School Leadership in Two Countries: Shared Leadership in American and Chinese High Schools. Wenlan Jing, Ph.D. candidate. Arizona State University

Transcription

1 School Leadership in Two Countries: Shared Leadership in American and Chinese High Schools Wenlan Jing, Ph.D. candidate Arizona State University Mary Lou Fulton College of Education Division of Educational Leadership and Policy Studies PO Box Tempe, AZ

2 Introduction School principals role in building an effective school has been widely acknowledged in current research. Research shows that there is substantially positive relationship between school leadership and student achievement. The positive relationship implies that it is possible to improve student achievement by enhancing school leaders leadership competence, though the influence of the enhanced leadership on learning outcomes may not be direct. While school leadership has reaped great attention from educational researchers in the U.S since the early 20 th century, in China, until 1980s, the essential role of school leadership had been less recognized. The change in attitude toward school leaders resulted from the implementation of Chinese Open door policy in 1978, and the abolishment of the old selection criterion of all leaders which paid equal or more attention to the political loyalty and attitude than professional skills. Today, leadership expertise is better understood by educators and stakeholders and the importance of school leadership in terms of improving students learning is a critical issue both in the U.S. and in China (Lewin, Little, Xu, & Zheng, 1994; Waters, Marzano, & McNulty, 2003). A new form of leadership, shared leadership, is one idea which recognizes the importance of school leaders in helping school future socioeconomic development, organizational development, and personal development. While the idea of getting subordinates involved in the leadership process is not entirely new, shared leadership is a relatively new term which was first addressed by scholars around mid-1990s (Avolio, Jung, Murry, & Sivasubramaniam, 1996; Pearce, 1997). This concept is a response to the complex tasks faced by leaders. While shared leadership research is still in its infancy, its 2

3 effectiveness has been supported by initial empirical examinations (Howe & Aditya, 1997; Pearce, Yoo, & Alavi, 2003; Avolio, Jung, Murry, & Sivasubramaniam, 1996). Similar to their counterparts in the business field, school leaders also encounter challenges produced by the increasingly complex context and tasks. American school principals are facing the challenges of increased accountability created by No Child Left Behind, whereas Chinese school principals are facing the demands of cultivating more talents for the rapid economic development. Research indicates that effective education leadership makes a difference in improving learning (Leithwood, Louis, Anderson, & Wahlstrom, 2004). Moreover, the schools management and professional development of teachers and administrations is limited by depending on single school leader to lead all aspect of the school (Levine, 2005). In research, both calling for establishing school leadership teams, which are less hierarchical and draw on the leadership talents of all members of the team. This study is designed to describe the shared leadership practices among administrator leadership teams at urban high schools in the U.S. and in China. The research questions are as follow: 1. Are the Chinese high school leaders behaviors linked to shared leadership reported by leaders associated with gender and position? 2. Are the American high school leaders behaviors linked to shared leadership reported by leaders associated with gender and position? 3. Among the Chinese high schools, what are the differences of leaders behaviors related to shared leadership between the schools reporting highest level of shared leadership and the schools reporting lowest level of shared leadership? 3

4 4. Among the American high schools, what are the differences of leaders behaviors related to shared leadership between the schools reporting highest level of shared leadership and the schools reporting lowest level of shared leadership? 5. Are there similarities and differences of high school leaders behaviors linked to shared leadership reported by leaders between China and the U.S.? Significance of the Study As a new leadership form which is suggested to be an important predictor of group effectiveness by initial empirical evidence, shared leadership could be a new perspective for school leaders to look at leadership, handle the increasingly complex environments and manage changes on a daily basis. Effective shared leadership in the top leaders team will encourage a shared leadership mindset at all levels of the organization. This study will focus on the practices of shared leadership of school top leaders principals and assistant principals. This study fills the gap of research on shared leadership in high school contexts, and also provides a cross-cultural perspective of shared leadership practices in the U.S. and China. Its significance lies in the following areas: (1) deepens American and Chinese school principals understanding of shared leadership; (2) provides evidence from school principals self assessment of competence related to shared leadership and serves as a reference for future school leadership professional development; (3) provides comparisons on practices of shared leadership between American and Chinese high school leaders. 4

5 Review of the Literature Defining Shared Leadership Shared leadership is a new type of leadership which emerged around mid-1990s (Avolio et al., 1996). Comparing to the traditional view of leadership which is the topdown influence of the leader on his/or her subordinates, shared leadership is an interactive group influence process. In other words, in this model, the leader is no longer the only source of influence. Accordingly, leaders role is dramatically changed from giving directions and orders to a designer, teacher and servant (Senge, 1990). While leaders do not have to make all the critical decisions by themselves, they are facing new challenges--creating an environment to develop people and encouraging people to get involved into the influencing process, i.e., play the leader s role. There is no unitary definition of shared leadership. Though shared leadership writers have similar idea about what shared leadership should be, the definitions they provide lack consistency. One possible reason is that the field of shared leadership is still in its infancy, and another reason lies in the lack of specificity to the concept of leadership itself. Conger and Pearce (2003) define shared leadership as follows: Shared leadership is a dynamic, interactive influence process among individuals in groups for which the objective is to lead one another to the achievement of group or organizational goals or both. This influence process often involves peer, or lateral, influence and at other times involves upward or downward hierarchical influence. (p.1) Shared leadership is also defined by some writers as a condition in which teams collectively exert influence (Cox, Pearce & Perry, 2003). Seers, Keller and Wilkinson (2003) view shared leadership as the extent to which more than one individual can effectively operate in a distinctively influential role within the same interdependent role 5

6 system. Some leadership theorists define shared leadership as an approach that considers the role of mutual influence among team members as another source of leadership for the group (Mayo, Meindel & Pastor, 2003) Despite these different perspectives reflected in these definitions, all writers agree that in shared leadership model, the influence process involves more than just top-down influence on subordinates by a leader, and the influence is distributed among group members. The context that shared leadership takes place can be a group, a team, an organization, or multiple parties in a system. The group members may consist of peers or subordinates. In this study, shared leadership is defined as follows: Shared leadership is an interactive group influencing process in which members influence each other to achieve the agreed upon goals. Related Concepts to Shared Leadership Besides shared leadership, similar concepts, such as distributed leadership and collaborative leadership, all underline the idea that leadership is an interactive influence instead of centrally determined. Distribute Leadership According to Spillane (2006), distributed leadership is not equal to shared leadership, and means more. The discussion of distributed leadership goes beyond the acknowledgement that multiple individuals take responsibility for leadership. In order to capture the complexity of leadership practice, distributed leadership approach explores the interactions among leaders, followers and their situation. By situations, he refers to tools and routines. Moreover, distributed leadership attempts to foster cooperation among all individual involved in leadership practice including formal, informal leaders and 6

7 potential leaders. Shared leadership can take place among peers, leaders and subordinates, and is not necessarily inclusive of everyone that has a hand in leadership practice. Collaborative Leadership Chrislip and Larson (1994) describe collaboration as a mutually beneficial relationship between two or more parties who work toward common goals by sharing responsibility, authority, and accountability for achieving results. The purpose of collaboration is to create a shared vision and joint strategies to address concerns that go beyond the purview of any particular party (p.5). The Turning Point program (university of Washington, 1997) which explores collaborative leadership in public health defines collaborative leadership as The processes, activities, and relationships in which a group and its members engage in collaboration (p. 2). Lafasto and Larson (2001) point out that in the last fifteen years, there is a dramatic change in both public and private organizations increasing team work and collaborative efforts. People with different views and perspectives come together, putting aside their narrow self-interests, discussing issues openly and supportively in an attempt to solve a larger problem or achieve a broader goal (p. XVII). While collaborative leadership can be applied to look at the relationship among parties with different functions either inside an organization or among different organizations, most of the collaborative leadership studies look at the collaborative relationships among different organizations. While these concepts may have slightly different emphases when applied to leadership research, they all adopt a social view of organizational life and emphasize the importance of human relations. Additionally, clarifying vision, sharing power, building 7

8 trust, and developing people are the common practices stressed by shared leadership, distributed leadership and collaborative leadership (Pearce et al, 2003; Gronn, 2000; Spillane, 2006; Lafasto & Larson, 2001). Effectiveness of Shared Leadership The effectiveness of shared leadership is usually examined through the comparisons of shared leadership and traditional leadership or vertical leadership. According to Bradford and Cohen (1984), shared responsibility for leadership functions and empowerment of subordinates is more effective than traditional leadership. Research indicates that shared leadership is an important predictor of group behavior, attitudes, creativity and effectiveness. Shared leadership has an effect on group behavior. Leadership behavior emanated by team members has been found to be more predictive of team performance than more traditional vertical leadership (House & Aditya, 1997). Research indicates that shared leadership is significantly related to team member satisfaction (Avolio et al. 1996). Pearce et al. (2003) found the relationship between shared leadership and social integration between the members of the team. Shared leadership is related to the group effectiveness (Shamir & Lapidot, 2003; Hooker & Csikszentmihalyi, 2003; Avolio et al., 1996). Pearce and Sims (2002) examined 71 change management teams at a large automotive manufacturing firm of United States and found that shared leadership is a predictor of the leadership effectiveness. Hooker and Csikszentmihalyi (2003) report that shared leadership can bolster the creative process. They found shared leadership can enhance the conditions for flow which is a condition for creativity to occur. 8

9 Very little empirical research on shared leadership in educational settings has been conducted. Turning Points (Center for Collaborative Education, 2001) designed for middle school change gives a guide for schools to develop a collaborative culture and create shared leadership. Developing a collaborative culture requires school leaders to empower teachers and administrators to work together. In a study of Successful school leadership project, Moller and Eggen (2005) found that in schools with distributed leadership, trust and power are closely interrelated. Trust creates the conditions and mobilizes people to action and collaboration. Shared Leadership in Education In recent work, educational leadership scholars suggest that viewing school leadership from a shared leadership perspective has the potential to provide useful insight into effective school leadership (Gronn, 2000). Researchers argue that this new type of leadership will help school leadership face the new challenges in today s schools. The increasingly complex information society creates new requirements for students: more advanced skills and knowledge. Raising standards for school promotion and graduation, mandating student testing, and demanding school accountability are part of the leadership challenges. In addition, curriculum standards, achievement benchmarks, programmatic requirements, and other policy directives from many sources generate complicated and unpredictable requirements for schools. No Child Left Behind (NCLB) represents the most significant federal policy initiative on education in decades and is designed to change how states and local school districts educate children (Farkas, Johnson & Duffett, 2003). The law uses annual testing for students to assess teaching and learning outcomes. Consequently, school principals role has been shifted from 9

10 supervisors to change agents which required advanced leadership skills. School leadership should no longer be referred as the principal s activity, rather a social activity woven into the threads of the organization. Shared leadership is a possible model which can help build more effective schools. Research shows that the key components of shared leadership are also critical for successful school leadership (Pearce & Conger, 2003; Tschannen-Moran, 2004; Short & Greer, 1997). One of the most important components of successful schools that leadership scholars suggest is a shared vision (Center for Collaborative Education, 2001) which guides the community toward a common purpose. Schools stakeholder, teachers, administrators, students and their families share a common vision regarding the school future. Tschannen-Moran s book (2004), Trust Matters: Leadership for successful schools, offers practical advice to educators on how to establish and maintain trust in school. She believes that building trust is vital because without trust, it is unlikely that schools can fulfill their fundamental mission of educating students to be productive citizens. The school principal is the key to building a trusting environment (Tschannen- Moran, 2004; Short & Greer, 1997). In school settings, empowerment is defined as a process whereby school participants develop the competence to take charge of their own growth and resolve their own problems (Short & Greer, 1997, p.134). With the essential competence, people have confidence in their skills and knowledge to act on a situation and improve it. To improve schools, school leaders need to provide more opportunities for meaningful and collective participation in these critical areas. 10

11 By developing people, school leaders are actually helping others to acquire further skills they need, so that people can quickly get accustomed to lead in a changing environment (Short & Greer, 1997). Theoretical Framework Hofstede s taxonomy has been used most often as a conceptual framework in the cross-cultural research on leadership (Hofstede, 1993). Through a comparison of the values of employees and managers in sixty-four national subsidiaries of IBM Corporation, Hofstede found that four dimensions explained 49% of the variance in the data. (1) Power Distance is defined as the degree of inequality among people which the population of a country considers as normal (p. 89); (2) Individualism is defined as the degree to which people in a country prefer to act as individuals rather than as members of groups (p. 89). Collectivism is the opposite of individualism; (3) Masculinity/Femininity is the degree to which tough values like assertiveness, performance, success and competition, which in nearly all societies are associated with the role of men, prevail over tender values like the quality of life, maintaining warm personal relationships, service, care for the weak, and solidarity, which in nearly all societies are more associated with women s roles (p. 90); (4) Uncertainty Avoidance is defined as the degree to which people in a country prefer structured over unstructured situations (p. 90). A fifth dimension was added based on another study using the Chinese Value Survey. It is labeled Long-term versus Shortterm Orientation. The long-term orientation means values oriented towards the future such as thrift. The short-term orientation means values oriented towards the past and present, for example, respect for tradition. While Chinese managers have higher values for power distance, uncertainty avoidance, collectivism, femininity, and long-term 11

12 orientation, American managers are more individualistic, masculine, short-term oriented, egalitarian, and comfortable with uncertainty (Fu & Taber, 1998; Hofstede, 1993). Shared leadership theories are drawn from literature. In contrast to the traditional vertical leadership, the leader is no longer the only source of influence. Accordingly, leaders role is dramatically changed from giving directions and orders to a designer, teacher and servant (Senge, 1990). By clarifying vision, sharing power and influences, building trust, and developing people, leaders create an environment to encourage people to get involved into the influencing process. Research Design Quantitative research method was employed to examine the practices of shared leadership of high school leaders in the United States and China by analyzing the leaders behaviors linked to shared leadership. Instrument Survey data were collected using an instrument: Collaborative leadership--selfassessment questionnaires. This assessment was intended to help individuals evaluate key behaviors that are important to each of the five practices of effective collaborative leaders: creating clarity: visioning and mobilizing, building trust, sharing power and influence, developing people, and self-reflection. These five practices provide a behavioral view of successful collaborative leaders and help individuals assess their collaborative leadership competence. School leaders were asked to rate the frequency of their own shared leadership behaviors. A seven-point scale was used to represent behavior frequency from seldom to almost always. An example of the survey questions is: 12

13 Behavior Frequency Behaviors Seldom Sometimes Often Almost Always I can describe a personal vision for the school that offers a future achievable with the assets available Since the instrument is designed and widely applied in public health leadership field, some of the survey items were rephrased to fit the working context of high school principals. In this study, copies of this tool were distributed to urban public high school principals and assistant principals in the United States and China. The anonymity of the instrument facilitated objective, honest and user-friendly feedback. Pilot-Test Procedures A pilot test was conducted to test the face validity of the instrument. The test was designed to evaluate how clear and how relevant the survey questions are. Participants came from a convenient group, a workshop regarding building trust in schools for school leaders. Five school leaders participated in the pilot-test. Results indicate that survey questions are clear and relevant. Three questions which scores on clarity were lower than the average were modified. The reliability of the instrument was evaluated using Cronbach s Alpha and three reversed questions were deleted based on the results. Data Collection Procedure Chinese participants were chosen from Harbin, China. Harbin is the capital of Heilongjiang province which locates in the northeastern China. This city is the seventhlargest city in China with 3.2 million populations in the urban area. Participants were 13

14 selected through three steps. First of all, a school list was made which included three categories of public urban high schools in Harbin, key high schools of the province, key high schools of the city, and ordinary high schools which are defined in light of their students academic performance. Secondly, six schools were randomly selected from each category. In total, 18 schools were selected for the study. Usually, there are one principal and 2 to 4 assistant principals in each school. Therefore, 54 to 90 school leaders were expected to fill out the research questionnaires. Finally, re-sample if the response rate was low and less than 50 questionnaires were returned. The author went back to Harbin, China to collect the data. A test was conducted to understand the Chinese high school leaders attitude toward participating in this research. Cold calls were made to a couple of the urban high schools in Harbin which were not originally selected to solicit for participation. Unfortunately, none of them were willing to get involved. In view of this result, the author decided to use good connections which are called guanxi in Chinese, to get people s trust and participation. A note was written on the cover letter of the survey to ask for support by the person who is in charge of Harbin high schools in the education department of the local government and is very familiar with most of the selected school principals. The author called the principals of the selected schools mentioning the name of the person who wrote the note. With the school leaders permission, copies of the questionnaire were brought to the schools and picked up after the principals were done. This time, 17 out of the18 schools agreed to participate in the research, filled out and returned the questionnaires. The rest one school initially agreed to do the survey, but never returned them even after several reminder calls were made. One more school was selected to substitute for that school and the author managed 14

15 to find a relative who has good relationship with the school s principal. With the support of the principal, questionnaires were successfully distributed and collected. Once again, the data collecting procedure confirmed the importance of guanxi in China and also reflected the power possessed by the government educational officials, as well as how Chinese people respect power and authority. Participants of the U.S. came from Phoenix, Arizona. Phoenix is the sixth-largest city in the country with 1.5 million populations and sprawls across 400 square miles. Survey participants were selected from city of Phoenix high school districts. Public schools in Phoenix can be categorized into excelling schools, highly performing schools, and performing schools based on AZ LEARNS achievement. Six schools were randomly selected from each category. In the next step, the principals or assistant principals of the 18 schools were contacted either by phone or and solicited for participation. Copies of the questionnaire were sent by mail to the schools if the school leaders were willing to help and the completed questionnaires were picked up from the schools. Cold calls were made to the 18 school leaders. It took the author more than a month just to get hold of the school leaders, because the survey was conducted in March, almost the end of an academic year. Schools were busy with the Arizona's Instrument to Measure Standards (AIMS), a Standards-Based test, which is designed to examine the progress of Arizona's students toward mastering Arizona's reading, writing and mathematics Standards. One of the 18 schools refused to participate and another school responded that they couldn t make the decision because all outside research must go through the district office. But later, although the district office was very supportive, the school principal still decided not to participate in the research. Three of the eighteen 15

16 schools agreed to participate in the study, but never really had the questionnaires filled out. While thirteen schools returned their questionnaires, the total number of questionnaires was less than 50. A convenient group was used to get more participants. The group was composed of current students and alumni of an Ed.D. program -- Dynamic Educational Leadership for Teachers and Administrators at Arizona State University. With the support of program s coordinator, the solicitation for participation was sent to twenty high schools by s. Seven of the twenty schools replied and returned 12 copies of the questionnaire. The data collecting process of this research indicates that good timing is very critical for a survey study to achieve a higher response rate. The difficulty of conducing research involving participants could be escalated if the timing is not right. Researchers who are interested in high school education should avoid conducting a research at the end of an academic year which is from February to May in Phoenix and from March to July in Harbin. Another thing worth mentioning is the different channels of accessing the contact information of school leaders in these two countries. In Phoenix, the contact information of school leaders, such as phone numbers and addresses, is open to public and can be easily accessed at the school or district website. In Harbin, however, not every school has a website and usually the contact information of school and school leaders can not be found on line. Good connection is absolutely essential to access the phone numbers of the school leaders and to get research participants. Results In this section, data was analyzed to reveal the characteristics of the leadership practices of high school principals and assistant principals which are related to shared 16

17 leadership. Five practices, visioning and mobilizing, building trust, sharing power and influence, developing people, and self-reflection were analyzed within each country and also compared across cultures, between China and the United States. Shared Leadership Behaviors: Association with Culture, Gender, and Position Nineteen schools with fifty-two valid questionnaires from the U.S. and eighteen schools with sixty-three valid questionnaires from China were included in the analysis. Means were calculated for the five leadership practices, which were used to examine whether the high school leaders practices related to shared leadership associate with culture, gender, and position. Association with Culture As mentioned earlier, Hofstede s cross-culture research (1993) reveals the existence of leadership differences between Chinese managers and American managers in business. He pointed out the important role of values, norms, and assumptions in the understanding and practices of leadership in each society. Is culture also associated with the leader practices related to shared leadership in schools? Table 1 shows the mean scores and the percentage differences for shared leadership practices by country rated by school leaders. For Chinese school leaders, among the five shared leadership practices, building trust has the highest behavior frequency and self-reflection has the lowest behavior frequency. Developing people is rated the second highest, which supports the findings of Weldon (2005) -- Chinese leaders believe that helping others succeed is of secondary importance. American school 17

18 leaders rated developing people the highest and visioning the lowest behavior frequency. Overall, American school leader rate themselves higher on shared leadership practices than the Chinese school leaders do. When each of the five practices is examined separately, the mean score of building trust is the only one practice that Chinese school leaders rate higher than American school leaders. The characteristics of the schools and school leaders shown in Table 2 indicate that Chinese school leaders have more years of administration experience before they were promoted to principals or assistant principals than the American school leaders. The means of years in the position for American and Chinese school leaders are very close, but the mean for Chinese school leaders total years of administration experience is twice as many as that for American school leaders. Table 2 also reveals that Chinese school leaders tend to stay at one school much longer than American school leaders. The number of current students which reflects the school size indicates that generally, Chinese high schools are larger than American high schools. The five most exhibited behaviors listed in Table 3 are more similar than different between Chinese and American school leaders. The five least exhibited behaviors in Table 4, however, are quite different for the school leaders in these two countries. But both the Chinese and American school leaders have the two items, use self-assessment tools to inform self reflections and share power as a means for increasing power in the five least exhibited behavior category. Association with Gender 18

19 Table 1 Means and Percentage Differences for Shared Leadership Behavior Frequencies: Culture differences Mean Shared leadership practice American schools Chinese schools Percentage difference % Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). Table 2 Characteristics of the Schools and School Leaders Variable Mean American schools Chinese schools Percentage difference % Number of years in the position Number of years at the school Number of years of admin experience Number of current students

20 Table 3 Five Most Exhibited Behaviors of American and Chinese High School Leaders School Behavior Mean American school leaders I take seriously my responsibility for coaching and mentoring others I use my personal power responsibly I "walk the talk", i.e., I do what I say I will do I demonstrate to my peers that I believe that trust is the foundation for successful collaboration I recognize the effect of my emotions on work performance Chinese school leaders I use my personal power responsibly I take seriously my responsibility for coaching and mentoring others I demonstrate to my peers that I believe that trust is the foundation for successful collaboration I "walk the talk", i.e., I do what I say I will do I establish my expectations for the people I mentor Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). 20

21 Table 4 Five Least Exhibited Behaviors of American and Chinese High School Leaders School Behavior Mean American school leaders I use self-assessment tools such as personality inventories to inform my self reflections I use influence to produce results whenever possible I shared power as a means for increasing power I create a mutually agreed-upon coaching plan, including criteria for success When exercising leadership. I rely significantly on peer problem-solving Chinese school leaders I shared power as a means for increasing power I use self-assessment tools such as personality inventories to inform my self reflections I seek feedback from all relevant constituencies about my behavioral impact I create a framework for action using systems thinking I ensure that processes for exercising school collaborative leadership are open to other administrators Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). 21

22 Table 5 Means and Percentage Differences for Shared Leadership Behavior Frequencies: Gender Differences Mean Shared leadership practice Women Men Percentage difference % American school leaders Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Chinese school leaders Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). Table 5 shows that for American school leaders, the frequency means of overall shared leadership behaviors are similar for men and women, though the rate for women is a little bit higher. Men and women rate themselves very similar on building trust and selfreflection. According to their self-evaluation, the female school leaders practice 22

23 clarifying vision and developing people more frequently than the male school leaders, but rate lower on sharing power than the male leaders. Chinese women and men have very close means of overall shared leadership behaviors. Furthermore, they all rate building trust the most exhibited practice and selfreflection the least exhibited practice. Chinese female school leaders rate building trust, sharing power, and developing people a little bit higher than male leaders, but rate clarifying vision and self-reflection a little bit lower than male leaders. For both countries, it seems that gender doesn t make much difference in these five shared leadership practices. Compared to American counterparts, Chinese male leaders did not rate higher on any of the five practices than American male leaders. Although Chinese female leaders also rate lower than American female leaders on the mean of the overall shared leadership behaviors, their average behavior frequencies on building trust and sharing power are higher than those of American female leaders. Table 6 illustrates that the number of male school leaders is larger than that of the female school leaders in both countries. In American schools, the ratio of the male principals to the female principals is similar to the ratio of the male assistant principals to the female assistant principals, approximately 60:40. In China, the percentage of the female assistant principals is a little bit higher than the percentage of the male assistant principals; however, the female principals only account for 26.7% of the total number of the Chinese principals. 23

24 Table 6 Gender of School Leaders Gender Frequency Percentage American school leaders Male Female Chinese school leaders Male Female American school principals Male Female Chinese school principals Male Female American school asst. principals Male Female Chinese school asst. principals Male Female

25 Association with Position In order to find out if people in different positions practice leadership differently, we compared the five shared leadership behavior frequencies between principals and assistant principals. Table 7 indicates that in the American high schools, the principals rate clarifying vision the most exhibited and self-reflection the least exhibited behavior, whereas clarifying vision is rated the least exhibited and developing people the most exhibited by the assistant principals. Chinese principals, similar to American principals, rate self-reflection the least exercised behavior, but rate building trust the most exercised behavior. Chinese assistant principals, like the American assistant principals, rate clarifying vision the least exercised. The frequency means of overall shared leadership behaviors for principals are higher than those for assistant principals in both countries, except for the practice of selfreflection in China (see Table 7). The Chinese assistant principals rate slightly higher in self-reflection than the Chinese principals. The percentage differences of the means of five practices between the principals and the assistant principals in Chinese schools are smaller than those of American schools. Moreover, the American principals and assistant principals have higher means of the five practices than the Chinese counterparts, except that building trust is rated slightly higher by the Chinese school leaders than their American counterparts, and sharing power is also rated a little bit higher by the Chinese assistant principals than the American assistant principals. 25

26 Table 7 Means and Percentage Differences for School leaders Shared Leadership Behavior Frequencies in China and the U.S. Mean Shared leadership practice Asst. Principals Principals Percentage difference % American schools Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Chinese schools Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). 26

27 Table 8 Means and Percentage Differences for Shared Leadership Behavior Frequencies of Chinese and American School Leaders Mean Shared leadership practice Chinese schools American schools Percentage difference % Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Asst. Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). Furthermore, the means of the five shared leadership practices for the assistant principals in both countries are very close, whereas the means for the principals in both counties are quite different (see Table 8). The mean of the American principals behavior frequencies of clarifying vision is 10.83% greater than that of the Chinese principals. 27

28 Self-reflection generates the second largest percentage difference between the Chinese and American principals, and the largest percentage difference between the Chinese and American assistant principals. Shared Leadership Behaviors: Differences among the U.S. Schools From this point on, we are going to compare the five practices related to shared leadership among the schools within each country. How do the five leadership practices differ among schools in the United States? We chose the schools which means of overall shared leadership behaviors ranked top three and treated the schools as one group, the top group. Similarly, the schools ranked bottom three were selected and regarded as one group, the bottom group. We compared the two groups on the five leadership practices and also the characteristics of the schools and the school leaders. The greatest difference between the top group and the bottom group lies in the practice of clarifying vision (see Table 9). The mean of behavior frequencies of clarifying vision for the top group is about 15.8% larger than that for the bottom group. Developing people is rated the most exhibited practice by the top group leaders, which behavior frequency mean is about 15.3% larger than that of the bottom group leaders. It is interesting that both the top group and the bottom group rate clarifying vision the least exhibited practice. The result, however, could be explained by the make-up of the groups. In the selected three schools of each group, there are only 3 principals, but 7 to 9 assistant principals. Moreover, from the earlier analyses, we understand that in the U.S., there is great difference between the principals and the assistant principals in practicing clarifying 28

29 vision. Therefore, the ratings for clarifying vision may not represent the real situation because of the uneven numbers of the principals and assistant principals. Table 9 Means and Percentage Differences for Shared Leadership Behavior Frequencies of the Top Group and the Bottom Group in the U.S Mean Shared leadership practice Bottom group Top group Percentage difference % Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). For further understanding the characteristics of behaviors related to shared leadership in the top group and the bottom group, we compare the behavior frequencies between the principals and assistant principals within each group. Table 10 shows that the difference on clarifying vision associated with position is the biggest for both groups with the principals having greater frequency means. Moreover, the difference is much bigger for the bottom group than the top group. The bottom group also has greater percentage difference associated with position on building trust and self-reflection than the top group. 29

30 Table 10 Means and Percentage Differences for School Leaders Shared Leadership Behavior Frequencies of the Top Group and the Bottom Group in the U.S. Mean Shared leadership practice Asst. Principals Principals Percentage difference % Top group Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Bottom group Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). When comparing the behaviors linked to shared leadership of the principals and assistant principals, it was also found that the differences of the assistant principals between the top group and the bottom group are greater than those of the principals (see Table 11), except for sharing power and developing people. As to the principals, the 30

31 frequency mean of developing people for the top group is 15.8% greater than that for the bottom group. For the assistant principals, their behaviors of clarifying vision and selfreflection produce greater difference between the top group and the bottom group than the other three leadership components. Table 11 Means and Percentage Differences for Shared Leadership Behavior Frequencies of the Top Group and the Bottom Group Leaders in the U.S. Mean Shared leadership practice Bottom group Top group Percentage difference % Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Assistant Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). 31

32 The characteristics of the U.S. schools and the school leaders in Table 12 indicate that the school leaders in the top group have more years of working experience and stay at one school longer than the school leaders in the bottom group. Moreover, the average school size of the schools in the top group is a little bit larger than the schools in the bottom group. Table 12 Characteristics of the U.S. Schools and the School Leaders for the Top Group and the Bottom Group Mean Variable Bottom group Top group Percentage difference % Number of years in the position Number of years at the school Number of years of admin experience Number of current students Shared Leadership Behaviors: Differences among the Chinese Schools How are leaders behaviors linked to shared leadership similar or different among the Chinese high schools? Similar to the U.S. school analyses, three schools that ranked top three on shared leadership behavior frequency and the three schools that ranked bottom three on shared leadership behavior frequency were selected and constituted the top group and the bottom group respectively. Table 13 shows that building trust, clarifying vision and sharing power have greater percentage differences between the bottom group and the top group. School leaders in the top group rate building trust the most exhibited and self-reflection the least exhibited behavior, whereas the leaders in the 32

33 bottom group rate developing people the most exhibited and clarifying vision the least exhibited behavior. Table 13 Means and Percentage Differences for Shared Leadership Behavior Frequencies of the Top Group and the Bottom Group in China Mean Shared leadership practice Bottom group Top group Percentage difference % Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). As we did for the U.S. schools, we also looked at the percentage differences of the leadership behaviors associated with position in the top group and the bottom group respectively. As shown in Table 14, the percentage differences related to position are smaller for both the top group and the bottom group, compared to the differences between the two groups in the U.S. The greatest difference between the principals and the assistant principals for both the top group and the bottom group lies in the behaviors of building trust. The bottom three schools also have greater difference in the behavior of sharing powering associated with position than the top group. It is very interesting to note that in the top group, except for building trust, the other four practices were rated a little bit higher by the assistant principals than the principals. 33

34 Table 14 Means and Percentage Differences for School Leaders Shared Leadership Behavior Frequencies of the Top Group and the Bottom Group in China Mean Shared leadership practice Asst. Principals Principals Percentage difference % Top group Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Bottom group Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). There is great percentage difference in school leaders behaviors between the top group and the bottom group (see Table 15). The frequency mean of building trust for the principals in the top group is 27.66% greater than that for the principals in the bottom group. Among the assistant principals, the behavior frequency of clarifying vision creates 34

35 the greatest difference between the top group and the bottom group. Moreover, the assistant principals from the top and the bottom group have greater mean percentage difference in clarifying vision, sharing power and developing people than the principals. Table 15 Means and Percentage Differences for Shared Leadership Behavior Frequencies of the Top Group and Bottom Group Leaders in China Mean Shared leadership practice Bottom group Top group Percentage difference % Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Assistant Principals Overall behaviors Visioning Building trust Sharing power & influence Developing people Self-reflection Note. Shared leadership behavior frequencies were rated on 7-point scales (1 = seldom, 7 = almost always). 35