﬁ cant main effects

or interaction effects were revealed for children’s

social orientation in conversations (German

context: M = -2.09, SD = 4.30; Costa Rican

context: M = -1.36, SD = 4.19; Mexican context:

M = -2.01, SD = 4.04).

For parents there were signiﬁ cant effects for

cultural context, F(2, 42) = 6.17, p < .01, η

² =

.23 and for children’s gender, F

(

2, 42)

= 5.10, p

< .05, η

² = .11. No other effects were revealed.

Post-hoc tests for cultural context demonstrated

that relative to nonsocial contexts, parents of

the German context (M = -.79, SD = 7.47)

referred to social contexts signiﬁ cantly less often

than parents of the Mexican context (M = 7.64,

SD = 9.60). The social orientation in parents of

the Costa Rican context lay in-between by not

differing from any of the two other contexts (M

= 3.47 SD = 8.21). Thus, our hypothesis was

conﬁ rmed that conversations in the German

families were least socially oriented. For children’s

gender, post-hoc tests revealed as expected

that across contexts, the social orientation was

signiﬁ cantly higher in conversations with girls

(M = 7.80, SD = 12.55) compared to boys (M

= 2.57, SD = 12.66). However, the standard

deviations indicate high intra-cultural and intra-

gender variability.

Correlations of parental and child variables

All correlations were conducted within each

context separately applying simple Pearson’s

correlations (two-tailed). Due to the small

context sizes, more sophisticated statistics like

regression analyses were statistically not feasible.

Because of high intercorrelations of the

parental structure variables, we only conducted

a correlation for overall elaborativeness of

parents with child memory elaborations. Child

elaborations were highly correlated with the

elaboration-repetition difference score across

contexts and parents (between rs = .74 and .87,

ps < .001), except for mother-child conversations

in the Mexican context (r = .18, n.s.).

A signiﬁ cant relation for maternal and paternal

elaborativeness was revealed within the Costa

Rican context (r = .51, p < .05), but not the

German (r = .38, n.s.) and Mexican (r =-.17, n.s.)

contexts. Thus, in contrast to the German and

Mexican families, children in the Costa Rican

families, experienced a consistently (higher or

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

lower) elaborative reminiscing style independent

of who they talked to, mother or father.

Child memory elaborations contributed during

mother-child and father-child conversations,

were not signiﬁ cantly correlated in any context.

However, the correlation coefﬁ cient was highest

in the Costa Rican context (r = .34, p = .17)

compared to the German (r = .11, n.s.) and

Mexican (r = .12, n.s.) contexts.

Discussion

In the present study we investigated the

elaborative structure and social orientation of

mother-child and father-child conversations

about shared past events in three urban middle-

class contexts when children were three years

of age. So far, cross-cultural studies have

mainly focused on mother-child conversations;

however, especially in urban middle-class

contexts, the function of fathers as socialization

agents has increased over the last decades (e.g.,

Lamb & Tamis-Lemonda, 2004). We investigated

contexts representing two different cultural

models: the model of psychological autonomy in

Western middle-class families from Osnabrück,

Germany; and the model of communal

psychological autonomy and relatedness: non-

Western middle-class families from San José,

Costa Rica and Guanajuato, Mexico.

In line with our hypotheses, families of the

three contexts did not vary in their overall

elaborativeness and children contributed a

similar amount of memory elaborations across

contexts. As expected, parents differed in

their use of speciﬁ c elaborative elements that

can be related to the elicitor and constructor

styles. Furthermore, conversations in German

families were least socially oriented compared

to conversations of the Latin American families

in line with the underlying cultural models.

Results also conﬁ rmed the hypothesis that

across contexts, conversations would be more

socially oriented with girls than boys. Regarding

differences between maternal and paternal

reminiscing, results were heterogeneous across

contexts. In the following, all results will be

discussed in detail.

Overall elaborativeness and different elaborative

elements

Overall, there were no cultural differences

in conversational length or elaborativeness, in

line with our hypotheses and previous studies

investigating mother-child reminiscing in Latin

American and Western Euro-American or

European mothers with similar socioeconomic

backgrounds (Melzi, 2000; Schröder, Keller,

Kärtner et al., 2013; however, different to

results of Melzi et al., 2011: Peruvian mothers

having longer conversations than Euro

American mothers). As expected, different

elaborative elements were predominantly used

in each cultural context, forming culture speciﬁ c

elaborative styles (as found for other cultural

contexts; see Melzi, 2000; Schröder et al., 2011;

Tõugu, Tulviste, Schröder, Keller, & DeGeer,

2011). In the following, we will illustrate the

differences among contexts with transcript-

excerpts (“M:” indicating maternal turns;

“F:” indicating paternal turns; “C:” indicating

children’s turns; within conversations: “xxx”

indicate ineligible words; “***” indicate a pause;

(italic words) in brackets refer to nonverbal

aspects or additional information).

Compared to the other contexts, the parental

reminiscing style of the Costa Rican families

was characterized by many open elaborative

questions, regular conﬁ rmations and few

elaborative statements. Following, the excerpt

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

of a Costa Rican mother-child conversation is

illustrated as an example:

M: Do you remember when we were bathing the

doll? Closed elaborative question

C: yees, I remember.

M: And with… what did we put on the doll? Open

elaborative question

C: My shampoo! (Laughs)

M: Your shampoo? Conﬁ rmation

And why did we use shampoo? Open elaborative

question

C: Me full of shampoo made.

M: Full with shampoo? Conﬁ rmation

C: Yes.

M: And what did you do? Open elaborative question

What did you help me with? Open elaborative question

C: The doll was scared of the water.

M: The doll was scared of the water? Conﬁ rmation

And the doll was shaking herself, right? Statement

elaboration

C: And made wet us.

M: Made us wet? Ah, yes. Conﬁ rmation (2)

(…) Associative talk (3 turns)

M: What happened? Open elaborative question

C: She just hurt herself.

M: But where did she hurt herself, where exactly?

Open elaborative question

C: (no response)

M: Did she cry? Open elaborative question

C: No, the doll is brave.

In this example the conversational course

is characterized by the mother asking an open

elaborative question and then conﬁ rming the

child’s contribution by repeating the child’s

answer. The mother is giving positive feedback

even when the child’s response is not relating to

what she had asked for. Thus, she follows what the

child wants to talk about and does not –as in the

German context- contribute her own thoughts

or memories through elaborative statements.

This style corresponds to the typical pattern of

a high elaborative style (see Fivsuh et al., 2006

for a review). Also, other studies found Latino

mothers to ask more open questions than Euro-

American mothers (Melzi, 2000). Furthermore,

parents were requesting more information than

providing information when looking at intra-

cultural differences. This style corresponds to

the “elicitors style” (or high-eliciting; Haden et

al., 2009) in Latino mothers when they distance

themselves as the audience and let their children

take the role of the narrator (Melzi, 2000; Melzi

et al., 2011).

In the Mexican families, parents also provided

many open elaborative questions but gave less

positive feedback (conﬁ rmations), and were

providing elaborative statements to a medium

degree compared to the other two contexts.

Following, the excerpt of a Mexican father-child

conversation is illustrated as an example:

F: Do you remember when you were playing with

them (meant are cousins)? Closed elaborative question

C: Mmmm.

F: What did you play? Open elaborative question

C: Mmmm.

F: To kick the ball. Elaborative statement

C: Mmm.

F: Not? Prompt (closed)

C: No.

F: What did you play? Repetition

C: With the scooter.

F: With the scooter? Conﬁ rmation

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

C: Yes.

F: And who went on the scooter? Open elaborative

question

C: Me.

F: And who else? Prompt (open)

C: Co Cechi.

F: Who from? From Ceci? Associative talk

C: (no response)

M: And the three little ones. Elaborative statement

C: Yes.

F: And where did you start from? Open elaborative

question

C: xxx there.

F: On the platform. Elaborative statement

And did you have lots of fun with them? Closed

elaborative statement

And what else were you playing? Open elaborative

question

C: With the ball.

Like in the Costa Rican mother-child example,

the father asks many elaborative questions. In

contrast to the conversational course of the

Costa Rican mother, he is though conﬁ rming

the child’s answer just one time. In the other

two cases he goes on with asking another

question without giving feedback to the child’s

contribution. Elaborative statements are made

by the father as well. Looking at intra-cultural

results, parents requested more information

than they provided themselves like in the Costa

Rican families. Thus, parents were also rather

taking an elicitor role (Melzi et al., 2011). Styles

still differed between the two Latin American

contexts conﬁ rming heterogeneity of Spanish-

speaking Latin American populations as

claimed for other narrative contexts (Caspe &

Melzi, 2008).

Compared to the other contexts, parents in

the German context made many elaborative

statements and gave regularly positive feedback

to their children’s contributions, but they asked

less open elaborative questions. Following, the

excerpt of a German mother-child conversation

is presented as an example:

M: On the playground, what was there? Open

question elaboration

C: A amaziling slide!

M: A very amazing slide? Conﬁ rmation

C: A tunnel-slide!

M: A tunnel-slide? Conﬁ rmation

C: Yes.

M: Aha. How did you crawl inside? Open question

elaboration

C: Yees. But I was, em, like em, like like backwards

sliding.

M: You can already slide backwards? Associative

talk (conﬁ rmation)

C: Yeeeeees.

M: (laughs) super! *** Associative talk (conﬁ rmation)

You know what I liked a lot? Closed elaborative

question

C: What?

M: The frog. Statement elaboration

C: In there?

M: No, not not of this game; Statement elaboration

but on the playground there was a huge frog.

Statement elaboration

Do you remember that Dad was sitting next to it?

Closed elaborative question

C: Yes. There also was, there is a frog inside.

M: Yes. *** Conﬁ rmation

And we saw Max and Moritz, huh? Statement

elaboration

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

C: Aaahaaa.

M: That was in the garden of fairy tales. Statement

elaboration

In this example, the mother provides many

elaborative statements and regularly conﬁ rms

the child’s contributions. In contrast to the

examples of the Latin American contexts, she

asks less open elaborative questions. This result

-mothers making more elaborative statements

and conﬁ rmations compared to other contexts-

has also been revealed for the German middle-

class context of Berlin (Tõugu et al., 2011).

Parents requested and provided information to

a similar degree. This characteristic corresponds

to the reminiscing pattern described by Melzi

and colleagues (Melzi, 2000; Melzi et al., 2011)

for Euro-American families: mothers being

“conarrators” (constructors) i.e. they also

provide information concerning their own

knowledge to structure the conversation as

they treat their children as equal interlocutors

to foster their autonomy. In the investigation by

Melzi (2000), however, mothers of the Euro-

American families were high in asking closed-

ended questions rather than making elaborative

statements as in the present study

. Interestingly,

when inspecting transcripts of the Osnabrück

context, we noticed that children were also

asking questions to their parents (as seen in

the example above). Thus, parents making

elaborative statements might be related to a

bidirectional effect of the way children involve

their parents in the conversation.

Although the context speciﬁ c styles have

exemplarily been described and illustrated

with individual excerpts, the respective style

characteristics represent the mean differences

for each element on a variable oriented level.

The described patterns must not necessarily

exist on an individual level, when determining

each parent’s pattern (being high or low in

different elements; like in other studies, Haden

at al., 2009; Schröder et al., 2013).

Child Memory Contributions

Overall, children’s provision of memory

elaborations did not differ among the three

contexts despite parents used the elaborative

elements to different degrees across contexts.

This might be due to reciprocity effects

between children’s and parent’s conversational

behavior. For example, some children might

contribute memory elaborations independently

-without being asked- or even ask questions to

their parents.

Children in the German families were

contributing more memory elaboration when

reminiscing with their mothers than with their

fathers, whereas children in the Costa Rican and

Mexican contexts provided an equal amount

of memory elaborations within mother-child

and father-child conversations. For the German

and Costa Rican contexts, this corresponds to

the differences in the parental reminiscing style:

mothers being more elaborative than fathers in

the German context, and mothers and fathers

being equally elaborative in the Costa Rican

context. Elaborativeness was highly correlated

with children’s memory contributions across

parents within these contexts in line with multiple

other studies across cultures (e.g., Farrant &

Reese, 2000; Reese & Fivush, 1993; Wang, 2001;

2007). Only for the Mexican context, the result

It has to be kept in mind though, that the coding scheme slightly

differed in that elaboration and different types of questions were

coded separately in the study by Melzi (2000), and maternal

requests in the study by Melzi et al. (2011) also contained

memory prompts, whereas in the present study only elaborative

utterances were coded for their type of being a question or

statement.

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

that children contributed an equal amount of

elaborations is surprising since mothers were

more elaborative than fathers were. Relating to

this ﬁ nding is the missing correlation of the

maternal overall elaborativeness and children’s

memory in this context, which is surprising.

Information about the frequency of mother-

child reminiscing and information about

situational motivation of children could possibly

explain this ﬁ nding.

Differences between mothers and fathers

Across cultural contexts, mother-child

conversations were longer and focused more on

the speciﬁ c event under discussion (in relation

to other types of talk) compared to father-

child conversations, that were overall shorter

and had a lower proportion of event talk; in

line with studies on parental conversational

behavior (Fivush et al., 2000; Pancsofar &

Vernon-Feagans, 2006). Furthermore, mothers

were more elaborative than fathers (in line with

Zaman & Fivush, 2013); however, this was only

true for German and Mexican families, whereas

parents in the Costa Rican context were similarly

elaborative (in line with Reese & Fivush, 1993;

Reese et al., 1996). Thus, overall our results

conﬁ rm the heterogeneous ﬁ ndings of mothers

being more elaborative than fathers in some

contexts, and mothers and fathers being equally

elaborative in others. In this respect, also the

correlation patterns are important to consider.

A correspondence for mothers’ and fathers’

elaborativeness was only revealed within the

Costa Rican context. Differences in the similarity

and correspondence of paternal reminiscing

style might be related to reminiscing traditions

within families. It is possible that in the Costa

Rican context, reminiscing might take place in

a family setting rather than dyadic settings. In

this way families may have established a “family

reminiscing style” being similar in mothers

and fathers, which might not be the case in

families from the other cultural contexts. It is

possible that in the Mexican families reminiscing

takes place in one-to-one settings, yet neither

mother nor father might be the most frequent

conversational partners for reminiscing, but

many other social partners (Eisenberg, 1985).

However, these are speculations that would need

to be studied empirically.

In general, it seems difﬁ cult to ﬁ nd a

satisfactory explanation why studies differ

in their ﬁ ndings about mothers and fathers

elaborative contributions. An interesting aspect

to explore could be the attitude concerning past

event talk among parents. Regular reminiscing

of parents with each other might also inﬂ uence

the parental consistency in their reminiscing

style with their children. Various other task-,

contextual, or individual factors might inﬂ uence

whether mothers and fathers do differ in their

way to reminisce with their children (see Fivush

& Buckner, 2003 for a review).

Parental elaborative reminiscing did not differ

overall as a function of children’s gender in line

with previous cross-cultural studies looking at

the maternal reminiscing style only (e.g., Fivush

& Wang, 2005; Melzi, 2000; Schröder et al., 2013).

On a structural level, an overall elaborative style

with boys and girls might reﬂ ect the importance

of educational achievement and academic

success in both European and Latino families

(e.g., Cristofaro & Tamis-LeMonda, 2008;

Keller, 2007), as an elaborative style is related

to school relevant competencies like language-

and literacy- development (e.g., Peterson, Jesso,

& McCabe, 1999; Reese, 1995; Reese, Leyva,

Sparks, & Grolnick, 2010; Sparks & Reese,

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

2013). In future studies it might be illuminating

to consider also the inﬂ uence of sibling’s gender.

Previous studies have revealed consistency

across siblings in the maternal reminiscing style

with children of the same gender (Haden, 1998).

Investigating gender speciﬁ city and consistency

across siblings in families with daughters and

sons (ideally across cultures and parents) could

be revealing.

Content of conversations: Social orientation

Cultural as well as gender differences, did

though exist with respect to the content referred

to during reminiscing. Conversations in the

German families were least socially oriented

and on average, parents referred more to

nonsocial- than social aspects reﬂ ected in a

negative difference score of social- to nonsocial

references. In contrast, parents of the Mexican

and Costa Rican contexts referred more often

to social than nonsocial contexts reﬂ ected in

their positive mean difference scores. The social

orientation (relative to nonsocial aspects) in

Mexican families was highest, and Costa Rican

families’ social orientation lay in-between. Thus,

in line with the contexts’ cultural models content

reminisced about, reﬂ ected a higher relatedness

orientation in the Latin American families

compared to the German families. In line with

previous studies (Buckner & Fivush, 2000;

Cristofaro & Tamis-Lemonda, 2008; Fivush et

al., 2003), conversations with girls were more

socially oriented than conversations with boys:

relative to nonsocial content, parents talked

about social aspects more with girls than boys

across contexts. This result also corresponds

with women being conceived of as more

relationally oriented than men, also reﬂ ected in

women’s autobiographical narratives as adults

(Fivush & Buckner, 2003 for a review). This

orientation was, however, not yet found in what

children referred to in conversations.

Limitations

A limitation of the present study is that

mother- and father- child conversations were

only assessed at one point in time. Thus, the

generalization of these conversations might

not be reliable for all families since the child’s

motivation in the speciﬁ c situation, might also

inﬂ uence the conversational structure and

content. It might be helpful in future studies to

control for the overall motivation of children

assessed by parental judgment (compared to

other reminiscing situations). Furthermore, in

contrast to studies cited here, the instructions

for the present assessments did not ask parents

to reminisce about speciﬁ cally emotional events

with their children as implemented in other

studies (e.g., Fivush & Wang, 2005; Melzi et

al., 2004; Zaman & Fivush, 2013). Thus, the

overall emotional signiﬁ cance of the discussed

events might vary among conversations and also

inﬂ uence the way of reminiscing.

Conclusions

Despite contexts having similar educational

background and two contexts having a Latin

American background, each cultural context was

unique in the way parents and children talked

about the past. This supports the importance

to look at culture beyond the level of country

as well as beyond ethnicity and educational-

economic status (e.g., also family type and

historical change; Keller & Kärtner, 2013).

In order to fully understand the socialization

milieu of middle-class families, it is important to

investigate parental behavior of both, mothers

and fathers, but also other socialization agents

that are important in the child’s life.

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

Reminiscing with children is an important

practice and relevant to many different

developmental areas such as language- and

literacy development, but also social-emotional

competencies (Reese et al., 2010; Van Bergen,

Salmon, Dadds, & Allen, 2010). Thus, ﬁ ndings

in reminiscing are of great practical relevance

as well. Cross-cultural studies like the present

can contribute to conceptualize culture sensitive

approaches adjusted to children growing up

in families with migration history and thus,

potentially different cultural backgrounds than

predominant in the host society. Looking at

gender differences might also have practical

relevance; for example, encouraging parents to

also talk with their sons about social-emotional

aspects more intensively as it is related to

emotion knowledge and might ﬁ nally even

prevent externalizing behaviors (Salmon, Dadds,

Allen, & Haws, 2009).

Reminiscing is a very important family

interaction across cultures that inﬂ uences

many different developmental areas, but is also

relevant to family bonds and the formation

of childhood memories that, in turn, are

fundamental for our identity.

References

Adams, S., Kuebli, J., Boyle, P., & Fivush, R. (1995).

Gender differences in parent-child conversations

about past emotions: A longitudinal investigation.

Sex Roles, 33, 309-323.

Buckner, J. P., & Fivush, R. (2000). Gendered themes

in family reminiscing. Memory, 8, 401-412.

Carmiol, A., & Sparks, A. (in press). Narrative

Development across Cultural Contexts: Finding

the Pragmatic in Parent-Child Reminiscing. In

D. Matthews (Ed.), Pragmatic Development in First

Language Acquisition. Amsterdam: John Benjamins.

Caspe, M., & Melzi, G. (2008). Spanish-speaking Latin

American mother-child narrative discourse. In A.

McCabe, A.L. Bailey, & G. Melzi (Eds.), Spanish-

language narration and literacy: Culture, cognition,

and emotion (pp. 6–33). New York: Cambridge

University Press.

Cohen, J. (1988). Statistical power analysis for the behavioral

sciences. Hillsdale, NJ: Lawrence Earlbaum

Associates.

Conway, M. A., & Pleydell-Pearce, C. W. (2000). The

construction of autobiographical memories in

the self-memory system. Psychological Review, 107,

261-288.

Cristofaro, T. N. & Tamis-LeMonda, C. S. (2008).

Lessons in mother-child and father-child personal

narratives in Latino families. In A. McCabe, A.L.

Bailey, & G. Melzi (Eds.), Spanish-language narration

and literacy: Culture, cognition, and emotion (pp. 54-

91). New York: Cambridge University Press.

Eisenberg, A. R. (1985). Learning to describe past

experiences in conversation. Discourse Processes, 8,

177-204.

Eisenberg, A. R. (1999). Emotion talk among

Mexican American and Anglo American mothers

and children from two social classes. Merrill-Palmer

Quarterly, 45, 267-284.

Farrant, K., & Reese, E. (2000). Maternal style and

children’s participation in reminiscing: Stepping

stones in children’s autobiographical memory

development. Journal of Cognition and Development,

1, 193-225.

Fivush, R., Berlin, L. J., McDermott-Sales, J., Menuti-

Washburn, J., & Cassidy, J. (2003). Functions

of parent-child reminiscing about emotionally

negative events. Memory, 11, 179–192.

Fivush, R., Brotman, M. A., Buckner, J. P., &

Goodman, S. H. (2000). Gender differences in

parent-child emotion narratives. Sex Roles, 42,

233-253.

Fivush, R. & Buckner, J. (2003). Constructing

gender and identity through autobiographical

narratives. In R. Fivush and C. Haden (Eds.)

Autobiographical Memory and the construction of a

Narrative Self: Developmental and Cultural Perspectives,

(pp. 149-67). Hillsdale, NJ: Erlbaum.

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

Fivush, R., & Fromhoff, F. A. (1988). Style and

structure in mother-child conversations about

the past. Discourse Processes, 11, 337-355.

Fivush, R., & Haden, C. A. (Eds.). (2003).

Autobiographical memory and the construction of a

narrative self: Developmental and cultural perspectives.

Mahwah, NJ: Lawrence Erlbaum Associates.

Fivush, R., Haden, C. A.; & Reese, E. (2006).

Elaborating on elaborations: Role of maternal

reminiscing style in cognitive and socioemotional

development. Child Development, 77, 1568-1588.

Fivush, R., & Wang, Q. (2005). Emotion Talk of

Mother-Child Conversations of the Shared Past:

The Effects of Culture, Gender, Event Valence.

Journal of Cognition and Development, 6, 489-506.

Haden, C. A. (1998). Reminiscing with different

children: Relating maternal stylistic consistency

and sibling similarity in talk about the past.

Developmental Psychology, 34, 99-114.

Haden, C. A., Ornstein, P. A., Rudek, D. J., &

Cameron, D. (2009). Reminiscing in the early

years: Patterns of maternal elaborativeness and

children’s remembering. International Journal of

Behavioral Development, 33, 118-130.

Han, J. J., Leichtman, M. D., & Wang, Q. (1998).

Autobiographical memory in Korean, Chinese,

and American children. Developmental Psychology,

34, 701-713.

Kärtner, J., Borke, J., Maasmeier, K., Keller, H., &

Kleis, A. (2011). Socio-cultural inﬂ uences on

the development of self-recognition and self-

regulation in Costa Rican and Mexican toddlers.

Journal of Cognitive Education and Psychology, 10, 96-

112.

Keller, H. (2007). Cultures of infancy. Mahwah, NJ US:

Lawrence Erlbaum Associates Publishers.

Keller, H. (2011). Culture and cognition:

Developmental perspectives. Journal of Cognitive

Education and Psychology, 10, 3-8.

Keller, H., Lamm, B., Abels, M., Yovsi, R., Borke,

J., Jensen, H., et al. (2006). Cultural models,

socialization goals, and parenting ethnotheories:

A multicultural analysis. Journal of Cross-Cultural

Psychology, 37, 155-172.

Keller, H. & Kärtner, J. (2013). Development – The

cultural solution of universal d e v e l o p m e n t a l

tasks. In M. Gelfand, C.-Y. Chiu & Y.-Y. Hong

(Eds.), Advances in Culture and Psychology (Vol. 3), (pp.

63-116). Oxford University Press.

Keller, H. & Otto, H. (2011). Different faces of

autonomy. In: X. Chen & K.H. Rubin (Eds.),

Socioemotional Development in Cultural Context. New

York: The Guilford Press.

Kuebli, J., & Fivush, R. (1992). Gender differences in

parent-child conversations about past emotions.

Sex Roles, 12, 683-698.

Lamb, M. E. & Tamis-Lemonda, C. S. (2004): The Role

of the Father. An Introduction. In: Lamb, M. E.

(Ed.): The Role of the Father in Child Development (4th

ed.). Hoboken, NJ (John Wiley & Sons, Inc.), S.

1–31.

Lamm, B. & Keller, H. (2012). Väter in verschiedenen

Kulturen. In H. Walter & A. Eickhorst (Eds.), Das

Väterhandbuch: Theorie, Forschung, Praxis (pp 77-88).

Gießen: Psychosozial-Verlag.

McCabe, A., Bailey, A. L., & Melzi, G. (Eds.) (2008).

Spanish-language narration and literacy: Culture, cognition,

and emotion. New York: Cambridge University Press.

Melzi, G. (2000). Cultural variations in the construction

of personal narratives: American and European

American mothers’ elicitation styles. Discourse

Processes, 30, 153-177.

Melzi, G. & Fernández, C. (2004). Talking about past

emotions: conversations between Peruvian mothers

and their preschool children. Sex Roles, 50, 641-657.

Melzi, G., Schick, A., & Kennedy, J. (2011). Narrative

Participation and Elaboration: Two Dimensions

of Maternal Elicitation Style. Child Development, 82,

1282-1296.

Mullen, M. K., & Yi, S. (1995). The cultural context of

talk about the past: Implications for the development

of autobiographical memory. Cognitive Development,

10, 407-419.

Pancsofar, N. & Vernon-Feagans, L. (2006). Mother

and father language input to young children:

contributions to later language development. Journal

of Applied Developmental Psychology, 27, 571-587.

Lisa Schröder, Heidi Keller and Astrid Kleis

Actualidades en Psicología, 27(115), 2013, 49-73

Peterson, C., Jesso, B., & McCabe, A. (1999).

Encouraging narratives in preschoolers: An

intervention study. Journal of Child Language, 26,

49-67.

Reese, E., & Fivush, R. (1993). Parental styles of

talking about the past. Developmental Psychology, 29,

596-606.

Reese, E., Haden, C. A., & Fivush, R. (1993).

Mother-child conversations about the past:

Relationships of style and memory over time.

Cognitive Development, 8, 403-430.

Reese, E., Haden, C. A., & Fivush, R. (1996). Mothers,

father, daughters, sons: Gender differences

in reminiscing. Research on Language and Social

Interaction, 29, 27-56.

Reese, E., Hayne, H., & MacDonald, S. (2008).

Looking back to the future: Māori and Pakeha

mother-child birth stories. Child Development, 79,

114-125.

Reese, E., Leyva, D., Sparks, A., & Grolnick, W.

(2010). Maternal elaborative reminiscing increases

low-income children’s narrative skills relative to

dialogic reading. Early Education and Development,

21, 318-342.

Rogoff, B. (2003). The cultural nature of human

development. New York: Oxford University Press.

Sachs, L. (2003). Angewandte Statistik. Anwendung

statistischer Methoden. (11th Edition). Berlin:

Springer-Verlag.

Salmon, K., Dadds, M. R., Allen, J., Hawes, D. J.

(2009). Can emotional language skills be taught

during parent training for conduct problem

children? Child Psychiatry and Human Development,

40, 485-498.

Schröder, L., Keller, H., Kärtner, J., Kleis, A.,

Abels, M., Yovsi, R. D., Chaudhary, N., Jensen,

H., Papaligoura, Z. (2013). Early reminiscing

in cultural context: Cultural models, maternal

reminiscing styles, and children’s memories.

Journal of Cognition and Development, 14, 10-34.

Schröder, L., Keller, H., Tõugu, P., Tulviste,

T., Lenk, M., Schwarzer, M., Rübeling, H.,

De Geer, B. (2011). Cultural expressions of

preschoolers’ emerging self: Narrative and iconic

representations. Journal of Cognitive Education and

Psychology, 10, 77-95.

Sparks, A. (2008). Latino mothers and their preschool

children talk about the past: Implications for

language and literacy. In A. McCabe, A.L. Bailey,

& G. Melzi (Eds.), Spanish-language narration and

literacy: Culture, cognition, and emotion (pp. 273-295).

New York: Cambridge University Press.

Sparks, A. & Reese, E. (2013). From reminiscing

to reading: Home contributions to children’s

developing language and literacy in low-income

families. First Language, 33, 89-109.

Tamis-LeMonda, C. S., Shannon, J. D., Cabrera, N.

J., & Lamb, M. E. (2004). Fathers and mothers at

play with their 2- and 3-year-olds: Contributions

to language and cognitive development. Child

Development, 75, 1806−1820.

Tõugu, P., Tulviste, T., Schröder, L., Keller, H., &

De Geer, B. (2011). Socialization of past event

talk: Cultural differences in maternal elaborative

reminiscing. Cognitive Development, 26, 142-154.

Van Bergen, P., Salmon, K., Dadds, M. K., & Allen, J.

(2009). The effect of mother training in emotion-

rich, elaborative reminiscing on children’s shared

recall and emotion knowledge. Journal of Cognition

and Development, 10, 162-187.

Wang, Q. (2001). ‘Did you have fun?’ American and

Chinese mother-child conversations about shared

emotional experiences. Cognitive Development, 16,

693-715.

Wang, Q. (2004). The emergence of cultural self-

constructs: Autobiographical memory and self-

description in European American and Chinese

children. Developmental Psychology, 40, 3-15.

Wang, Q. (2006). Relations of maternal style and

child self-concept to autobiographical memories

in Chinese, Chinese immigrant, and European

American 3-year-olds. Child Development, 77, 1794-

1809.

Wang, Q. (2007). ‘Remember when you got the big,

big bulldozer?’ mother-child reminiscing over

time and across cultures. Social Cognition, 25, 455-

471.

Cultural and Gendered Reminiscing

Actualidades en Psicología, 27(115), 2013, 49-73

Recibido: 21 de mayo de 2013

Aceptado: 02 de setiembre de 2013

Wang, Q., & Fivush, R. (2005). Mother-child

conversations of emotionally salient events:

Exploring the functions of emotional reminiscing

in European American and Chinese families.

Social Development, 14, 473-495.

Wang, Q., Leichtman, M. D., & Davies, K. I. (2000).

Sharing memories and telling stories: American

and Chinese mothers and their 3-year-olds.

Memory, 8, 159-178.

Zaman, W. & Fivush, R. (2013) Gender differences

in elaborative parent-child emotion and play

narratives. Sex Roles, DOI: 10.1007/s11199-013-

0270-7.