Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire

G Model

ARTICLE IN PRESS

EARCHI 799 1–11

Early Childhood Research Quarterly xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Early Childhood Research Quarterly

Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire

1

2

3

Q1

4 5

Enrica Ciucci a,∗ , Andrea Baroncelli b , Monica Toselli a a b

Department of Education and Psychology, University of Florence, Via di San Salvi 12, Complesso di San Salvi, Padiglione 26, 50135 Florence, Italy Department of Health Sciences, University of Florence, Via di San Salvi 12, Complesso di San Salvi, Padiglione 26, 50135 Florence, Italy

6

7 22

a r t i c l e

i n f o

a b s t r a c t

8 9 10 11 12 13

Article history: Received 10 January 2014 Received in revised form 17 April 2015 Accepted 26 April 2015 Available online xxx

14

21

Keywords: Meta-emotion philosophy Early childhood teacher Emotional self-efficacy Emotional styles Teachers’ emotional competence Individual teachers’ characteristics

23

Introduction

15 16 17 18 19 20

Q2 24Q3 25 26 27 28 29 30 31 32 33

Meta-emotion philosophy has been theoretically conceptualized as an emotional connection between parents and children which reflects parents’ awareness, acceptance, and regulation of their own emotions and their awareness, acceptance, and regulation of their children’s emotions. The main aim of the current study was to develop and test the psychometric properties of a self-report questionnaire assessing metaemotion philosophy in early childhood teachers (the Crèche Educator Emotional Style Questionnaire – CEESQ). CEESQ consisted of two sections: the first referred to children’s emotions (CEESQ-Children’s Emotions) and the second to early childhood educators’ emotions (CEESQ-Individual Emotions). Participants were 306 early childhood teachers, recruited from 58 day care centers in the center and south of Italy. Exploratory and confirmatory factor analyses for the first section identified three dimensions: two styles used by early childhood teachers to deal with children’s emotions (i.e., Coaching and Dismissing), and teachers’ Self-efficacy as Emotional Socializers. Similar analyses for the second section identified two dimensions: Emotional Self-efficacy (i.e., teachers’ awareness and ability to deal with their own emotions), and Denial of Emotion (i.e., teachers’ lack of acceptance of their own emotions). Both CEESQ sections demonstrated an equivalence in structure considering having/not having own children, years of working experience, academic education level. No main or interaction effects pertaining to teachers’ individual characteristics emerged in any of the CEESQ dimensions. Lastly, both Emotional Self-efficacy and Self-efficacy as an Emotional Socializer were positively associated with a Coaching style, with Selfefficacy as an Emotional Socializer playing a partial mediation role. Administering CEESQ and discussing its scores with teachers could work as a prompt in order to discuss and mentalize teachers’ emotional work as emotional socializers. © 2015 Published by Elsevier Inc.

In this paper, we tested the psychometric properties of a new instrument being developed to measure early childhood teachers’ meta-emotion philosophy. The concept of parents’ meta-emotion philosophy (Gottman, Katz, & Hooven, 1996, 1997) encompasses an organized set of feelings and thoughts (a philosophy of emotional expression and regulation) about their own and their children’s emotions, and an approach to their own and their children’s emotions. Meta-emotion philosophy relates to parenting behaviors, to children’s ability to regulate emotions, and to children’s emotional and social competence (Gottman et al., 1996, 1997).

To date, the construct of meta-emotion philosophy has only been studied in parents and not in other meaningful socializing agents, for instance, early childhood teachers. Research recently evidenced that early childhood teachers offer important socialization opportunities for children (Denham, Bassett, & Zinsser, 2012), and they perform many emotion-laden caregiving tasks (Ahn, 2005; Ahn & Stifter, 2006; Hyson, 2002; Mill & Romano-White, 1999). Considering that like parents, caregivers also play a role as socializers of emotional competence in children, we think that it is important to further understand teachers’ perceptions about their own philosophy regarding how they socialize children to express and control their emotional responses, as well as how the teachers recognize, accept, and regulate their own individual emotions. Meta-emotion philosophy in caregivers and its measurement

∗ Corresponding author. Tel.: +39 055 2755012; fax: +39 055 6236047. E-mail addresses: enrica.ciucci@unifi.it (E. Ciucci), a.baroncelli@unifi.it (A. Baroncelli), monica.toselli@unifi.it (M. Toselli).

The meta-emotion philosophy construct (Gottman et al., 1996, 1997) provides an interesting conceptual framework for

http://dx.doi.org/10.1016/j.ecresq.2015.04.006 0885-2006/© 2015 Published by Elsevier Inc.

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

34 35 36 37 38 39 40 41 42 43 44 45 46

47

48 49

G Model EARCHI 799 1–11 2 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

representing the way through which parents’ thoughts and feelings about emotions in themselves and in their children are connected to how they socialize emotions in their children. That is to say, the meta-emotion model assumes that adults’ beliefs and feelings about emotions are expressed through their behaviors toward children’s emotions. The two main parenting styles concerning children’s emotional socialization are emotion coaching and emotion dismissing (Gottman et al., 1996, 1997; Hooven, Gottman, & Katz, 1995; Lagacé-Séguin & Coplan, 2005; Lunkenheimer, Shields, & Cortina, 2007; Ramsden & Hubbard, 2002). Coaching children’s emotions is a style in which parents are aware of both their own and their children’s feelings, accept them empathically, and effectively help children to process and regulate emotions. Dismissing children’s emotions is a parenting style characterized by a lack of awareness of both their own and their children’s emotions, and by a deficiency in effectively solving emotional problems which results in devaluating, minimizing, and ignoring children’s negative emotions. According to this approach, parental meta-emotion philosophy is conceptualized as an emotional connection between parents and children which reflects parents’ awareness, acceptance, and regulation of their own emotions and their awareness, acceptance, and regulation of their children’s emotions. Gottman et al. (1996, 1997) developed a meta-emotion interview specifically focused on parents’ awareness, acceptance, and regulation of their own, as well as their children’s emotions. High levels of these three processes define a coaching style, whereas low levels define a dismissing style. In order to overcome some limits of this interview (e.g., 90 min required to administer it, specialized training to conduct it), LagacéSéguin and Coplan (2005) and Hakim-Larson, Parker, Lee, Goodwin, and Voelker (2006) created and validated two self-report questionnaires specifically focused on emotional parenting styles, referring exclusively to the parental socializing of children’s emotions. The aim of the current study was twofold. First, we aimed to develop a self-report questionnaire based on Gottman and colleagues’ (1996, 1997) original interview in which the three processes of meta-emotion philosophy (i.e., awareness, acceptance, and regulation of emotions) were referred to children’s emotions in terms of emotional styles as well as to the emotions of the caregivers. Second, in order to fill a sizeable gap in the literature, we focused on early childhood teachers, considering the important role they play as emotional socializers of children’s emotions. As stated, the construct of meta-emotion philosophy had only been referred to parents and not to other meaningful socializing agents, for instance, early childhood teachers. Even though the role of early childhood teachers as socializers of emotions has to a large extent been neglected, recent research has stressed that they offer important socialization opportunities for children (Denham et al., 2012), they spend a significant amount of time with the children (Innocenti Research Centre, 2008; Phillips & Adams, 2001), and perform many emotion-laden care-giving tasks (Ahn, 2005; Ahn & Stifter, 2006; Hyson, 2002; Mill & Romano-White, 1999). As Denham et al. (2012) asserted, early childhood teachers work as socializers of children’s emotional competence as much as parents do, mainly by modeling or scaffolding, or by reacting to children’s emotions. Furthermore, early childhood teachers can react to children’s emotions by showing empathy or physical comfort, or distract and help children to cope with feelings by problem-solving, or ignore and give negative responses to the children’s emotions (Ahn, 2005; Ahn & Stifter, 2006; Hyson, 2002; Meyer & Turner, 2007). In fulfilling the aim of the current study, we considered it crucial to identify the nature of early childhood educators’ beliefs about children’s emotions, in terms of which practices they consider pertinent to their professional role, how they feel to be able to implement them (i.e., self-efficacy), and finally, how they perceive

the impact of their behaviors on children’s emotional development. We were driven by the parenting literature in which parental selfefficacy and perceived impact emerge as important variables in explaining parental functioning across various domains (e.g., emotional availability, nurturance, physical childcare behaviors, play, discipline), as well as parental satisfaction (Bandura, 1986, 1997; Boivin et al., 2005; Bornstein, 2002; Coleman & Karraker, 1997, 2003; Teti & Gelfand, 1991). Moreover, our interest in teachers’ beliefs about children’s emotions was justified by a previous study by Hyson and Lee (1996) who developed a tool (the Caregiver’s Beliefs About Feelings Survey – CBAF) for preschool educators, focused on dimensions like “Beliefs about adult labeling and talking about emotions with children” and “Beliefs about affectionate bond with children”. Their model, specifically concerning beliefs, does not identify coherent behavioral styles among educators toward children’s emotions, nor does it evaluate teachers’ self-efficacy or perceived impact. To summarize, in the present study, we tested the psychometric properties of a new tool developed in order to measure early childhood teachers’ meta-emotion philosophy. The tool comprised two sections: the first referred to children’s emotions and the second to early childhood educators’ emotions. The first section was created in order to identify the styles used by early childhood teachers to deal with children’s emotions, self-efficacy as emotional socializers, and their perceived impact in this task. The second section was created in order to identify the individual attitude of early childhood teachers toward their own emotions. In so doing, we were inspired by instruments addressing parental meta-emotion philosophy (Gottman et al., 1996, 1997; Katz & Windecker-Nelson, 2004; Lagacé-Séguin & Coplan, 2005), and parental beliefs about their role as parents (Boivin et al., 2005; Coleman & Karraker, 2003; Teti & Gelfand, 1991). The impact of teachers’ emotions in teaching A further rationale for this study derived from a review of the extant research, mainly focusing on primary school teachers, about the impact of emotions in teaching. Several studies have provided evidence that teachers’ emotional competence (i.e., being aware of their own emotions, accurately perceiving others’ emotions, managing emotions and using them to facilitate thought and action; Mayer, Salovey, Caruso, & Sitarenios, 2001) is related to their perception of effectiveness in their role and in their practices (Day & Leitch, 2001; Penrose, Perry, & Ball, 2007; Perry & Ball, 2008; Sutton & Wheatley, 2003; Winograd, 2003; Zembylas, 2007). According to Bandura’s postulation (1997), experiences of efficacy as well as emotional states enhance or impair beliefs about the self, because they are a source of information on individual selfefficacy. Moreover, teachers’ efficacy is affected by the variance in their emotionality and emotion regulation (Sutton & Wheatley, 2003). Research about teachers’ emotions recently recognized that the emotional labor teachers are required to perform when they regulate their own feelings in teaching has an impact on the emotional quality of the interactions between teachers and students, parents and colleagues, on the sense of the teachers’ personal and professional identities, and finally on the caring ethics of teaching (Schutz & Zembylas, 2009; Sutton, 2004; Yin & Lee, 2012). In other words, teachers can be seen as emotional workers, who intentionally use their emotional experience to fit professional and ethical norms in teaching. To date, few studies have investigated the impact of teachers’ emotions on the quality of the caregiver-child relationship and on the socialization of emotions in preschool children (de Schipper, Riksen-Walraven, Geurts, & Derksen, 2008; Ersay, 2007). In particular, Ersay (2007) found that preschool teachers who were more aware of their own emotions were also less likely to ignore or

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147

148

149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179

G Model EARCHI 799 1–11

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

180 181 182 183 184 185 186 187

188

minimize their students’ emotions, and more likely to accept and encourage all their students’ emotions, including the negative ones. A key advance made by the present study was to analyze the associations between individual differences in early childhood teachers’ emotional competence (i.e., awareness, acceptance, and regulation of own emotions), their emotional styles with the children’s emotions (i.e., coaching and dismissing styles), and their self-efficacy as emotional socializers. The role of teachers’ individual characteristics

226

A further question worthy of exploration is how teachers’ individual experience with their own children and professional characteristics (i.e., academic education level, years of teaching) affect their attitudes and responses to the children’s emotions, their self-efficacy as emotional socializers, and their emotional processes. As far as we know, there is no research about the role of individual experience with own children in teachers’ practices, so this issue was investigated in an exploratory manner. The link between caregivers’ academic education and/or professional training and children’s outcomes (e.g., emotional, social and academic functioning) is perhaps the area most considered in the research about quality in childcare (for a review, see Huntsman, 2008). Generally, the weight of evidence favors the idea of a positive link between teachers’ level of academic education, qualifications and training, and childcare quality, which, in turn, positively affects children’s development (Blau, 2000; Burchinal, Howes, & Kontos, 2002; Hyson & Lee, 1996; Layzer & Goodson, 2006; Manlove, Vazquez, & Vernon-Feagans, 2008; NICHD, 1996, 2000). Similarly, the amount of professional experience (i.e., years of teaching) may be a key ingredient in teachers’ sense of efficacy as well as in their emotional competence as socializers of children’s emotions (Penrose et al., 2007; Sutton, 2004). Specifically, the longer the period spent in a position at work, the greater the likelihood that teachers will demonstrate positive efficacy and the ability to regulate their own emotions. In Italy, as well as in other countries, early childhood teachers hold various levels of academic qualifications. Educators can work after obtaining a specific high school degree for teaching in preschools, but they may also have obtained an university degree in Educational Sciences, which requires four more years of study. Moreover, in Italy, early childhood educators regularly follow compulsory refresher courses, organized by the institutions where they work. A range of different courses is offered by the schools themselves, while it is compulsory to attend a specific number of hours. More work experience could mean more attending training courses. However, this link is in need of indepth exploration, as it was not confirmed by Hyson and Lee (1996).

227

The current study

189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225

228 229 230 231 232 233 234 235 236 237 238 239 240 241

On the basis of the above-reported literature, the main aim of the current study was to develop a multidimensional self-report questionnaire to assess meta-emotion philosophy in early childhood teachers (the Crèche Educator Emotional Style Questionnaire – CEESQ) and test its psychometric properties. A further aim was to explore the influence of teachers’ individual experience with their own children as well as their professional characteristics (i.e., academic education level, years of teaching) on the CEESQ scales. Consistently with Ersay’s results (2007), we hypothesized that early childhood teachers who were aware of their own emotions and more prone to accept and regulate their own emotions were also more likely to recognize, accept, and regulate the children’s emotions. We also hypothesized that a high level of academic education and more years of teaching were positively related to teachers’

3

self-efficacy as emotional socializers and to an adaptive emotion coaching style. No specific hypotheses were put forward about the link between teachers’ individual experience of bringing up their own children and the dimensions measured by the CEESQ.

242 243 244 245

Method

246

Participants and procedure

247

Participants were 306 early childhood teachers from 58 day care centers in the center and south of Italy. In Italy, early education teachers interact with infants aged 0–3 years old. Children can attend day centers for 4–8 h, 5 days a week, depending on parental needs. Day centers are regulated or strictly controlled by the municipalities/local councils. Teachers were recruited on refresher courses held at their place of work. The centers were chosen through convenience sampling, when they approached the research team to follow refresher courses. All the course participants spontaneously accepted to take part in the research and no economic incentives were given. All of the teachers were female and Italian. This exclusively female sample gives a good representation of early childhood educators in Italy, who are almost exclusively women. Their ages ranged between 25 and 61 (mean age = 43.40 years, SD = 9.20 years); 203 out of the 302 (67.22%) teachers declared that they had at least one child. The number of years that the answering participants had been teaching (244 out of 306) ranged from 1 to 35 (mean length = 14.77 years, SD = 9.63 years). The high rate of missing data for this item (n = 62) may be attributed to difficulties in calculating their length of service as they may have worked for on and off as supply teachers. As the highest level of education, 190 out of 301 (63.12%) reported a high school degree, and 111 out of 301 (36.88%) a university degree. The study took place in each day care center, in a group format, within a single session. The early childhood teachers were requested to fill out the questionnaire anonymously, on individual sheets, without any time limitation. The CEESQ was administered by trained assistants. Measure Preliminary instrument development and final version of the CEESQ The Crèche Educator Emotional Style Questionnaire – CEESQ – was developed to assess meta-emotion philosophy in early childhood teachers. It consisted of two scales with items assessing the frequency of teachers’ behaviors and beliefs with regard to both the children’s and their own emotions. Each item is rated on a 5point Likert scale ranging from 1 (rarely or never) to 5 (very often). The first scale – CEESQ-Children’s Emotions – comprised items representing coaching and dismissing emotional styles, self-efficacy as emotional socializers and teachers’ perceived impact about the effect of their behavior on the children’s emotional development. An original pool of 26 items was adopted. Coaching and Dismissing reflected adult behavioral tendencies toward the regulation of children’s emotions. Two other theorized emotional styles described by parenting literature (i.e., Laissez-faire and Disapproving) were not included in our scale, owing to their lack of social desirability in this context, as resulting from teachers’ specific training and knowledge about this issue. Early childhood teachers’ self-efficacy refers to teachers’ perceived ability to carry out tasks associated with the role of emotional socializer, specifically within the context of day care. Early childhood teachers’ perceived impact relates to teachers’ evaluation of the effect of their behavior on the children’s emotional development.

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276

277

278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301

G Model EARCHI 799 1–11 4

302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

These domains were selected because they reflected significant bodies of theory or research on early childhood emotional socialization (for parental meta-emotion philosophy: Gottman et al., 1996, 1997; Katz & Windecker-Nelson, 2004; for parents’ and teachers’ beliefs on children’s development: Boivin et al., 2005; Coleman & Karraker, 2003; Teti & Gelfand, 1991). The items on emotional styles were inspired by the Maternal Emotional Style Questionnaire (MESQ, by Lagacé-Séguin & Coplan, 2005) which mainly focused on parents’ behaviors in dealing with the emotions of sadness and anger in their own child. Thus, some items were modified to refer to children in general (i.e., When my child is angry. . . was transformed into When a child is angry. . .), or to vary the emotions, in order to have a more balanced distribution of the four basic emotions (i.e., fear, happiness, sadness and anger) across the two emotional styles. Only three items were strictly derived from the MESQ: When my child is angry, my goal is to make him/her stop; I help my child get over sadness quickly so that he/she can move on; When my child is angry, it’s an opportunity for getting close. Items on early childhood teachers’ self-efficacy as emotional socializers and early childhood teachers’ perceived impact were inspired by the same parental dimensions as the Parental Cognitions and Conduct Toward the Infant Scale (PACOTIS, Boivin et al., 2005) which measures parents’ belief about their role as parents. Teachers’ self-efficacy referred to their perceived capability to manage the children’s emotional lives, specifically within the context of the day care center. This included perceived efficacy in discerning children’s emotional states, helping children to comprehend their own emotions, managing the expression of positive and negative emotion, and coping with them (e.g., I feel that I am very good at. . .). Coaching style was expressed by items such as Children’s sadness is an emotion worth exploring; When a child is angry, it’s an opportunity for getting close; dismissing was expressed by items such as I try to change the negative mood of a child into a cheerful one; When a child is afraid, I try to distract him/her. Perceived impact of early childhood teachers was expressed by items such as The contribution of early childhood teachers to the emotional development of young children is fundamental at the care center; teachers’ self-efficacy as emotional socializers was expressed by items such as I am able to stay close to an angry child; I feel able to help children cope with their fears and their anger. The second CEESQ scale – Individual Emotions – comprised items on self-evaluation in the three components of the metaemotion philosophy (Gottman et al., 1996, 1997): awareness, acceptance and regulation of own emotions. Awareness of own emotions was expressed by the ability to talk regarding and to differentiate emotions; acceptance of own emotions referred to the meaning and value attributed to the emotions. Regulation of own emotions referred to the ability to control the strength and quality of emotions. An original pool of 15 items was adopted. The items on these three emotional processes were inspired by the focus of the Maternal Emotional Interview (MEI, Gottman et al., 1996, 1997). Awareness was expressed by items such as When my mood changes, I easily recognize my emotions. Acceptance was expressed by items such as I don’t like the emotions I experience (reversed score). Regulation of one’s own emotions was expressed by items such as When I’m angry, I have some control over my emotions. To assess the face validity of the CEESQ, three early childhood teachers following a post-graduate course were requested to evaluate whether the items reflected the early childhood teachers’ beliefs and behaviors regarding children’s emotions as well as their own emotional processes. According to their feedback, 5 out of the 26 original items in the first scale were discarded because they were judged too vague or ambiguous. The final version of the CEESQ-Children’s Emotions consisted of 21 items, while the CEESQIndividual Emotions consisted of 15 items. The full scale is available as online supplementary material.

Data analyses The data analysis strategy used in the current study refers to various steps; in order to make our paper easier to read, we grouped all the analyses performed into five main steps which will be described below. Step 1: Adequacy of observed items and exploratory factor analyses First of all, we performed an exploratory factor analysis (EFA) on the 21 items from the CEESQ-Children’s Emotions scale. Before performing the EFA, several methods were employed to check the adequacy of our data. We initially inspected the skewness and kurtosis scores for each observed variable, and scores ranging between −1.00 and +1.00 were considered to indicate a normal distribution. Kaiser-Meyer-Olkin’s sampling adequacy criteria (KMO) with individual measures of sampling adequacy for each item (MSA) tested whether there was a significant number of factors in the data: the range of KMO is from .00 to 1.00; values lower than .50 were considered unacceptable, .50 to .60 miserable, .60 to .70 mediocre, .70 to .80 middling, .80 to .90 meritorious, and greater than .90 marvelous; items with an MSA below .50 do not belong to a latent group and they can be removed from the factor analysis (Kaiser, 1974). Bartlett’s sphericity test (1950) was used to test the hypothesis that correlations between variables were greater than expected by chance; in this, the p-value should be significant. To determine how many factors should be extracted, we computed eigenvalues and inspected the relative scree diagram (Cattel, 1966). After that we ran an EFA; to avoid any possible distortion due to data distributions, we followed Costello and Osborne’s (2005) suggestion of using the principal axis method, and we allowed for correlations between latent factors by performing a promax rotation. Only items with a factor loading greater than |.30| were considered. For each factor that emerged, internal consistency was inspected using Cronbach’s alpha; values lower than .60 were considered insufficient, .60 to .69 marginal, .70 to .79 acceptable, .80 to .89 good, and ≥.90 excellent (Barker, Pistrang, & Elliott, 1994). Step 2: Confirmatory factor analyses A confirmatory factor analysis (CFA) approach was used to test the CEESQ-Children’s Emotions structure more closely. Two models were tested: Model 1 was a single general factor fitting the covariance matrix for all the CEESQ-Children’s Emotions items emerging from the EFA; Model 2 was a factor model in which the latent variables emerging in the EFA were tested. CFA was carried out using the robust maximum likelihood method, which is preferable to avoid distortions due to data distribution (Schermelleh-Engel, Moosbrugger, & Müller, 2003). To compare the fit of each model, several fit indices were used to overcome the limitations of each index (Hooper, Coughlan, & Mullen, 2008; Marsh, Balla, & Hau, 1996). From the family of absolute fit indices (which determine how well an a priori model reproduces the sample data), we chose the relative chi-square (2 /df), the root mean square error of approximation (RMSEA; Steiger & Lind, 1980), and the standardized root mean square residual (SRMR; Bentler, 1995). From the comparative (or incremental) fit indices family (which compares the tested model to a baseline one) we chose the comparative fit index (CFI; Bentler, 1990). From the family of parsimony fit indices (which indicates which one is the most parsimonious when different models are compared) we adopted the Akaike Information Criterion (AIC; Akaike, 1987). A good model fit is indicated by a relative chi-square value between 0 and 2 (between 2 and 3 is acceptable), an RMSEA and SRMR of .08 or lower, and a CFI exceeding .90; moreover, the model with the minimum AIC values is regarded as the best fitting (Byrne, 1994; Hooper et al., 2008). Finally, the list of possible modi- Q4 fication indices for Model 2 was inspected and covariance between

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

368

369 370 371 372

373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402

403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430

G Model EARCHI 799 1–11

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

431 432

error variables was allowed for greater modification indices within each factor.1

5

Results

476

Step 1A. Adequacy of observed items and exploratory factor analyses for CEESQ-Children’s Emotions 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450

451 452 453 454 455 456 457 458

459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475

Step 3: Invariance of factor structure In order to examine the invariance of the CEESQ-Children’s Emotions factor structure among different sub-samples, we tested three multi-group procedures. In particular, we aimed to explore whether factor structure was different for early childhood teachers with/without at least one child of their own, for teachers with different amounts of teaching experience (divided at the median, with fewer than or equal to 12 years of teaching and over 12 years), and for early childhood teachers with a high school versus university level of education. To investigate this, we compared a baseline model which allowed the factor coefficients and the correlations between latent factors to differ across groups (Model B) to a model in which factor loadings were fixed so as to be equal across groups (Model L), and to a model in which factor loadings and correlations between latent factors were fixed so as to be equal across groups (Model L + C). Steps 1–3 were re-performed for the 15 items in the second section of the CEESQ (i.e., CEESQ-Individual Emotions).

Step 4: The influence of early childhood teachers’ individual characteristics on the CEESQ scales After presenting the descriptive statistics for the CEESQ scales, a subsequent set of analyses aimed to explore the differences in the CEESQ dimensions that considered having/not having children, years of teaching, and academic education level through a 2 × 2 × 2 MANOVA for the CEESQ-Children’s Emotions factors and through a 2 × 2 × 2 MANOVA for the CEESQ-Individual Emotions factors.

Step 5: Correlation and mediation analyses Moreover, correlations (Spearman’s r) were employed in order to analyze bivariate associations between emerging factors, and a suggested mediation effect was tested. A mediator is a third variable (e.g., C) that explains how (or why) two other variables (e.g., A and B) are related, and, in our sample, in order to investigate it, we followed a linear regressions approach (Fairchild & McQuillin, 2010; Holmbeck, 1997): we tested the significance of the “A predicting C” path (regression 1); we tested the significance of the “A predicting B” path (regression 2, step 1); and we tested the significance of the “C predicting B” path, and the relative effect of A on B (regression 2, step 2). In order to check whether the reduction in the beta value for the “A predicting B” path was significant after the introduction of the hypothesized mediator (i.e., C), a Sobel test (Sobel, 1982; Soper, 2013) was performed. Steps 1–3 were conducted using Lavaan Package of R Software (Rosseel, 2012), and steps 4–5 using IBM SPSS version 20.

1 Note 1: The high correlation that emerged between Coaching and Self-efficacy as an Emotional Socializer led us to investigate whether they could be collapsed into a single latent variable. To do this, we compared our factor structure (Model 2-modified) with a model in which items of both Coaching and Self-efficacy as an Emotional Socializer were considered as indicators of a single factor (Model 3). The goodness-of-fit of Model 3 was far from sufficient (2 = 430.92, p < .001, df = 135, 2 /df = 3.19, CFI = 74, RMSEA = 09 (95% CI .08-.09), SRMR = 09, AIC = 12,399.98), and the chi-square difference test was significant (2 = 216.33, df = 18, p < .001); moreover, the AIC value of Model 2-modified was lower than that of Model 3 (i.e. 12,189.07 versus 12,399.98). We concluded by rejecting the hypothesis that Model 3 had a better fit than Model 2-modified, and retained the three-factor structure for the CEESQ-Children’s Emotions.

477 478

Considering the 21 items of the CEESQ-Children’s Emotions, skewness scores ranged between |.015| and |.914| and kurtosis scores ranged between |.022| and |1.000|, with the exception of two items (4 and 21) that had kurtosis values of 1.434 and 1.899 respectively. The results from our dataset indicated that the MSA values of all the indicators ranged between .63 and .80, and the KMO index was middling (.80). The result of Bartlett’s sphericity test was 2 = 1591.92, df = 210; p < .001. Six eigenvalues greater than 1.00 were found, but the scree diagram suggested that three factors (eigenvalues = 4.40, 2.63 and 1.73) could be an adequate number to overcome deleterious results due to either over- or underextraction. Table 1 includes rotated factor loadings: items 13 and 15 failed to reach factor loading >|.30|, moreover item 5 was considered ambiguous because it had a factor loading on factor 3 of |.33| and a factor loading on factor 1 of |.23|; for these reasons, items 5, 13, and 15 were excluded from subsequent analyses. To summarize, factor 1 included 7 items reflecting teachers’ awareness, acceptance, and ability to deal with children’s emotions, and beliefs about the impact of early childhood teachers’ role in children’s emotional development (i.e., Children’s sadness is an emotion worth exploring; I accept children’s fear even if it seems unmotivated; When a child is feeling a negative emotion, it’s an opportunity to use my educational skills; The contribution of early childhood teachers to the emotional development of young children is fundamental at the care center), and it was labeled “Coaching Style.” It accounted for 12% of variance and its factor loadings ranged from |.38| to |.82|. The Cronbach’s alpha for this subscale was acceptable (.75). Factor 2 comprised 6 items relating to teachers’ self-efficacy as emotional socializers (i.e., I feel able to help children cope with their fears and their anger; I easily recognize the emotions that a child is experiencing) and it was labeled “Self-efficacy as an Emotional Socializer.” It accounted for 9% of variance and its factor loadings ranged from |.44| to |.77|. The Cronbach’s alpha for this subscale was acceptable (.78). Factor 3 consisted of 5 items reflecting a dismissing emotional style (i.e., I try to change the negative mood of a child into a cheerful one; When a child is afraid, I try to distract

Table 1 Factor loadings of EFA on CEESQ-Children’s emotions.

480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514

Q5

Item

Coaching

Self-efficacy as emotional socializer

Dismissing

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

.03 .02 −.08 .09 −.23 .44 .82 .08 .38 .14 .66 .61 .11 .01 .07 .55 −.19 .14 .39 −.09 −.16

.13 .11 .05 −.44 −.11 −.02 .25 −.02 −.07 −.61 −.12 −.11 −.23 −.74 −.07 −.04 −.06 −.51 −.11 −.75 −.59

.63 .74 .44 .04 .33 .15 −.01 .57 .09 .06 −.07 −.03 .03 −.09 .27 .01 .55 .03 .05 −.12 −.04

.77

.78

.70

Cronbach’s alpha

479

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

G Model EARCHI 799 1–11 6 515 516 517 518 519 520 521

522 523

524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550

551 552

553 554 555 556 557 558 559 560 561 562 563

564 565

566 567 568 569 570 571 572 573 574

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

him/her) and it was labeled “Dismissing Style.” It accounted for 12% of variance and its factor loadings ranged from |.44| to |.74|. Once again, the Cronbach’s alpha for this subscale was acceptable (.70). Together, these three factors accounted for 33% of the variance. The correlations between latent factors were: Coaching–Self-efficacy as Emotional Socializer = .50; Coaching–Dismissing = .06; Self-efficacy as an Emotional Socializer–Dismissing = .30. Step 2A. Confirmatory factor analyses for CEESQ-Children’s Emotions Initially, two models were tested: Model 1 was a single general factor fitting the covariance matrix for all the CEESQ-Children’s Emotions items emerging from the EFA; Model 2 was a threefactor model in which the three latent variables emerging in the EFA were tested (with the Self-efficacy as an Emotional Socializer factor correlated to the others). With regard to Model 1, the fit turned out to be not acceptable (2 = 636.88, p < .001, df = 135, 2 /df = 4.72, CFI = .56, RMSEA = .11 (95% CI .10–.12), SRMR = .11, AIC = 12,634.77). A significant chi-square difference test along with a lower AIC indicated that Model 2 showed a significantly improved fit compared to Model 1 (2 = 34,549, df = 2, p < .001), but the CFI still did not reach a completely acceptable value (2 = 291.39, p < .001, df = 133, 2 /df = 2.19, CFI = .86, RMSEA = .06 (95% CI .05–.07), SRMR = .06, AIC = 12,246.03). To avoid this problem, the list of possible modification indices for Model 2 was inspected and covariance between error variables was allowed for greater modification indices within each factor. Model 2-modified showed a significantly improved fit compared to Model 2 (2 = 76.80, df = 16, p < .001), and all fit indices achieved good values (2 = 214.59, p < .001 df = 117, 2 /df = 1.83, CFI = .91, RMSEA = .05 (95% CI .04–.06), SRMR = .06, AIC = 12,189.07). Factor loadings ranged from .42 to .81 for the Coaching dimension, from .39 to .74 for the Self-efficacy as an Emotional Socializer dimension, and from .33 to .72 for the Dismissing dimension (all p values were <.001). The correlations between latent factors were: Coaching–Self-efficacy as an Emotional Socializer = .60, p < .001; Self-efficacy as an Emotional Socializer–Dismissing = 14, p > .05. Step 3A. Invariance of factor structure for CEESQ-Children’s Emotions The emerging factor structure proved to be invariant across subjects with/without children (Model B–Model L: 2 = 19.41, df = 15, p > .05; Model B–Model L + C: 2 = 21.41, df = 17, p > .05), with/without teaching experience of more than 12 years (Model B–Model L: 2 = 22.59, df = 15, p > .05; Model B–Model L + C: 2 = 25.43, df = 17, p > .05), and with a high school versus university level of education (Model B–Model L: 2 = 12.39, df = 15, p > .05; Model B–Model L + C: 2 = 17.87, df = 17, p > .05). The means and standard deviation for the CEESQ-Children’s Emotions are provided in Table 2. Step 1B. Adequacy of observed items and exploratory factor analyses for CEESQ-Individual Emotions As for the CEESQ-Individual Emotions, the skewness scores for the observed items ranged between |.007| and |.816| and the kurtosis scores ranged between |.026| and |.919|, with the exception of item 5 that had a skewness value of 1.437 and a kurtosis value of 2.291. The MSA values of all the indicators ranged between .61 and .91, and the KMO index was middling (.87). The result of Bartlett’s sphericity test was 2 = 1303.38, df = 105; p < .001. Three eigenvalues greater than 1 were found, but the scree diagram suggested that two factors (eigenvalues = 4.86 and 1.79) were

sufficient. Table 3 includes the rotated factor loadings. Item 14 was considered ambiguous because it presented high loadings on both factors, and it was excluded from subsequent analyses. The first factor comprised 10 items relating to self-efficacy about appraisal, expression, regulation, and use of own emotions (i.e., When my mood changes, I easily recognize my emotions; I am able to express what I feel; I am able to prevent my fears taking over; I can describe the strategies I use in order to cope with negative emotions) and it was labeled “Emotional self-efficacy.” It accounted for 9% of variance and its factor loadings ranged from |.50| to |.68|. The Cronbach’s alpha for this subscale was good (.86). The second factor comprised 4 items relating to denying or being afraid of one’s own emotions (i.e., When I feel euphoric, I have the feeling of losing control; I don’t like the emotions I experience; I perceive my negative emotions as something to defend myself against) and it was labeled “Denial of Emotion.” It accounted for 10% of variance and its factor loadings ranged from |.39| to |.56|. The Cronbach’s alpha for this subscale was near-marginal (.59). Together, these two factors accounted for 19% of the variance. The correlation between the two latent variables was −.28. Step 2B. Confirmatory factor analyses for CEESQ-Individual Emotions Model 1 was a single general factor model fitting the covariance matrix for all the CEESQ-Individual Emotions items emerging from the EFA. Model 2 was a two-factor model in which the two correlated latent variables were tested. The goodness of-fit indices for Model 1 were not sufficient (2 = 231.53, p < .001, df = 77, 2 /df = 3.01, CFI = .83, RMSEA = .08 (95% CI .07–.09), SRMR = .07, AIC = 9984.25). Model 2 showed a significantly improved fit compared to Model 1 (2 = 80.47, df = 1, p < .001) and reached near-sufficient values (2 = 171.06, p < .001, df = 76, 2 /df = 2.25, CFI = .90, RMSEA = .06 (95% CI .05–.08), SRMR = .06, AIC = 9915.29). Model 2-modified (with the inclusion of modification indices) showed a significantly improved fit compared to Model 2 (2 = 38.56, df = 2, p < .001), and all the goodness-of-fit indices were good (2 = 132.50, p < .001, df = 74, 2 /df = 1.79, CFI = .94, RMSEA = .05 (95% CI .04–.06), SRMR = .05, AIC = 9873.39). Factor loadings ranged from .40 to .73 for the Emotional Self-efficacy dimension, and from .41 to .56 for the Denial of Emotion dimension (all p values were <.001). The correlation between latent factors was −.45, p < .001. Step 3B. Invariance of factor structure for CEESQ-Individual Emotions Factor structure proved to be invariant across early childhood teachers with/without children (Model B–Model L: 2 = 12.76, df = 12, p > .05; Model B–Model L + C: 2 = 13.86, df = 13, p > .05), with/without more than 12 years of teaching experience (Model B–Model L: 2 = 3.62, df = 12, p > .05; Model B–Model L + C: 2 = 3.85, df = 13, p > .05), and with a high school versus university education (Model B–Model L: 2 = 9.23, df = 12, p > .05; Model B–Model L + C: 2 = 10.66, df = 13, p > .05). The means and standard deviation for the CEESQ-Individual Emotions are provided in Table 4. Step 4. The influence of early childhood teachers’ individual characteristics on the CEESQ Scales The descriptive statistics of the five latent factors of the CEESQ-Children’s Emotions and CEESQ-Individual Emotions are summarized in Table 5. All the factors are normally distributed, as suggested by the inspection of the skewness and kurtosis values. Two 2 × 2 × 2 (having own children × years of teaching

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594

595 596

597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615

616 617

618 619 620 621 622 623 624 625 626 627

628 629

630 631 632 633 634

G Model

EARCHI 799 1–11

Without child/children (n = 99)

With child/children (n = 203)

Teaching experience up to 12 years (n = 135)

Teaching experience greater than 12 years (n = 109)

With high school degree (n = 190)

With university degree (n = 111)

M

SD

M

SD

M

SD

M

SD

M

SD

M

SD

M

SD

11 12 7 16 9 6 19 Coaching

3.95 3.85 4.24 3.78 3.71 4.05 3.72 3.90

.81 .83 .76 .84 .81 .83 .98 .54

3.91 3.80 4.28 3.63 3.80 4.19 3.53 3.88

.76 .88 .73 .92 .78 .81 1.04 .54

3.97 3.85 4.24 3.86 3.74 4.01 3.79 3.92

.83 .81 .80 .80 .79 .86 .96 .53

3.93 3.76 4.24 3.69 3.77 4.14 3.76 3.90

.79 .87 .80 .87 .76 .79 .93 .51

3.94 3.91 4.26 3.87 3.72 3.99 3.64 3.91

.83 .79 .75 .82 .82 .90 1.07 .57

3.95 3.83 4.21 3.76 3.65 4.05 3.66 3.87

.81 .83 .78 .86 .82 .86 .98 .55

3.96 3.90 4.29 3.79 3.79 4.03 3.83 3.94

.81 .82 .74 .81 .80 .78 .99 .51

10 18 14 20 4 21 Self-efficacy as emotional socializer

3.48 3.56 3.27 3.34 2.98 3.01 3.27

.68 .69 .67 .64 .76 .63 .47

3.47 3.51 3.17 3.28 3.01 2.88 3.22

.70 .71 .72 .68 .83 .54 .51

3.46 3.53 3.28 3.35 2.94 3.05 3.27

.69 .69 .67 .63 .77 .67 .46

3.45 3.53 3.21 3.29 2.99 2.95 3.23

.70 .70 .69 .63 .75 .63 .47

3.49 3.51 3.28 3.38 2.94 3.05 3.27

.69 .69 .70 .66 .84 .63 .48

3.47 3.55 3.26 3.36 2.98 3.06 3.28

.65 .70 .70 .63 .83 .66 .46

3.51 3.56 3.27 3.28 2.96 2.91 3.25

.74 .66 .62 .65 .66 .57 .47

2 1 8 17 3 Dismissing

2.45 2.31 2.80 2.54 2.32 2.48

1.02 .96 .94 1.06 .86 .66

2.36 2.38 2.94 2.67 2.42 2.56

1.08 .92 .95 1.03 .83 .70

2.53 2.38 2.74 2.42 2.31 2.47

1.03 .99 .91 1.05 .85 .65

2.48 2.37 2.83 2.50 2.40 2.52

1.09 1.00 .96 1.03 .86 .70

2.44 2.39 2.79 2.52 2.28 2.49

1.00 .91 .88 1.08 .84 .63

2.44 2.34 2.84 2.71 2.37 2.54

1.07 .97 .99 1.06 .91 .67

2.48 2.24 2.70 2.22 2.25 2.38

.94 .94 .84 1.00 .76 .62

ARTICLE IN PRESS

Full Sample (n = 306)

Item

E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

Table 2 Means and standard deviations for CEESQ-Children’s emotions by teachers’ individual characteristics.

7

G Model

ARTICLE IN PRESS

EARCHI 799 1–11

655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679

.89 .79 .79 .94 .89 .77 .83 .80 .75 .87 .56

.93 .80 .98 .81 .59 1.93 1.63 3.35 2.03 2.23 .96 .81 1.16 1.01 .67 2.05 1.63 3.09 2.31 2.27 .98 .86 1.10 1.00 .64 2.07 1.70 3.23 2.25 2.31 .90 .80 1.08 .95 .63 1.82 1.56 3.04 2.18 2.15

M

3.31 3.17 3.22 3.24 3.32 3.47 3.47 3.58 3.49 3.57 3.37 .82 .79 .70 .86 .81 .89 .85 .82 .91 .87 .57

SD M

With high school degree (n = 190)

3.23 3.17 3.27 3.17 3.49 3.22 3.45 3.61 3.34 3.77 3.39 .78 .81 .76 .85 .93 .95 .86 .82 .86 .83 .56

SD M

3.19 3.06 3.17 3.17 3.36 3.27 3.42 3.58 3.28 3.66 3.32 .92 .82 .79 .94 .84 .80 .83 .85 .85 .95 .59

SD M

3.36 3.31 3.35 3.21 3.50 3.40 3.50 3.63 3.49 3.68 3.44 .83 .78 .72 .91 .82 .87 .88 .83 .85 .87 .54

1.01 .80 1.12 .98 .66 2.05 1.62 3.22 2.15 2.26 .81 .81 1.05 .87 .59

654

1.88 1.65 3.09 2.30 2.23

653

.95 .80 1.10 .95 .64

652

Zero-order correlations (Pearson’s r) between factors are reported in Table 5. Coaching turned out to be positively related to both Self-efficacy as Emotional Socializer (r = .46, p < .001) and Emotional Self-efficacy (r = .36, p < .001), and two other positive associations were found between Self-efficacy as Emotional Socializer and Emotional Self-efficacy (r = .54, p < .001), and between Dismissing and Denial of Emotion (r = .17, p < .01); Emotional Selfefficacy was negatively correlated with Denial of Emotion (r = −.30, p < .001). The above-presented correlation coefficients suggested that we should make an in-depth exploration of the CEESQ factors by examining the multivariate associations between Coaching, Self-efficacy as an Emotional Socializer and Emotional Self-efficacy. In particular, we tested whether Self-efficacy as an Emotional Socializer could be a mediator between Emotional Self-efficacy and Coaching style. Following a linear regression procedure, we first implemented a regression in which Self-efficacy as an Emotional Socializer was predicted by Emotional Self-efficacy (regression 1); we next implemented a regression in which Coaching was predicted by Emotional Self-efficacy (regression 2, step 1); finally we added a second step to the previous regression by including Self-efficacy as an Emotional Socializer in predicting Coaching (regression 2, step 2). See Fig. 1. The results revealed that all the considered paths were positive and significant (ˇs = .54,.36, .38, respectively, all p < .001); in particular, “Emotional Self-efficacy predicting Coaching” decreased after the introduction of Selfefficacy as an Emotional Socializer into the regression model, but it still remained positive and significant (ˇ = .15, p < .05); the relative Sobel test turned out to be significant (Sobel test = 5.49, p < .001), indicating a partial mediation effect. To sum up, in early childhood teachers the level of Emotional Self-efficacy was directly

2.00 1.62 3.18 2.20 2.25

651

3 5 7 11 Denial of emotion

650

SD

647

3.23 3.13 3.23 3.19 3.43 3.32 3.49 3.62 3.41 3.70 3.38

646

M

645

.88 .81 .78 .87 .91 .84 .76 .81 .86 .87 .62

644

SD

643

3.32 3.24 3.31 3.23 3.45 3.31 3.40 3.60 3.36 3.66 3.40

642

M

641

.85 .79 .74 .89 .85 .86 .84 .82 .85 .88 .56

640

SD

639

3.25 3.17 3.26 3.20 3.43 3.31 3.46 3.61 3.40 3.70 3.39

638

M

Step 5. Correlation and mediation analyses

637

8 12 4 9 2 10 15 6 13 1 Emotional self-efficacy

649

636

Item

648

experience × academic education level) MANOVAs were performed to explore differences in the CEESQ-Children’s Emotions and CEESQ-Individual Emotions factors. As for the CEESQ-Children’s Emotions factors, no main effects emerged: having children (Pillai’s Trace = .02; F(3, 230) = 1.31; 2 = .02; p > .05), years of teaching experience (Pillai’s Trace = .02; F(3, 230) = 1.36; 2 = .02; p > .05), and academic education level (Pillai’s Trace = .01; F(3, 230) = 1.03; 2 = .01; p > .05). No interaction terms emerged. Considering the CEESQ-Teacher’s Emotions, once again no main effects emerged for having own children (Pillai’s Trace = .00; F(2, 231) = .16; 2 = .00; p > .05), years of teaching experience (Pillai’s Trace = .01; F(2, 231) = .79; 2 = .01; p > .05), or academic education level (Pillai’s Trace = .00; F(2, 231) = .51; 2 = .00; p > .05). No interaction terms emerged.

635

Teaching experience greater than 12 years (n = 109)

.59

Teaching experience up to 12 years (n = 135)

.86

Cronbach’s alpha

With child/children (n = 203)

.14 −.24 .50 −.21 .51 .09 .56 −.12 .03 −.21 .39 −.14 .10 .30 −.24

Without child/children (n = 99)

Denial of emotion

.52 .54 −.07 .61 −.04 .68 −.03 .66 .65 .50 .03 .62 .66 .46 .51

Full Sample (n = 306)

Emotional self-efficacy

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Table 4 Means and standard deviations for CEESQ-Individual emotions by teachers’ individual characteristics.

Item

With university degree (n = 111)

Table 3 Factor loadings of EFA on CEESQ-Individual emotions.

SD

E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

8

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

G Model

ARTICLE IN PRESS

EARCHI 799 1–11

E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

9

Table 5 Descriptive statistics and zero-order correlations (Pearson’s r) of CEESQ dimensions in the full sample (n = 306). Latent factor

Numbers of items

1 – Coaching 2 – Self-efficacy as emotional socializer 3 – Dismissing 4 – Emotional self-efficacy 5 – Denial of emotion

Cronbach’s alpha

M

SD

Skew.

Kurt.

Expected range

Observed range

1

2

7 6

.77 .78

3.90 3.27

.54 .47

−.122 .449

−.325 .534

1.00–5.00 1.00–5.00

2.43–5.00 1.83–4.67

1.00 .46***

1.00

5 10

.70 .86

2.48 3.39

.66 .56

.059 .581

−.080 .303

1.00–5.00 1.00–5.00

1.00–4.60 1.90–5.00

.02 .36***

4

.59

2.25

.64

.390

.563

1.00–5.00

1.00–4.75

.10

.10 .54*** −.03

3

4

1.00 .11 **

.17

5

1.00 −.30***

1.00

Notes: ** p < .01; *** p < .001.

680 681 682 683

684

685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711

associated to the level of the adopted Coaching style; however, the level of Emotional Self-efficacy was indirectly and positively associated to Coaching style through the level of Self-efficacy as an Emotional Socializer. Discussion The main aim of the current study was to develop and test the psychometric properties of a self-reported rating scale for early childhood teachers. This was designed first to assess their emotionrelated beliefs (i.e., Self-efficacy as an Emotional Socializer) and emotional styles with regard to children’s emotions (i.e., Coaching and Dismissing styles), and second the attitude toward their own emotions. The EFA and CFAs performed on the CEESQ-Children’s Emotions scale revealed three factors: two emotional styles (i.e., Coaching and Dismissing), and Self-efficacy as an Emotional Socializer. Consistently with the literature about parents (Gottman et al., 1996, 1997; Lagacé-Séguin & Coplan, 2005), we identified the two emotional styles in early childhood teachers. We ascertained that teachers’ coaching style involves the awareness, acceptance, and regulation of children’s emotions, particularly the negative ones. Moreover, teachers’ dismissing emotions describes a style defined by a low awareness of children’s emotions, a view of negative emotions as inappropriate, and an attitude of ignoring or dismissing children’s emotions, above all the negative ones. Teachers’ self-efficacy referred to their perceived capability of managing the children’s emotional lives within the context of the day care center, specifically of differentiating emotional states, helping emotional comprehension, and managing the expression of positive and negative emotions in children. Effectively, the present study demonstrates that Coaching and Dismissing are two discrete but not antithetic styles. This evidence suggests that teachers can adopt each of them depending on the specific situation. In other words,

Self-efficacy as emotional socializer (1) .54***

Emotional self-efficacy

.38*** (3) .36*** / .15* (2) / (3)

Coaching style

(1) Regression 1: R = .29, ΔF(1,304) = 124.22, p < .001. (2) Regression 2, Step 1: R2 = .13, ΔF(1,304) = 44.34, p < .001. (3) Regression 2, Step 2: R2 = .23, ΔR2=.10, ΔF(1,303) = 45.09, p < .001. 2

Fig. 1. Mediation model between Emotional self-efficacy, Self-efficacy as emotional socializer and Coaching style in early childhood teachers.

the two styles cannot be considered either totally positive or negative. However, Coaching style is related to Self-efficacy as an Emotional Socializer, and where this relationship is found, this identifies a more satisfied teacher. The CEESQ-Individual Emotions revealed two distinct though related dimensions: Emotional Self-efficacy, namely the awareness and ability to deal with own emotions, and Denial of Emotion, namely a lack of acceptance of one’s own emotions. The relationship between these two dimensions is quite obviously a negative one. Importantly, both CEESQ sections demonstrated an equivalence in structure even when a number of individual characteristics were considered (i.e., having/not having own children, years of working experience, academic education level). As for our second aim, no effects emerged of teachers’ individual characteristics on the CEESQ dimensions. That is to say, having/not having children, professional experience, and schooling are individual experiences which do not influence such a personal construct as meta-emotion philosophy. It may be a construct which is not linked to the kind of experiences – parenthood, years of work, and academic level – considered in our study. Significant relationships emerged between the two parts of the CEESQ (i.e., on children’s and on teachers’ emotions) and the relative dimensions constituting the two scales. Specifically, early childhood teachers who were more aware of their own emotions and more likely to accept and regulate their own emotions, were also more likely to recognize, accept, and regulate the emotions of the children in the day care centers; on the other hand, a dismissing style was positively associated with denying one’s own emotions. These results are in line with the findings obtained by Ersay (2007) on preschool teachers, which, however, referred to teachers who teach older children. Our results are also consistent with the emotional connection between children’s and parents’ emotions theorized by Gottman et al. (1996, 1997) through their meta-emotion philosophy. Meta-emotion philosophy in parents as well as early childhood or preschool teachers may be an individual attitude that is not tied to relational variables like the age of the children they take care of, but is more individually modulated. Moreover, the tested mediation model led us to make an indepth exploration of the connection between the professional and individual emotional worlds. Coaching style was explained by professional (i.e., Self-efficacy as an Emotional Socializer) as well as individual emotional self-efficacy (i.e., Emotional Self-efficacy). In particular, our results suggested that a trait-like individual emotional self-efficacy influences how teachers act within their professional context (i.e., coaching style toward children’s emotions), but when we added a measure of professional self-efficacy it did not completely explain the professional behavior, because individual emotional self-efficacy continues to play a significant role. To conclude, the current results are an advancement on

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761

G Model EARCHI 799 1–11 10 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811

812

813 814 815 816 817 818

819

820 821 822

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

past work as they show that early childhood teachers’ emotional experience processes (i.e., awareness and regulation of own emotions) may play an important role in teaching situations and professional lives. However the cross-sectional nature of our data prevents us to assume causal effects, and the above mentioned relationships should be tested in future longitudinal studies. We are aware that these results need to be interpreted in light of some limitations. First, exploratory and confirmatory factor analyses were both conducted on the same dataset and this may represent a potential limitation. Moreover, concurrent and discriminant validity were not tested, nor was temporal stability; future studies could fill this gap. Further, all different constructs were assessed in one measurement moment, using the same selfreport instrument, and as we specified above causation between CEESQ-scales cannot be tested. Finally, we are aware that beliefs and practices are “culturally loaded” (Hyson & Lee, 1996) and context-dependent, so our findings request further cross-cultural inquiries. It may be that these beliefs and practices are also gender-loaded, hence future studies could add information on male teachers. Nevertheless, this study forms a first step in investigating the meta-emotion philosophy of early childhood educators and a starting point for future research. For instance, the items in the CEESQ were formulated using the broad terms “child” or “children,” without reference to their gender or age; according to Ahn and Stifter (2006), both of these are important when considering beliefs and behaviors about children, and future works could include them. Moreover, it would be interesting to investigate the impact of early childhood teachers’ meta-emotion philosophy on children’s outcomes (e.g., emotional, social and academic functioning) as well as on parents’ and colleagues’ behavior. To conclude, we believe that a self-report instrument such as the CEESQ questionnaire may provide an in-depth set of information about our own emotional worlds: even though self-reports are criticized for being subject to response biases, we agree with many theorists in stressing that self-perception about one’s own functioning and limitations may exert a significant influence on both behavior and psychological health, and constitutes a critical component required in the motivation to change (Bandura, 1986, 1997; Zimmerman, 2001). Administering the CEESQ could be useful to evaluate teachers’ meta-emotion philosophy and in training sessions aimed at early childhood teachers. CEESQ scores could work as a prompt in order to discuss and mentalize teachers’ emotional work as emotional socializers. We foster a kind of training specifically aimed at teachers’ emotional competence (e.g., reflective supervision, mindfulness, stress reduction; Emde, 2009). Paying attention to the contribution of the early childhood teacher as an emotional socializer could make childcare settings more favorable to children’s emotional growth (Denham et al., 2012). Acknowledgments The authors would like to thank all early childhood teachers for their participation in this study. They also acknowledge the valuable assistance of (in alphabetical order) Elisabetta Bernardini, Francesca Cocco, Francesca Grandi, Alessandra Nardelli, Giulia Piccardi, Roberta Salti and Elisabetta Savelli for data collection and coding. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ecresq.2015.04. 006

References Ahn, H. J. (2005). Teachers’ discussions of emotion in child care centers. Early Childhood Education Journal, 32, 237–242. http://dx.doi.org/10.1007/s10643004-1424-6 Ahn, H. J., & Stifter, C. (2006). Child care teachers’ response to children’s emotional expression. Early Education and Development, 17, 253–270. Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52, 317–332. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: W.H. Freeman. Barker, C., Pistrang, N., & Elliott, R. (1994). Research methods in clinical and counseling psychology. Chichester, UK: Wiley. Bartlett, M. S. (1950). Tests of significance in factor analysis. British Journal of Psychology, 3, 77–85. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Bentler, P. M. (1995). EQS structural equations program manual. Encino, CA: Multivariate Software. Blau, D. M. (2000). The production of quality in child-care centers: Another look. Applied Developmental Science, 4, 136–148. Boivin, M., Pérusse, D., Dionne, G., Saysset, V., Zoccolillo, M., Tarabulsy, G. M., et al. (2005). The genetic-environmental etiology of parents’ perceptions and selfassessed behaviours toward their 5-month-old infants in a large twin and singleton sample. Journal of Child Psychology and Psychiatry, 46, 612–630. Bornstein, M. H. (Ed.). (2002). Handbook of parenting (Vol. 1–4). Mahwah, NJ: Erlbaum. Burchinal, M., Howes, C., & Kontos, S. (2002). Structural predictors of child care quality in child care homes. Early Childhood Research Quarterly, 17, 87–105. Byrne, B. M. (1994). Structural equation modeling with EQS and EQS/Windows. Thousand Oaks, CA: Sage Publications. Cattel, R. B. (1966). The scree test for the number of factors. Multivariate Behavior Research, 1, 245–276. Coleman, P. K., & Karraker, K. H. (1997). Self-efficacy and parenting quality: Findings and future applications. Developmental Review, 18, 47–85. Coleman, P. K., & Karraker, K. H. (2003). Maternal self-efficacy beliefs, competence in parenting, and toddlers’ behavior and developmental status. Infant Mental Health Journal, 24, 26–148. Costello, A. B., & Osborne, J. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment Research & Evaluation, 10. Retrieved from: http://pareonline.net/getvn. asp?v=10&n=7 Day, C., & Leitch, R. (2001). Teachers’ and teacher educators’ lives: The role of emotion. Teaching and Teacher Education, 17, 403–415. Denham, S. A., Bassett, H. H., & Zinsser, K. (2012). Early childhood teachers as socializers of young children’s emotional competence. Early Childhood Education Journal, 40, 137–143. http://dx.doi.org/10.1007/s10643-012-0504-2 de Schipper, E. J., Riksen-Walraven, J. M., Geurts, S. A. E., & Derksen, J. J. L. (2008). General mood of professional caregivers in child care centers and the quality of caregiver–child interactions. Journal of Research in Personality, 42, 515–526. Emde, R. N. (2009). Facilitating reflective supervision in an early child development center. Infant Mental Health Journal, 30, 664–672. http://dx.doi.org/10. 1002/imhj.20235 Ersay, E. (2007). Preschool teachers’ emotional experience traits, awareness of their own emotions and their emotional socialization practices. Pennsylvania State University. Retrieved from: https://etda.libraries.psu.edu/paper/7660/2957 Fairchild, A., & McQuillin, S. (2010). Evaluating mediation and moderation effects in school psychology: A presentation of methods and review of current practice. Journal of School Psychology, 45, 54–84. Gottman, J. M., Katz, L. F., & Hooven, C. (1996). Parental meta-emotion philosophy and the emotional life of families: Theoretical models and preliminary data. Journal of Family Psychology, 10, 243–268. Gottman, J. M., Katz, L. F., & Hooven, C. (1997). Meta-emotion: How families communicate emotionally. Mahwah, NJ: Erlbaum. Hakim-Larson, J., Parker, A., Lee, C., Goodwin, J., & Voelker, S. (2006). Measuring parental meta-emotion: Psychometric properties of the emotion-related parenting styles self-test. Early Education and Development, 17, 229–251. Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: Examples from the child–clinical and pediatric psychology literatures. Journal of Consulting and Clinical Psychology, 65, 599–610. Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural equation modeling: Guidelines for determining model fit. The Electronic Journal of Business Research Methods, 6, 53–60. Hooven, C., Gottman, J. M., & Katz, L. F. (1995). Parental meta-emotion structure predicts family and child outcomes. Cognition and Emotion, 9, 229–264. Huntsman, L. (2008). Determinants of quality in childcare: A review of the research evidence. In Centre for Parenting and Research. Ashfield, NSW: Department of Community Services. Hyson, M. (2002). Emotional development and school readiness: Professional development. Young Children, 57, 76–78. Hyson, M. C., & Lee, K. (1996). Assessing early childhood teachers’ beliefs about emotions: Content, contexts, and implications for practice. Early Education and Development, 7, 59–78. Innocenti Research Centre. (2008). The child care transition: A league table of early childhood education and care in economically advantaged countries. Florence,

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

823

824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907

G Model EARCHI 799 1–11

ARTICLE IN PRESS E. Ciucci et al. / Early Childhood Research Quarterly xxx (2015) xxx–xxx

908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946

Italy: United Nations Children’s Fund. Retrieved from: http://www.unicefirc. org/publications/pdf/rc8.eng.pdf Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 31–36. Katz, L. F., & Windecker-Nelson, B. (2004). Parental meta-emotion philosophy in families with conduct-problem children. Journal of Abnormal Child Psychology, 32(4), 385–398. Lagacé-Séguin, D. G., & Coplan, R. J. (2005). Maternal emotional styles and child social adjustment: Assessment, correlates, outcomes and goodness of fit in early childhood. Social Development, 14, 613–636. Layzer, J. I., & Goodson, B. D. (2006). The ‘quality’ of early care and education settings. Evaluation Review, 30, 556–576. Lunkenheimer, E. S., Shields, A. M., & Cortina, K. S. (2007). Parental emotion coaching and dismissing in family interaction. Social Development, 16, 232–248. Manlove, E. E., Vazquez, A., & Vernon-Feagans, L. (2008). The quality of caregiving in child care: Relations to teacher complexity of thinking and perceived supportiveness of the work environment. Infant and Child Development, 17, 203–222. Marsh, H. W., Balla, J. R., & Hau, K. T. (1996). An evaluation of incremental fit indexes: A clarification of mathematical and empirical properties. In G. A. Marcoulides, & R. E. Schumacker (Eds.), Advanced structural equation modeling techniques (pp. 315–353). Mahwah, NJ: Erlbaum. Mayer, J. D., Salovey, P., Caruso, D. R., & Sitarenios, G. (2001). Emotional intelligence as a standard intelligence. Emotion, 1, 232–242. http://dx.doi.org/10.1037//15283542.1.3. 232-242 Meyer, D. K., & Turner, J. C. (2007). Scaffolding emotions in classrooms. In P. A. Schutz, & R. Pekrun (Eds.), Emotion in education (pp. 243–258). San Diego, CA: Elsevier. Mill, D., & Romano-White, D. (1999). Correlates of affectionate and angry behavior in child care educators of preschool-aged children. Early Childhood Research Quarterly, 14, 155–178. NICHD Early Child Care Research Network. (1996). Characteristics of infant child care; factors contributing to positive caregiving. Early Childhood Research Quarterly, 11, 296–306. NICHD Early Child Care Research Network. (2000). Characteristics and quality of child care for toddlers and preschoolers. Applied Developmental Science, 4, 116–135. Penrose, A., Perry, C., & Ball, I. (2007). Emotional intelligence and teacher self efficacy: The contribution of teacher status and length of experience. Issues In Educational Research, 17, 107–126. Perry, C., & Ball, I. (2008). Identifying the underlying dimensions of teachers emotional intelligence. Problems of Education in the 21st Century, 7, 89–98. http://dx. doi.org/10.1177/1025382307088091

11

Phillips, D., & Adams, G. (2001). Child care and our youngest children. The Future of Children, 11, 35–51. Retrieved from: http://www.jstor.org/stable/1602808 Ramsden, S. R., & Hubbard, J. A. (2002). Family expressiveness and parental emotion coaching: Their role in children’s emotion regulation and aggression. Journal of Abnormal Child Psychology, 30, 657–667. Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48, 1–36. Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the fit of structural equation models: Test of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23–74. Schutz, P. A., & Zembylas, M. (2009). Introduction to advances in teacher emotion research: The impact on teachers’ lives. In P. A. Schutz, & M. Zembylas (Eds.), Advances in teacher emotion research: The impact on teachers’ lives (pp. 3–11). Dordrecht, The Netherlands: Springer. Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290–312. Soper, D. S. (2013). Sobel test calculator for the significance of mediation [Software]. Retrieved from: http://www.danielsoper.com/statcalc Steiger, J. H., & Lind, J. M. (1980). Statistically based tests for the number of common factors. Paper presented at the annual meeting of the Psychometric Society, Iowa City, IA. Sutton, R. E. (2004). Emotional regulation goals and strategies of teachers. Social Psychology of Education, 7, 379–398. Sutton, R., & Wheatley, K. (2003). Teachers’ emotions and teaching: A review of the literature and directions for future research. Educational Psychology Review, 15, 327–358. http://dx.doi.org/10.1023/A:1026131715856 Teti, D. M., & Gelfand, D. M. (1991). Behavioral competence among mothers of infants in the first year: The mediational role of maternal self-efficacy. Child Development, 62, 918–929. Winograd, K. (2003). The functions of teacher emotions: The good, the bad and the ugly. Teachers College Record, 105, 1641–1673. Yin, H., & Lee, J. C. (2012). Be passionate, but be rational as well: Emotional rules for Chinese teachers’ work. Teaching and Teacher Education, 28, 56–65. Zembylas, M. (2007). Theory and methodology in researching emotions in education. International Journal of Research and Method in Education, 30, 57–72. Zimmerman, B. J. (2001). Theories of self-regulated learning and academic achievement: An overview and analysis. In B. J. Zimmerman, & D. H. Schunk (Eds.), Self-regulated learning and academic achievement: Theoretical perspectives (2nd ed., Vol. 1–4, pp. 1–37). Mahwah, NJ: Erlbaum.

Please cite this article in press as: Ciucci, E., et al. Meta-emotion philosophy in early childhood teachers: Psychometric properties of the Crèche Educator Emotional Styles Questionnaire. Early Childhood Research Quarterly (2015), http://dx.doi.org/10.1016/j.ecresq.2015.04.006

947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985