Pedagogical and Social Climate in School Questionnaire

Method

This study was carried out in 10 schools in Western and Southern Bulgaria (post-communist country in Eastern Europe, bordering with Romania, Greece, Macedonia, Serbia, Turkey, and the Black Sea) with a sample of 307 teachers. Thirty-two percent of the teachers had a bachelor’s degree in education, 68% a master’s degree, and 0, 7% high school degree. Eighty-three percent were teachers at a high school level and 17% at middle school. The selected schools were representative of geographically and ethnically mixed populations across the country and were sampled as part of a larger ongoing study in these regions. All contacted schools agreed to participate, and the teachers’ response rate was very high (up to 95%) due to long-term collaboration with the research team and incentives to participants. PESOC is composed of 67 items rated on a 4-point Likert-type scale (completely agree–completely disagree), which form 11 subscales (see Table 1). Items were translated from English into Bulgarian by four bi-lingual speakers while adhering to standard guidelines to ensure linguistic equivalence by adopting a culturally sensitive translation of items in a committee approach to assure good translation, language features, and cultural applicability of item contents in the local culture (van de Vijver & Leung, 1997). In addition, two separate focus groups of teachers were involved in assessing the appropriateness of the translated items. Teachers filled in PESOC in the school and were assisted by four trained research assistants.

OPEN IN VIEWER

Table 1.

Sample.

	Descriptives
Age, M (SD)	46.79 (9.15)
Gender, n
Male	53
Female	245
Teacher PESOC subscale, M (SD)
Student Expectations	3.18 (0.52)
Unity	3.27 (0.41)
Student Approach	3.23 (0.42)
Basic Assumptions	3.15 (0.52)
School Home	3.11 (0.48)
Teacher Collaboration	3.19 (0.43)
Teacher Development	3.21 (0.52)
Teaching	3.09 (0.50)
Student Evaluation	3.19 (0.52)
Principal’s Leadership	3.22 (0.43)
School Management	3.30 (0.56)

Note. n = total number; PESOC = Pedagogical and Social Climate in School.

Results and Discussion

Preliminary analyses tested for normality and missing data. At the scale level, kurtosis and skewness values were approximately normal and in the range of ±1. At the item level, we found missingness in the range of 1% to 6%. We used full-information maximum likelihood (FIML) estimation (Enders & Bandalos, 2001). We therefore performed the following steps. First, we examined reliability of PESOC using coefficient alpha (recommended value > .70) at scale and overall levels (Cronbach & Shavelson, 2004). The analyses indicated good internal consistency values for each scale and the overall score (see Table 2). Second, bivariate Pearson correlations among all PESOC scales were explored showing that all subscales correlated positively and significantly (Table 3). Third, we estimated a uni-dimensional model of PESOC via CFA in a structural equation model using Mplus (Muthén & Muthén, 2010). Model fit was tested with the Comparative Fit Index (CFI; recommended value > .90) and the root mean square error of approximation (RMSEA; recommended value < .08; Hu & Benter, 1999). The CFA testing a single-factor model controlling for the clustered nature of data by school showed a good fit, χ²(44, N = 306) = 125.36, p < .001, CFI = .941, and RMSEA = .078. Because our sample’s characteristics assume an analysis with clustered data, we used the TYPE=COMPLEX command in Mplus, which uses a sandwich estimator considered in the main substantive analyses that teacher responses are grouped (e.g., by school), and that responses are potentially interdependent. Clustered data are those that can be classified into a number of distinct groups within a study, and our analysis considers the fact that it is critical to take data clustering into account (Sally, James, & Bryce, 2010). In addition, in line with prior work (Hultin et al., 2015) and to provide a stronger evidence of PESOC factorial structure, we tested a two-factor model (Factor 1 = expectations for students, unity among teachers, approach to students, basic assumptions about students’ ability to learn, school–home relations, teacher cooperation, teachers’ professional development, teaching activities, evaluation of students; Factor 2 = principal’s pedagogical leadership and school management). This model showed a slightly improved and acceptable fit with a bi-dimensional factor structure, χ²(43, N = 306) = 106.79, p < .001, CFI = .954, and RMSEA = .070 (Figure 1). The two latent factors were significantly correlated, and the magnitude of the correlation was high (.77), which suggests a strong relation between the two school climate indicators (Tabachnick & Fidell, 2001). We also performed chi-square difference test, indicating that the second “larger” model with more freely estimated parameters fits the data better than the first “smaller” model (Kline, 1998). Therefore, we opted for a two-factor model also in line with prior psychometric investigations on the instrument (Hultin et al., 2015).

OPEN IN VIEWER

Table 2.

Scales, Item Examples, and Internal Consistency Coefficients of Teacher PESOC.

Scale (total items)	Example item	Coefficient alpha
1. Student Expectations (4)	Our principal has high demands and expectations for pupil academic results	.78
2. Unity (10)	There is a high degree of unity among teachers regarding the schools goals	.88
3. Student Approach (6)	It is a well-established policy that teachers show respect for pupils	.78
4. Basic Assumptions (3)	It is well known by the staff that knowledge goals are the most important task of this school	.70
5. School Home (6)	Parents are always informed if a child behaves badly in school	.83
6. Teacher Collaboration (8)	I have the support of my colleagues in selecting teaching content and methods	.88
7. Teacher Development (4)	In this school, you feel that you can develop as an educator	.87
8. Teaching (6)	There is a policy in this school that no pupil will leave this school without gaining basic knowledge and skills	.84
9. Student Evaluation (2)	Regular evaluation of pupil results is used as the basis for planning teaching in this school	.73
10. Principal’s Leadership (13)	In this school, the principal has the final responsibility for teaching quality	.91
11. School Management (3)	There is always a member of school management accessible to discuss and give advice on issues regarding teaching and education	.92
Overall		.97

Note. n = total number; PESOC = Pedagogical and Social Climate in School.

OPEN IN VIEWER

Table 3.

Correlations Among Teacher PESOC Subscales.

Scale	1	2	3	4	5	6	7	8	9	10	11
1. ES	—
2. U	.60	—
3. AS	.51	.61	—
4. BA	.50	.53	.60	—
5. SH	.49	.56	.64	.53	—
6. TC	.49	.60	.61	.51	.64	—
7. TD	.51	.65	.63	.51	.62	.69	—
8. T	.53	.55	.60	.53	.59	.61	.68	—
9. SE	.46	.50	.51	.50	.55	.56	.61	.74	—
10. PL	.57	.56	.58	.47	.49	.54	.51	.57	.58	—
11. SM	.43	.39	.37	.24	.33	.45	.41	.38	.42	.60	—

Note. ES = Expectations for Students; U = Unity Among Teachers; AS = Approach to Students; BA = Basic Assumptions About Students’ Ability to Learn; SH = School–Home Relations; TC = Teacher Cooperation; TD = Teachers’ Professional Development; T = Teaching Activities; SE = Students’ Evaluation; PL = Principal’s Leadership; SM = School Management.

All correlations are significant at p < .01.

OPEN IN VIEWER

Figure 1.

Estimates and residual variance of the two-factor model of the teacher PESOC.

Adaptation of the school climate measures to different languages is needed to allow valid evaluations and cross-cultural comparisons. This study is the first to examine the psychometric properties of PESOC in a group of individuals who speak Bulgarian. This is the first study of the factorial validity and reliability of PESOC with data from different schools in an international context (e.g., Bulgaria) outside of the country where the instrument was developed. Consistent with earlier investigations in Sweden (Hultin et al., 2015), a bi-dimensional factorial structure provided good fit to the data. We were also able to confirm a high level of internal consistency at single and overall scale levels, and that school climate characteristics are significantly and positively related to each other.

Although the Bulgarian version of PESOC demonstrated adequate reliability and factor validity for research purposes, some limitations need to be acknowledged. This is a first-time study in Bulgaria, and replications of our findings are needed prior to generalizing results further in a Bulgarian or any other international school context. Future studies could also explore additional indicators of test–retest stability and predictive validity. In addition, multi-method studies of PESOC (e.g., teacher and student ratings) are important to complement these teacher-reported only study findings. Future research would also extend school climate analysis to mental health outcomes of youth. Given strong evidence that the severity of the mental health problems in the population is related to school setting (Bond et al., 2007), there is need to carry out further studies with a wide measurement net that includes important educational and public health outcomes. Additional promising extension of the present study would be to link school climate domains to teachers’ perceived school satisfaction, which would yield richer information for schools (Zullig et al., 2010).

Nevertheless, this study provides new and preliminary evidence for the validity, reliability, and utility of this scale in a Bulgarian-speaking sample. The Bulgarian version of PESOC adds to the existing school climate literature and provides diversity to a research pool that has been dominated by Western samples. School climate factors can affect student academic success and positive socio-emotional developmental outcomes (Cohen, McCabe, Michelli, & Pickeral, 2009). Measurement tools on these factors such as the one detailed here play important role for educational policy and teaching practices. Therefore, PESOC may be useful in school climate–related interventions.