Closeness and conflict in teacher-child relationships in preschool: the role of child personality types

ABSTRACT The positive impact of warm and supportive relationships between children and preschool teachers on children’s development has been highlighted in many studies. However, research gaps exist regarding the prediction of teacher-child relationships in preschool. In particular, little is known about associations between child personality and relationships with preschool teachers, although previous evidence indicates links between personality and social relationships in general. The study aims to address this research gap and identifies associations between child personality types and teacher-child relationships by using a sample of 163 children attending 65 preschool classes in Austria. After controlling for child and teacher characteristics, the results show that preschool teachers perceived less closeness with children characterized as ‘overcontrollers’ and ‘undercontrollers’ compared to children described as ‘resilients’. Furthermore, the preschool teachers perceived their relationships as more conflictual with ‘undercontrollers’ compared to ‘resilients’ and ‘overcontrollers’. Considering children's personality types seems to be important to improve teacher-child relationships.


Introduction
An increasing body of evidence shows that interactions and relationships for children not only within but also outside the family are important for a child's wellbeing and development (Thompson, 2014).In early childhood, this is particularly true for the relationship with teachers or caregivers (Burchinal et al., 2008;Cameron, 2012;Pianta, Downer, & Hamre, 2016).Many studies demonstrate that a warm and supportive teacher-child relationship in ECCE institutions has a lasting positive effect on child development (Ahnert & Harwardt, 2008;O'Connor & McCartney, 2007;Sabol & Pianta, 2012).Moreover, high-quality teacher-child relationships can serve as a protective factor for behavioural adjustment problems and have been shown to be supportive for children with learning difficulties (Sabol & Pianta, 2012).Furthermore, high-quality teacher-child relationships can buffer children from negative consequences resulting from previous low-quality caregiving experiences and they seem to be protective for children with specific demographic factors designated as risky such as minority status and low parental educational background (Sabol & Pianta, 2012).
With regard to the factors upon which a good teacher-child relationship in early childhood depends, individual characteristics of the persons involved are shown to be theoretically relevant (Pianta, Hamre, & Stuhlman, 2003).Furthermore, there is a wide range of corresponding empirical evidence for the influence of various individual characteristics on the quality of teacher-child relationships in early childhood such as teacher's work experience (e.g.Hamre, Pianta, Downer, & Mashburn, 2008), teacher's personality traits like self-efficacy (e.g.Mashburn, Hamre, Downer, & Pianta, 2006), child's gender (e.g.Choi & Dobbs-Oates, 2016), and child's language skills (e.g.Sullivan, Hegde, Ballard, & Ticknor, 2015).Because the professional discussion emphasizes the active role of children in their development (Copple & Bredecamp, 2009;Magnusson & Stattin, 2006) and the claim that teachers should pay attention to the individual needs of the children (Copple & Bredecamp, 2009), knowledge of relevant child characteristics is of particular interest for professional relationship building with children.In this context, a rarely focused-upon characteristic is the child's personality.In general, the influence of personality on relationships is grounded theoretically (Back et al., 2011;McCrae & Costa, 1996) and empirically (Jensen-Campbell, Knack, & Rex-Lear, 2009).Furthermore, there is even some particular evidence that temperament traits of children such as shyness or effortful control play a role in affecting the quality of teacher-child relationships in early childhood (Justice, Cottone, Mashburn, & Rimm-Kaufman, 2008;Rudasill & Rimm-Kaufman, 2009).However, to our knowledge there are no studies specifically on the influence of personality types of preschoolaged children to date.Even more, there is a lack of research from German-speaking countries, which would be important, as the broader cultural context might also play a role in shaping relations between personality and social relations (Jensen-Campbell et al., 2009).Against this background, this study examines how the child's personality predicts the preschool teacher-child relationship in Austrian preschools.

Teacher-child relationships: theoretical conceptualization
In the present study, we refer to the 'Conceptual Model of Child-Teacher-Relationships' (Pianta et al., 2003), which is an update and extension of Pianta's (1999) model of relationships.Studies of relationships between teachers and children have multiple roots in educational psychology and developmental psychology, but one of the greatest influences on today's view of the effects of relationships on development is attributed to attachment theory (Pianta et al., 2003).The model from Pianta et al. (2003) is additionally based on principles of systems theory, for example referring to Bronfenbrenner and Morris (1998).
The key point of this integrative model is that relationships are not simply determined by their interactions but are characterized by the complex interaction of the following components: features of individuals, representational models, information exchange processes, and external influences.Features of individuals (e.g.gender, ethnicity, personality, self-efficacy, biographical or work experience) are described as the most basic element of relationships (Pianta et al., 2003).Representational models are defined as a 'set of feelings and beliefs that has been stored about a relationship that guides feelings, perceptions, and behaviour in that relationship' (Pianta et al., 2003, p. 210).Information exchange processes refer to reciprocal exchanges between teacher and child, in which the more important element is how information is exchanged (e.g.tone of voice) as compared to what information is exchanged.Finally, external influences highlight the fact that cultural or structural aspects of the institution must also be considered (e.g.laws and rules, including those about the professional-child ratio) (Pianta et al., 2003).
This paper focuses on relationships between preschool teachers and children from the perspective of the two components exchange processes and representational models.These two elements are considered in the widely used 'Student-Teacher Relationship Scale' (STRS; Pianta, 2001), which captures teachers' perceived processes and representations about the relationship with a child along the three dimensions of 'closeness' (degree to which the teacher perceives the relation to a child as affectionate and warm), 'conflict' (extent to which a teacher is struggling with a child, viewing the child as angry or unpredictable), and 'dependency' (level of perceiving the child as requesting help if not needed and reacting strongly to separation from teacher).
While there is a lot of research on the dimensions 'closeness' and 'conflict', fewer studies have focused on 'dependency' (Verschueren & Koomen, 2021).This is also evident in the studies on the influence of individual characteristics on teacher-child relationships presented in the next section.

Individual characteristics as predictors of closeness and conflict in teacher-child relationships
Although, according to Pianta et al. (2003), individual characteristics of teachers and children can be viewed in a broader sense as components of relationships, they can also be viewed as predictors of the procedural and representational qualities of relationships, such as closeness or conflict.In this line of reasoning, many studies have investigated features of teachers and children as influences on relationship quality.
One feature of teachers that seems to be important is teachers' self-efficacy, as it can be a protective factor against poor teacher-child relationship quality (Zee & Koomen, 2016).Hamre et al. (2008) showed that higher levels of teachers' self-efficacy are associated with fewer teacher-child conflicts in prekindergarten classrooms.Furthermore, Mashburn et al. (2006) found significant positive relations between teachers' self-efficacy and teacher-child closeness in prekindergarten classes.In addition, a recent study of Vatou, Gregoriadis, Tsigilis, and Grammatikopoulos (2022) showed that preschool teachers' social self-efficacy was positively related to teacher-child closeness and negatively related to teacher-child conflicts.However, some studies did not find significant associations between teachers' self-efficacy and teacher-child relationship quality (for an overview, see Zee & Koomen, 2016).
Previous research also shows that preschool teachers' stress predicts relationships and interactions between children and their teachers (e.g.Penttinen, Pakarinen, von Suchodoletz, & Lerkkanen, 2020;Whitaker, Dearth-Wesley, & Gooze, 2015).Whitaker et al. (2015) found that more workplace stress corresponded with more teacher-child conflicts and, to some extent, to less teacher-child closeness.Furthermore, Penttinen et al. (2020) showed that higher teaching-related stress predicted lower scores for the quality of teacher-child interactions in the dimensions 'emotional support' and 'classroom organization'.
Sometimes teachers' work experience is also considered as a predictor of teacher-child relationship quality (e.g.Chung, Marvin, & Churchill, 2005;Hamre et al., 2008;Mashburn et al., 2006).While Mashburn et al. (2006) found a significant negative effect between teachers' years of work experience and closeness, other studies did not find significant associations between years of experience and teacher-child relationships (e.g.Chung et al., 2005;Hamre et al., 2008).Other teacher-related predictors of teacher-child relationships that are sometimes mentioned are teachers' education or teachers' ethnicity (e.g.Chung et al., 2005;Hamre et al., 2008;Mashburn et al., 2006).
Because language skills provide resources for interactions, which are a central aspect of relationships in the presented model (Pianta et al., 2003), several studies (Justice et al., 2008;Rudasill, Rimm-Kaufman, Justice, & Pence, 2006;Sullivan et al., 2015) have examined this child feature in terms of its predictive influence on the teacher-child relationship.Results show a closer teacher-child relationship if children had a better comprehension of language (Justice et al., 2008) or were native English speakers just as the teachers were, in comparison to children who were English Language Learners (Sullivan et al., 2015).However, for the relationship dimension of conflict, Sullivan et al. (2015) point out that teachers had higher conflicting relationships with English native speakers.
Various aspects of family background have also been studied as predictors of relationships.Ladd, Birch, and Buhs (1999), for example, used a composite measure for family background (parents' socio-economic status, ethnicity, and educational attainment) and showed that kindergarten children from more advantaged and nonminority families had fewer conflicts and were closer with their teachers.Jerome et al. (2009), however, found no influence of maternal education on the level of conflict and closeness in the teacher-child relationship in kindergarten, though their data point to a closer relationship with teachers when children came from families with a higherquality home environment.Mayer and Beckh (2018) showed a significant effect of the quality of the mother-child relationship on the quality of the teacher-child relationship, although this influence appeared to be partially due to the child's problem behaviour.
Furthermore, there is evidence that the child's ethnicity plays a role in determining the quality of the teacher-child relationship in early childhood, even if the findings are limited and partly inconsistent.For example, teachers report perceiving more conflict with Black children than with White children (Jerome et al., 2009) or with children with a migration background (Mayer & Beckh, 2018).Saft and Pianta (2001) point out the relevance of the ethnic match between teachers and students: If teachers and children were from the same ethnic group, teachers reported more closeness as well as less conflict.While Choi and Dobbs-Oates ( 2016) and Ewing and Taylor (2009) found no relationship between ethnic match and relationship quality with the STRS, Howes and Shivers (2006) study shows that ethnic match between caregivers and toddlers also matters when it comes to good relationship quality in terms of secure attachment.
McCrae and Costa's (1996) five-factor theory of personality provides hints for links between child personality traits and teacher-child relationships by considering the traits as 'basic tendencies', whereas interpersonal relationships are described as their concrete manifestations and classified as 'characteristic adaptations' (McCrae & Costa, 1996, p. 69).More specifically, this might reflect the presumption that individuals evolve interpersonal relationships that are consistent with their personality traits (McCrae & Costa, 1996).More recently the theoretical framework PERSOC (composed of PERsonality and SOCial relationships; Back et al., 2011) has been developed to illustrate the underlying processes describing the interplay of personality and social relationships in different contexts such as pedagogical settings.One central assumption is that the interdependency between individual dispositions (ID, e.g.own perception of Conscientiousness) and relationship dispositions (RD, e.g.trust in relevant others) is the foundation for links between personality and social relationship.This is because ID and RD can have the same representational roots or because important or several similar RD to different persons can not only influence ID, but can themselves evolve into an ID (Back et al., 2011).
A growing amount of empirical research is focusing on relations between personality traits and social relationships such as family relationships, peer relationships, and work relationships (Back, 2015).However, empirical research specifically targeting relations between child personality traits in the form of the Big Five and teacher-child relationships in preschools is rare.At least there is some evidence for associations between children's temperament, which has been discussed as partly overlapping with personality traits (De Pauw, Mervielde, & van Leeuwen, 2009; Shiner & DeYoung, 2013), and teacher-child relationships (Rudasill et al., 2020).In a study on first-graders, Rudasill and Rimm-Kaufman (2009) showed that less shy children and children with lower effortful control had more conflictive relationships with teachers.Furthermore, lower shyness and higher effortful control corresponded with closer relationships with teachers (Rudasill & Rimm-Kaufman, 2009).Similar findings on relations between children's shyness and closeness have been reported by Justice et al. (2008).Furthermore, a meta-analysis of temperamental characteristics (including 19 studies on preschool and school-aged children) revealed that children's shyness was negatively related to the closeness of teacher-child relationships (Nurmi, 2012).

Study aims
Numerous studies have identified individual characteristics of preschool teachers and children as predictors of teacher-child relationships (e.g.Hamre et al., 2008;Jerome et al., 2009;Justice et al., 2008;Mashburn et al., 2006;Rudasill & Rimm-Kaufman, 2009;Whitaker et al., 2015).However, little is known about associations between child personality and teacher-child relationships in preschool, although theoretical frameworks underlie the interplay of personality and social relationships in different contexts (e.g.Back et al., 2011).The current study focuses on the personality of preschool children in the form of the frequently studied personality types 'resilient', 'overcontrolled', and 'undercontrolled' and examines associations with teacher-child closeness and teacher-child conflict in Austrian preschools.The following research question will be addressed: Do the child personality types 'resilient', 'overcontrolled', and 'undercontrolled' predict teacher-child closeness and teacher-child conflict?Various other predictors (e.g.child gender, teachers' self-efficacy) considered important are also taken into account as control variables.

The Austrian preschool system
For a better understanding of the specificity of the Austrian preschool system, a brief overview is given below.In Austrian preschools, the attendance rates for three-to five-year-old children are around 93% (Statistik Austria, 2021).Attendance of the final preschool year has been obligatory since 2009 (Smidt, 2018).In contrast to many other countries, the normal vocational qualification of preschool teachers is non-academic (5-year training or a shortened 2-year training for students with university entrance qualification) (Smidt, 2018).However, in recent years, some academic courses for early childhood education (advanced trainings) have been introduced (Holzinger & Reicher-Pirchegger, 2020).In 2009, an educational plan was implemented with specific guiding principles (e.g.diversity, empowerment) as well as educational domains (e.g.emotions and social relationships) (Charlotte-Bühler-Institut, 2009).There is a strong separation between preschool education and elementary school education (Smidt, 2018).

Participants
Participants for this study were derived from the first wave of the longitudinal project 'Quality of Children's Interactions in Preschool', which was funded by the Austrian Science Fund (FWF).Data was collected from April to June 2019 in Tyrol, a federal state of Austria.In the present study, a subsample of 163 three-and four-year-old children (84 girls) from 65 preschool classes (from 65 randomly selected preschools) was used.Up to four children from each preschool class (M = 2.51, SD = 0.99) participated in the study.These number allowed the use of different instruments for the assessment and observation of children's interactions and relationships (not all of these instruments are part of the current study, see e.g.Smidt & Embacher, 2023).The children had a mean age of 49.62 months (SD = 4.31, ranging from 36.76-57.63months) and were at the end of their first preschool year.A total of 14.72% of the children had an immigration background (and a family language other than German).Overall, 65 preschool teachers (all female) participated in the study and completed a paper-pencil questionnaire.Furthermore, the parents of the children participated through paper-pencil questionnaires or telephone interviews.Of the parents who participated, 88.16% were mothers and 11.18% were fathers; in one case (0.66%), both mother and father took part in the survey.Half of the participating parents (50.31%) had a university entrance qualification.Before data collection began, preschool teachers and parents were informed about the study, and written parental informed consent was acquired for all children.In terms of ethical considerations, participation was completely voluntary.Furthermore, the anonymity of all participants was assured throughout data processing.

Teacher-child relationships
Teachers' perception of their relationship with a particular child was measured through preschool teachers' self-report using a short form of the 'Student-Teacher Relationship Scale' (STRS; Pianta, 2001).The STRS is a frequently used instrument to assess teacher-child relationships in preschools and elementary schools (see Nurmi, 2012).The short form consists of 15 items rated on a fivepoint scale ranging from 1 ('definitely does not apply') to 5 ('definitely applies') (Pianta, 1992).A German version of this scale was used in the NUBBEK study (Tietze et al., 2015).The 15 items are assigned to the subscales 'closeness' (eight items; e.g.'I share an affectionate, warm relationship with this child.','This child openly shares his/her feelings and experience with me'; α = .74)and 'conflict' (seven items, e.g.'This child and I always seem to be struggling with each other', 'This child easily becomes angry at me'; α = .80).Based on the manual, a sum of the individual items was calculated for each subscale (Pianta, 2001).For descriptive results on teacher-child closeness and conflict, see Table 1.

Control variables
Because previous studies found associations between various child characteristics as well as teacher characteristics and teacher-child relationships (e.g.Ewing & Taylor, 2009;Hamre et al., 2008;Jerome et al., 2009;Justice et al., 2008;Mashburn et al., 2006;Rudasill & Rimm-Kaufman, 2009;Whitaker et al., 2015; see chapter 'Individual characteristics as predictors of teacher-child relationships'), children's language skills, gender, and immigration background, parents' educational level, and preschool teachers' self-efficacy, overcommitment, and work experience were included as control variables (see Table 1 for descriptive statistics).Language skills of the children were assessed with the 'Sprachentwicklungstest für drei-bis fünfjährige Kinder [language development test for three-to five-year-old children]' (SETK 3-5;Grimm, 2015).For this purpose, a mean score (T-value) of three subtests (understanding sentences, morphological rule formation, phonological working memory; α = .73)was computed.Children's gender, immigration background (family language other than German) and parents' educational level were captured through telephone interviews or paper-pencil interviews with parents.Regarding parents' educational level, a distinction was made between low (0 = no university entrance qualification) and high (1 = university entrance qualification).Preschool teachers' selfefficacy, overcommitment, and work experience (in years) were assessed through teacher questionnaires.Self-efficacy was captured with the 'Kurzskala zur Erfassung allgemeiner Selbstwirksamkeitserwartungen [short scale for measuring general self-efficacy beliefs]' (ASKU; Beierlein, Kemper, Kovaleva, & Rammstedt, 2013).This short scale for measuring general self-efficacy beliefs consists of three items (e.g.'I can rely on my own abilities in difficult situations.', α = .84)and is rated on a five-point scale from 1 ('strongly disagree') to 5 ('strongly agree') (Beierlein et al., 2013).Preschool teachers' personal style of coping with work demands was captured with the scale on 'overcommitment' of the 'Effort-Reward Imbalance (ERI) Questionnaire' (Rödel, Siegrist, Hessel, & Brähler, 2004;Siegrist et al., 2004).This scale consists of six items (e.g.'I get easily overwhelmed by time pressures at work.', α = .68)rated on a four-point scale from 1 ('strongly disagree') to 4 ('strongly agree') (Siegrist et al., 2004).Higher sum scores indicate a higher overcommitment.It is expected that persons who exhibit a specific pattern of coping with work demands characterized by an excessive workrelated commitment and a high need for approval ('overcommitment') have an increased susceptibility to stress responses (Siegrist et al., 2004;Siegrist & Li, 2016).

Data analyses
Associations between teacher-child relationships and child personality types were analysed using hierarchical regression analyses, with teacher-child closeness and teacher-child conflict as dependent variables.In a first step, the child personality types 'overcontrolled' and 'undercontrolled' ('resilient' was used as the reference category) served as independent variables.In a second step, children's language skills, gender, and immigration background, parents' educational level, and preschool teachers' self-efficacy, overcommitment, and work experience were added as control variables.As preschool children were clustered in preschool classes, robust standard errors were calculated in Stata (Williams, 2000).To determine effect sizes, correlation coefficients (r) were imputed using information about corresponding beta coefficients (β) (Lenhard & Lenhard, 2016;Peterson & Brown, 2005).According to Cohen (1988), thresholds are r = .10(small effect), r = .30(medium effect), and r = .50(large effect).

Results
The results of the regression analyses show that a classification as 'overcontroller' (β = -.20,transformed into r = -.25) or 'undercontroller' (r = -.40) was negatively related to teacher-child closeness (see Table 2).These small to medium effects indicate that preschool teachers perceived less closeness to 'overcontrollers' and 'undercontrollers' compared to 'resilients'.The small to medium effects of 'overcontrollers' (r = -.21) and 'undercontrollers' (r = -.33)remained after inclusion of the control variables.Furthermore, the gender of the children (r = -.20)tended to be negatively related to teacher-child closeness, with a small effect size.This means that preschool teachers tended to rate teacher-child closeness lower for boys than for girls.Immigration background (r = -.31) was also negatively related to teacher-child closeness.This medium effect indicates that preschool teachers experienced less teacher-child closeness among children with an immigration background compared to children with no immigration background.Teachers' self-efficacy (r = .21)was positively related to teacher-child closeness, with a small effect size.This indicates that a higher selfefficacy of preschool teachers was associated with a greater closeness between teachers and children.
Regarding the prediction of teacher-child conflict, a classification as 'undercontroller' (r = .33)was positively related to teacher-child conflict.This medium effect indicates that preschool teachers perceived the teacher-child relationship to be more conflictual with 'undercontrollers' than with 'resilients'.However, after inclusion of the control variables, the positive relation between teacher-child conflict and 'undercontrollers' (r = .24)only tended to be significant, with a small effect size.Furthermore, a small positive effect was found between overcommitment of the preschool teachers (r = .26)and teacher-child conflict.This means that a higher overcommitment of the preschool teachers was associated with higher levels of teacher-child conflict.When 'overcontrollers' were used as a reference category (without table), a medium positive effect of 'undercontrollers' (β = .37,r = .42,p < .001)regarding teacher-child conflict was found.This medium effect of 'undercontrollers' (β = .29,r = .34,p = .007)remained after inclusion of the control variables and indicates that preschool teachers perceived a more conflictual relationship with 'undercontrollers' than with 'overcontrollers'.
All regression models were significant (p < .001)and the total amount of explained variance (adjusted R 2 ) ranged between 9% and 19% (see Table 2).

Discussion
The present study aimed to investigate associations between child personality and teacher-child relationships in preschool classes in Tyrol, Austria.After characteristics of the children and the preschool teachers were accounted for, the findings revealed that teacher-child closeness with 'overcontrollers' and 'undercontrollers' was lower compared to 'resilients'.Moreover, preschool teachers perceived the teacher-child relationship as more conflictual with 'undercontrollers' compared to 'resilients' and 'overcontrollers'.The effect sizes were small to medium.

Child personality and teacher-child relationships in preschool
Teacher-child relationships have been considered important for children's outcomes and functioning (e.g.Sabol & Pianta, 2012).With regard to closeness, the findings of the present study show that child personality types matter for teacher-child relationships in preschool.After several variables were controlled for, and with small to medium effect sizes, preschool teachers perceived less closeness to children described as 'overcontrollers' and 'undercontrollers' compared to children described as 'resilients'.This pattern of results underpins theoretical frameworks postulating relations between personality traits and relationships with social partners (Back et al., 2011;McCrae & Costa, 1996) and indicates that these relations can also be applied to teacher-child relationships in preschool.With regard to 'overcontrollers', the findings are rather in line with previous research focusing on children's temperament, where greater shyness corresponded with less close relationships with pedagogues (e.g.Justice et al., 2008;Nurmi, 2012;Rudasill & Rimm-Kaufman, 2009).This fits with research where 'overcontrollers' were rated as shyer than 'resilients' and 'undercontrollers' (Asendorpf & van Aken, 1999;Hart et al., 2003).The findings of the present study can raise awareness that a combination of relatively pronounced values in certain personality traitsparticularly low Extraversion (e.g. to be less sociable and outgoing) and high Neuroticism (e.g. to be anxious and shy)may serve as challenges for the development of close relationships between preschool teachers and 'overcontrollers'.A corresponding explanation connects to research illustrating that 'overcontrollers' more frequently have internalizing problems (Asendorpf & van Aken, 1999;Hart et al., 2003), which may make it more difficult to develop close relationships with preschool teachers.
Results of a meta-analysis, where internal problem behaviour corresponded with less close teacher-child relationships (Nurmi, 2012), and more recent findings that children's internalizing problems corresponded with a lower quality of teacher-child relationship (Zatto & Hoglund, 2019), also point in this direction.
In addition, preschool teachers perceived less close relationships with 'undercontrollers' compared to children labelled as 'resilients'.This finding somewhat fits with research indicating that children with lower effortful control (e.g.low inhibitory control, low concentration skills) had more conflictual relationships with their preschool teacher (Rudasill & Rimm-Kaufman, 2009).Because 'undercontrollers' tend to have partly comparable traits (e.g.low Conscientiousness, which is considered to be related to effortful control; Shiner & DeYoung, 2013), similar associations to conflictual relationships are probably to be expected.When asking for the underlying mechanisms for this relation, previous research indicating that 'undercontrollers' are more prone to show externalizing behaviour problems than 'resilients' and 'overcontrollers' (Hart et al., 2003;Robins et al., 1996) can provide some support for the assumption that those behaviour problems may hinder the development of close relationships between 'undercontrollers' and preschool teachers.Research provides hints for this assumption (Nurmi, 2012).
The findings of the present study also show that preschool teachers perceived the teacher-child relationship with children identified as 'undercontrollers' as somewhat more conflicted than that with 'resilients' andmore pronouncedmore conflicted than that with 'overcontrollers'.The effect sizes were small to medium.These findings fit with previous research showing that children's externalizing behaviour problems corresponded with more conflictual relationships with preschool teachers (Jerome et al., 2009;Mejia & Hoglund, 2016).Thus, it appears that children's problem behaviour becomes evident in more conflictual relationships with preschool teachers.

Control variables
With regard to the control variables, findings are mixed.Only some of those considered were substantially related to teacher-child relationships and effect sizes were mostly small.Consistent with previous research (e.g.Choi & Dobbs-Oates, 2016;Tsigilis et al., 2018), teacher-child closeness was slightly lower for boys than for girls.However, unlike in other studies (e.g.Jerome et al., 2009;Silva et al., 2011), conflictual teacher-child relationships were not higher for boys in the present study.Further, immigration background (family language other than German) corresponded with lower teacher-child closeness.This finding somewhat fits with research showing that children with better language comprehension (Justice et al., 2008) revealed a closer teacher-child relationship as well as with findings in which closer teacher-child relationships were reported when children and their teachers were both native speakers (Sullivan et al., 2015).However, we did not find any influence of immigration background on conflictual teacher-child relationships as previousand rather inconsistentresearch might suggest (e.g.Justice et al., 2008;Sullivan et al., 2015).The result that preschool teachers' self-efficacy is slightly positively related to teacher-child closeness corresponds with previous research (Mashburn et al., 2006).Finally, a higher overcommitment of the preschool teacher was associated with higher levels of teacher-child conflict, with a small effect size.Considering that overcommitment relates to increased susceptibility to stress responses (Siegrist et al., 2004;Siegrist & Li, 2016), this finding is rather consistent with findings showing that experiences of stress correspond with more teacher-child conflict (Whitaker et al., 2015).

Study limitations and implications for research and practice
The study has some limitations.First, although scholarly work describes many potential predictors of the teacher-child relationship (Pianta et al., 2003), we used only individual characteristics of children and preschool teachers as predictors and captured the teacher-child relationship only in the form of conflict and closeness.The dimension 'dependency' was not included as it was not measured by the short form of the STRS used in the present study.Thus, more predictors (e.g.external influences), respectively, a broader understanding of the teacher-child relationship as suggested by Pianta et al. (2003), should be considered in future research.Second, the well-tested and frequently used STRS captures teacher-child relationships based on preschool teachers' self-reports; a child's perspective on relationships (see e.g.Gregoriadis, Grammatikopoulos, Tsigilis, & Verschueren, 2021) has not been considered.Third, some research suggests that additional information about the prediction of teacher-child relationships could be added by considering interaction effects.For instance, in the present study, children's language skills (used as a control variable) did not predict teacherchild relationships but effects would probably appear in interaction with personality (Rudasill et al., 2006).Future analyses should account for this.Fourth, the present study focuses on children's personality types based on the assessments of the preschool teachers.Although teacher assessments of children's personalities can be considered valid (Asendorpf, 2007;Müller et al., 2016), possible bias in the assessment of the preschool teacher may not be excluded.Fifth, future studies on the prediction of teacher-child relationships should also consider the personality types of the preschool teachers because it can be assumed that the personalities of all actors involved may play a role (Asendorpf, 2002).
Based on the fact that high-quality teacher-child relationships are considered important for child development (Ahnert & Harwardt, 2008;O'Connor & McCartney, 2007;Sabol & Pianta, 2012), the findings of the present study underpin that children identified as 'undercontrollers' and 'overcontrollers' require special attention from pedagogues in terms of building good relationships.This is significant, as the notion of the awareness of individual differences between children is in line with best-practice standards on developmentally appropriate practices (Copple & Bredecamp, 2009).Regarding the improvement of educational practices, recent research in Austrian preschools suggests, for example, that social-constructivist approaches such as scaffolding (Acar, Hong, & Wu, 2017;Bodrova & Leong, 2018) could support interactions between 'undercontrollers' as well as 'overcontrollers' and their preschool teachers and peers (Smidt & Embacher, 2023).The interaction processes are a central element of teacher-child relationships (Pianta et al., 2003), on which improvement efforts can concretely be basedin contrast to external influences or underlying representations or personality traits that would require a longer societal or deeper personal change.
Further, focusing more attention on the importance of personality in teacher-child relationships in the education, training and continuing education of preschool teachers may help preschool teachers reflect on their educational practice including their relationships with children.Another way to support preschool teachers in building good relationships with all children could focus on the reduction of work-related stress, as this kind of stress can influence both closeness and conflict relationships (see Whitaker et al., 2015).In this respect, interventions on the level of the organization (e.g.sufficient staff) and the person (e.g.stress management training) (Bamberg, Busch, Ducki, & Glaveris, 2003; see also Embacher, in press, for an overview), for instance, seem to be promising.In summary, considering children's personality typesalong with various other characteristics -

Table 1 .
Descriptive results and intercorrelations of the considered variables.

Table 2 .
Prediction of teacher-child relationships through child personality types.