Introduction

The first three years of life is a prototype period for our species, since it is during this time that the groundwork is laid for adult life (Wallon, 1980). Feeding has been shown to be a key element during this period for physiological, psychological and social development (Léniz et al., 2021; Valsiner, 1987), and it is fundamental to determine not only what food is eaten, but also how, in what quantities and in which environments (McIsaac et al., 2022; Tovar et al., 2019). Transitioning from breast or bottle to solid food constitutes an important learning process that involves moving from total feeding dependency to a certain degree of autonomy (Belza et al., 2020; Black & Hurley, 2017; Vincze, 2018). In this process, a turning point occurs from the second year of life onwards, when food placed on the table goes from being complementary to being the principal source of nutrition, a circumstance that results in an overlap between the social–cultural and the intrinsically nutritional dimensions (Hallam et al., 2016; Hughes et al., 2015; Ishiguro, 2016; Rodríguez et al., 2017).

Children’s eating behaviour and their development will differ in accordance with the specific combination of factors involved, namely the environment and, specifically, the conditions promoted by adult caregivers. Consequently, there is increasing interest in defining parenting practices and styles linked to feeding, and in analysing their influence on children’s subsequent eating habits (Frankel et al., 2014; Shloim et al., 2015; Ventura & Birch, 2008). Parenting practices are defined as the way in which parents behave towards their children in order to achieve a specific goal, in this case, for example, to encourage them to eat their food. The concept of feeding style, on the other hand, refers to the emotional climate created by the adult caregiver during mealtimes (Hodges et al., 2013; Hughes et al., 2011; Power et al., 2018). Research has shown that the ideal situation is for adult mediation not to interfere with questions such as the amount of food eaten, since self-regulation is a key preventive factor for excess weight gain and child obesity (Hoffmann et al., 2016; Kiefner-Burmeister et al., 2014; Léniz et al., 2021; McIsaac et al., 2022). However, a sound body of literature has demonstrated how babies’ innate capacity to self-regulate their food intake declines as they grow and the influence of their environment increases (Faith et al., 2004; Frankel et al., 2014; Monnery-Patris et al., 2019). By their first birthday, it appears that young children stop paying attention to their own internal hunger and satiety signals and start being guided by external signals, mainly those conveyed by their adult caregivers.

In this sense, responsive feeding (Benjamin-Neelon, 2018; Black & Hurley, 2017; Heller & Mobley, 2019; Hodges et al., 2013) refers to parenting practices that encourage children to feed themselves autonomously and in response to their own physiological needs. For children to become aware of these needs and then be able to express them, it is vital to establish an appropriate environment, emotional climate and routine. According to various authors, this sensitive and responsive care style during meals supports healthy physical and psychological development during early childhood, including the self-regulation of hunger (Ainsworth & Bell, 1969; Heller & Mobley, 2019; Hughes et al., 2015; Malek-Lasater et al., 2022).

However, the research carried out in this field has some limitations. Firstly, studies to date have been based on caregiver self-reports of their own behaviour (Dev et al., 2020; Heller & Mobley, 2019). This prevents us from determining the exact type of feeding style adopted, since several studies have pointed out that self-reports on participants’ own feeding behaviours often differ from observed feeding behaviours (Hodges et al., 2013; Hughes et al., 2011; Lewis & Worobey, 2011; Power et al., 2018). Systematic observation, on the other hand, allows us to assess spontaneous behaviour in its everyday context (Anguera, 1979, 2003; Portell et al., 2015). This methodological option offers valuable procedural resources for the study of daily life in the classroom and the adult–child relationships that are established therein (Anguera, 2001; Escolano-Pérez, 2020; Terroba et al., 2021). Actually, the literature on child education and development advocates direct, systematic observation in the learning context itself (Early Head Start National Resource Center, 2013; Escolano-Pérez, 2020). Secondly, there are limits linked to context, since most studies to date have focused on the family environment (Malek-Lasater et al., 2022; Swindle et al., 2018; Tovar et al., 2019). Although schooling from 0 to 3 years has increased significantly this century and despite the amount of times children eat meals today in nursery schools (in many cases breakfast, lunch and afternoon snack), research in this field is still confusing, particularly in relation to the implementation of responsive feeding practices in those early education centres (Malek-Lasater et al., 2022; Swindle et al., 2018). Some studies have observed feeding times in nursery schools (Hallam et al., 2016, Ishiguro, 2016, Rodríguez et al., 2017; Malek-Lasater et al., 2022) and, despite working with small samples, have reported enlightening results regarding the role played by adults in this learning process. All remark on the need to explore best feeding practices designed to improve everyday interactions in the nursery setting.

The Pikler-Lóczy education (see compilation in Herrán, 2018) is an international benchmark in the field of high-quality institutional care and education for children aged 0–3 years. This model is based on four main principles (David & Appell, 2010; Falk, 2018): (1) valuing children’s autonomous activity, (2) establishing a privileged, close and continuous affective relationship with the educator, (3) fostering self-awareness and awareness of one’s environment and (4) ensuring good physical health. This education was designed and has been implemented and developed for decades in a centre in Budapest (Falk & Pikler, 1972; Pikler, 1984; Tardos, 2007) that used to be a children’s home (1946–2011), and is today the Emmi Pikler Nursery School.

Over the past decade, several authors have highlighted the interest and unique nature of the Pikler-Lóczy education (Herrán, 2013; Herrán et al., 2014; McCall et al., 2010; Salamon & Harrison, 2015) and recent research has used the observational methodology to explore the everyday behaviour of educators at the school (Belasko et al., 2019, 2022; Belza et al., 2019, 2020; Sagastui et al., 2020, 2021). Studies have shown that the habitual behaviour of educators ensures the stability of the environment and the consistency of educational behaviours, thanks to the use of two dimensional patterns, one that orders the physical, tangible and material plane of everyday routine through a sequence of instrumental actions, and another, relational one, inherent to this education model, which is made up of specific verbal, spatial and gestural actions.

Continuing with this avenue of research, the present study aimed to use systematic and direct observation to analyse and explain some of the specific feeding behaviours engaged in by educators, focusing not only on nutritional aspects, but also on the educational dimension of the group eating activity in which children engage on a daily basis. To this end, the study analysed both dimensions of behaviour at this nursery school during two educators’ breakfast time with their respective group of children. The first specific aim, focused on the first plane, was to identify basic common behavioural patterns that order the sequence of instrumental actions carried out during the breakfast activity. We expected both educators to carry out similar actions with a view to establishing a daily routine designed to encourage children’s self-regulation. The second specific aim, focused on the second plane (relational), was to compare the behaviour of the two educators. We expected to find a shared relational pattern designed to support instrumental actions, with both educators exemplifying a comprehensive responsive feeding style for early childhood education (0–3 years).

Materials and Methods

This work forms part of a line of research into the Pikler-Lóczy education developed under the scope of Observational Methodology, which enabled us to study spontaneous, habitual behaviour in its natural context (Anguera, 1979, 2003; Portell et al., 2015). It is therefore an ideal option for assessing educational interventions (Alcover et al., 2019; Escolano-Pérez et al., 2019; Terroba et al., 2021).

Based on the proposal by Anguera et al. (2001) regarding observational designs, a nomothetic, follow-up and multidimensional (N/F/M) design was used for the study of two educators. The two educators (units of study) were observed as they cared for their respective groups of children in the classroom. A total of 25 breakfast sessions were videotaped over the space of 3 months, ten in Educator 1’s classroom and 15 in Educator 2’s classroom; each session was recorded in a continuous, uninterrupted fashion (follow-up criterion). Several aspects of the behaviour of both educators were taken into consideration simultaneously (multidimensional criterion).

Participants

Two educators from the Emmi Pikler Nursery School with over 30 years' professional experience were observed first thing in the morning, during breakfast with the children in their groups in the area of their respective classrooms used for mealtimes. Educator 1’s group comprised 4 boys and 8 girls, aged between 26 and 47 months at the beginning of the observation period. Educator 2’s group comprised 5 girls and 5 boys aged between 20 and 36 months. Children’s participation in the breakfast activity varied from session to session, since it is optional.

Breakfast is offered to all children who arrive before 9 o´clock in the morning, although they may, if they wish, go straight to play in the area designated for that activity, which is separated from the eating area by a guardrail. If the child decides to have breakfast, they sit at the table and are offered a drink (water, milk, infusion or juice) and something to eat (every day there is bread with margarine and/or cereal available, as well as other items which vary from day to day, such as cold meats, cream cheese, kohlrabi and peppers). The educator places the food chosen by each child on the table, along with their respective utensils, which are prepared beforehand and stored in the cupboard–sideboard. When the child finishes, they move to the play area by opening the guardrail. The session ends when the educator collects up all the used breakfast utensils and puts the uneaten food away.

The research project was approved by the Ethics Committee at the authors’ University (INA0139), as well as by the management team of the nursery school. Informed consent was obtained from both the educators and the children's parents.

Instruments

Observation Instrument

The observation instrument used in this study follows the Giving breakfast in the Emmi Pikler Nursery School (GBEPNS) field format (Belza, 2020; Belza et al, 2019, 2020) and was created ad hoc to observe breakfast activity at the Emmi Pikler Nursery School. We decided to construct this type of multidimensional instrument as a result of the complexity of the situation observed. The instrument is susceptible to multiple coding (Anguera et al., 2007) and enables all perceivable behaviours to be registered and grouped into pre-established dimensions and sub-dimensions in accordance with specific objectives.

A recent review of studies on responsive feeding (Heller & Mobley, 2019) identified verbal expression, physical availability, visual attention, expressions of affection and gestures as elements that make up a responsive style. In the present study, as well as recording actions involving the handling of utensils and food by educators during breakfast (instrumental plane), the dimensions of the field format also gather information about the relational plane, including verbal, proxemic and kinesic behaviours. The dimensions were ordered in accordance with the proposal made by Weick (1968) and modified by Anguera (1979), regarding handling, verbal and postural responses, moving around and exchanging looks and gestures. These are aspects that are widely recognised as valid for explaining the significance of any educational action (Castañer et al., 2013; Sagastui et al., 2020). Table 1 presents the hierarchical system used to order all the dimensions selected for this study, along with their definitions. In the field format, each sub-dimension folds out into a list of observed behaviour codes, along with their name and definition.

Table 1 Macro-dimensions, dimensions and sub-dimensions of the field format GBEPNS
Full size table

Recording and Analysis Software

The information gathered using the observation instrument was systematically recorded and uploaded to a database for subsequent analysis by two computer programs: (1) the free software SDIS-GSEQ (version 4.1.2), updated to GSEQ [https://www.mangold-international.com/en/products/software/gseq.html] (Bakeman & Quera, 2011), and (2) the free software HOISAN, version 1.6.3 [www.menpas.com] (Hernández-Mendo et al., 2012). This last program was used to control the quality of the data.

Procedure

Firstly, 25 sessions were compiled in which the educators observed served breakfast to their group. All sessions corresponded to breakfast time and were videotaped in each educator’s own classroom during the established timetable: 08:30 to 09:30. Breakfast sessions can last between 10 and 40 min, depending on how many children are eating at the same time and at what time each starts.

Then, each session was recorded. The record is the transcription in codes of what was observed during the session, using the codes defined in the observation instrument (Anguera & Izquierdo, 2006). This phase is especially important, since it enables the transformation of the qualitative data record into a code matrix that can be analysed in a quantitative fashion (Anguera et al., 2018), since this data matrix contains information about the frequency (real number of occurrences), order and sequence of behaviours. As a result, certain categorical data analysis techniques can be used that are adapted to the profile of observational methodology and produce a set of quantitative results with explanatory power (Anguera et al., 2021).

Data Quality Control

Before conducting the relevant analyses, it is first necessary to verify the quality of the data collected (Arana et al., 2016). To this end, Krippendorff's intra- and inter-observer canonical concordance (2013) was calculated using the HOISAN program. Furthermore, 10% of the observation period was registered twice at different moments to enable a comparison and then registered once again by an external observer. In both cases the result was satisfactory, since the inter-observer agreement score was 0.89, and the score for the two records taken at different moments was 0.92.

Data Analysis

The lag sequential analysis technique was applied to analyse the data, using the software GSEQ. This technique was initially proposed by Bakeman (1978) and is designed to detect regularities in the succession of registered behaviours or behavioural patterns with a significantly higher probability of appearance than would be expected by chance. To this end, and based on a type of behaviour considered to be a possible initialiser or trigger for what comes next (called a criterion behaviour), a table was compiled of conditional or transition frequencies in the successive lags, along with the corresponding conditional likelihoods (Bakeman & Gottman, 1989; Bakeman & Quera, 2011; Quera, 2018). The unconditional likelihoods were also calculated on the basis of the frequency of each of the categories in the register, divided by the total sum of the row (N). Conditional likelihoods depend on the order of occurrence of the behaviours, and the unconditional ones on their frequency throughout the register. Once the observed and expected conditional likelihood values have been found, they are compared by means of a binomial test, with the result being a series of residuals adjusted to each of the established lags, which are then subjected to the adjustment proposed by Allison and Liker (1982) to obtain the hypergeometric z score. This enables us to identify the significant activating behaviours in each lag (adjusted residuals > 1.96 and > 2.58, with a significance level of p < 0.05 and p < 0.01, respectively). This existence of a cohesive force (associative activating relationship) between the different behaviours, which is stronger than would be triggered just by chance, enables us to reveal a behavioural pattern.

Results

The principal aim of the present study was to verify the existence of a basic shared pattern of instrumental behaviour, or in other words, to identify the similar sequence of instrumental actions deployed by both educators during breakfast time. To this end, the picks up/puts down (code D206) action was selected as the criterion or initialising behaviour of the chain of actions, and six instrumental actions linked to breakfast management were selected as the conditional behaviours: again picks up/puts down (D206), transfers (D209), offers (D204), serves (D207), clears up (D212) and no action (D225).

The adjusted residuals from lags 1 to 5 were calculated to explore the succession of behaviours or intentional concatenation of actions. The scores obtained in the significant adjusted activator residuals (> 2.58 with a significance level of p < 0.01) were used to identify the behavioural patterns of Educator 1 and Educator 2, for subsequent comparison. Table 2 presents the results of both. The little variation observed effectively reflects the basic shared sequence used by both educators to serve breakfast.

Table 2 Behavioural patterns of the two educators based on significant adjusted residuals of the lag sequential analysis
Full size table

The second specific aim is to detect similarities between the two educators on the relational plane and, in other words, to find shared ways of supporting and making sense of the instrumental actions in order to ensure the healthy management of the breakfast activity. A sequential analysis of lag 0 (or co-occurrence) was performed to identify the relational behaviours engaged in simultaneously with the instrumental behaviours selected in the previous analysis (no action, picks up/puts down, transfers, offers, serves, clears up). At this point, those six behaviours were considered criterion behaviours, whereas the others corresponding to the relational dimension were considered conditional. The results (see Table 3) are explained below following the same order as the relational sub-dimensions in the field format: verbal behaviour, proxemic behaviour and kinesic behaviour.

Table 3 Concurrence of educators’ instrumental and relational behaviours based on significant adjusted residuals
Full size table

Verbal Behaviour

Verbal behaviour was very similar in both educators. Moments in which they engaged in no instrumental action (D225) led to chatting (P103), whereas when they served (D207) they did it describing, informing or explaining (P145) what it was they were doing or serving at that moment. It should be noted that only when they picked up/put down (D206) utensils and food they did not speak.

Proxemic Behaviour

Proxemic behaviour was analysed from two perspectives: static (selection of place and posture in the breakfast area of the classroom) and dynamic (series of displacements inherent to the breakfast activity).

In general, the results indicate the permanent availability of the educators, since even when they engaged in no action (D225), they sit down (P305) and remained beside (P410) the child. Moreover, in the alone static criterion, the results were identical for the two educators, with the most significant findings for both being offers (D204) and serves (D207) food by bending towards (P303) in order to lower themselves to the same height as the face of the child they were caring for, who was sitting beside them. In the static with child criterion, the educators positioned themselves next to or opposite (P410) their charge.

Kinesic Behaviour

The present study also analysed gaze, hand gestures (with direct contact), emblems, illustrators and regulators.

In relation to gaze or visual attention, similarities were observed between the two educators in the majority of actions. Both educators transferred (D209) the food while looking at the child they were attending to at the time (P606) and were about to serve. However, when they were serving (D207) and clearing up (D212), they looked at what they were doing (P604). What is curious about this section is that the results revealed a different use of joint attention (P605). Educator 2 used joint attention principally to offer (P204) an item of food. In other words, she used it to establish a joint focus with each child on an object, a third element about which she was making a proposal. Educator 1, on the other hand, offered the child in question food while looking straight at them (P602) and used joint attention (P605) during moments in which there was no action (D225). In other words, she used it to establish joint focus on the topics of the conversation initiated during those moments.

No results were obtained regarding hand gestures, although the analyses of emblems, illustrators and regulators were rich. The results confirmed that both educators used the same gestures during various instrumental actions. For example, both used a centripetal gesture (P1004) in the absence of action (D225), indicating waiting or giving the child time (P908). Both also positioned themselves strategically (P907) when transferring (D209) food in order to deepen the focus of attention (P1007) on the child when offering it (D204), pointing to or showing them (P901) the food in question in order to make them do (P1103) or respond to the offering.

Discussion

The results of the present study demonstrate that the interactive behaviour observed in the two educators working at the Emmi Pikler Nursery School during breakfast time was very similar. A pattern of specific behaviour common to both of them was observed: they ordered their instrumental actions to manage and progress through the breakfast task in exactly the same way, since these actions are vital to establishing a clear, unequivocal daily routine. Moreover, the relational dynamics observed were also part of both educators’ habitual actions, with close similarities being observed in their verbal, proxemic and kinesic behaviour. The results therefore indicate the existence of a shared feeding style, inherent to this educational approach, which includes certain key elements of responsive feeding.

The direct and systematic observation of these two educators enabled the identification of specific behaviours that make up the feeding style developed at this school. It therefore overcomes one of the limitations of previous studies, which were unable to precisely describe feeding practices just as they occurred in everyday sessions. We will now discuss the significance of these feeding practices, or in other words, the specific educator behaviours observed, relating them to the premises of the Pikler-Lóczy education and the ideas expressed in more recent studies on responsive feeding.

First, the sequence of instrumental actions observed indicates that educators serve children food several times during the breakfast activity. In other words, as Belza et al. (2019) pointed out, during breakfast, educators pass round the group more than once (Wallon, 1980), picking up and putting down utensils and food, transferring them and offering, serving or clearing up. More than serving a predetermined quantity of food, educators serve breakfast little by little as many times as necessary, in accordance with the children’s desires and tastes, until they are full. Another striking finding is the fact that, at certain moments, the absence of action was a common approach used by both educators, who often engaged in no instrumental action at all, but rather simply accompanied the children with an active, alert presence. This makes sense in the framework of responsive feeding (Black & Hurley, 2017; Heller & Mobley, 2019), since it facilitates recognition of the children's signals, with the educators accompanying them without intervening directly, but remaining close and interested in their activity. Consequently, the results suggest that the procedure followed by the educators is designed to foster children's capacity for regulating their own hunger, since the progression and completion of the breakfast activity is modelled in accordance with the children's readiness and feeding behaviour (Hodges et al., 2013).

Second, recent studies have classified feeding styles based on combinations of the responsiveness and demandingness dimensions (Frankel et al., 2014; Hughes et al., 2011; Power et al., 2018). Responsiveness refers to how willing adults are to respond to the needs expressed by the child in their care, whereas demandingness refers to how they encourage them to eat through suggestion, example or imposition. In this case, both the proxemic behaviour and gaze of the educators indicate two characteristic situations linked to breakfast time and reflect the specific nature of the two dimensions that together make up the feeding style deployed in this nursery school. Firstly, alert availability with no instrumental action (responsiveness dimension) is established by the educators remaining close, without moving or doing anything other than sitting beside the children and focusing their gaze on them in order to look after their possible needs. Secondly, a request for cooperation (demandingness dimension) is made through the posture they adopt when serving, bending forwards and positioning themselves beside or opposite the child and adjusting their gaze to the appropriate level of interaction (Sagastui et al., 2020, 2021), linking instrumental actions to the children’s capacity for action and joint attention (Belza et al., 2019; Ishiguro, 2016).

Third, one of the most revealing findings of this study was the absence of results regarding hand gestures with direct physical contact. This suggests that in the Emmi Pikler Nursery School, emotionality is deliberately kept under control (Herrán, 2013). Although this may seem strange in relation to young children, studies that have analysed this aspect in the field of responsive feeding report mixed results (Hodges et al., 2013; Hoffmann et al., 2016; Hughes et al., 2011). Indeed, Hughes et al. (2015) argue that an inadequate emotional climate may have a negative influence on children’s attention to their internal hunger and satiety cues. This finding supports the strategy adopted by observed educators not to associate direct intervention with the children in their care with explicit expressions of positive or negative affect, but rather to opt for a constant communicative flow, characterised by language rich in verbal and gestural exchanges (Belza et al., 2020). According to the Pikler-Lóczy education (2018), the basis of adult–child relationships is rooted in mutual attention and trust, with the adult recognising the child as an active companion capable of influencing the events that concern them. Consequently, both educators serve food while explaining and describing the situation, in order to enable the child to understand what is happening and feel they are participating in the activity.

Fourth, the results also reveal that communicative exchange is not just verbal in nature, but also intentionally gestural (Belza et al., 2020; Rodríguez et al., 2017). The verbal discourse is linked to unequivocal gestures, such as, for example, pointing to or showing food when offering it, holding objects while they are served and explaining what they are, and centripetal gestures associated with waiting and giving the child the time they need. Through this unified communicative model, the two educators guide, or rather regulate, the activity, avoiding forcing children to eat either more or less than they want to as a result of their intervention. Several studies have found that mothers tend to be more responsive to their children's hunger cues than to satiety cues (Heller & Mobley, 2019). In this case, our results confirm that the demandingness dimension exercised by the two educators is not oriented towards eating, but rather towards ensuring that it is the child who decides when to eat and when to stop. To this end, before serving them, both educators strive to deepen the child’s focus of attention, showing them the food in order to prompt a response to the specific proposal being made and encouraging them to indicate their acceptance or rejection of the food in question. In other words, they do not make them eat in the ways reported in other studies (Hughes et al., 2011), either directly (spoon-feeding) or indirectly (asking them to try a bite, reasoning with them about why they should eat more or trying to trick them through games).

Finally, it is important to note that, although the repertoire of shared behaviours is fairly extensive, a certain degree of diversity was also observed. Differences were mainly found on the relational plane, although some small nuances were also observed on the instrumental one. This may be due to the individual personalities of the two educators, the structural and organisational characteristics of the two classrooms or, most likely, to the specificities of the children in each group (age, maturity level, personality, impulsiveness, etc.). Explaining this variability is beyond the scope of the analyses carried out in the present study and is something that future research may wish to explore.

Conclusions

In sum, the results reveal that the feeding style shared by both educators in our study is based on interlinking, time and time again, the same succession of actions required for all children to eat breakfast as they see fit, with ongoing adult availability and attention that fosters the establishment of a true dialogue between all participants around the table. This interlinking helps educators provide individual attention to several children, almost at the same time, since they serve one after the other, respecting each one’s habits, tastes and pace, thereby collaborating in the establishment of each child’s individual and group identity by the third year of life (Wallon, 1980). Moreover, by acting in this way, the educators are able to listen to each child in order to respond appropriately to their feeding cues. They therefore illustrate how to implement responsive feeding in the everyday activities of a nursery school.

The present study provides evidence of a high-quality early education model. The actions of the educators working at the school reflect the aforementioned fundamental principles of the Pikler-Lóczy education (David & Appell, 2010; Falk, 2018), in the form of basic premises that are applied during children’s mealtimes.

  1. 1.

    Do not interrupt the child’s autonomous activity: once the appropriate utensils have been provided, the transition to autonomous feeding involves letting the child eat at their own pace, without constant intervention, but remaining close in order to adapt to and cooperate with what they want to do by themselves and what they want the adult to do for them.

  2. 2.

    Constant interaction: the continuous flow of communication establishes a close educator–child relationship, in which the adult explains what is happening or what is going to happen next, making the child feel secure and preparing them, while organising their own activities in such a way as to enable them to perceive the child’s cues and respond to, reinforce and make sense of them.

  3. 3.

    ‘Not one spoonful too many’: as well as avoiding forcing, distracting or exonerating the child, in order to encourage them to gain a greater awareness of their internal hunger and satiety signals, the entire physical/material and human/relational environment should be kept stable and free from outside improvisations.

  4. 4.

    Respect the child’s natural capacity to self-regulate: this is both the starting point and outcome of all the above, and a means of ensuring good physical health, since, in the field of nutrition, it is considered a protective factor against child obesity.

These four premises involve a radical change in the way adults view early childhood feeding. Thanks to the tools provided by the observational methodology, we have been able to explore the significance of this educational intervention and show how it is possible to develop healthy feeding practices in children’s everyday lives, with no interference or bias. However, it should be pointed out that, in order to obtain high-quality data that can be analysed quantitatively and have sufficient explanatory power, it is necessary to observe habitual behaviour in a direct, ongoing manner, complying with certain formal aspects and limiting the object of study to a specific situation. Given the specific area under study, this involves working with a small number of participants, about whom a great deal of information is gathered.

In this sense, the observation instrument used was developed specifically for the context being observed. It was intentional and extremely precise in order to record every aspect of the educators’ behaviour and the environment they generated, right down to the very last detail. However, after having demonstrated the relationship between the different components of responsive feeding and their integration into everyday educational behaviour, it is now important to think about adapting the instrument to other contexts. A simplified, versatile and easy-to-use observation instrument may be useful not only for evaluation purposes, but also as a training tool for professionals and parents, to help them improve their performance in this area and enhance their everyday interactions with their young charges/offspring.

Finally, one limitation of the present study is the fact that the behaviour of the children themselves was not observed. It would be interesting to conduct a longitudinal study to determine how this type of early feeding affects children’s eating habits later in life, as indeed Pikler and Falk did (1972) with the children brought up at Lóczy.