Skip to content
Frontiers
Search

Brief Research Report

Frontiers in Developmental Psychology ·

Volume 4 - 2026 | https://doi.org/10.3389/fdpys.2026.1895134

Short-term familiarity shapes social preference but not word learning in preschoolers

  • Jina Ahn1*corresponding author
  • Erica A. Cartmill2,3
  • Catherine M. Sandhofer1
  1. 1Department of Psychology, University of California, Los Angeles, Los Angeles, LA, United States
  2. 2Program in Cognitive Science, Indiana University, Bloomington, IN, United States
  3. 3Santa Fe Institute, Santa Fe, NM, United States
Article metrics0 viewsView details
PDFCiteMetrics

Abstract

Children learn language in social environments shaped by interactions with caregivers and other adults. Prior research suggests that even brief familiarity with a social partner can facilitate some types of learning, but it remains unclear whether short-term familiarity influences children's word learning. Across two studies, we examined whether short-term familiarity with an experimenter affected children's performance on a novel-label generalization task. In Study 1, younger children (22–37 months) were taught and tested by an experimenter with whom they had completed a 10-min interactive free-play session or a different unfamiliar experimenter. Children performed similarly in the familiar and unfamiliar conditions, and a Bayes-factor analysis provided moderate evidence favoring no difference between conditions. In Study 2, older children (49–69 months) completed blocked training-testing with both familiar and unfamiliar experimenters and a resource-allocation measure of social preference. Children showed a social preference for the familiar experimenter, but their generalization performance did not differ significantly between experimenters; the Bayesian evidence regarding a familiarity effect was inconclusive, although it favored no difference. Together, these findings suggest that brief familiarity can shape children's expressed social preference without producing a corresponding measurable advantage on a structured word-generalization task with equivalent linguistic input.

Keywords

1 Introduction

Children acquire language through social interaction with caregivers, teachers, peers, and others with whom they have varying degrees of familiarity (Dickinson and Tabors, 2001; Mashburn et al., 2009). In everyday life, children regularly encounter and converse with many different adults, including individuals outside their immediate caregiving network (Goldenberg et al., 2022). All adults that children encounter serve as potential models for language learning, but children do not learn from all adults equally. Speaker characteristics like accent, gender, and age have been found to influence children's selection of models in imitation tasks (Buttelmann et al., 2013; Wang et al., 2021; Zmyj et al., 2012). However, it is unclear whether the child's familiarity with a model similarly influences word learning.

Research on children's imitation indicates that even brief familiarity with an interaction partner can affect learning. For example, toddlers were more likely to imitate novel actions demonstrated directly to them by a familiar experimenter than an unfamiliar experimenter.

However, when children observed the experimenters demonstrate the actions to a third person, imitation did not differ by experimenter familiarity (Shimpi et al., 2013). Another study found that even brief prior interaction with an experimenter increased 2-year-olds' subsequent preference for engaging with and smiling toward that experimenter (Slaughter et al., 2008), and children tend to prefer both interacting with and learning from social partners with whom they have shared positive social experiences (Krcmar, 2010; Seehagen and Herbert, 2012). Together, these findings suggest that brief familiarity can influence children's imitation in direct interactive contexts as well as their social preference.

Word learning itself is often characterized as a fundamental social process. Theories of early language acquisition emphasize joint attention, social referencing, and sensitivity to communicative intent as central mechanisms supporting word–referent mapping (Baldwin, 1995; Kuhl, 2007; Tomasello, 2000). From this perspective, social cues provided by a learning partner help children determine what a speaker intends to label and whether the information is relevant or trustworthy. If word learning depends on such socially mediated processes, familiarity with an interaction partner could plausibly shape word-learning outcomes in ways similar to its effect on imitation.

Building on evidence from imitation and social preference, short-term familiarity may also support word learning by shaping children's engagement and attention. Familiarity may increase engagement by reducing children's wariness and increasing their willingness to interact with and respond to a social partner. Separately, familiarity may allow children to devote more attention to the labels and their referents rather than to monitoring an unfamiliar partner. Either process could potentially support word–referent mapping and retention (Ramírez-Esparza et al., 2017).

On the other hand, familiarity may not play a central role in word learning. Unlike imitation, which often depends on close attention to a partner's actions and engagement with that partner, word learning can also occur in non-interactive contexts. For example, children can acquire new words from overhearing conversations between others, even without direct engagement (Akhtar et al., 2001), although speech directed toward children has been shown to be more conducive to learning (Shneidman and Goldin-Meadow, 2012). These findings suggest that word learning may be relatively robust to variation in social familiarity, especially the short-term familiarity brought about by a single interaction. Although familiarity clearly shapes children's social preferences and imitation behavior, its role in word learning remains an open question. In particular, it remains unclear whether children's preference for a familiar vs. an unfamiliar social partner is associated with differences in their acquisition and generalization of words. Thus, we examined whether short-term familiarity with an experimenter influences children's generalization of newly-learned object labels. Study 1 focused on children aged 22 to 37 months, comparing children assigned to either a familiar or unfamiliar experimenter condition. Study 2 focused on older children aged 49 to 69 months and compared each child's generalization performance across the familiar and unfamiliar experimenters. Study 2 also included a resource allocation task to measure children's expressed partner preference.

2 Study 1

Based on prior work showing that brief familiarity can influence children's partner choice and imitation behavior (e.g., Kuhl, 2007; Olson and Spelke, 2008; Shimpi et al., 2013; Slaughter et al., 2008), we asked whether short-term familiarity would also affect word learning. If short-term familiarity facilitates word learning, children taught by a familiar experimenter should show better generalization of newly learned object labels than children taught by an unfamiliar experimenter. Alternatively, if early word learning is relatively robust to variation in social familiarity, children's generalization performance may be similar across familiar and unfamiliar experimenters.

Study 1 investigated this question in younger children (22 to 37 months of age). We selected children in this age range because prior research has shown that brief familiarity with an adult can influence imitation and social engagement at this age (Shimpi et al., 2013; Slaughter et al., 2008). Using a between-subjects design, children were taught novel object labels by either a familiar or an unfamiliar experimenter. Short-term familiarity was established through a 10-min interactive free-play session before the word-learning task.

3 Research method

3.1 Participants

Participants were 36 children (Mage = 30.33 months, range = 22–37 months; 15 females and 21 males). Families were recruited through preschools and daycares in Los Angeles County, and data were collected in 2023. All children were learning English as a first language, and all could understand instructions in English.

Parents reported children's racial or ethnic identity as White (n = 23), Asian (n = 5), African American (n= 1), and Hispanic or Latino (n = 5). Two parents chose not to report their child's racial or ethnic identity. Most parents (n = 35) reported holding a college or graduate degree.

Nine additional participants began testing but were excluded either due to their inability to pass the practice trials (n = 3) or experimenter error (n = 6). Before the experiment, parents or legal guardians provided signed parental permission in accordance with approval from the UCLA Institutional Review Board (#10-001578). Children selected books to donate to their classrooms as compensation for participation.

3.2 Experimental materials

Materials used during the free-play warm-up phase included a wooden fruit and vegetable toy set, children's books, and a magnetic drawing board.

During an initial practice phase, familiar objects (bear, shoe, flower, cup, spoon, baby bottle, cow, and horse) were used to confirm that children understood the task procedure. These items were selected based on age-appropriate vocabulary from the MacArthur–Bates Communicative Development Inventories (Fenson, 2007).

For the word learning phase, novel objects were created using a 3D printer and painted in distinct colors. A total of 56 novel objects were constructed, all designed to be unfamiliar to children. There were eight target categories, each consisting of four exemplars that shared the same shape but differed in color (32 target objects in total; Supplementary Figure S1). Three of these exemplars (target objects) were used during the training phase, and the remaining exemplar (the last target object) was reserved for the testing phase. An additional eight objects served as distractors during training and testing, and 16 objects served as unfamiliar objects in the testing phase. Neither distractor nor unfamiliar objects shared shape or color with target exemplars, and none were assigned labels. The distractor object was a previously seen but unlabeled object, allowing us to distinguish selection of the target exemplar from selection based solely on familiarity with a particular object. Each target object category was assigned to one of eight novel labels (“Wiso,” “Toma,” “Bink,” “Koba,” “Zeb,” “Dax,” “Modi,” and “Leam”) selected from the Novel Object and Unusual Name (NOUN) Database (Horst and Hout, 2016). Labels were phonetically consistent with English.

3.3 Experimenters

The design involved three experimenters, each playing a different role. To minimize bias, experimenters alternated between roles as Experimenter A (who conducted the warm-up phase for all children and the training and testing phases in the Familiar condition) and Experimenter B (who conducted only the training and testing phases in the Unfamiliar condition). The recorder gave the child instructions and recorded their responses. All experimenters were female, in their 20s, and native English speakers.

3.4 Design and procedure

Study 1 used a between-subjects design. Participants were randomly assigned to one of two conditions: the Familiar condition, in which the same experimenter from the warm-up phase conducted all subsequent phases, or the Unfamiliar condition, in which one experimenter conducted the warm-up phase and left, then a different experimenter conducted the training and testing phases. Equal numbers of participants were assigned to each condition.

3.4.1 Warm-up phase

The recorder escorted the child into the testing room, where the child was seated on the floor facing Experimenter A and the recorder. The child then engaged in a 10-min free play session with Experimenter A. The warm-up session was not scripted. Consistent with brief free-play familiarization procedures used in prior work (Shimpi et al., 2013; Slaughter et al., 2008), Experimenter A followed the child's lead in play with the warm-up materials, took turns during play, and offered positive comments on the child's speech and actions, with the goal of establishing positive social rapport. The session lasted exactly 10 min for every child. None of the novel objects used in the word-learning task were present, and none of the novel labels were produced during the warm-up session. The 10-min duration was selected because previous studies have shown that a 10-min free-play interaction can establish short-term familiarity with a new adult (Shimpi et al., 2013; Slaughter et al., 2008). In the Familiar condition, Experimenter A conducted all subsequent phases. In the Unfamiliar condition, Experimenter A informed the child they were leaving, and Experimenter B (a previously unfamiliar adult) entered to conduct the remaining phases.

3.4.2 Word learning - practice phase

The practice phase was used to ensure that children understood the word-learning task procedure prior to the training phase. The recorder presented children with four familiar objects (e.g., a bear, shoe, cup, and spoon), labeled each object in turn (e.g., “This is a bear...this is a shoe...”), and then asked the child to place one into the bucket (e.g., “Can you put the shoe in the magic bucket?”). The recorder requested three of the four objects. To proceed to the training and testing phases, children were required to complete two of the three requests successfully.

3.4.3 Word learning - training phase

After the child successfully completed the practice phase, the experimenter introduced the word-learning training phase. During each of the eight training trials, the experimenter sequentially presented three target exemplars, labeling each exemplar twice (e.g., “Look, it's a Wiso”; Figure 1). Children were given approximately 10 s to handle and explore each target exemplar before being asked to return it to the bucket. After all three target exemplars were presented, labeled, and offered to the child, the experimenter presented the distractor object without providing a label and directed the child's attention to it (e.g., “Look at this!”). The child was given approximately 10 s to interact with the distractor object before the trial concluded.

Example of the training and testing phases.
Figure 1. Example of the training and testing phases.

3.4.4 Word learning - testing phase

After each training trial, a corresponding testing trial followed immediately. Testing trials consisted of a forced-choice generalization task (Figure 1). In each testing trial, children were asked to choose between four objects: two unfamiliar objects (never-before-seen and unlabeled), one distractor (familiar but unlabeled), and one target object in a different color (familiar shape previously labeled during training). Children were given an opportunity to handle and explore all four objects. After the objects were returned to the tray, the experimenter prompted the child to select the target object using its label (e.g., “Can you find the Wiso and put Wiso in the magic bucket?”). The child then selected one object and placed it into the experimenter's bucket. The experimenter responded with neutral, non-evaluative feedback (e.g., “Thank you”) regardless of the child's accuracy and then proceeded immediately to the next training trial.

4 Results and discussion

All statistical analyses were conducted with R (version 4.3.1; R Core Team, 2023). On each testing trial, children's responses were scored as correct if they selected and placed the target object into the experimenter's bucket. Across the eight testing trials, children received a generalization score ranging from 0 to 8, reflecting the number of correct target selections.

First, we asked whether the generalization of novel object label scores in each familiarity condition differed from chance. Chance performance was 25% because children had four choices in each testing trial. On average, children performed above chance in both the Familiar (M = 4.39, SD = 2.06, t(17) = 4.92, p < 0.001, d = 1.16) and Unfamiliar conditions (M = 4.39, SD = 1.94, t(17) = 5.21, p < 0.001, d = 1.23).

To examine whether children's generalization performance differed by experimenter familiarity, we conducted an independent samples t-test. As shown in Figure 2, there was no significant difference between the Familiar and Unfamiliar conditions (t(34) = 0.00, p = 1.000). To determine whether this non-significant result provided evidence for no difference between conditions or remained inconclusive, we conducted a post hoc Bayes factor analysis (Dienes, 2014; Lakens et al., 2020). We used a Bayesian independent-samples t-test with a default JZS Cauchy prior (r = 0.707; Rouder et al., 2009), using the BayesFactor package in R (Morey and Rouder, 2026). The analysis yielded moderate evidence favoring no difference between the familiar and the unfamiliar conditions (BF01 = 3.11). These data indicate that the observed outcomes were roughly 3.11 times more likely under the null model than under the specified alternative model, providing moderate evidence favoring the null model. Thus, the frequentist and Bayesian analyses converged in indicating no measurable familiarity-related difference in generalization performance in Study 1. This finding contrasts with previous studies showing that short-term familiarity influenced imitation in direct interactive contexts or social partner preference (Shimpi et al., 2013; Slaughter et al., 2008). The familiarity benefit previously observed in those contexts was not observed on the word-generalization task used here.

Number of correct generalizations (out of eight trials) by familiarity conditions. Each dot represents one participant. No significant differences were observed between the Familiar and Unfamiliar conditions.
Figure 2. Number of correct generalizations (out of eight trials) by familiarity conditions. Each dot represents one participant. No significant differences were observed between the Familiar and Unfamiliar conditions.

Study 1 did not include a direct measure of children's social preference. Thus, although short-term familiarization did not produce a measurable advantage in word generalization, Study 1 could not determine whether it established a preference for the familiar experimenter. Study 2 examined both expressed social preference and word generalization with familiar and unfamiliar experimenters.

5 Study 2

Children exercise selectivity by choosing whom to learn from, and the cues they rely on change with development. Younger children (e.g., 2-year-olds) tend to rely on a range of social cues, such as familiarity and affiliation, when selecting learning models, whereas older children increasingly prioritize cues related to an informant's reliability or credibility (Corriveau and Harris, 2009; Tong et al., 2020). Preschool-aged children demonstrate a stronger preference for individuals with whom they have previously shared positive social experiences through resource allocation, favoring familiar or socially affiliated individuals when forced to choose between recipients (Lee et al., 2018). For example, when resources are limited and cannot be divided equally, children preferentially allocate resources to ingroup or familiar individuals (Lee et al., 2018). Although preschoolers typically value fairness norms, they tend to allocate more resources to familiar and socially favored individuals when equality cannot be maintained (Olson and Spelke, 2008; Paulus and Moore, 2014). Children also use patterns of resource distribution to signal or interpret social relationships, inferring affiliative ties from preferential allocations (Liberman and Shaw, 2017). Together, these findings indicate that resource allocation reflects children's underlying social preferences and perceived relationships.

Study 2 extended this investigation by including a measure of expressed social preference and having each child complete word-learning trials with both the familiar and unfamiliar experimenters. Prior research has used forced-choice resource allocation to assess selective social behavior in 5-year-old children, including procedures in which resources could not be divided equally between two recipients (Yang and Park, 2022). Preschool-aged children have also successfully completed blocked novel-word training-and-testing tasks in which training on multiple categories preceded generalization testing (Goldenberg and Sandhofer, 2013). We therefore selected older children (49 to 69 months) because we anticipated that the present all-or-none preference task, together with the expanded two-experimenter word-learning procedure, would be completed more reliably by children in this age range. Study 2 used a within-subjects design so that each child's generalization performance with the familiar and unfamiliar experimenters could be compared directly and so that we could explore whether children's expressed social preference was associated with their experimenter-specific generalization performance. Training and testing were blocked by experimenter to maintain a clear association between each experimenter and the labels they taught.

6 Research method

6.1 Participants

Participants were 34 children (Mage = 55.35 months, range = 49–69 months, 17 females and 17 males). Families were recruited through preschools and daycares in Los Angeles County, and data were collected between 2024 and 2025. All children were learning English as a first language, and all could understand instructions in English.

Parents reported children's racial or ethnic identity as White (n = 12), Asian (n = 6), African American (n = 1), Other (n = 5), multiracial (n = 2), and Hispanic or Latino (n = 7). One parent chose not to report racial or ethnic identity. Most parents (n = 31) reported holding a college or graduate degree.

Five additional participants began testing but were excluded either due to their inability to pass the practice trials (n = 3) or experimenter error (n = 2). Before the experiment, parents or legal guardians provided signed parental permission in accordance with approval from the UCLA Institutional Review Board (#10-001578). Children selected books to donate to their classrooms as compensation for participation.

6.2 Experimental materials and experimenters

Materials and experimenter roles were the same as Study 1. Materials for the resource-allocation task included a set of star-shaped toys, each approximately 4 cm across, and two identical buckets.

6.3 Design and procedure

Study 2 used a within-subjects design, in which each child participant interacted with both Experimenter A (familiar) and Experimenter B (unfamiliar) throughout the social preference and word-learning tasks. Short-term familiarity with Experimenter A was established through the same 10-min free play warm-up session used in Study 1. Following the warm-up phase, children completed a social preference phase and a word-learning phase consisting of practice, training, and testing trials.

6.3.1 Warm-up phase

The warm-up procedure was the same as Study 1. At the conclusion of the warm-up phase, the recorder informed the child that a new person, Experimenter B (unfamiliar), whom the child had not previously met, would enter the room to play a game with them.

6.3.2 Preference phase

Immediately following the warm-up phase, Experimenter B entered the testing room, greeted the child, and sat beside the recorder, forming a triangular seating arrangement with Experimenter A and the recorder (Supplementary Figure S2). The recorder began the preference phase by giving the child a set of stars and stated, “Look, we have a bunch of stars and two buckets.” Both Experimenter A and Experimenter B then simultaneously placed identical buckets in front of themselves while maintaining neutral facial expressions. After the child observed both buckets, the recorder administered a forced-choice instruction: “Please put the stars in the bucket of the person you want to play with more.” The experimenter who received the stars from the child responded neutrally by saying, “Thank you.” The recorder recorded which bucket the child selected as a measure of the child's expressed social preference for one experimenter over the other. Following the preference phase, the session transitioned directly to the word learning task.

6.3.3 Word learning - practice phase

The practice procedure was identical to that used in Study 1 (Section 3.4.2).

6.3.4 Word learning – training and testing phases

The training and testing procedures were the same as Study 1, except for the order of trials. In Study 1, each training trial was immediately followed by the corresponding testing trial for that object label. In Study 2, each experimenter conducted all four training trials before the four corresponding testing trials began. The order of training and testing trials in which Experimenter A (familiar) and Experimenter B (unfamiliar) conducted the training and testing trials was counterbalanced across participants.

7 Results and discussion

7.1 Preference phase

In the preference phase, we examined whether brief prior social interaction with Experimenter A would influence children's social preference relative to the unfamiliar Experimenter B. Because each child made a single forced choice between the two experimenters, we conducted a chi-square goodness-of-fit test to determine whether the number of children choosing each experimenter differed from chance levels. Children's choices differed significantly from chance, χ2(1, N = 34) = 9.53, p = 0.002, V = 0.53. Twenty-six children (76.5%) gave the stars to the familiar experimenter, whereas eight children (23.5%) gave the stars to the unfamiliar experimenter (Table 1). Thus, a brief, 10-min positive social interaction was sufficient to shape children's expressed social preferences toward the familiar experimenter.

Table 1. Number and percentage of children who preferred each experimenter in the preference phase (N = 34).

Experimenter typeNumber of childrenPercentage (%)
Familiar2676.5%
Unfamiliar823.5%

7.2 Word learning phase

In the word learning phase, we assessed children's generalization of novel object labels to new exemplars when taught by a familiar and unfamiliar experimenter. On each testing trial, children's responses were scored as correct if they selected and placed the target object into each experimenter's bucket. Each child received a separate generalization score for each experimenter, ranging from 0 to 4 and reflecting the number of correct target selections across the four testing trials.

One-sample t-tests showed that children selected the correct target object more often than expected by chance with both the familiar experimenter (M = 2.59, SD = 1.31, t(33) = 7.09, p < 0.001) and the unfamiliar experimenter (M = 2.24, SD = 1.39, t(33) = 5.17, p < 0.001) (Figure 3).

Children's mean number of correct generalizations (out of four trials) by each experimenter. Each child completed four testing trials with both a familiar and an unfamiliar experimenter in this within-subjects design. Error bars represent the standard error of the mean. No significant differences were observed between the Familiar and Unfamiliar experimenters.
Figure 3. Children's mean number of correct generalizations (out of four trials) by each experimenter. Each child completed four testing trials with both a familiar and an unfamiliar experimenter in this within-subjects design. Error bars represent the standard error of the mean. No significant differences were observed between the Familiar and Unfamiliar experimenters.

A paired-samples t-test found no significant difference between children's generalization scores with the familiar and unfamiliar experimenters, t(33) = 1.26, p = 0.216, d = 0.22, indicating that any familiarity effect, if present, is small. Thus, although children showed a social preference for the familiar experimenter in the preference phase, their word learning performance did not differ as a function of experimenter familiarity.

To evaluate the non-significant difference between the familiar and unfamiliar experimenter conditions, we conducted a paired-samples Bayesian t-test using the same prior and implementation described in Study 1. Because each child contributed scores to both familiarity conditions, Study 2 used a paired-samples Bayesian t-test. The comparison between the familiar and unfamiliar conditions yielded BF01 = 2.63, indicating that the observed outcomes were roughly 2.63 times more likely under the null model than under the alternative model. However, since BF01 did not reach the standard threshold of 3, the evidence remains inconclusive, although it numerically favors the null model.

In addition, we conducted exploratory analysis using a linear mixed model, to examine whether children's generalization performance differed as a function of their social preference in the preference phase (i.e., whether children who favored the familiar experimenter differed in their generalization scores from those who favored the unfamiliar experimenter when learning from each experimenter) (Supplementary Table S1). The results showed no main effects of experimenter familiarity (b = −0.35, SE = 0.32, t(32) = −1.07, p = 0.294) or preference group (b = 0.70, SE = 0.54, t(60.7) = 1.30, p = 0.199). There was no significant interaction between experimenter familiarity and preference group (b = −0.03, t(32) = −0.04, p = 0.966). These findings suggest that children's generalization of novel object labels did not significantly differ based on experimenter familiarity, regardless of whom they favored in the social preference phase.

Children preferentially allocated resources to the familiar adult, indicating that a 10-min interaction was sufficient to shape their expressed social preference. However, children's generalization scores did not differ significantly between the familiar and unfamiliar experimenters, and the Bayesian evidence regarding a familiarity effect was inconclusive. Thus, short-term familiarity affected children's expressed social preference, but did not produce a statistically reliable advantage in their performance on the word-generalization task.

One potential explanation for children's preference for the familiar experimenter is that brief positive interaction provided social cues that shaped subsequent partner choice. Previous research shows that children's responses to social partners are influenced by facial expressions, emotional tone, eye contact, and speech prosody (Nishida and Lillard, 2007; Smith et al., 2013; Vouloumanos and Werker, 2007). Interactions with unfamiliar individuals, in contrast, are sometimes associated with more reserved or inhibited behavior in early childhood (Asendorpf, 1991; Walker et al., 2015). Moreover, when children receive positive social cues, such as smiles or verbal affirmations, during prior interactions, they are more likely to attend to and engage with social partners (Vernetti et al., 2017). The findings of Study 2 are consistent with prior work, suggesting that even a brief positive interaction can shape subsequent social choices.

Children's allocation behavior is also consistent with evidence that preschool-aged children use resource allocation to express social preference, particularly when resources cannot be divided equally. In such situations, children preferentially allocate resources to favored or prosocial individuals (Baumard et al., 2012; Kenward and Dahl, 2011; Lee et al., 2018; Liberman and Shaw, 2017; Olson and Spelke, 2008). Resource allocation can therefore function as a behavioral indicator of social preference rather than a simple fairness judgment. In Study 2, children were required to allocate all the stars to one experimenter, making the task a direct measure of partner preference. Consistent with prior findings, even a brief prior interaction was sufficient to bias children's allocation toward the familiar adult.

8 General discussion

Across two studies, we examined whether short-term familiarity with a social partner influences children's word learning. In Study 1, children between 22 and 37 months of age showed no difference in the generalization of novel object labels when taught by a familiar vs. an unfamiliar experimenter, and the Bayes-factor analysis provided moderate evidence favoring no difference between conditions. In Study 2, children between 49 and 69 months of age expressed a social preference for the familiar experimenter in the resource-allocation task, indicating that the familiarity manipulation was sufficient to shape social preference. However, children's generalization performance did not differ significantly between experimenters, and the Bayesian evidence regarding a familiarity effect was inconclusive, although it favored no difference. Together, these findings suggest that short-term familiarity can influence children's partner preference without providing a measurable benefit on a structured word-generalization task with equivalent linguistic input.

The present findings contrast with prior research showing brief familiarity can influence imitation learning in direct interactive contexts (Shimpi et al., 2013). One possibility is that direct imitation and word learning place different task demands on children. Direct imitation requires children to monitor and reproduce a partner's actions and coordinate their actions with that partner (Nielsen, 2006; Shimpi et al., 2013; Slaughter et al., 2008). Thus, familiarity-related engagement may be especially relevant to imitation. For word learning, familiarity could influence performance through two related but distinct processes: engagement with the social partner and attention to labels and their referents. Familiarity may increase engagement by reducing children's wariness and increasing their willingness to interact with and respond to a social partner. Separately, familiarity may allow children to devote more attention to the labels and their referents rather than to monitoring an unfamiliar partner. Because the present studies did not measure either process directly, they cannot determine whether attention or engagement differed across familiarity conditions.

In both studies, the familiar and unfamiliar experimenters provided equivalent explicit labels within clear referential contexts. This task structure may have left fewer opportunities for familiarity-related differences in engagement or attention to influence children's generalization performance. This interpretation remains tentative because the studies did not manipulate the clarity of the linguistic or referential information. Familiarity may play a greater role in word learning under conditions involving ambiguous reference, conflicting testimony, or more complex pragmatic inference.

8.1 Limitations and future research

Although the present findings indicate that brief prior social interaction affected children's social preferences without producing measurable differences in word learning performance across both Studies 1 and 2, several limitations warrant consideration. First, the absence of a significant difference in word learning in either study should not be interpreted as definitive evidence of equivalence between conditions. Although the Bayes-factor analysis in Study 1 provided moderate evidence favoring no difference, the evidence in Study 2 remained inconclusive. Future research with larger samples or equivalence-testing approaches would help clarify whether familiarity has a negligible effect on generalization performance.

Second, the familiarity manipulation consisted of a brief 10-min interaction conducted within a familiar preschool setting. Because children were tested in an environment in which they were already comfortable, the unfamiliar experimenter may not have been perceived as strongly novel or socially distant. This context may have reduced the salience of the familiarity contrast and attenuated potential effects on learning across both studies. Future studies could examine whether stronger familiarity manipulations or less familiar testing environments produce different outcomes.

Third, both studies used a structured word-learning task that provided explicit labeling and unambiguous referential cues. Under such conditions, learning may be relatively insensitive to social variation. Familiarity may exert a stronger influence in contexts involving ambiguity, conflicting testimony, or more complex pragmatic inference. In addition, because the target was the only test object sharing the shape of previously handled and labeled exemplars, the task does not fully distinguish label-based generalization from recognition and generalization of a trained shape. Future studies could distinguish these possibilities by pitting exemplars from two trained categories against one another or by including a separate recognition trial. Because testing occurred immediately after training and measured receptive generalization only, the present studies also could not detect familiarity effects that might emerge in consolidation, longer-term retention, or production. Additionally, Study 1 did not include a direct measure of social preference. Future studies with younger children should incorporate a preference measure to better examine the relationship between social partner preference and generalization performance. Studies 1 and 2 also differed in study design, trial structure, the number of items taught by each experimenter, and participant age. Therefore, the studies do not provide a direct developmental comparison. Using matched designs across ages would allow stronger conclusions about developmental change. Future research should also consider whether individual differences in temperament, particularly shyness and behavioral inhibition, moderate children's responses to unfamiliar adults. Finally, the present research did not directly measure children's engagement, attention, or affect during learning. Incorporating behavioral or eye-tracking measures in future work would help clarify whether these processes mediate familiarity-related effects on children's social preferences or generalization performance.

9 Conclusion

The present research suggests that children's generalization performance may be relatively robust to short-term familiarity with a social partner under structured learning conditions. Across two studies, familiarity did not produce statistically reliable differences in children's performance on the novel-label generalization task. In Study 2, children showed an expressed social preference for the familiar experimenter following a brief interaction, but this preference was not accompanied by a statistically reliable advantage in generalization performance. Together, these findings suggest that brief familiarity can shape children's expressed partner preference without producing a corresponding measurable advantage on a structured word-generalization task in which linguistic and referential cues are clear.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. The author(s) acknowledge the use of ChatGPT (Version 5.1, OpenAI) to generate Supplementary Figure S2, a schematic illustration of the experimenter and the participant seating positions in Study 2. The author(s) reviewed the figure and verified that it accurately represents the study procedure.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fdpys.2026.1895134/full#supplementary-material

Example of a target category consisting of four exemplars. Supplementary material at the source

Picture of a seating arrangement. Supplementary material at the source

Descriptive statistics for children's generalization scores by experimenter familiarity and preference group. Supplementary material at the source

Footnotes

Edited by: Mark Antoniou, Western Sydney University, Australia

Reviewed by: Kin Chung Jacky Chan, Durham University, United Kingdom

Nils Frederik Tolksdorf, Heidelberg University, Germany

Statements

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author CS declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by a UCLA Council on Research Grant to Dr. Catherine M. Sandhofer.

Data availability

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics approval

The studies involving humans were approved by University of California, Los Angeles, Institutional Review Board (#10-001578). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin.

Author contributions

JA: Writing – review & editing, Project administration, Formal analysis, Writing – original draft, Methodology, Conceptualization. EC: Conceptualization, Writing – review & editing. CS: Funding acquisition, Conceptualization, Writing – review & editing.

Acknowledgements

Special thanks to the preschools and families who participated in the studies. Thank you to the LCD Lab's undergraduate research assistants who recruited participants and collected data and a thank you to the members of the UCLA Language and Cognitive Development Lab for helpful comments and suggestions.

References

  1. Akhtar N., Jipson J., Callanan M. A. (2001). Learning words through overhearing. Child Dev. 72, 416 – 430. doi: doi:10.1111/1467-8624.00287
  2. Asendorpf J. B. (1991). Development of inhibited children's coping with unfamiliarity. Child Dev. 62, 1460 – 1474. doi: doi:10.2307/1130819
  3. Baldwin D. A. (1995). “Understanding the link between joint attention and language,” in Joint Attention: Its Origins and Role in Development,eds. C. Moore and P. J. Dunham (Hilsdale, NJ: Lawrence Erlbaum Associates), 131 – 158.
  4. Baumard N., Mascaro O., Chevallier C. (2012). Preschoolers are able to take merit into account when distributing goods. Dev. Psychol. 48, 492 – 498. doi: doi:10.1037/a0026598
  5. Buttelmann D., Zmyj N., Daum M., Carpenter M. (2013). Selective imitation of in-group over out-group members in 14-month-old infants. Child Dev. 84, 422 – 428. doi: doi:10.1111/j.1467-8624.2012.01860.x
  6. Corriveau K., Harris P. L. (2009). Preschoolers continue to trust a more accurate informant 1 week after exposure to accuracy information. Dev. Sci. 12, 188 – 193. doi: doi:10.1111/j.1467-7687.2008.00763.x
  7. Dickinson D. K., Tabors P. O. (2001). Beginning Literacy with Language: Young Children Learning at Home and School. Baltimore, MD: Paul H. Brookes Publishing Co.
  8. Dienes Z. (2014). Using Bayes to get the most out of non-significant results. Front. Psychol. 5: 781. doi: doi:10.3389/fpsyg.2014.00781
  9. Fenson L. (2007). MacArthur-Bates Communicative Development Inventories: User's Guide and Technical Manual, 2nd Edn. Baltimore, MD: Paul H. Brookes Publishing Co. doi: doi:10.1037/t11538-000
  10. Goldenberg E. R Sandhofer C. M. (2013). Who is she? Changes in the person context affect categorization. Front. Psychol. 4: 745. doi: doi:10.3389/fpsyg.2013.00745
  11. Goldenberg E. R., Repetti R. L., Sandhofer C. M. (2022). Contextual variation in language input to children: a naturalistic approach. Dev. Psychol. 58, 1051 – 1065. doi: doi:10.1037/dev0001345
  12. Horst J. S., Hout M. C. (2016). The Novel Object and Unusual Name (NOUN) database: a collection of novel images for use in experimental research. Behav. Res. Methods 48, 1393 – 1409. doi: doi:10.3758/s13428-015-0647-3
  13. Kenward B., Dahl M. (2011). Preschoolers distribute scarce resources according to the moral valence of recipients' previous actions. Dev. Psychol. 47, 1054 – 1064. doi: doi:10.1037/a0023869
  14. Krcmar M. (2010). Can social meaningfulness and repeat exposure help infants and toddlers overcome the video deficit? Media Psychol. 13, 31 – 53. doi: doi:10.1080/15213260903562917
  15. Kuhl P. K. (2007). Is speech learning ‘gated' by the social brain? Dev. Sci. 10, 110 – 120. doi: doi:10.1111/j.1467-7687.2007.00572.x
  16. Lakens D., McLatchie N., Isager P. M., Scheel A. M., Dienes Z. (2020). Improving inferences about null effects with Bayes factors and equivalence tests. J. Gerontol. B 75, 45 – 57. doi: doi:10.1093/geronb/gby065
  17. Lee K. J. J., Esposito G., Setoh P. (2018). Preschoolers favor their ingroup when resources are limited. Front. Psychol. 9: 1752. doi: doi:10.3389/fpsyg.2018.01752
  18. Liberman Z., Shaw A. (2017). Children use partial resource sharing as a cue to friendship. J. Exp. Child Psychol. 159, 96 – 109. doi: doi:10.1016/j.jecp.2017.02.002
  19. Mashburn A. J., Justice L. M., Downer J. T., Pianta R. C. (2009). Peer effects on children's language achievement during pre-kindergarten. Child Dev. 80, 686 – 702. doi: doi:10.1111/j.1467-8624.2009.01291.x
  20. Morey R. D., Rouder J. N. (2026). BayesFactor: computation of Bayes factors for common designs. R package version 0.9.12-4.8. doi: doi:10.32614/CRAN.package.BayesFactor
  21. Nielsen M. (2006). Copying actions and copying outcomes: social learning through the second year. Dev. Psychol. 42, 555 – 565. doi: doi:10.1037/0012-1649.42.3.555
  22. Nishida T. K., Lillard A. S. (2007). The informative value of emotional expressions: ‘social referencing' in mother–child pretense. Dev. Sci. 10, 205 – 212. doi: doi:10.1111/j.1467-7687.2007.00581.x
  23. Olson K. R., Spelke E. S. (2008). Foundations of cooperation in young children. Cognition 108, 222 – 231. doi: doi:10.1016/j.cognition.2007.12.003
  24. Paulus M., Moore C. (2014). The development of recipient-dependent sharing behavior and sharing expectations in preschool children. Dev. Psychol. 50: 914. doi: doi:10.1037/a0034169
  25. R Core Team (2023). R: A Language and Environment for Statistical Computing (Version 4.3.1) [Computer software]. Vienna: R Foundation for Statistical Computing.
  26. Ramírez-Esparza N., García-Sierra A., Kuhl P. K. (2017). The impact of early social interactions on later language development in Spanish–English bilingual infants. Child Dev. 88, 1216 – 1234. doi: doi:10.1111/cdev.12648
  27. Rouder J. N., Speckman P. L., Sun D., Morey R. D., Iverson G. (2009). Bayesian t tests for accepting and rejecting the null hypothesis. Psychon Bull. Rev. 16, 225 – 237. doi: doi:10.3758/PBR.16.2.225
  28. Seehagen S., Herbert J. S. (2012). Selective imitation in 6-month-olds: the role of the social and physical context. Infant Behav. Dev. 35, 509 – 512. doi: doi:10.1016/j.infbeh.2012.05.002
  29. Shimpi P. M., Akhtar N., Moore C. (2013). Toddlers' imitative learning in interactive and observational contexts: the role of age and familiarity of the model. J. Exp. Child Psychol. 116, 309 – 323. doi: doi:10.1016/j.jecp.2013.06.008
  30. Shneidman L. A., Goldin-Meadow S. (2012). Language input and acquisition in a Mayan village: how important is directed speech? Dev. Sci. 15, 659 – 673. doi: doi:10.1111/j.1467-7687.2012.01168.x
  31. Slaughter V., Nielsen M., Enchelmaier P. (2008). Interacting socially with human hands at 24 months of age. Infancy 13, 185 – 195. doi: doi:10.1080/15250000701795721
  32. Smith N. A., Gibilisco C. R., Meisinger R. E., Hankey M. (2013). Asymmetry in infants' selective attention to facial features during visual processing of infant-directed speech. Front. Psychol. 4: 00601. doi: doi:10.3389/fpsyg.2013.00601
  33. Tomasello M. (2000). The social-pragmatic theory of word learning. Pragmatics 10, 401 – 413. doi: doi:10.1075/prag.10.4.01tom
  34. Tong Y., Wang F., Danovitch J. (2020). The role of epistemic and social characteristics in children's selective trust: three meta-analyses. Dev. Sci. 23: e12895. doi: doi:10.1111/desc.12895
  35. Vernetti A., Smith T. J., Senju A. (2017). Gaze-contingent reinforcement learning reveals incentive value of social signals in young children and adults. Proc. R. Soc. B. 284: 20162747. doi: doi:10.1098/rspb.2016.2747
  36. Vouloumanos A., Werker J. F. (2007). Listening to language at birth: evidence for a bias for speech in neonates. Dev. Sci. 10, 159 – 164. doi: doi:10.1111/j.1467-7687.2007.00549.x
  37. Walker O. L., Degnan K. A., Fox N. A., Henderson H. A. (2015). Early social fear in relation to play with an unfamiliar peer: actor and partner effects. Dev. Psychol. 51, 1588 – 1596. doi: doi:10.1037/a0039735
  38. Wang Z., Fong F. T. K., Meltzoff A. N. (2021). Enhancing same-gender imitation by highlighting gender norms in Chinese pre-school children. Brit. J. Dev. Psychol. 39, 133 – 152. doi: doi:10.1111/bjdp.12356
  39. Yang L., Park Y. (2022). Group membership trumps shared preference in five-year-olds' resource allocation, social preference, and social evaluation. Front. Psychol. 13: 866966. doi: doi:10.3389/fpsyg.2022.866966
  40. Zmyj N., Aschersleben G., Prinz W., Daum M. (2012). The peer model advantage in infants' imitation of familiar gestures performed by differently aged models. Front. Psychol 3: 00252. doi: doi:10.3389/fpsyg.2012.00252

Citation

Ahn, J., et al. (2026). Short-term familiarity shapes social preference but not word learning in preschoolers. Frontiers in Developmental Psychology, Volume 4. https://doi.org/
Received
Accepted
Published
Volume
4 - 2026

Copyright

Copyright © 2026 Ahn, Cartmill and Sandhofer.

License: https://creativecommons.org/licenses/by/4.0/

*Correspondence: Jina Ahn

Article metrics

Views
0
Downloads
0
Citations
0
Shares
0

Related articles

PerspectiveFriluftsliv (open-air life) as a cultural site of emotion socialization in early childhood in NorwayEvi Petersen, Natália Meireles Santos da Costa, Tuva Beyer BrochReviewScreen exposure and cognitive outcomes in school-aged children and adolescents: a narrative review through the lens of Cognitive Load TheoryHiya Al Kalbani, Erin Kitt-LewisOriginal ResearchGratitude in childhood: ethnotheories from the Turkana of Kenya and Amhara of EthiopiaPamela Akinyi Wadende, Missaye Mulatie Mengstie