1. Introduction
Findings indicating that at least 20% of teachers in the USA, the UK, China, Australia, and several European countries leave their profession in the first three years of their career [
1,
2]. The negative emotional and psychological factors experienced in the organization are noteworthy among the reasons for leaving the teaching profession [
3]. Previous studies have argued that teachers have a moderate quality of work life [
4] and moderate job satisfaction [
5,
6]. Moreover, considerable amount of workload for teachers results in a high level of stress and potential for burnout [
7,
8,
9].
The literature suggests a relation between organizational happiness of preschool teachers and their capacity to offer quality education [
10]. Also, teachers’ organizational happiness reflected on children [
11]. Moreover, organizational happiness of teachers may affect the perception of social justice in society [
12]. Furthermore, the quality of work life directly affects job satisfaction, motivation, and the quality of work performed by teachers [
13]. Previous studies have emphasized that further research is essential to guide the reinforcement of teachers’ organizational happiness in practice [
14].
Sustainable schools can be defined as the product of the processes involving followers’ happiness and their joint participation in the leadership roles [
15]. In other words, this approach assumes that leadership, in this context, consists of the reciprocal learning processes that enable followers to construct and negotiate meanings leading to a shared purpose of schooling. Leadership is also concerned with learning that facilitates constructive change [
16]. Learning has a direction toward a shared purpose, and change in the school context is a collective endeavor; therefore, people can achieve this more effectively in the presence of colleagues. The learning process within an organization must be shared; otherwise, shared purposes and actions are never achieved. Leadership requires the distribution of power and authority [
16,
17]. Shared learning, purposes, actions, and responsibilities necessitate the realignment of power and authority. Ultimately, it is stated that distributed leadership can be a beneficial approach for achieving sustainable purposes and actions in schools through positively affecting organizational happiness [
18].
Previous studies also argue that sustainable school conditions are associated with teacher well-being [
19,
20]. Considering that organizational happiness and quality of work life are two indicators of teacher well-being in an organization, perceiving high levels of organizational happiness and quality of work life by the teachers may facilitate sustainable education environment [
20]. Given these findings, two points gain significance: First, determining ways to support the organizational happiness and quality of work life variables in the organizational environment. Second, identifying ways to provide such support to teachers at higher levels. According to [
21], as cited from the study by [
22], a collective administration of leadership processes by the actors in the school rather than a single person is recommended to provide a healthy school environment. In this context, the distributed leadership approach comes to the fore. Numerous studies on distributed leadership in the literature have attracted our attention [
23,
24,
25,
26,
27,
28,
29,
30]. These studies have argued that the distributed leadership approach can affect job satisfaction [
31,
32], organizational trust [
33], student achievement [
34], and teacher effectiveness [
35]. We believe that the examination of the effect of the distributed leadership approach of administrators on the organizational happiness and quality of work life of teachers would be useful in this context.
The collegial approach is one of the theoretical perspectives that explains organizational behavior in educational institutions [
36]. Major assumptions of this approach are decision making based on a process of discussions, agreements and consensus and sharing the power among the members of the organization who are considered to have a joint perception of the organizational aims [
37]. Distributed leadership is related to the collegial approach, and has been the focus of attention of scholars in the 21st century [
37]. This leadership style, as mentioned in [
36], is one of the most significant approaches within the context of educational leadership in the past decade. In the context of organizational behavior in educational institutions, distributed leadership is an approach in which collaborative working is undertaken among followers who trust and respect each other’s contribution and happens most effectively when people at all levels engage in action. However, it is also considered that the collegial approach is strongly normative and may not have a positive contribution to teachers and other employees in real school life [
38]. Furthermore, the positive [
23,
30,
35] and negative [
24,
25,
26,
27,
28,
29] effects of distributed leadership within organizations are discussed in the literature. It is believed that an examination of the effects of the distributed leadership approach on the organizational happiness and quality of work life of teachers will be useful in order to test the validity of both collegial theory and distributed leadership in real school life in the context of organizational behavior.
According to [
39], in organizations where a distributed leadership approach is adopted, it is found that the organizational happiness perceptions of employees are high depending on the responsibility of the employees. In addition, it was stated in other study [
40] that individuals who are not happy in their working environment have low perceptions of quality of work life. In the literature, it is concluded that distributed leadership positively affects organizational happiness and there is a positive correlation between organizational happiness and quality of work life. Based on these findings, it is thought that distributed leadership may affect the quality of work life of followers through organizational happiness. In this context, it is argued that the structural model where organizational happiness is a mediating variable in the possible relationship between distributed leadership and quality of work life can be tested.
This study mainly aims to go one step ahead in investigating the linear relations among the variables of the distributed leadership characteristics of administrators, quality of work life of employees, and organizational happiness of employees in the literature [
5,
31,
33,
41]. The study explains how these linear relations are realized, in other words, how the independent variable predicts the dependent variable. Furthermore, it argues that researching the processes through which the relations among the variables are realized would make a theoretical contribution to the literature. The mediation effect of organizational happiness on the possible relation between distributed leadership and quality of work life can be tested through structural modeling with mediating variable created to examine the situation. After these arguments, the first and second research questions are formulated to be tested.
In the literature, the notion of distributed leadership was interpreted in the light of descriptive and correlational findings. However, no studies have comprehensively investigated how these findings are reflected in practice [
25,
26,
27,
28,
29,
31,
32,
33,
34,
35,
42]. However, conducting in-depth qualitative studies on how distributed leadership characteristics is practiced can lead to a better understanding of the distributed leadership within organizational life. For this reason, the third research question is formulated.
Based on these concerns, we conclude that a mixed methods approach to distributed leadership could contribute to a better understanding of the actors involved in the organizational structure. This study investigates distributed leadership, organizational happiness, and the quality of work life in preschools. In this context, we investigated the answers to the following research questions:
- (1)
Do school administrators’ distributed leadership characteristics predict teachers’ organizational happiness and quality of work life?
- (2)
Is there a mediating effect of the organizational happiness of teachers on the relation between the school administrators’ distributed leadership characteristics and teachers’ quality of work life?
- (3)
What are school administrators’ thoughts on the effects of their behaviors on organizational happiness and quality of work life of teachers?
2. Materials and Methods
Adopting a sequential explanatory design, we utilized a mixed methodology where the quantitative paradigm has a dominant and the qualitative paradigm has a supplemental status in data collection, analysis, and interpretation.
2.1. Universe, Sample, and Study Group
The universe of this study comprised the preschools in Northern Cyprus (TRNC) as there have been no previous leadership studies conducted in preschools in the TRNC. As a result of the limited number of studies, it is evident that more in-depth research should be conducted to explore the relationships between teachers and administrators in preschools [
43]. It is believed that a large proportion of the inter-organizational experiences and relations between the administrators and teachers at preschools in TRNC remains unexplored [
43]. Furthermore, in the TRNC teachers’ law, only the tasks expected from teachers are specified and the articles of this law are based on the fulfillment of unilateral orders [
44]. There is no provision that supports an environment of cooperation and trust within the organization and is based on sharing the duties and responsibilities of school administrators with teachers. In the light of these arguments, conducting this research in TRNC preschools may facilitate the understanding of the actual situation in kindergartens in the context of leadership.
It was thought that this research, which focuses the relationships between administrators and teachers, should be evaluated on the basis of data obtained from both administrators’ and teachers’ perspectives. As previously mentioned, in this research, distributed leadership was one of the variables investigated. In the quantitative section, some of the data were collected through a distributed leadership scale that was only administered to the preschool teachers. The reason that the school principals were not invited to participate in the quantitative part was that some of the administrators may have had a tendency to give socially desirable responses rather than accurate answers about themselves. It was thought that this could have violated the assumption of “participants giving honest and sincere responses to measurement instruments. In terms of the TRNC universe, it was argued that reaching all the preschool teachers would be more qualified in terms of the generalizability of the research results. The first and second research questions were correlational, and according to time, accessibility and economic criteria, the data were obtained via the quantitative method as it was difficult to reach the whole preschool teacher universe by using the qualitative method. Finally, to answer the third research question, teachers were not included in this section, since it was necessary to examine the administrative behaviors in greater depth.
No sampling was performed for the quantitative component of the study, and all class teachers and subject-matter teachers teaching at preschools affiliated with the TRNC Ministry of National Education in the 2018–2019 academic year (N = 290) were invited to participate in the study. Two hundred and forty-one teachers voluntarily participated in the study; however, data of only 208 teachers were suitable for analysis. Teachers and administrators at private schools were not included in the study because those working at private schools could refrain from making negative comments about the institution they work at. This could have violated the assumption of “participants giving honest and sincere responses to measurement instruments”.
A large number of the teachers who participated in the study were in the 31–40 age group (38.9%), followed by those in the 21–30 (28.4%) and those in the 41–50 (24.0%). Most of the participants were females (76.4%), and males were the minority (23.6%). Majority of the participants (81.3%) had a bachelor’s degree. However, a small number of teachers (1.9%) had an associate degree (
Table 1).
In terms of the schools they graduated from, most teachers (73.1%) were graduates of Atatürk Teacher Training Academy and the rest (26.9%) had graduated from departments at private universities in Turkey and TRNC. Of the participants, 26.4% had 6–10 years of teaching experience, 23.1% had 11–15 years of experience, and 7.2% had 26 years of experience and above. The study group in the qualitative component of the study was determined through stratified sampling. Based on stratified sampling, participants were included in the study from six districts of TRNC considering their availability and accessibility.
Similarly, two administrators from each district (total 12: 7 females and 5 males) were included in the study, with whom a discussion was held during to the researchers’ visits to schools for the quantitative data collection, and their consent for participating in the study was obtained. The ages of the school administrators varied between 45 and 54 years. They had between 2 and 16 years of experience as school administrators. Two participants had graduated from the department of child development, and 10 participants had graduated from the department of classroom teaching (
Table 2).
2.2. Data Collection Tools and Data Collection
The first and second research questions were investigated through quantitative data, which was collected through three scales with the participation of the teachers. A total of three scales were given to those teachers who were willing to participate in the research voluntarily; one scale measuring the distributed leadership perceptions of teachers about their own school administrators, and two scales measuring their perceptions of quality of work life and organizational happiness. The third research question was investigated through interviews with the school administrators. Administrators’ thoughts on the effects of their behaviors on organizational happiness and quality of work life of teachers were examined.
The Distributed Leadership Scale was developed by [
45]; Quality of Work Life Scale developed by [
46] and adapted into Turkish by [
47]; and Organizational Happiness Scale was developed by [
48]. 868 teachers from Gaziantep participated in the development of the Distributed Leadership Scale. Expert opinion was sought for content and face validity. Explanatory factor analysis (EFA) and confirmatory factor analysis (CFA) were used for construct validity. EFA revealed that the scale explained 62.5% of the total variance and comprised five factors, i.e., formal structure, unity of purpose, collaboration and trust, sharing responsibilities, and incentive and initiative. CFA revealed that all goodness of fit values ((Chi-square fit index (χ2), Chi-square fit index/degree of freedom (χ2/df), Standardized Root Mean Square Residual (SRMR), Comparative Fit Index (CFI), Goodness of Fit Index (GFI), Adjusted Goodness of Fit Index (AGFI), Root Mean Square Error of Approximation (RMSEA), Root Mean Square Residual (RMR), Normed Fit Index (NFI), and Non-normed Fit Index (NNFI)) were acceptable [
49]. In the validity analysis, Cronbach’s alpha and split-half reliability test coefficients were 0.87 and 0.92, respectively. Even though the draft scale included 52 items, 20 items were removed on the basis of EFA, CFA, and reliability analyses. The final form of the scale comprised 32 items, 1 of which was a reverse item. The five-point Likert-type scale was scored between totally disagree (1) and totally agree (5). The arithmetic means of all responses to the items were taken as the score obtained on the scale [
45].
Regarding the Quality of Work Life Index developed by Van Laar, Edwards, and Easton (2007) and adapted into Turkish by [
47], 31 English teachers in Kilis participated in the linguistic validity study, 324 teachers participated in the construct validity study, and 118 teachers participated in the test-retest reliability study. To ensure linguistic validity, the items were translated from English into Turkish and then from Turkish into English. The correlation coefficient between the Turkish form and English form was 0.77. In addition, based on the results of the EFA, the scale was used without removing any items from the English form and changing the number of factors. The results of the CFA indicated that goodness of fit values were acceptable. In the reliability analysis, the reliability coefficient for each sub-dimension was higher than 0.7, and the Cronbach’s alpha coefficient of the whole scale was 0.95. Accordingly, there were 6 factors, i.e., satisfaction with job and career (6 items), overall well-being (6 items), controlling work (3 items), working conditions (3 items), stress at work life (2 items), work–family life balance (3 items), and 23 items (3 of which were reverse items). The arithmetic means of the responses to the items in the five-point Likert-type scale were interpreted as the score obtained from the scale [
47].
An item pool was created in the first phase of the scale development study for the Organizational Happiness Scale, which was developed to measure organizational happiness in the school environment, and a draft scale with 52 items was formed by considering expert opinions. The items with a factor load below 0.3 were removed from the scale based on the EFA conducted with 352 teachers. In addition, a structure comprising 38 items and 5 factors, including management processes, professional attitude, communication with stakeholders, professional commitment, and economic conditions, explained 51% of the total variance. The following CFA indicated that goodness of fit values were good and acceptable. Cronbach’s alpha coefficient was used to measure reliability. Cronbach’s alpha coefficient for each sub-dimension was higher than 0.7, and Cronbach’s alpha coefficient for the whole scale was 0.92. The arithmetic means of the responses to the items in the five-point Likert-type scale were accepted as the score obtained from the scale [
48].
The qualitative data of the study were collected through an interview form comprising three semi-structured questions developed by the researcher.
2.3. Validity and Reliability
Prior to the data analysis, the model fit between the determined structure of the scales and Turkish Cypriot culture was examined for assessing the construct validity. Due to the lack of a large research universe (N = 208), the samples in the CFA and the implementation were not different; instead, all the data from the universe underwent CFA, and data analysis started consequently. Because the CFA and reliability analysis on the scales were prerequisites for structural equation modeling (SEM), these analyses are detailed in the findings section under the heading “testing measurement models.”
Expert opinion was consulted to ensure the validity and reliability of the interview form, and the form was reshaped on the basis of these opinions. To measure the reliability of the form, the inter-rater reliability rate was calculated on the basis of
) the formula by [
50]. The analysis revealed that the inter-rater reliability rate was 100%. The interview form got its final form after the pilot study; no guidance was provided during the interviews to ensure validity; and direct excerpts from the interviews were used to represent meaningful data while interpreting the findings to enhance reliability. The interviews, which took 15 minutes on average, were recorded using a voice recorder to ensure no data loss.
2.4. Data Analysis
SEM was used to test the research hypotheses in the quantitative component of the study. In this approach, a causal and relational structure is formed between the observed and latent (unobservable) variables. This structure is defined as a model and is analyzed using multivariate statistical methods [
49]. The observed and latent variables within the model created in the SEM analysis are assessed together. SEM is simultaneously defined as multiple regression analysis [
51].
In the study, SEM was preferred over multiple regression analysis because multiple regression analysis can only measure the extent to which an independent variable can predict two dependent variables in two different stages instead of doing so at the same time and in one stage. This measure would not be reliable [
51]. Testing the whole model in a single stage instead of through individual coefficients would cause less deviation in the measurement error and p-value [
49]. However, SEM was thought to be favorable in terms of decreasing the measurement error and p-value [
51,
52].
Quantitative data were analyzed using SPSS 24 and AMOS 24 package programs. The multivariate normality tests and reliability analysis tests were performed on the data collected. The significance value was accepted as (
p < 0.05), and the goodness of fit indices were examined through CFA on the measurement models and structural model. The regression weights and R² values were analyzed to interpret the model. A bootstrap analysis was performed on 500 samples to verify a mediating effect of the mediator variable in the mediation model [
52].
Qualitative data were analyzed through content analysis in four stages. In the first stage, coding the data, the audio recordings obtained from the interviews were transcribed. Thereafter, each word, sentence, or paragraph considered to be meaningful within the raw data were named and coded by the researcher in one or more words. No predetermined conceptual structure was provided to guide the data obtained from the 12 administrators. Instead, because the opinions of the administrators about the effect of their behaviors on teachers’ organizational happiness and quality of work life were collected, the analysis started with an inductive analysis of the data. Codes were formed considering the similarities and differences in the data. In the second stage, the codes were grouped under certain categories and themes that contain the common aspects of all relevant codes were created. The code and themes obtained in the third stage were arranged in a way that the reader would understand. The findings were interpreted in the last stage. In the analysis of qualitative data, focus was placed to ensure that the findings were meaningful and consistent for internal validity [
53].
4. Conclusions
In this study, regarding the first research question, the direct effects of the distributed leadership characteristics of administrators as perceived by the teachers on the quality of work life and organizational happiness of teachers were examined. The results indicated that distributed leadership significantly and directly positively affects teachers’ quality of work life and organizational happiness.
Another main objective of this study was to go one step ahead in investigating the linear relations among the variables of distributed leadership characteristics of administrators, quality of work life of employees, and organizational happiness of employees. The study aimed to explain how these linear relations are realized, in other words, how the independent variable predicts the dependent variable. In the structural model with a mediating variable developed to investigate this situation, the mediation effect of the organizational happiness of the teachers in the relation between distributed leadership of administrators and teachers’ quality of work life was assessed as the second research question. According to the findings obtained from the analysis of the mediation model, change in the quality of work life was realized by the positive and significant effect of the distributed leadership on organizational happiness and the positive effect of consequent organizational happiness on the quality of work life (a full mediation effect). Accordingly, a theoretical model which determines how and in which ways distributed leadership affects teachers’ quality of work life was tested and verified. In this context, we believe that this study contributes to the literature.
A review of previous studies on distributed leadership indicates that descriptive and correlational studies were conducted more commonly [
25,
30,
31,
32]. These studies sought to answer the question of “what” by looking for a relation or effect between distributed leadership and other variables. However, no studies have investigated how the predictive variable (distributed leadership characteristics) is displayed by administrators in real life. In the qualitative component of this study, the opinions of school administrators about the effects of their behaviors on teachers’ organizational happiness and quality of work life were discussed as the third research question, in addition to concrete examples showing “through which behaviors” (how) distributed leadership characteristics are displayed in the real school environment. This was determined by identifying the examples of behaviors of administrators in the school environment that bear the characteristics of distributed leadership.
In the literature review, it was found that the earlier studies investigating distributed leadership in national and international settings did not include the organizations where preschool education was conducted. With these findings, it can be said that there are distributed leadership characteristics in preschools. The finding that distributed leadership behaviors positively affect organizational happiness and the quality of work life of preschool teachers can contribute to the international literature on distributed leadership.
Sustainable schools were defined as the product of the processes involving followers’ happiness and their joint participation in the leadership roles. Shared learning, actions and responsibilities demand the realignment of power and authority. Related to this, leadership requires the distribution of power and authority [
16,
17]. Ultimately, it was argued that distributed leadership may be beneficial for creating sustainable school environments by achieving organizational happiness and action in schools [
18].
Suggestions for Further Studies and Practices and Limitations
This study investigated the relations among distributed leadership, organizational happiness, and the quality of work life. According to the results of the mediating model, undetermined variables were effective on 54.3% of the change in the quality of work life. It is recommended that theoretical models are developed and tested in the future studies to understand these undetermined dynamics. In addition, a positive and significant relation was observed between distributed leadership and organizational happiness (a = 0.598, t = −18.270,
p < 0.001). Reference [
41] indicated a connection between the organizational happiness of teachers and their ability to offer quality education. Investigating the extent to which distributed leadership can predict the capacity of teachers to offer quality teaching through the organizational happiness of teachers in future studies will enable testing the extent to which distributed leadership can affect quality education output and its indirect effect on students in practice.
In this study, distributed leadership characteristics of administrators as perceived by teachers positively and significantly predicted teachers’ organizational happiness and quality of work life. Accordingly, disseminating distributed leadership within the organization would be useful in enhancing teachers’ organizational happiness and quality of work life. To achieve this, in-service training sessions on distributed leadership can be organized for school administrators.
This study was conducted with class teachers, subject-matter teachers, and 12 school administrators at preschools in TRNC. This can be seen as a limitation in the generalizability of the findings. Replicating similar studies in organizations that offer education at primary and secondary levels might provide improved generalizability, a more comprehensive discussion of the findings, and more reliable judgments through testing of the mediation model again. In addition, in this study, the sub-dimensions of distributed leadership were not investigated in terms of the extent to which they can predict teachers’ quality of work life and organizational happiness. Investigating which dimension(s) of distributed leadership can better predict teachers’ quality of work life and organizational happiness in future studies will allow the discussion of the quantitative findings obtained in this study in greater detail. However, reviewing the literature on organizational happiness, only a limited number of studies have mentioned the lack of consensus on the foundations and dimensions of the concept of organizational happiness. Further research in this area may allow us to explore the dimensions of organizational happiness more reliably and better understand organizational happiness.