Olweus (Reference Olweus1994) describes bullying as ‘A student is being bullied or victimized when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other students’ (p. 1173). Therefore, repetition, the intent to harm others and power imbalance are regarded as the three pillars of bullying. Although ample research has been conducted with students at primary school (e.g., Mundy et al., Reference Mundy, Canterford, Kosola, Degenhardt, Allen and Patton2017), secondary school (e.g., Cross et al., Reference Cross, Runions, Resnicow, Britt and Gray2018) and high school (e.g., Topcu & Erdur-Baker, Reference Topcu and Erdur-Baker2018), research on preschool peer bullying is quite limited.
One of the reasons for the dearth of knowledge on preschool bullying may be the developmental characteristics of the early childhood period. Preschool children tend to have difficulties with verbally expressing themselves, and they also may not be able to read or write. An additional reason could be the inability of the preschool child to comprehensively understand their personal experiences and attribute meanings to their own mental states as well as others. Hughes and Leekham (Reference Hughes and Leekam2004) note that the preschool period is the time when the abilities of theory of mind (Premack & Woodruff, Reference Premack and Woodruff1978) exist in a more sophisticated way compared to infancy and toddlerhood. Preschool students can understand the main mental states such as emotions, empathy and beliefs, but this understanding is still limited during these years. Due to this fact, they may not be able to accurately name the negative social interactions they may be involved in. With regard to bullying experiences, preschool students may not be able to consider how bullying perpetration and bullying victimisation are harmful personal experiences. Another reason can be the lack of the psychometrically sound instruments that aim to unravel the bullying involvement of the preschool child. This study therefore aims to fill this gap with a valid and reliable measurement instrument tool to assess preschool peer bullying, namely the Preschool Peer Bullying Scale-Teacher Form (PPBS-Teacher Form).
Preschool teachers can be significant stakeholders to provide data about peer bullying in early childhood education (Crick et al., Reference Crick, Casas and Ku1999). This suggestion is supported by the previous research conducted with preschool teachers. For example, Yalçıntaş-Sezgin (Reference Yalçıntaş-Sezgin2018) reported that preschool teachers were good at identifying the content of bullying episodes in their classroom. Moreover, Perren and Alsaker (Reference Perren and Alsaker2006) underline the fact that preschool teachers have essential roles in detecting, preventing and intervening in peer bullying incidents. In addition, Kirves and Sajaniemi (Reference Kirves and Sajaniemi2012) found that teacher reports about the bullying behaviours in their classroom were in parallel with the reports of the preschool students. Therefore, teacher reports of bullying behaviours in their classroom can be regarded as important sources of data to investigate early childhood bullying. This current study focused on teachers’ evaluations of preschool bullying.
Before conducting this research, the researchers of this study did an extensive search on the extant literature on preschool bullying. They realised that a need for a psychometrically validated, quantitative instrument was evident to extend the available knowledge on preschool bullying. To fill this void, the PPBS-Teacher Form was developed by considering labelling theory (Becker, Reference Becker1963), and its validity and reliability were tested in this current research. By means of the PPBS-Teacher Form, teachers will be able to validly and reliably assess peer bullying engagement at preschool level without individually labelling children as bullies or victims.
In line with the existing literature (Helgeland & Lund, Reference Helgeland and Lund2017; Kirves & Sajaniemi, Reference Kirves and Sajaniemi2012), the terms ‘preschool peer bullying’ and ‘early childhood peer bullying’, as well as ‘preschool children’ and ‘preschool students’, were alternately used in this present investigation. There is a worldwide controversy regarding the age range of the preschool children (Aktan & Akkutay, Reference Aktan and Akkutay2014). Previous studies presented in the introduction section of this research involved children ranging in age between 36 months to 72 months. In line with this information, this present investigation did not involve teachers working with preschool children aged below 36 months.
Types of Preschool Peer Bullying
According to Tanrikulu (Reference Tanrikulu2018), physical, verbal and relational bullying are the three types of traditional bullying that exist in the early childhood classroom. While physical bullying involves hitting, biting, kicking, spitting, pushing, pinching, slapping, hair pulling or throwing an object (Özdemir & Tepeli, Reference Özdemir and Tepeli2015; Rose et al., Reference Rose, Richman, Fettig, Hayner, Slavin and Preast2016), verbal bullying includes threatening, ridiculing, using bad language, revealing secrets and making gossip (Gültekin-Akduman, Reference Gültekin-Akduman2012). Relational bullying consists of ignoring, distancing a friend, deterring a friend from playing with a peer, spoiling a game and damaging friendships (Jansen et al., Reference Jansen, Verlinden, Dommisse-van Berkel, Mieloo, van der Ende, Veenstra, Verhulst, Jansen and Tiemeier2012). Reviewing the past literature, Tanrikulu (Reference Tanrikulu2018) found that the prevalence of early childhood peer bullying experiences (i.e., bullies, victims, as well as bully-victims who not only bully others but also concurrently become victimised) ranged from 4% to 25%. In short, physical, verbal or relational bullying exist in the preschool classroom and with different prevalence rates.
Associations Between Preschool Peer Bullying and Behavioural, Psychological and Social Problems
Convincing evidence suggests associations between early childhood bullying victimisation and behavioural, psychological and social problems. According to Crick et al. (Reference Crick, Werner, Casas, O’Brien, Nelson, Grotpeter and Markon1999), preschool peer bullying victimisation (participants ranged in age between 37 months and 66 months) was significantly linked to problematic social relationships, rejection among peers, poor prosocial behaviours, internalising behaviour problems, hyperactivity and distraction problems. Ostrov et al. (Reference Ostrov, Woods, Jansen, Casas and Crick2004) found that being a victim of preschool bullying was positively related to psycho-social adjustment problems, including peer rejection and exclusion and lack of prosocial behaviours, among children between 44 months and 66 months. Perren and Alsaker (Reference Perren and Alsaker2006) reported that victims of early childhood bullying (children ranging in age from 60 months to 95 months) displayed submissive social behaviours, had poor leadership skills and often did not have playmates, in addition to being alone, introverted, less sociable and less collaborative with other children. All in all, the research findings reported above highlight the fact that peer bullying victimisation has negative impacts on preschool children.
Some longitudinal studies have also addressed the problems associated with peer preschool bullying victimisation. Children who are victims of early childhood peer bullying suffer from loneliness, school adaptation problems and low school success (Kochenderfer & Ladd, Reference Kochenderfer and Ladd1996), as well as crime involvement in early adolescence and late adolescence (Sourander et al., Reference Sourander, Jensen, Rönning, Elonheimo, Niemelä, Helenius, Kumpulainen, Piha, Tamminen, Moilanen and Almqvist2007). In short, convincing evidence has been accumulated in support of the negative impacts of early childhood bullying victimisation, which can continue into adolescence. Hence, preschool peer bullying is a detrimental concern, and psychometrically validated assessment tools are required to create effective prevention and intervention strategies against early childhood peer bullying. With the help of psychometrically validated assessment tools, the nature and the extent of preschool bullying could be better understood. Thus, this current research aims to develop and report the psychometric qualities of the PPBS-Teacher Form.
Assessment of Preschool Peer Bullying
For a complete understanding about the assessment of preschool bullying, published research articles were reviewed. There were two criterion for the reviewing process. The articles had to present qualitative or quantitative data regarding preschool peer bullying, and they had to be published in English or have an abstract in English. The obtained manuscripts were examined and grouped into three categories as (1) qualitative assessments, (2) quantitative assessments and (3) mixed design assessments.
Qualitative assessments
Studies with qualitative design can be classified into three groups. The first group researched how preschool children whose ages ranged between 36 months and 60 months conceptualised peer bullying (Gillies-Rezo & Bosacki; Reference Gillies-Rezo and Bosacki2003; Helgeland & Lund, Reference Helgeland and Lund2017; Özözen-Danacı & Çetin, Reference Özözen-Danacı and Çetin2016; Swit, Reference Swit2018). The second group investigated teachers’ conceptualizations of preschool peer bullying as well as their classroom strategies to prevent bullying (Tepetaş et al., Reference Tepetaş, Akgun and Akbaba-Altun2010; Yalçıntaş-Sezgin, Reference Yalçıntaş-Sezgin2018). The last group explored teachers’ and students’ understanding of the nature of peer bullying in preschool as well as teachers’ classroom prevention and intervention strategies against early childhood bullying (Karachaliou & Theodotou, Reference Karachaliou and Theodotou2012).
Reliability and validity problems in qualitative assessments
Bullying studies in preschool-level children with a qualitative research design have limitations that were pointed out by Mays and Pope (Reference Mays and Pope1995). Studies with a qualitative research methodology lack standardisation, although they share similar goals to unravel early childhood bullying. Moreover, studies with a qualitative research design have been conducted with a small number of participants, which limits the external validity of their findings. Also, these studies were highly prone to peers as well as teachers labelling students, and labelling students as bullies or victims may harm the classroom climate by damaging the peer relations of the early childhood students.
Furthermore, studies with a qualitative research design are prone to some internal validity issues, as reported by Fraenkel et al. (Reference Fraenkel, Wallen and Hyun2015). For example, the characteristics of the limited number of children and teachers as research participants in a qualitative research design may distort the findings, and this may result in personally biased findings concerning the peer bullying experiences of the children in early childhood education. In addition, findings of the studies with a qualitative research design may be negatively affected by data collector bias since researchers themselves carried out face-to-face interviews with children and teachers.
History can also be a threat for the findings because teachers may have had some negative experiences with the students or the parents that may hinder them from objectively reporting peer bullying instances in their classrooms. Lastly, that some students and teachers are interviewed by researchers may lead to a Hawthorne effect, which arises when research participants feel that someone attends and recognises their needs. In the case of a Hawthorne effect, research participants in qualitative research studies may underestimate or overestimate the bullying incidents in their classrooms since they are under the attention of the researcher. Such an effect may bias the findings of studies with a qualitative research design.
Quantitative assessments
Quantitative data regarding the peer bullying experiences of preschool students have been mainly obtained from teachers. Repo and Sajaniemi (Reference Repo and Sajaniemi2015) examined teachers’ institutional and pedagogical practices of prevention of preschool bullying, whereas Yörük (Reference Yörük2016) investigated the nature and the extent of preschool bullying from the perspective of the teachers. Jansen et al. (Reference Jansen, Zwirs, Verlinden, Mieloo, Jaddoe, Hofman and Tiemeier2017) asked teachers to assess the bullying involvement of their students and explored the predictors of peer bullying perpetration and victimisation in a preschool classroom. On the other hand, Cameron and Kovac (Reference Cameron and Kovac2016) included the reports of parents as well as teachers and inquired about their views on the nature and extent of early childhood bullying. Camodeca et al. (Reference Camodeca, Caravita and Coppola2015) considered the reports of teachers and students and unravelled the links between preschool peer bullying involvement and social competence. Moreover, Sourander et al. (Reference Sourander, Jensen, Rönning, Elonheimo, Niemelä, Helenius, Kumpulainen, Piha, Tamminen, Moilanen and Almqvist2007) based their study on the reports of teachers and students as well as parents by investigating the longitudinal linkages between preschool bullying and later criminal involvement. Another research study acquired data from the parents about their children’s bullying involvement and examined the prevalence of preschool bullying and its correlations with psycho-social factors (Ilola et al., Reference Ilola, Lempinen, Huttunen, Ristkari and Sourander2016).
Reliability and validity problems in quantitative assessments
The first group of quantitative studies on preschool bullying implemented a one-item strategy to assess preschool bullying, and it did not report any evidence concerning the validity and the reliability of this strategy (Cameron & Kovac, Reference Cameron and Kovac2016; Ilola et al., Reference Ilola, Lempinen, Huttunen, Ristkari and Sourander2016). The second group of researchers measured preschool bullying by creating items based on the extant peer bullying literature, but they did not provide any information about the psychometric validity and reliability of their self-compiled questionnaires (Repo & Sajaniemi, Reference Repo and Sajaniemi2015; Sourander et al., Reference Sourander, Jensen, Rönning, Elonheimo, Niemelä, Helenius, Kumpulainen, Piha, Tamminen, Moilanen and Almqvist2007). The third group made use of the already existing assessment instruments. Camodeca et al. (Reference Camodeca, Caravita and Coppola2015) utilised the bullying involvement section of the Participant Roles Questionnaire (Belacchi & Farina, Reference Belacchi and Farina2010), and provided internal reliability as well as the results of a confirmatory factor analysis. Using the items of Perren and Alsaker (Reference Perren and Alsaker2006), Jansen et al. (Reference Jansen, Zwirs, Verlinden, Mieloo, Jaddoe, Hofman and Tiemeier2017) only reported internal reliability. Yörük’s (Reference Yörük2016) study was the only one in the last group with an independent instrument to examine peer bullying involvement of the preschool children. Yörük developed The Peer Bullying and Victimization Determination Scale (Preschool Teacher Form) and reported the internal reliability and the results of the exploratory and the confirmatory factor analysis, and there was a total of 69 items in his scale.
Another problem with using quantitative studies to measure preschool peer bullying involvement is labelling. Quantitative studies asked teachers, parents and students to report a specific child’s bullying involvement and to label the child as a bully or a victim. Such an assessment strategy can be risky in preventing peer bullying in preschool because teachers, parents and students may develop an attitude towards the labelled child and behave according to the findings revealed by the quantitative assessment tool. As labelling theory (Becker, Reference Becker1963) proposes asking who the bullies or the victims are, this may have an undesirable impact on the attitudes of teachers, parents and students, and following such assessments, teachers, parents and students may begin to feel pity for the labelled victims and feel angry with the labelled bullies.
Quantitative research also bears some internal validity risks, as reported by Fraenkel et al. (Reference Fraenkel, Wallen and Hyun2015). As an example, unfavourable circumstances may have happened between teachers and students, between teachers and parents of the students, or among the students in a certain classroom. These previously experienced problems may have a negative consequence on the evaluations of the teachers and the students (history). Furthermore, asking parents to assess their own child’s bullying or victimisation experiences may negatively impact the attitude of the parents. Being questioned about their own child’s bullying involvement may make the parents feel uncomfortable and distort their assessment (attitude of subjects). In addition, quantitative studies demand that the teachers evaluate all the students individually in their class in terms of bullying engagement. Such an evaluation procedure may make the teachers tired, and this may discourage them from making accurate assessments (instrument decay). A similar problem may occur for the data collectors as well. For instance, data collectors in Camodeca et al.’s (Reference Camodeca, Caravita and Coppola2015) investigation read the items of a peer nominations questionnaire to the 320 preschool students. This may make the data collectors feel exhausted, which may result in inaccurate assessments.
Mixed-design assessments
Considering the two criteria mentioned above, the researchers of this present study found three studies on preschool bullying using a mixed research method. The first was Perren and Alsaker’s (Reference Perren and Alsaker2006) investigation examining the bullying involvement of 344 preschool students aging from 60 to 95 months. In the quantitative part of the study, teachers evaluated the peer bullying perpetration and victimisation experiences of each student in their class using an eight-item measurement tool. In the qualitative part of the study, face-to-face interviews were conducted with each student, and students were shown photographs containing peer bullying behaviours and asked to classify their peers as bullies or victims. The results of Perren and Alsaker’s research revealed distinct behavioural patterns between victims, bullies and bully-victims. While victims experienced socialisation problems, bullies were aggressive and demonstrated leadership skills, and bully-victims were both lonely and less collaborative.
The second study was Uysal’s (Reference Uysal2011) research working with children aged between 60 and 72 months. In the quantitative part, the Storied Hypothetical Situations Form was applied face to face to 121 preschool children, and they were asked which of their friends exhibited physical, verbal and/or relational bullying. In the quantitative part of Uysal’s study, teachers were the participants. Teachers rated each of their student in their classrooms by considering physical aggression, relational aggression and positive social behaviour subdimensions of the Preschool Social Behavior Scale-Teacher Form. In the qualitative part of the study, 14 students were observed in their classrooms. Uysal’s research findings indicated that bullying occurred in free-time activities, verbal bullying were the most common bullying type, bullies and bully victims frequently hit others, and the findings obtained from the observations were mostly in parallel with teacher reports.
The last study was from Kirves and Sajaniemi (Reference Kirves and Sajaniemi2012). In the quantitative part, they collected data regarding the prevalence of peer bullying behaviours in the classrooms of 770 preschool teachers through a self-developed measurement tool. Teachers evaluated the peer bullying engagement of each student in their class one by one. In the qualitative part, face-to-face interviews were conducted with 61 preschool children (36–72 months), 24 parents and 29 preschool teachers, and the participants were asked about the nature of early childhood peer bullying. Kirves and Sajaniemi found that peer bullying was an everyday topic for the preschool children, more than 10% of the preschool children were engaged in bullying incidents, and relational peer exclusion was the most common type of bullying in early childhood level.
Reliability and validity problems in mixed-design assessments
The quantitative sections of these studies with a mixed-research method were not standardised and thus vulnerable to subjective and biased findings. Although Perren and Alsaker (Reference Perren and Alsaker2006) reported the internal reliability value, they did not provide any information about other psychometric properties of their quantitative instrument. Similarly, Uysal (Reference Uysal2011) reported the internal reliability value, but did not give any information about the other psychometric properties of the Preschool Social Behavior Scale-Teacher Form. Uysal also did not report any of the psychometric features of the self-developed Storied Hypothetical Situations Form. Likewise, Kirves and Sajaniemi (Reference Kirves and Sajaniemi2012) did not report the psychometric properties of the measurement tool in their study.
In addition, preschool teachers assessed the bullying involvement of all 344 students in Perren and Alsaker’s (Reference Perren and Alsaker2006) study, and of 121 students in Uysal’s (Reference Uysal2011) research, and of 6910 students in Kirves and Sajaniemi’s (Reference Kirves and Sajaniemi2012) investigation. Such a data collection strategy may create problems that impact the internal validity of the findings (Fraenkel et al., Reference Fraenkel, Wallen and Hyun2015). For instance, while teachers may be very energetic at the beginning of the assessment procedure, they may become tired towards the end. This may adversely affect the validity of assessments (instrument decay). Also, a teacher may have experienced negative situations with a student or with a student’s parent before the measurement. This may lead the teacher to become biased towards the student and label the student based on the previously experienced negative situations (history).
Considering the qualitative sections of the studies with mixed-research design, there may be some internal validity threats that have been explained by Fraenkel et al. (Reference Fraenkel, Wallen and Hyun2015). For example, if some interpersonal problems between the teachers and the parents or among students formerly happened, the assessment would be biased. Moreover, conducting daily qualitative interviews with many preschool students may make the researchers feel fatigued after a while, which may distort the data. In addition, the face-to-face interviews with preschool students in Perren and Alsaker’s (Reference Perren and Alsaker2006) research may create a Hawthorne effect. Due to the photographs containing bullying behaviours, the students may develop a negative attitude, which may affect the findings. Furthermore, the researcher had to share the same class with the students to make observations in Uysal’s (Reference Uysal2011) study, raising the risk of data collector bias, which may adversely affect the results. In addition, an adult continuously taking notes during the observations may create a Hawthorne effect, which may negatively affect the validity of the data.
Theoretical Background
The theoretical background of this research rests upon labelling theory (Becker, Reference Becker1963). It posits that social groups anticipate certain emotional, behavioural, cognitive and interpersonal characteristics for the individuals bearing a label, who in turn conforms to the characteristics associated with their label. After some time, individuals begin to reinforce their label by acting in accordance with their labelled characteristics. Bullying involvement can be conceptualised from the standpoint of labelling theory (Becker, Reference Becker1963). As significant individuals in children’s primary social groups, peers, parents and teachers may label schoolchildren as bullies or victims, and such labelling may entail certain emotional, cognitive and interpersonal traits applied to the labelled children. After some time, students labelled as bullies or victims may act in accordance with their labelled roles. Labelling students as bullies or victims implies that students exhibit behavioural patterns that make them bullies or victims, and these patterns can become continuous over time and may be difficult to change.
In educational settings, labelling may be useful to support students with special educational needs (Henley et al., Reference Henley, Ramsey and Algozzine2010). For example, the use of ‘students with autism’ as a label can help professionals to develop and implement teaching strategies or assessment methods based on the specific needs of the students with autism. On the other hand, labelling schoolchildren may pose some risks. Henley et al. (Reference Henley, Ramsey and Algozzine2010) argue that (a) students may be discriminated based on their labels; (b) labelling may prevent teachers from understanding the causes of the problem behaviours of the students; (c) teachers labelling their students may exaggerate troublesome student behaviours that can be handled, and feel helpless to correct behaviour; (d) labelling may lead teachers to assume that parents are the only responsible people to help students; (e) labelling may impair classroom activities offered by teachers and cause labelled children to receive ineffective schooling; and (f) labels may negatively impact teacher expectations regarding students’ academic success, behaviours and interpersonal interactions. In addition, confronted with teachers’ labels, students may feel desperate and behave in accordance with the teacher’s labelling to comply with the required characteristics of the label. Supporting these arguments, Yeager et al. (Reference Yeager, Johnson, Spitzer, Trzesniewski, Powers and Dweck2014) found that labelling can cause youngsters to suffer stress, health problems and low school success, and noted that labelling could have longitudinal unfavourable impacts on young children.
Evidence suggests that labelling children as bullies or victims can create serious problems, including children normalising peer bullying incidents, or dissociating themselves from the victim, or transferring the responsibility for stopping a bullying incident to others (Thornberg, Reference Thornberg2010). In addition, Graham and Juvonen (Reference Graham and Juvonen1998) found that while students who viewed themselves as a victim were at risk of loneliness, social anxiety and low self-worth, students who were labelled as a victim were at risk of rejection by peers. Drawing from the arguments of Henley et al. (Reference Henley, Ramsey and Algozzine2010), students labelled as a victim by their teachers in the classroom may face discrimination or experience increasing levels of victimisation, or may suffer from increasing levels of academic, behavioural and interpersonal problems because of the expectations of the teacher towards the victimised child.
The available literature on early childhood peer bullying necessitates peers (e.g., Uysal, Reference Uysal2011), teachers (e.g., Yalçıntaş-Sezgin, Reference Yalçıntaş-Sezgin2018) and parents (e.g., Ilola et al., Reference Ilola, Lempinen, Huttunen, Ristkari and Sourander2016) individually evaluating a preschool student’s bully or victim role experiences. Yet, this may inevitably create the risk of labelling. Considering that children bearing labels as bullies or victims experience unfavourable consequences (Graham & Juvonen, Reference Graham and Juvonen1998; Thornberg, Reference Thornberg2010), measuring peer bullying experiences of preschool students as a group seems to be a better alternative than individual assessments. Therefore, the researchers of the current study designed the PPBS-Teacher Form to examine peer bullying experiences of preschool students as a group. In this way, the PPBS-Teacher Form aims to decrease the risk of labelling while investigating peer bullying in early childhood education.
The Present Study
Existing studies entail certain limitations, including labelling the preschool child, validity and reliability problems, or having too many items. The goal of the present study was to develop a measurement tool free from labelling children with strong reliability and validity evidence. For this purpose, the PPBS-Teacher Form was developed and its psychometric properties are reported in this article.
Method
Procedure
Two independent sets of data were collected, including the preliminary study sample and the cross-validation sample. The first author of this research began working with the preliminary study sample in fall semester of 2016–2017, whereas she began the cross-validation sample in the spring semester of 2016–2017. Ethical approval was obtained separately for the preliminary study and the cross-validation study from the Ethics Committee of Gaziantep University. Ethical clearance was also granted by the provincial Directorate of National Education in Gaziantep, Turkey for the preliminary study, and in Istanbul, Turkey for the cross-validation study. All participants in both samples were verbally informed about the goal of the study, duration of the study, voluntary participation, confidentiality and the right to withdraw from the study at any time. To attract more attention from participants, this information was also included at the top of the first page of each survey administered. Participants filled out paper-based surveys in their own classrooms. While survey completion took approximately 15 minutes for the preliminary study sample, it took about 10 minutes for the cross-validation sample.
Participants
Participants in the preliminary study sample and the cross-validation samples were recruited by a convenience sampling strategy. The preliminary study sample included 306 preschool teachers working in private or public preschools in Gaziantep. However, 27 cases were excluded because 24 of them included a high percentage of the missing values, and 3 cases were outliers. This resulted in a final preliminary study sample of 279 participants.
The cross-validation sample was collected from 254 preschool teachers working in private or public preschools in Istanbul. Seven cases were deleted because of the high percentage of missing values, which resulted in 247 participants. All of the participants in the preliminary and the cross-validation samples were teachers working with preschool children aged from 36 months to 72 months. Demographic information of the participants in the preliminary and the cross-validation samples are given in Table 1.
Demographic information of the preliminary and the cross-validation samples
As presented in Table 1, the mean age of the preliminary study sample was 32.57 with a mean of 9.21 years’ duration in their profession. The mean age of the cross-validation sample was 30.51 with a mean of 8.22 years’ duration in their profession. The majority of the participants in both samples were females holding an undergraduate degree. The number of males was quite low in both samples. Although most of the participants in the preliminary study sample worked in public preschools, most of the participants in the cross-validation sample worked in private preschools.
Data Collection Instrument
Developed by the authors of the current research, the PPBS-Teacher Form was the data collection instrument. The PPBS-Teacher Form asks teachers to report the frequency of bullying behaviours of the students they observe in their classroom, and to assess the students as a group. Items are positively worded, and a 5-point Likert-type response format is utilised: 1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always. The highest score to be obtained from PPBS-Teacher Form is 70 whereas the lowest score is 14. Higher scores indicate more peer bullying behaviours in the preschool classroom. Example items are ‘Behaving ignorantly to a classmate/s in the classroom’, ‘Hitting, slapping, punching, pinching or kicking a classmate/s’ and ‘Verbally threatening a classmate/s’.
The development of the PPBS-Teacher Form was carried out in four steps (DeVellis, Reference DeVellis2016). The extant literature on bullying in preschool was initially reviewed. The review showed that physical bullying, verbal bullying and relational bullying are the three main bullying categories at preschool level, as also reported by Tanrikulu (Reference Tanrikulu2018). The second step involved creating an item pool consisting of 28 items that were categorised under three subdimensions of physical bullying, verbal bullying and relational bullying. As the third step, opinions of preschool teachers were acquired for the item pool. Individual interviews lasting for about 30 minutes were held, with four preschool teachers evaluating the item pool in terms of face validity, content, choice of words, spelling, grammatical structure and understandability. Considering their feedback, the items were modified.
After this modification, the item pool was reassessed by a PhD candidate conducting research on preschool bullying. In accordance with the feedback obtained, the items were refined again with regard to face validity, content, choice of words, spelling, grammatical structure and understandability. Seeking evidence for validity and reliability of the PPBS-Teacher Form was the last step.
Data Analysis
To unravel the factor structure of the PPBS-Teacher Form, an exploratory factor analysis (EFA) was employed for the preliminary study sample data. To validate the factor structure, a confirmatory factor analysis (CFA) was tested. The reliability of the PPBS-Teacher Form was tested by Cronbach’s alpha interitem coefficients. Data were analysed using SPSS 21 and AMOS 21.
Results
Exploratory Factor Analysis
An EFA was performed on the data of the preliminary study sample. As suggested by Pallant (Reference Pallant2011), the suitability of the dataset for the EFA was checked by considering the sample size and the proportion of the variance among the items. Nunnally (Reference Nunnally1978) recommends that there should be at least 10 participants for each item of a scale. There were 28 items, and the data were obtained from 279 participants items in the preliminary study, complying with Nunnally’s (Reference Nunnally1978) criterion. Considering the proportion of the variance among the items, Kaiser-Meyer-Olkin value was .93, and Barlett’s test value of sphericity was significant (p < .05), complying with the criteria suggested by Tabachnick and Fidell (Reference Tabachnick and Fidell2013). Hence, the data from the preliminary study sample were suitable for an EFA.
An EFA with maximum likelihood extraction method and direct oblimin rotation technique was performed to extract the factor structure of the PPBS-Teacher Form. In the first step, the items in the form were examined, and 16 items with the highest variance were selected. These were items numbered 1, 2, 3, 4, 6, 8, 11, 14, 15, 20, 21, 23, 24, 22, 25 and 26. However, the difference in variance values was quite low between items 11 and 20 (the difference was .02) as well as 15 and 24 (the difference was .02), which aimed to assess the same bullying behaviour. Since item 20 and item 24 better expressed the behaviours that were intended to be measured, the authors of this present research excluded item 11 and item 15. As a result, the items of the PPBS-Teacher Form were the 14 items numbered 1, 2, 3, 4, 6, 8, 14, 20, 21, 22, 23, 24, 25 and 26. The EFA with direct oblimin rotation technique was repeated with the 14 items, revealing a three-factor structure (Table 2) with eigenvalues greater than 1 (Tabachnick & Fidell, Reference Tabachnick and Fidell2013).
Total variance explained
As seen in Table 3, the 14 items loaded on three factors before rotation. Nevertheless, a closer look at the factors showed that the items basically loaded to the first two factors. Hence, an additional EFA with direct oblimin rotation technique was run by forcing the items to two factors. Table 3 demonstrates that the items loaded on two factors with a clear pattern after the rotation. While nine items loaded on the first factor, five factors loaded to the second factor. Considering the item contents, the first factor was named as physical-relational bullying whereas the second factor was named as verbal bullying. The correlation between physical-relational bullying and verbal bullying was .54. While physical-relational bullying explained 40.43% of the variance, verbal bullying explained 9.66% of the variance. The two factors explained 50.09% of the total variance.
Factor loadings before and after the rotation
Confirmatory Factor Analysis
Items numbered as 1, 2, 3, 4, 6, 8, 14, 20, 21, 22, 23, 24, 25, 26 in the EFA procedure were renumbered ranging from 1 to 14 in the CFA procedure. A CFA was conducted on the data of the cross-validation sample. The results of the CFA were evaluated by considering the fit indices of normed fit index (NFI), incremental fit index (IFI), relative fit index (RFI), comparative fit index (CFI), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), standardised root mean residual (SRMR), root mean. square error of approximation (RMSEA) and χ2/df as per good practice (Tabachnick & Fidell, Reference Tabachnick and Fidell2013). The two-factor structure of the PPBS-Teacher Form with 14 items was tested by the data obtained from the cross-validation sample. Results of the CFA showed an inadequate fit to the data (χ2 = 313,85 df =76, p = .00; χ2 /df = 4.130; GFI = .829, CFI = .815, TLI = .77, SRMR = .064, RMSEA = .113). For this reason, modification indices were checked, and the error covariance between item 4 and item 5, item 13 and item 14, item 7 and item 8, and item 6 and item 7 were freely estimated. As these item pairs assessed the same construct, freely estimating their error covariances was theoretically justifiable (Hayduk, Reference Hayduk1990; Kaplan, Reference Kaplan1989). After the modification process, results of the CFA improved and showed an adequate fit to the data (χ2 = 220.58, df = 72, p = .00, χ2/df = 3.064; GFI = .885, CFI = .884, TLI = .85, SRMR = .054, RMSEA = .092). These fit indices were acceptable in accordance with the criteria suggested by Erkorkmaz et al. (Reference Erkorkmaz, Etikan, Demir, Özdamar and ve Sanioğlu2013). The CFA model with modifications is shown in Figure 1. In addition, as can be seen in Table 4, the factor loadings of the CFA model ranged between .50 and .75.
The CFA model with modifications and standardised estimates.
Note: P&R, physical bullying and relational bullying; V, verbal bullying.
Factor loadings of the CFA after the modification
Internal Reliability Analysis
Internal reliability of the 14 items of the PPBS-Teacher Form was computed for the data from the preliminary study sample as well as the cross-validation sample. As presented in Table 5, Cronbach’s alpha values of the PPBS-Teacher Form and its two factors ranged from acceptable to good considering the suggested criteria (Tabachnick & Fidell, Reference Tabachnick and Fidell2013).
Internal reliability analysis results
Discussion
The PPBS-Teacher Form was designed with consideration of labelling theory (Becker, Reference Becker1963), and its validity and reliability were tested in the present research. The EFA results indicated that the PPBS-Teacher Form with 14 items had two factors that were named as physical-relational bullying and verbal bullying, explaining 50.09% of the total variance. The two-factor structure was validated by the CFA, revealing adequate fit indices. Internal reliability analyses revealed Cronbach’s alpha values ranging from acceptable to good for the PPBS-Teacher Form. Therefore, based on the teacher assessments, the PPBS-Teacher Form is a valid and reliable measurement instrument to examine peer bullying among preschool children.
Before discussing the strengths, limitations and implications for future research, the two-factor structure of the PPBS-Teacher Form should be discussed. While the first factor was named as physical-relational bullying, the second factor was named as verbal bullying. Before naming these factors, the authors of this current research discussed the factor structure detected as a result of the EFA with two researchers studying early childhood bullying, as well as three preschool teachers. The researchers and the preschool teachers agreed that naming the second factor as verbal bullying was appropriate. In terms of the first factor, they were asked why physical and relational bullying were considered together by the teacher participants. According to the researchers and the preschool teachers, physical acts have a strong and immediate impact on the relations among the preschool children. For instance, hitting someone will cause the victim to become angry with the bully by negatively affecting the relational issues between the victim and the bully in the preschool classroom. The researchers and the preschool teachers reasoned that this fact could be the reason why physical and relational bullying were considered together by the teacher participants. This was why the authors of this present investigation decided to name the first factor as physical-relational bullying. In line with this decision, Juliano et al. (Reference Juliano, Werner and Cassidy2006) reported that physical bullying has a detrimental impact on social relationships among kindergarten children. Yet, our decision should still be considered with caution since some researchers claim that physical and relational forms of bullying in preschool are two distinct types (e.g., Crick et al., Reference Crick, Werner, Casas, O’Brien, Nelson, Grotpeter and Markon1999). Hence, future research is needed to explore whether preschool teachers have a similar understanding to the participants of this current study.
Strengths, Limitations and Implications for Future Research
The PPBS-Teacher Form is a standardised scale with 14 items and two factors, and its validity and reliability were evidenced by two independent samples in this present investigation. Compared to the internal validity threats of the existing qualitative, quantitative and mixed-design measurement studies, the PPBS-Teacher Form is free from data collector bias, history effect and instrument decay. Another strength of the PPBS-Teacher Form is its design in terms of labelling theory (Becker, Reference Becker1963). The available studies require peers (e.g., Perren & Alsaker, Reference Perren and Alsaker2006), teachers (e.g., Kirves & Sajaniemi, Reference Kirves and Sajaniemi2012), as well as parents (e.g., Cameron & Kovac, Reference Cameron and Kovac2016) to rate the peer bullying involvement of preschool children one by one. As suggested by Graham and Juvonen (Reference Graham and Juvonen1998), assessment of individual students’ bullying engagement may create the risk of labelling as bullies or victims. However, the PPBS-Teacher Form asks preschool teachers to evaluate the peer bullying involvement of preschool children as a group, not as individuals. This strategy does not allow for individually labelling students as bullies and victims because preschool teachers need to consider all students’ peer bullying involvement in their classroom. This assessment method can protect preschool children from the risk of being labelled as bullies or victims, and thus the risk of normalising peer bullying, loneliness, social anxiety, low self-worth and rejection problems (Graham & Juvonen, Reference Graham and Juvonen1998; Thornberg, Reference Thornberg2010).
The PPBS-Teacher Form offers some practical solutions. For example, school counsellors and school psychologists can collaborate with preschool teachers to use the PPBS-Teacher Form to examine bullying involvement of the students in a preschool. In this way, school counsellors and school psychologists can have an objective understanding about the nature and the extent of bullying involvement in a specific school, and they can produce evidence-based data for the school administration to be able to develop and implement school-based antibullying policies. In addition, the PPBS-Teacher Form can be utilised by school counsellors or school psychologists to develop educational programs to raise school-based awareness programs for students, teachers and parents. School counsellors and school psychologists can also use the data obtained from the PPBS-Teacher Form to hold seminars to inform parents about the problem of bullying in a specific school. Since the data will involve all the students as a whole, school counsellors and school psychologists do not have to worry about labelling a specific child, and parents will also not have to worry that their children are being labelled by the school personnel.
Despite the dearth of the extant research on early childhood bullying, a considerable amount of information has been accumulated about the nature and the extent of bullying among preschool children (e.g., Perren & Alsaker, Reference Perren and Alsaker2006; Tanrikulu, Reference Tanrikulu2018). For this reason, future research should focus on the prevention and intervention strategies to curb peer bullying at preschool level. Considering that there is only one available study aiming to prevent early childhood bullying with an evidence-based strategy (Alsaker & Valkanover, Reference Alsaker and Valkanover2012), the importance of the prevention and intervention research is very high. As a valid and reliable measurement tool, the PPBS-Teacher Form can help preschool teachers, school counsellors and psychologists as well as researchers to test the effectiveness of their prevention and intervention programs by obtaining data at a classroom level without labelling individual children. Researchers can also utilise the PPBS-Teacher Form to conduct descriptive, cross-sectional, action or longitudinal research designs. Future studies testing the PPBS-Teacher Form’s validity and reliability with different samples are welcomed. Although the PPBS-Teacher Form was performed in Turkish in this present investigation, its English translation is ready for future international researchers who want to test its psychometric features and use it with culturally different samples.
Some limitations exist in this current study. First, the findings cannot be generalised due to the convenient sampling strategy used in this research. Next, the findings may be biased as self-report data were obtained from the participants. The individuals who participated in this current research may have worried about creating a negative image of their classroom.
Moreover, the majority of the participants of this study were women, and this fact should be considered while interpreting the results. In other words, the results of this investigation mainly reflect female teachers’ perspectives on preschool bullying. The authors of this present research could not find any study examining the possible impact of teachers’ gender on reporting bullying incidents. However, Yeh (Reference Yeh2006) found that gender of the victim affected the teacher’s intervention in bullying incidents. In other words, while teachers intervened more in male students’ physical bullying victimisation cases, teachers intervened more in female students’ relational bullying victimisation cases. Such a difference concerning the interventional behaviours of female and male teachers may exist. Therefore, researching the impact of teachers’ gender can certainly contribute to future knowledge about preschool bullying.
Furthermore, a test-retest reliability score could not be calculated because the majority of the participants of this study did not wish to fill out the PPBS-Teacher Form twice due to the heavy workload in their classrooms. Therefore, future investigations examining the test-retest reliability of the PPBS-Teacher Form are welcomed. Lastly, since the time the participants of this study could allocate to filling out the PPBS-Teacher Form was very limited, the convergent as well as the divergent validity of the PPBS-Teacher Form could not be examined. Hence, future studies should explore this.
Conclusion
The findings of this present study reveal that the PPBS-Teacher Form is a valid and reliable measurement instrument tool to assess preschool peer bullying without labelling students as bullies or victims. The PPBS-Teacher Form extends the available measurement tools to assess peer bullying among preschool children with a focus on teacher evaluations. Reliable and valid assessments of preschool bullying with the help of the PPBS-Teacher Form can help prevent and provide interventions for peer bullying among preschool children.
Preschool is a significant period when prevention and intervention strategies should begin against peer bullying. Prevention and intervention strategies beginning at preschool can protect preschool students from the negative impacts of early childhood bullying. Yet, labelling preschool children as bullies or victims is a serious concern in the existing literature on preschool bullying. Students labelled as bullies or victims at the preschool level may adopt these roles in their future education and may continue to suffer from the unfavourable impacts of peer bullying. By paying attention to this risk, the PPBS-Teacher Form aims to measure the preschool peer bullying without labelling children. Such a strategy could contribute to the available literature by drawing attention to the risks of labelling preschool students as bullies or victims and could advance the extant research on early childhood bullying.
Acknowledgments
This paper is based on the research conducted as a part of master’s degree studies of the first author at Gaziantep University, Turkey, under the supervision of the second author. The first author is sincerely thankful to Associate Prof. Ibrahim Tanrikulu for his excellent academic mentoring.
