28
Journal of Educational Administration Relation of principal transformational leadership to school staff job satisfaction, staff turnover, and school performance James Griffith Article information: To cite this document: James Griffith, (2004),"Relation of principal transformational leadership to school staff job satisfaction, staff turnover, and school performance", Journal of Educational Administration, Vol. 42 Iss 3 pp. 333 - 356 Permanent link to this document: http://dx.doi.org/10.1108/09578230410534667 Downloaded on: 18 November 2014, At: 03:40 (PT) References: this document contains references to 68 other documents. To copy this document: [email protected] The fulltext of this document has been downloaded 7085 times since 2006* Users who downloaded this article also downloaded: Kenneth Leithwood, Doris Jantzi, (2000),"The effects of transformational leadership on organizational conditions and student engagement with school", Journal of Educational Administration, Vol. 38 Iss 2 pp. 112-129 http://dx.doi.org/10.1108/09578230010320064 Femke Geijsel, Peter Sleegers, Kenneth Leithwood, Doris Jantzi, (2003),"Transformational leadership effects on teachers’ commitment and effort toward school reform", Journal of Educational Administration, Vol. 41 Iss 3 pp. 228-256 Kerry Barnett, John McCormick, Robert Conners, (2001),"Transformational leadership in schools – Panacea, placebo or problem?", Journal of Educational Administration, Vol. 39 Iss 1 pp. 24-46 Access to this document was granted through an Emerald subscription provided by 394654 [] For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information. About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services. Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. Downloaded by UNIVERSITI UTARA MALAYSIA At 03:40 18 November 2014 (PT)

09578230410534667

Embed Size (px)

DESCRIPTION

article

Citation preview

Page 1: 09578230410534667

Journal of Educational AdministrationRelation of principal transformational leadership to school staff job satisfaction, staffturnover, and school performanceJames Griffith

Article information:To cite this document:James Griffith, (2004),"Relation of principal transformational leadership to school staff job satisfaction, staffturnover, and school performance", Journal of Educational Administration, Vol. 42 Iss 3 pp. 333 - 356Permanent link to this document:http://dx.doi.org/10.1108/09578230410534667

Downloaded on: 18 November 2014, At: 03:40 (PT)References: this document contains references to 68 other documents.To copy this document: [email protected] fulltext of this document has been downloaded 7085 times since 2006*

Users who downloaded this article also downloaded:Kenneth Leithwood, Doris Jantzi, (2000),"The effects of transformational leadership on organizationalconditions and student engagement with school", Journal of Educational Administration, Vol. 38 Iss 2 pp.112-129 http://dx.doi.org/10.1108/09578230010320064Femke Geijsel, Peter Sleegers, Kenneth Leithwood, Doris Jantzi, (2003),"Transformational leadershipeffects on teachers’ commitment and effort toward school reform", Journal of Educational Administration,Vol. 41 Iss 3 pp. 228-256Kerry Barnett, John McCormick, Robert Conners, (2001),"Transformational leadership in schools –Panacea, placebo or problem?", Journal of Educational Administration, Vol. 39 Iss 1 pp. 24-46

Access to this document was granted through an Emerald subscription provided by 394654 []

For AuthorsIf you would like to write for this, or any other Emerald publication, then please use our Emerald forAuthors service information about how to choose which publication to write for and submission guidelinesare available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.comEmerald is a global publisher linking research and practice to the benefit of society. The companymanages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well asproviding an extensive range of online products and additional customer resources and services.

Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committeeon Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archivepreservation.

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 2: 09578230410534667

*Related content and download information correct at time of download.

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 3: 09578230410534667

Relation of principaltransformational leadership toschool staff job satisfaction,staff turnover, and school

performanceJames Griffith

US Department of Education, Washington, DC, USA

Keywords Principals, Schools, Transformational leadership, Job satisfaction,Employees turnover

Abstract In the present study, the direct effect of principal transformational leadership to schoolstaff turnover and school performance was examined, in addition to its indirect effect throughschool staff job satisfaction. Survey data were obtained from elementary school staff and students,and school-aggregated student achievement test scores were obtained from school archives. Resultsshowed that staff reports of principal behaviors could be described in terms of the threecomponents of transformational leadership: inspiration or charisma, individualized consideration,and intellectual stimulation. Principal transformational leadership was not associated directly witheither school staff turnover or school-aggregated student achievement progress. Rather, principaltransformational leadership showed an indirect effect, through staff job satisfaction, on schoolstaff turnover (negative) and on school-aggregated student achievement progress (positive).Finally, higher levels of school staff job satisfaction were associated with smaller achievement gapsbetween minority and non-minority students. This result was more evident among schools havinghigher levels of principal transformational leadership. Results are discussed in relation to the roleof transformational leadership in school performance and in recruiting, training, and evaluatingschool principals.

Conceptions of school leadershipClassic studies of school administration (Halpin, 1966) described schoolleadership as providing “structure” and “consideration”. Structure referred tothe extent administrators provided staff and materials necessary for effectiveinstruction and student learning. Consideration referred to the extentadministrators developed mutual trust and respect, and shared norms andvalues among school staff necessary for positive and productive socialrelations. In a series of in-depth interviews of teachers, Blase (1987) providedmore detailed behaviors of school leadership. Effective school principals, asdescribed by teachers, had clear and well-articulated goals; delegated tasks toothers; encouraged staff to participate in decision-making; incorporated others

The Emerald Research Register for this journal is available at The current issue and full text archive of this journal is available at

www.emeraldinsight.com/researchregister www.emeraldinsight.com/0957-8234.htm

This research was previously conducted when working in a large metropolitan school district.

Principaltransformational

leadership

333

Received July 2003Revised October 2003

Accepted October 2003

Journal of EducationalAdministration

Vol. 42 No. 3, 2004pp. 333-356

qEmerald Group Publishing Limited0957-8234

DOI 10.1108/09578230410534667

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 4: 09578230410534667

in problem-solving; treated staff fairly and equitably; and provided staffsupport in difficult situations. Moreover, these principal behaviors were shownto have positive outcomes on students and parents, to include school staff’sexperience of work. Blase et al. (1986) reported that principals’ initiatingstructure and displaying consideration were associated with more satisfyingwork conditions, higher job satisfaction, and less job stress. Staff’s experienceof job stress was seen as principals’ lack of consideration and was related toteacher dissatisfaction. The expanded study of leadership, most notably insocial psychology, offers further clarification of the components of principalleadership and their relation to outcomes.

One leadership theory that has gained recent and widespread attention istransformational leadership. Effective leadership, according to Burns (1978),involved the leader’s ability to make group members become less interested inthemselves and more interested in the group. To develop and build groupmembers’ commitment to common goals and purpose, transformational leadersthrough interpersonal relations appeal to broad humanmoral and psychologicalneeds. Moral needs include a sense of goodness, righteousness, duty, andobligation, and psychological needs include esteem, autonomy, andself-actualization. Bass (1985, 1990, 1996) further elaborated on processesenabling transformational leaders to alter the behaviors and attitudes ofindividual members. First, charisma or inspiration is the ability of leadersto provide a clear sense of mission, which leaders in turn convey to membersand develop a sense of loyalty and commitment. Second, individualizedconsideration is the leader’s treatment of each member as a unique individualand the leader’s willing delegation of projects to individual members,which stimulate and create learning experiences. Third, intellectualstimulation is the leader’s provision of opportunities for group membersto rethink traditional procedures and to examine situations in new and novelways.

Over the last decade, there has been increasingly more evidence to supporttransformational leadership theory. Studies have shown that the leaderbehaviors are adequately described by the three broad processes oftransformational leadership and occur in different organizational settings,such as industrial, military, political, and educational ( Bass, 1985; Bass et al.,1987; Waldman et al., 1987). Studies have also shown that transformationalleadership is associated with effective leaders. Group members’ commitment,extra effort and motivation in their jobs, and commitment to the organizationall have been positively associated with the underlying processes oftransformational leadership, namely, inspiration, consideration, andstimulation (Kane and Tremble, 2000; Koh et al., 1995). In addition, groupmembers perceived transformational leaders as effective, reported that theywould expend effort for the leader, were satisfied with the leader, andperformed well (Bass, 1985; Bass et al., 1987; Waldman et al., 1987).

JEA42,3

334

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 5: 09578230410534667

School leadership as transformational leadershipRecent research supports the notion that principal leadership might be modeledas transformational leadership. For example, Leithwood and Jantzi (1999)factor-analyzed school staff responses to survey items and summarilydescribed emergent factors in terms of the components of transformationalleadership. These researchers also reported a direct effect of principaltransformational leadership on school conditions, such as school goals,planning, and structure, which in turn showed a direct effect on classroomconditions such as instruction, policies, and procedures. The direct effect oftransformational leadership on student outcomes, such as identification andparticipation, was either negligible or statistically nonsignificant. Hallingeret al. (1996) also reported few direct effects of principal leadership on studentachievement. Rather, the effect of principal leadership (e.g. instructional focus,provision of resources for instruction and staff, and accessibility) on schooleffectiveness (i.e. aggregated student achievement) occurred largely throughprincipal actions, such as providing a clear school mission and optimizingstudent learning by grouping practices that shaped the school’s learningclimate. Other research results support the notion that principal behaviorsprimarily affect broad school conditions, such as climate and work conditions( Blase et al., 1986; Bossert et al., 1982) and that the relation of principalleadership to organizational outcomes – such as, employee turnover andschool-aggregated student achievement progress – is best described indirectlythrough school staff’s satisfaction with their work environment ( Blase et al.,1986; Hallinger and Heck, 1996). Figure 1 portrays these proposed relations.

The proposition that principal behaviors have stronger relations to outcomesassociated with staff, such as job satisfaction, than student outcomes hasintuitive appeal. The work of staff, classroom instruction, is more directlyrelated to student learning and achievement than the work of principals. Schoolstaff spendmore timewith students. By comparison, principals spendmore time

Figure 1.The relation of

leadership to workenvironment and

organizationalperformance

Principaltransformational

leadership

335

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 6: 09578230410534667

with school staff – providing direction and guidance, assessing and providingneeded resources, and observing and evaluating job performance – than withstudents. Thus, principal behaviors more directly affect school staff,specifically, their satisfaction and commitment to work and working relationswith one another. The principal’s relationship with school staff likely influencesjob satisfaction, which in turn relates to staff job performance. As an example, ina series of studies, Dinham and Scott (1998, 1999, 2000, 2003) reported moderatelevels of teacher satisfaction with a school leadership domain, consisting ofteacher perceptions of administrative and educational support, and schoolreputation. Job satisfaction among teachers is also likely discerned by students,affecting school performance measures based on student learning andachievement. For example, Steele (1997) has recently documented howbehaviors and expectations of school staff and students evoke negativestereotypes and performance anxieties among minority students, which wererelated to lower levels of student motivation to learn and to perform. Theprincipal’s relationship with school staff also likely relates to communicationamong all staff, cooperation and collaboration, mutual trust and understanding,and engagement of staff in their individual and group tasks, all of which areplausibly associated with organizational or school performance.

Transformational leadership seems appropriate for examining leader-driveninterventions to reduce the achievement gap between the minority andnon-minority students. In the United States, African-American and Hispanicchildren have scored as much as 15 points lower on standardized achievementtests than white children (Flynn, 1984; Loehlin et al., 1975; Neisser et al., 1996).This disparity has given rise to numerous studies attempting to explain thedifferences in achievement among children from different racial and ethnicbackgrounds (Ceci et al., 1999; Grissmer et al., 1999). Results of some studiessuggest that low achievement among minority students stem from lessdesirable work environments, school staff turnover, and less experienced stafffound in schools having proportionately more disadvantaged students ( Bryket al., 1990; Lee, 2001). Earlier research has associated employee jobsatisfaction, commitment, motivation, and effort to transformational leaders( Bass, 1985; Bass et al., 1987; Waldman et al., 1987). Thus, it would be expectedthat transformational leadership would be associated with less achievementdisparity between the minority and non-minority students.

Study purposeThe present study extends current research on transformational leadership byexamining whether principal behaviors can be described in terms oftransformational leadership. This study also examines the direct effect ofprincipal transformational leadership on school outcomes, such as school staffturnover and school performance, and the indirect effect on these outcomesthrough school staff job satisfaction. Specific research questions addressed areas follows.

JEA42,3

336

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 7: 09578230410534667

. Can principals’ behaviors be described in terms of the component parts oftransformational leadership?

. Do principals who display transformational leadership have school staffwith higher levels of job satisfaction and less school staff turnover?

. Do principals who display transformational leadership have school staffwith higher levels of job satisfaction and higher-performing schools?

. Does principal transformational leadership relate directly or indirectly(through staff job satisfaction) to school staff turnover and schoolperformance?

. Do relations among principal transformational leadership, staff jobsatisfaction, school staff turnover, and school performance vary by thestudent disadvantaged school population?

. Do schools in which staff with higher levels of job satisfaction havesmaller minority-non-minority achievement gaps? Do schools in whichprincipals display transformational leadership have smallerminority-non-minority achievement gaps?

MethodSchools in the studySchools under study were all elementary schools in a large metropolitan area,suburban school district. The schools varied in the sociodemographic make-upof the school structural, student population, and staff characteristics (Table I ).

School characteristic M SD Minimum Maximum

School structural characteristicsSchool enrollment 500.1 124.25 272 991Percentage of school utilization 97.8 14.91 60.85 131.7Class size 24.1 1.24 20.0 27.3Student to faculty ratio 11.3 1.85 6.9 15.9

School student populationPercentage of Students enrolled in FARMSprogram 31.5 22.4 1.9 92.9

Student racial/Ethnic identificationAfrican-American 20.1 13.8 2.1 72.2Asian-American 12.7 6.8 0.8 34.7Hispanic 16.5 14.1 1.7 61.7White 50.3 23.4 1.3 94.0

School staff characteristicsPercentage of staff ,1 year at school 23.78 8.68 2.9 42.0Number of staff per school who completedsurveys (N ) 25.06 6.46 8 44

Note: N ¼ 117 schools

Table I.School structural,

student population, andstaff characteristics for

schools under study

Principaltransformational

leadership

337

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 8: 09578230410534667

Analysis approachA structural equation model (SEM) (Arbuckle, 1997) was used to examine thedirect effect of principal transformational leadership on school staff turnoverand school performance. The SEM was also used to examine the indirect effectof staff job satisfaction on relations between principal transformationalleadership and school staff turnover and between principal transformationalleadership and school performance. Hierarchical liner modeling ( HLM )(Scientific Software International, 2000) was used to examine the cross-leveleffect of school staff job satisfaction and principal transformational leadershipon achievement disparities between the minority and non-minority students orthe variability in minority-achievement slopes across the schools ( Hofman andGavin, 1998).

Structural equation modelingThe school district in which the study was conducted administered annualsurveys to school staff and students. Surveys provided data to monitor schooland workplace environments for purposes of improvement. Survey items wereidentified by reviewing research literature relating to effective schools,school/organizational climate, employee opinion surveys, and educationalresearch incorporating surveys of school staff. Data obtained from surveys ofschool staff provided data for the SEM and HLM analyses. Surveys weremailed to the homes of all elementary school staff. School staff completed andreturned the surveys in postage-paid envelopes to a central research office. Ofthe 8,535 school-based employees surveyed in elementary schools, 3,291 staffmembers or 39 percent completed questionnaires. The median completion rateacross the schools was 38 percent. Each school, on an average, had 25 schoolstaff who completed the surveys, with a range of 8-44 staff. The distributions ofresponding staff members across job positions, seniority, and race/ethnic groupcategories were comparable to those of all school-based staff members.

Transformational leadership. The three components of transformationalleadership served as predictor variables of staff job satisfaction andorganizational performance. Survey items were chosen to represent the threecomponents of transformational leadership (Bass, 1985, 1990, 1996; Burns,1978):

(1) Charisma or inspiration. The ability of leaders to provide a clear sense ofmission, which they in turn convey to followers and develop a sense ofloyalty and commitment.

(2) Individualized consideration. The leader’s willing delegation of projectsto followers to stimulate and create learning experiences and the leader’streatment of each follower as unique individuals.

(3) Intellectual stimulation. The leader’s provision of opportunities forfollowers to rethink traditional procedures and to examine situations innew and novel ways.

JEA42,3

338

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 9: 09578230410534667

Table II displays the arrangement of survey items to represent thethree components of transformational leadership. Staff responses to itemscomprising of each component or scale showed adequate reliability. Cronbach’sa ranged from 0.86 to 0.96, and item-total correlations ranged from 0.57 to 0.84.

Job satisfaction. Earlier, job satisfaction was described as an importantmediating variable of the relation of leadership to organizational performance.Spector (1997) described job satisfaction as “a global feeling about the job or asa related constellation of attitudes about various aspects of facets of the job.The global approach is used . . . to find out which parts of the job produce

Scales/Sample itemsNumber of

items Cronbach’s a

Range ofitem-totalcorrelations

Charisma/Inspiration 4 0.86 0.67-0.73School goals and how to achieve are wellunderstoodGoals gives me sense of directionPrincipal encourages staff to talk aboutinstructionPrincipal gets staff to work together

Individualized consideration 10 0.94 0.57-0.83I can talk openly with principalPrincipal treats me with respectPrincipal supports me in matters of discipline,unreasonable demands from parentsPrincipal considers my suggestionsPrincipal lets me know when I am doinga good jobPrincipal lets me know what is expectedPrincipal provides feedback on jobperformancePrincipal understands my daily activitiesThere are school procedures to orient new staffNew staff get help from colleagues

Intellectual stimulation 4 0.89 0.66-0.84Principal encourages me to come with newideasI have opportunities to help develop schoolimprovement planPrincipal makes an effort to involve me indecisions affecting my workStaff are involved in decision-making thataffect their work

Staff job satisfaction 3 0.82 0.61-0.75I look forward to going to work each dayI would recommend my school as a goodplace to workI get a lot of satisfaction from my work

Table II.Survey items from each

scale and scalepsychometric properties

Principaltransformational

leadership

339

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 10: 09578230410534667

satisfaction or dissatisfaction” (Spector, 1997, pp. 2-3). Here, the interest was todetermine the relation of principal transformational leadership on school staffjob satisfaction, which in turn is presumed to affect job performance andturnover. To represent job satisfaction, three survey items were chosen fromthe ongoing school district staff surveys to form a scale of job satisfaction(Table II). The content of selected survey items comprising of the measure ofjob satisfaction is similar to the item content of other job satisfaction measuresin the management and organizational research literature, such as: “This is agood place to work” and “I often don’t think about leaving this job” (Barlinget al., 2003; Cook et al., 1981; Ostroff, 1992). The scales showed adequatereliability. The Cronbach a coefficient was 0.82, and the item-total correlationsranged from 0.61 to 0.75.

School staff turnover. Staff turnover served as one of the two outcomes in theSEM. Archival data were obtained on the percentage of school staff who hadbeen in the school for 1 year or less. These values served as the measure SchoolStaff Turnover. Such data gave an approximation of voluntary turnover (Homand Griffeth, 1995), realizing that some staff may have left for reasons ofretirements, medical leave, etc.

Organizational performance. School performance served as the second of thetwo outcomes in the SEM. School performance was calculated by averaging theperformance progress on standardized test scores for students in each school.To determine performance progress, initial (grade 3) test scores were regressedon current (grade 5) test scores. Residuals were used to determine how manyscale score points that each student’s test score was above or below the averagescore observed for a 5-year cohort of students with the same initial (grade 3)score. Values for students in each school were then summed up. Positive valuesindicated that on an average students performed better than students withsimilar initial performance levels. Negative values indicated that on an averagestudents performed worse than students with similar initial performance levels.Resulting school values were called School Achievement Progress and servedas the school performance measure. The advantage of this measure is that itrepresents an individual “change score” in which personal characteristics arekept constant across time, and individual changes are considered in relation toother students who had similar initial test scores [1].

Hierarchical linear modelingData obtained from surveys of students provided variables for the HLManalyses. Specifically, student responses provided variable values for derivingwithin-school minority-non-minority achievement gaps. Students enrolled ingrade 5 in all elementary schools were surveyed. Survey packets were sent tothe schools for distribution to students. Student questionnaires weregroup-administered by teachers to all students in their classrooms. Of the29,910 students in the 117 elementary schools, 25,087 completed the survey,

JEA42,3

340

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 11: 09578230410534667

representing an 84 percent completion rate. The median completion rate acrossthe 117 schools was 87 percent. Each school, on an average, had 79 studentswho completed the surveys, with a range of 35-146 students. The race andgrade composition of the student respondent sample was similar to that ofstudents enrolled in grade 5 in the elementary schools.

Level-1 predictor variables. Students provided responses to severalbackground questions in the survey. These variables were entered as Level-1predictor variables and served as controls in HLM analysis. Gender wasdetermined by student responses to, “Are you a boy or girl?” Girls were codedas 1s and boys as 0s. Minority status was determined by student responses to,“What is your racial/ethnic background?” Response categories included:African-American, American-Indian, Asian-American, Hispanic, White, andothers. Students who identified themselves as either African-American orHispanic were classified as minority (coded as 1s). Students who identifiedthemselves as either Asian-American or White were classified as non-minority(coded as 0s). Highest Expected Educational Level was determined by studentresponses to, “What is the highest educational level you will complete?”Response options were coded as not high school diploma ¼ 1; high schooldiploma ¼ 2; 2-year college degree, trade, or business degree ¼ 3; 4-yearcollege degree ¼ 4; and master’s or doctorate degree ¼ 5. Finally, studentsindicated how long they had been students at the school, called Years at School.

Level-1 outcome variable. The outcome of interest in HLM analyses wasstudent academic achievement, specifically, differences between the minorityand non-minority students within schools. Actual student achievement, eithergrade point average or scores on standardized achievement tests, could not beused in the analysis that included individual-level survey data. Concerns of theteachers’ union about uses of student perceptions to evaluate teacherperformance precluded collection of student identification numbers that couldbe matched to individual student archival data, which included test scores andgrade point averages. Student responses to the question, “This year, what didyou get on your report card?,” served as the outcome variable in Level-1 HLManalysis. Responses were coded as 1 ¼ mostly Es; 2 ¼ equal number of Ds andEs; mostly Ds ¼ 3; equal number of Cs and Ds ¼ 4; mostly Cs ¼ 5; equalnumber of Bs and Cs ¼ 6;mostly Bs ¼ 7; equal number of As and Bs ¼ 8; andmostly As ¼ 9: Responses were used as variable values for grade pointaverage (GPA). Despite problems typically associated with self-reportedperformance (namely, social desirability), the relation between aggregated,school-level self-reported GPA and performance on standardized achievementtests was reasonably high (r ð115Þ ¼ 0:66 for the school district’scriterion-referenced reading test scores; r ð115Þ ¼ 0:42 for the school district’scriterion-referenced math test scores). Because one research purpose was toexamine the relation of job satisfaction and principal transformationalleadership to the minority-minority achievement gap, it was important to

Principaltransformational

leadership

341

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 12: 09578230410534667

demonstrate that minority students reported lower GPAs than other students.The mean GPA for minority students was significantly lower (about onestandard deviation) (M ¼ 6:97; SD ¼ 1:49) than non-minority students’self-reported GPAs (M ¼ 7:99; SD ¼ 1:20) (tð6; 637Þ ¼ 30:97; p , 0:001;two-tailed).

Level-2 predictor variables. To discern the effects of staff job satisfaction andprincipal transformational leadership on the minority-non-minorityachievement gap, school mean values on job satisfaction and principaltransformational leadership were regressed on the within-slopes of the relationof minority status to student self-reported GPA (slopes-as-outcomes). TheLevel-2 predictor variable coefficients showed for staff job satisfaction andprincipal transformational leadership would indicate the effect of each on theminority-non-minority GPA gap. Rather than school means on each componentof principal transformational leadership, a summary score was used in HLManalyses. This was done for two reasons. First, measures of the components ofprincipal transformational leadership showed a high degree of intercorrelation(rs . 0:90), and if entered separately as predictor variables, would lead toproblems of multi-collinearity. Second, results from the SEM showed that theprincipal transformational leadership could be described as one latent variable.Using factor weights from the SEM, school values for the principaltransformational leadership were calculated and used in the HLM Level-2equation. To control the possible confounds of the school’s disadvantagedstudent population, the percentage of the school’s student population enrolledin the Free and Reduced-price Meals ( FARMS) program was also included as aLevel-2 predictor variable.

Justification for school aggregationSEM and HLM analyses relied on school-aggregated responses of school staffto surveys. Several statistical techniques were used to determine whetherindividual-level responses could be aggregated to represent group-level orschool characteristics. To justify group-level analyses, the aggregated datashould show ( Bliese, 2000) the following.

. Non-independence. A common statistical procedure to assessnon-independence is a one-way analysis of variance (ANOVA) in whichthe group serves as the independent variable and the ratings or scales ofinterest serve as the dependent variable (Kenny and Judd, 1986).Significance F-values indicate greater between-school than within-schooldifferences in individual responses and gives evidence for group-leveleffects (Dansereau and Alutto, 1990, pp. 206-207; Kenny and Judd, 1986).The intraclass correlation coefficient, the ICC(1), indicates the proportionof variance in individual responses explained by group membership(Bliese, 2000). The ICC(1) has values that range from 0.0 to 1.00. Typicalvalues range from 0.0 to 0.5, with a median value of 0.12 ( James, 1982).

JEA42,3

342

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 13: 09578230410534667

. Group-mean reliability. The ICC(2) is a common statistical assessment ofthe reliability of group means ( James et al., 1984; McGraw and Wong,1996; Ostroff, 1992). When the ICC(2) value is high, then responses from afew group members may be used to estimate the group’s mean response.When values of ICC(2) are close to 1.00, the better group means can bereliably differentiated. Ostroff (1993) suggests that values of ICC(2) be0.60 or higher indicate mean score reliability.

. Within-group agreement. Within-group agreement is commonly assessedby the rwg statistic ( James et al., 1984). The rwg has values ranging from0.00 to 1.00. The rwg is a ratio of the observed variation in responses to theresponses distributed evenly across the response categories. Lindell(2003) recommended values of rwg to be at least 0.50 to justify group-levelaggregation.

Statistical procedures provided evidence that individual staff responses couldbe aggregated to the school level and represent statistically reliable schoolattributes (Table III). All scales showed differences between schools ornon-independence (significant F-values). School membership accounted for13-14 percent of the variance in school staff scale scores (ICC(1) values). Staffscales showed group-mean reliability (ICC(2) values). All scales showedadequate within-group agreement, as indicated by the rwg values.

ResultsPrincipals as transformational leadersThe first research question was, “Can principals’ behaviors be described interms of the component parts of transformational leadership?” A confirmatoryfactor analysis was performed to determine how well ratings given to items byindividual respondents conformed to scales representing the three components.Overall, respondent ratings could be summarily described by their initialarrangement in three scales. Several fit statistics showed “good” fit; allexceeded 0.90: general fit index ðGFIÞ ¼ 0:97; normed fit index ðNFIÞ ¼ 0:98;Tucker-Lewis index ðTLIÞ ¼ 0:97; and comparative fit index ðCFIÞ ¼ 0:98: [2]The root mean squared error of approximation (RMSEA) also showed a

F ICC(1) ICC(2) rwgScales Significant F-valuea .0.12b .0.60c .0.50d

Charisma/Inspiration 5.16*** 0.13 0.81 0.58Individualized consideration 5.15*** 0.13 0.81 0.64Intellectual stimulation 5.09*** 0.13 0.80 0.53Staff job satisfaction 4.97*** 0.14 0.80 0.58

Notes: N ¼ 117 schools; school staff, N ¼ 1; 791; a Kenny and Judd (1986); b James et al. (1984);c Ostroff (1993); d Lindell (2003) and *** p , 0:001

Table III.Statistics for

determining theappropriate level of

analysis

Principaltransformational

leadership

343

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 14: 09578230410534667

reasonably good fit at 0.09. The chi-squared value, however, was statisticallysignificant ( x2ð134Þ ¼ 4037:13; p , 0:001), normally indicating a poor fit.However, large sample sizes often result in statistically significant chi-squaredvalues, even though the data adequately fit the model. Here, school staffnumbered 1,791.

Analytic approachSEM (Arbuckle, 1997) was performed to examine the relations among principalleadership, job satisfaction, school staff turnover, and school achievementprogress. Two structural equations models were tested. The first model(labeled Model A in Figure 2) included a test of the relations among principaltransformational leadership, school staff job satisfaction, and school staffturnover. The second model (labeled Model B in Figure 3) included a test of therelations among principal transformational leadership, school staff jobsatisfaction, and student achievement progress.

In both models, principal transformational leadership was represented as alatent variable consisting of school mean values on the scales ofcharisma/inspiration, individual consideration, and intellectual stimulation.

Figure 2.Model A: relationof principaltransformationalleadership to staff jobsatisfaction and schoolstaff turnover

Figure 3.Model B: relationof principaltransformationalleadership to staff jobsatisfaction to schoolachievement progress

JEA42,3

344

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 15: 09578230410534667

School mean values on the scales of staff job satisfaction served as the measureof school staff job satisfaction. The percentage of school staff new to the schoolserved as the measure of school staff turnover. School values on relativestudent achievement across two time periods (described earlier) served as themeasure of school achievement progress.

Relation of principal transformational leadership to staff job satisfaction andschool staff turnoverThe second research question was, “Do principals who displaytransformational leadership have school staff with higher levels of jobsatisfaction and less school staff turnover?” Model A (Figure 2) portrays therelations to be examined by this question. Data provided an “acceptable” fit toModel A (Figure 2). The chi-squared value was not statistically significant(x2ð5Þ ¼ 3:12; p , 0:68), and the several fit indices were well above 0.90:GFI ¼ 0:99; NFI ¼ 1:00; TLI ¼ 1:00; and CFI ¼ 1:00: The RMSEA alsoshowed a “good” fit at a value of 0.00. The three components oftransformational leadership contributed nearly equally to the principaltransformational leadership. The principal transformational leadership(standardized b . 0:90) showed a strong, positive and significant relation tostaff job satisfaction, which in turn showed a moderate, negative andsignificant relation to school staff turnover. Thus, schools in which principalswere perceived as transformational leaders had school staff who were moresatisfied with their jobs and had less staff turnover.

Relation of principal transformational leadership to staff job satisfaction andstudent achievement progressThe third research question was, “Do principals who display transformationalleadership have school staff with higher levels of job satisfaction andhigher-performing schools?” Model B (Figure 3) portrays the relations to beexamined by this question. Data provided a “good” fit to Model B. Thechi-squared value, was not statistically significant (x2ð5Þ ¼ 5:91; p , 0:32),and the several fit indices were well above 0.90: GFI ¼ 0:98; NFI ¼ 0:99;TLI ¼ 1:00; and CFI ¼ 1:00: The RMSEA also showed a “good” fit at a valueof 0.04. Again, the three components of transformational leadership contributednearly equally to the principal transformational leadership (as indicated by thestandardized bs). The principal transformational leadership showed a strong,positive and significant relation to the school staff job satisfaction, which inturn showed a moderate, positive and significant relation to the schoolachievement progress. Thus, schools in which principals were perceived astransformational leaders had school staff who were more satisfied with theirjobs and had greater achievement progress.

Principaltransformational

leadership

345

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 16: 09578230410534667

Indirect versus direct effects of principal transformational leadershipThe fourth research question was, “Does principal transformational leadershiprelate directly or indirectly (through staff job satisfaction) to employee turnoverand organizational performance?” The direct effect of principal leadership toboth school staff turnover and school achievement progress was tested byincluding these paths in the respective models. The data did not fit either model.For Model A, the chi-squared value was significant (x2ð4Þ ¼ 13:73; p , 0:01),and the RMSEA value was greater than 0.10, at a value of 0.15. In addition, thepath from the principal transformational leadership to school staff turnover wasnot statistically significant (standardized b ¼ 0:04; t ¼ 0:17). Other indiceswere, however, in the acceptable ranges: GFI ¼ 0:95; NFI ¼ 0:97; TLI ¼ 0:95;and CFI ¼ 0:98: For Model B, values for the fit statistics were very similar(x2ð4Þ ¼ 13:72; p , 0:01; GFI ¼ 0:95; NFI ¼ 0:97; TLI ¼ 0:95; and CFI ¼0:98; RMSEA ¼ 0:15). The path from the principal transformational leadershipto school achievement progress was not statistically significant (standardizedb ¼ 20:16; t ¼ 20:74). Formal tests showed statistically and significantindirect effects of the principal transformational leadership through staff jobsatisfaction on both school staff turnover (Sobel z-test¼22.12, p , 0:03) andschool achievement progress (Sobel z-test ¼ 2.38, p , 0:02) (Sobel, 1982).

Model fit for schools having high and low percent FARMS studentsThe fifth research question was, “Do relations among principaltransformational leadership, staff job satisfaction, school staff turnover, andschool performance vary by the student disadvantaged school population?” Toaddress this question, the sample of schools was split into two groups based onthe median percentage of students enrolled in the FARMS program. For eachgroup of schools, data were fit to Models A and B using SEM. Results arepresented in Table IV.

Data from schools having low percentages of FARMS students provided abetter fit to both models than schools having high percentages of FARMSstudents. For low-percentage FARMS schools, the chi-squared value was

Fit statisticModel x 2 df p GFI NFI TLI CFI RMSEA

Model ALeadership-staff job satisfaction-school staff turnoverLow percent FARMS ðN ¼ 59Þ 2.92 5 0.71 0.98 0.99 1.00 1.00 0.00High percent FARMS ðN ¼ 58Þ 13.70 5 0.02 0.91 0.95 0.93 0.96 0.18

Model BLeadership-staff job satisfaction-school achievement progressLow percent FARMS ðN ¼ 59Þ 4.06 5 0.54 0.98 0.98 0.96 1.00 0.00High percent FARMS ðN ¼ 58Þ 12.53 5 0.03 0.92 0.95 0.94 0.97 0.16

Table IV.Data-to-model fitstatistics for schoolshaving low and highpercentages of FARMSstudents

JEA42,3

346

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 17: 09578230410534667

statistically non-significant, and the RMSEA values were 0.00s. The remainingfit statistics also showed that the data provided a better fit for low-percentageFARMS schools. However, the path from staff job satisfaction to schoolachievement progress was statistically significant and positive amonghigh-percentage FARMS schools (standardized b ¼ 0:30; t ¼ 2:38; p , 0:05)than in low-percentage FARMS schools (standardized b ¼ 0:14; t ¼ 1:04; ns).This result suggests that principals who display characteristics oftransformational leaders may have more beneficial effects on studentachievement for schools having more disadvantaged students than schoolshaving less disadvantaged students. The next analysis further examines thisproposition.

The relation of staff job satisfaction and principal transformational leadershipto within-school minority/non-minority achievementThe sixth and final research questions were, “Do schools in which staff havehigher levels of job satisfaction have smaller minority achievement gaps?Do schools in which principals display transformational leadership havesmaller minority achievement gaps?” To provide results to address thesequestions, HLM analyses were performed. HLM allows examining the relationsof individual- and group-level characteristics on specified outcomessimultaneously. Here, in the Level-1 analysis, student self-reportedbackground characteristics, to include minority status, were used to predictthe self-reported GPA. Level-1 output included within-school differences inGPAs between theminority and non-minority students (i.e. slopes-as-outcomes).In the Level-2 analysis, school mean staff job satisfaction was used to predictthese slopes or the GPA gap between the minority and non-minority students.The percentage of students enrolled in FARMSwas also considered, mainly as acontrol variable.

Before proceeding with the analysis, an unconditional model was performedto determine the extent to which self-reported GPA varied across the schools.The resulting chi-squared value (x2ð116Þ ¼ 670:57; p , 0:001) showed thatschools differed in self-reported GPAs. In addition, school GPAs showedadequate reliability (0.82). A full HLM was then performed (Table V).

At level 1, student characteristics accounted for 14 percent of the variance instudent self-reported GPA, with minority status showing the strongest relation,followed by gender and expected educational level. At Level-2, within-schoolslopes served as the outcome and percentage of FARMS students and staff jobsatisfaction as predictor variables. Only staff job satisfaction showedstatistically significant relations to the within-school GPA gaps between theminority and non-minority students. Self-reported GPAs differed between theminority and non-minority students by 1/2-point on the 7-point response scale(see coefficient for mean minority GPA slope), and this difference was reducedby one-half in schools having more satisfied teachers (see coefficient for staff

Principaltransformational

leadership

347

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 18: 09578230410534667

job satisfaction). Similar results were observed when percentage of FARMSstudents and principal transformational leadership were entered as predictorvariables (see the last rows in Table V).

To discern the effect of the combination of staff job satisfaction andprincipal transformational leadership on the minority-non-minority GPA gap,the main effects of staff job satisfaction and principal transformationalleadership and their interaction (i.e. product of school values on eachvariable) were entered. The interaction term, while in the predicted direction,did not meet the traditional standards of statistical significance

Predictor variableUnstandardized

coefficientStandarderror t-ratio p-value

Level 1 (student)GPA intercept 7.63 0.028 270.17 ,0.000Gender (female)a 20.341 0.036 29.44 ,0.000Minorityb 20.646 0.047 213.77 ,0.000Highest expected educational levelc 0.295 0.021 14.37 ,0.000

Years at school 0.045 0.011 4.09 ,0.000Percent variance explained level-1variables 14.3 percent

First HLM Level 2 (school)GPA intercept 7.63 0.028 273.43 ,0.000Percent FARMS 20.014 0.001 210.97 ,0.000Staff job satisfaction 0.060 0.108 0.56 ,0.574

Mean minority-non-minority GPA slope 20.648 0.042 215.54 ,0.000Percent FARMS 0.004 0.002 1.88 ,0.060Staff job satisfaction 0.363 0.153 2.36 ,0.019

Percent variance explained byLevel-2 variables: 69.6 percent

Second HLM Level 2 (school)GPA intercept 7.63 0.028 273.58 ,0.000Percent FARMS 20.014 0.001 212.10 ,0.001Transformational leadership 0.004 0.098 0.043 ,0.966

Mean minority-non-minority GPA slope 20.646 0.046 213.77 ,0.000Percent FARMS 0.003 0.002 1.51 ,0.132Transformational leadership 0.322 0.163 1.99 ,0.047

Percent variance explained byLevel-2 variables 69.6 percent

Notes: N ¼ 117 schools; N ¼ 7; 910 students; student-level predictors are grand-mean centered,except for minority status, which is group-mean centered; school-level predictors are grand-meancentered; aGender was coded as girl ¼ 1; boy ¼ 0; bMinority status was coded as minority(African-American or HispanicÞ ¼ 1; white and Asian-American ¼ 0; cHighest ExpectedEducational Level was coded as not high school diploma ¼ 1; high school diploma ¼ 2; 2-yearcollege degree, trade, or business degree ¼ 3; 4-year college degree ¼ 4; and master’s ordoctorate degree ¼ 5

Table V.HLM: relation of staffjob satisfaction towithin-schoolminority-non-minorityGPA gap(“slopes-as-outcomes”)

JEA42,3

348

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 19: 09578230410534667

(standardized b ¼ 0:588; tð112Þ ¼ 1:51; p , 0:13). The interaction, whenplotted, was consistent with the expectation (Figure 4). Schools were dividedinto two groups based on a median-split of school-level principaltransformational leadership factor score (based on the earlier SEM). A plotof the relation of staff job satisfaction to the minority GPA gap showed thegap narrowed in schools having more satisfied teachers, and this effect wasmore evident (steeper slope) in schools having higher levels of principaltransformational leadership.

DiscussionResults here add to the evidence that the theory of transformational leadershipdescribes effective leadership in a variety of settings, including publiceducational settings. Staff reports of principal behaviors could be described interms of three components of transformational leadership: inspiration orcharisma, individualized consideration, and intellectual stimulation. Principaltransformational leadership was not associated directly with either school staffturnover or school performance. Rather, principal transformational leadershipshowed indirect effects, through staff job satisfaction, on both school staffturnover (negative) and school performance ( positive).

Role of transformational leadership in organizational performanceResults observed here have strong intuitive appeal. Job tasks of the principalpertain more to interactions with staff than with students, and thus, principals

Figure 4.Relation of staff job

satisfaction towithin-school

minority-non-minorityGPA differences

(slopes-as-outcomes inHLM analysis) for

schools having low andhigh levels of principal

transformationalleadership

Principaltransformational

leadership

349

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 20: 09578230410534667

spend more of their time with school staff than with students. The nature oftheir interactions with staff would be expected to affect staff’s workexperiences, including working relations among staff. For instance, principals,by including staff in the planning, problem-solving, decision-making andimplementing of school programs, likely leads to greater job satisfaction,commitment, and motivation among staff (Dinham and Scott, 1998, 1999, 2000,2003). Also, there is a likely better communication among staff, greater mutualtrust and understanding, greater cooperation and collaboration, and moreactive engagement of staff. In turn, higher levels of job satisfaction andcooperative working relationships would be expected to lead to a betterimplementation of school programs and their intended effects. The extent towhich school performance is based on the outcomes of such programs, the morethe school would be expected to be effective. Such linkages are not only evidentin recent educational studies of principal leadership (Hallinger et al., 1996;Leithwood and Jantzi, 1990), but also in organizational and managementstudies of the relation of employee perceptions of their jobs, workplace, andleaders to organizational performance (Koys, 2001; Ostroff, 1992; Ryan et al.,1996).

These linkages appear to be important considerations in reducing theminority-non-minority achievement gap. In HLM analyses, both staff jobsatisfaction and principal transformational leadership were associated withsmaller achievement gaps between the minority and non-minority students.Gaps tended to be smallest in schools having more satisfied staff who alsoviewed their principals as transformational leaders. These relations may haveto do with the positive affect evident in staff’s satisfaction with their work. Thisaffect may translate into positive classroom and school climate, which paststudies have found conducive to positive learning and achievement. Indeed,several contemporary explanations of minority students’ under-performance,such as Comer’s (1988) School Project and Steele’s (1997) “stereotype threat”theory, rely heavily on behavioral and affective aspects of the teacher-studentrelationship for effective student learning, and have increasingly accumulatedempirical support. Furthermore, the positive benefits of communal, expressiveschool environments on student learning have been documented ineducational studies, in particular, for schools having high percentages ofsocio-economically-disadvantaged students (Battistich et al., 1995; Brookoveret al., 1978; Shouse, 1996). Though speculative, this may explain why smallerachievement gaps between the minority and non-minority students wereobserved in schools having higher levels of job satisfaction and principaltransformational leadership. It seems, then, that transformational leadership ismore directly related to organizational processes associated with employeebehaviors, morale, and satisfaction, which in turn are related to the quality ofservice delivery and organizational performance (Schneider, 1990; Schneideret al., 1998; Tornow and Wiley, 1991).

JEA42,3

350

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 21: 09578230410534667

Role of transformational leadership in organizational effectiveness and leaderselectionMost of the existing measures of school effectiveness use student achievementtest scores aggregated to school level (Malen and Fuhrman, 1991; Office ofEducational Research and Improvement, 1988). Yet, much of the organizationaland management literature, in addition to the educational administrationliterature, portrays principal leadership as integral to school effectiveness(Sergiovanni, 1995). Indeed, results here support this perspective: principaltransformational leadership had an indirect effect on school-aggregatedstudent achievement progress through staff’s satisfaction with work.Consideration should be given to the use of staff perceptions of the qualityof principal leadership as additional measures of school effectiveness.Such measures broaden the perspective of what constitutes effectiveschools and include important measures of school processes andimplementation traditionally viewed as necessary for school improvementand effectiveness (Brown, 1982). Consideration should also be given to theuse of transformational leadership for the selection, training, and evaluationof leaders, in particular, in educational settings. The components oftransformational leadership might serve as the basis for rating and selectingapplicants for principals. Personal interviews and paper-and-pencilassessments (Bass, 1985) might be structured to indicate candidates’inspiration, individualized consideration, and intellectual stimulation.Similarly, the three components of transformational leadership might serveas the basis for developing the content of training and continuing education forcurrent and prospective principals. Finally, criteria for job evaluations might bedeveloped around the components of transformational leadership. Trainingcould be prescribed for principals in areas having lower ratings.

Study limitationsThe present study does not go without limitations. The given data wereobtained at one point in time, results cannot reveal causality among observedrelations. Higher levels of student achievement may result in school staff’smore positive evaluation of the principal and school environment (e.g. lesseffort to instruct, more immediate evidence of being effective, etc.). The relationof leadership to organizational performance and turnover may also berecursive. Improved student performance and less staff turnover may result inpositive perceptions about the principal, which, in turn, increases staffcommitment. Unmeasured aspects of the school environment may also explainrelations among leadership, performance, and turnover, such as materialsupport, community support, etc. An area for future research is to sort out theserelations by using alternative sources to obtain data on leadership,organizational climate, performance, and turnover across several time periods.

Change scores were used in the SEM analyses and have limitations.Discernible change is, in part, determined by the extent the pretest and

Principaltransformational

leadership

351

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 22: 09578230410534667

post-test scores correlate. For example, as the correlation between pretestand post-test scores decreases, then the relative error in change scores alsodecreases. Changes are more likely to be detected, but their meaning issomewhat obscure. The lack of correlation between the tests may indicate thatthe tests assess different constructs. Conversely, as the correlation between thepretest and post-test scores increase, then the relative error in change scoresincreases, reducing the likelihood of detecting change (Collins and Horn, 1991).In addition, change scores may not always mean the same thing. A smallchange that occurs when initial achievement is high may not have the samemeaning as the same change when initial achievement is more moderate.Change scores are also susceptible to statistical regression. Initial large initialvalues tend to have later smaller values and initial small values tend to havelater larger values (Finkel, 1995).

A final study limitation relates to the outcomes used in the HLM analyses.Students reported their GPAs. Retrospective reporting of GPAs maypotentially contribute to the unreliability of, such as failing memory, socialdesirability, and so on. As reported earlier, school aggregated GPAs did,however, correspond well with school means on standardized achievementtests.

Notes

1. This approach to school-aggregated progress in student achievement is consistent with thecurrent methods to derive measures of school effectiveness (Fitz-Gibbon, 1996). Theadvantage of this measure achievement is that each school serves as its own comparison or“control” across time. Additionally, the analysis more closely approximates a correlationaldesign allowing for causal inference. That is, current school processes are regressed onchanges in student achievement from the previous test administration to the current testadministration (Allison, 1990; Finkel, 1995).

2. Kenny (2001) has recommended an “acceptable” model fit to the data when the chi-squaredstatistic test is nonsignificant, values for the CFI and NFI are between 0.90 and 0.95, valuesfor the (RMSEA) are between 0.05 and 0.10. A “good” model fit to the data, according toKenny, is when the chi-squared statistic test is non-significant, values for the CFI and NFIare greater than 0.95, and values for the RMSEA are less than 0.05.

References

Allison, P.D. (1990), “Change scores as dependent variables in regression analysis”, in Clogg, C.C.(Ed.), Sociological Methodology 1990, Basil Blackwell, Oxford, pp. 93-114.

Arbuckle, J.L. (1997), AMOS Users’ Guide Version 3.6, SmallWaters Corporation, Chicago, IL.

Bass, B.M. (1985), Leadership and Performance beyond Expectation, Free Press, New York, NY.

Bass, B.M. (1990), “From transactional to transformational leadership: good, better, best”,Organizational Dynamics, Vol. 3 No. 3, pp. 26-40.

Bass, B.M. (1996), A New Paradigm for Leadership: An Inquiry into TransformationalLeadership, US Army Research Institute, Alexandria, VA.

Bass, B.M., Avolio, B.J. and Goodheim, L. (1987), “Quantitative description of world-classindustrial, political, and military leaders”, Journal of Management, Vol. 13, pp. 7-19.

JEA42,3

352

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 23: 09578230410534667

Battistich, V., Solomon, D., Kim, D., Watson, M. and Schaps, E. (1995), “Schools as communities,poverty levels of student populations, and students’ attitudes, motives, and performance: amultilevel analysis”, American Educational Research Journal, Vol. 32 No. 3, pp. 627-58.

Barling, J., Kelloway, E.K. and Iverson, R.D. (2003), “High-quality work, job satisfaction, andoccupational injuries”, Journal of Applied Psychology, Vol. 88 No. 2, pp. 276-83.

Blase, J.J. (1987), “Dimensions of effective school leadership: the teacher’s perspective”, AmericanEducational Research Journal, Vol. 24 No. 4, pp. 589-610.

Blase, J.J., Derick, C. and Strathe, M. (1986), “Leadership behavior of school principals in relationto teacher stress, satisfaction, and performance”, Journal of Humanistic Counseling,Education, and Development, Vol. 24 No. 4, pp. 159-71.

Bliese, P. (2000), “Within-group agreement, non-independence, and reliability: implications fordata aggregation and analysis”, in Klein, K.J. and Kozlowski, S.W. (Eds),Multilevel TheoryResearch, and Methods in Organizations: Foundations, Extension, and New Directions,Jossey-Bass, San Francisco, CA, pp. 349-81.

Bossert, S., Dwyer, D., Rowan, B. and Lee, G. (1982), “The instructional management role of theprincipal”, Educational Administration Quarterly, Vol. 18, pp. 34-64.

Brookover, W.B., Schweitzer, J.H., Schneider, J.M., Beady, C.H., Flood, P.K. andWeisenbaker, J.M.(1978), “Elementary school social climate and school achievement”, American EducationalResearch Journal, Vol. 15, pp. 301-18.

Brown, M.C. (1982), “Administrative succession and organizational performance: the successioneffect”, Administrative Science Quarterly, Vol. 27, pp. 1-16.

Bryk, A., Lee, V. and Smith, J. (1990), “High school organization and its effects on teachers andstudents: an interpretative summary of the research”, in Clune, W.H. and Witte, J.F. (Eds),Choice and Control in American Education, Volume 1: The Theory of Choice and Control inEducation, Falmer Press, New York, NY.

Burns, J. (1978), Leadership, Harper and Row, New York, NY.

Ceci, S.J., Rosenblum, T.B. and Kumpf, M. (1999), “The shrinking gap between high- andlow-scoring groups: current trends and possible causes”, in Neisser, U. (Ed.), The RisingCurve: Long-term Gains in IQ and Related Measures, American Psychological Association,Washington, DC, pp. 287-302.

Collins, L.M. and Horn, J.L. (1991), Best Methods for the Analysis of Change, AmericanPsychological Association, Washington, DC.

Comer, J. (1988), “Educating poor minority children”, Scientific American, Vol. 259, p. 42.

Cook, J.D., Hepworth, S.J., Wall, T.D. and Warr, P.B. (1981), The Experience of Work: ACompendium and Review of 249Measures and Their Use, Academic Press, New York, NY.

Dansereau, F. and Alutto, J.A. (1990), “Levels-of-analysis issues in climate and culture research”,in Schneider, B. (Ed.), Organizational Climate and Culture, Jossey-Bass, San Francisco, CA,pp. 193-236.

Dinham, S. and Scott, C. (1998), “A three domain model of teacher and school executive careersatisfaction”, Journal of Education Administration, Vol. 36 No. 4, pp. 362-78.

Dinham, S. and Scott, C. (2000), “Moving into the third, outer domain of teacher satisfaction”,Journal of Educational Administration, Vol. 38 No. 4, pp. 379-96.

Finkel, S.E. (1995), Causal Analysis with Panel Data, Sage, Newbury Park, CA.

Fitz-Gibbon, C.T. (1996), Monitoring Education: Indicators, Quality and Effectiveness, Cassell,London.

Principaltransformational

leadership

353

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 24: 09578230410534667

Flynn, J. (1984), “The mean IQ of Americans: massive gains”, Psychological Bulletin, Vol. 95,pp. 29-51.

Grissmer, D.W., Williamson, S., Kirby, S.N. and Berends, M. (1999), “Exploring the rapid rise inblack achievement scores in the United States (1970-1990)”, in Neisser, U. (Ed.), The RisingCurve: Long-term Gains in IQ and Related Measures, American Psychological Association,Washington, DC, pp. 251-85.

Hallinger, P., Bickman, L. and Davis, K. (1996), “School context, principal leadership, and studentreading achievement”, The Elementary School Journal, Vol. 96 No. 5, pp. 527-49.

Hallinger, P. and Heck, R.H. (1996), “Reassessing the principal’s role in school effectiveness: areview of empirical research, 1980-1995”, Educational Administration Quarterly, Vol. 32No. 1, pp. 5-44.

Halpin, A.W. (1966), Theory and Research in Administration, Macmillan, New York, NY.

Hofman, D.A. and Gavin, M.B. (1998), “Centering decisions in hierarchical linear models:implications for research in organizations”, Journal of Management, Vol. 24 No. 5,pp. 623-41.

Hom, P. and Griffeth, R. (1995), Employee Turnover, South-Western Publishing, Cincinnati, OH.

James, L.R. (1982), “Aggregation bias in estimates of perceptual agreement”, Journal of AppliedPsychology, Vol. 67, pp. 219-29.

James, L.R., Demaree, R.G. and Wolf, G. (1984), “Estimating within-group interrater agreement”,Journal of Applied Psychology, Vol. 78, pp. 306-9.

Kane, T.D. and Tremble, T.R. (2000), “Transformational leadership effects at different levels ofthe US Army”, Military Psychology, Vol. 12 No. 2, pp. 137-60.

Kenny, D.A. (2001), SEM: Measures of Fit, available at: http://users.rcn.com/dakenny/fit.htm

Kenny, D.A. and Judd, C.M. (1986), “Consequences of violating the independence assumption inanalysis of variance”, Psychological Bulletin, Vol. 99, pp. 422-31.

Koh, W.L., Steers, R.M. and Terborg, J.R. (1995), “The effects of transformational leadership toteacher attitudes and student performance in Singapore”, Journal of OrganizationalBehavior, Vol. 16, pp. 319-33.

Koys, D.J. (2001), “The effects of employee satisfaction, organizational citizenship behavior, andturnover on organizational effectiveness: a unit-level, longitudinal study”, PersonnelPsychology, Vol. 54 No. 1, pp. 101-14.

Lee, J. (2001), “Statewide educational staffing”, in Minority Achievement in Maryland at theMillennium: A Special Report, Maryland State Department of Education, Baltimore, MD,pp. 15-25.

Lindell, M.K. (2003), “A guide on how to use rwg”, unpublished.

Leithwood, K. and Jantzi, D. (1999), “Transformational school leadership effects: a replication”,School Effectiveness and School Improvement Journal, Vol. 10 No. 4, pp. 451-79.

Loehlin, J.C., Lindsay, G. and Spuhler, J. (1975), Race Differences in Intelligence, Freeman,New York, NY.

Malen, B. and Fuhrman, S.H. (1991), “The politics of curriculum and testing: introduction andoverview”, in Fuhrman, S.H. and Malen, B. (Eds), The Politics of Curriculum and Testing,Falmer Press, London, pp. 1-9.

McGraw, K.O. and Wong, S.P. (1996), “Forming inferences about some intraclass correlationcoefficients”, Psychological Bulletin, Vol. 109 No. 1, pp. 30-46.

JEA42,3

354

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 25: 09578230410534667

Neisser, U., Boodoo, G., Bouchard, T.J., Boykin, A.W., Brody, N., Ceci, S.J., Halpern, D.F.,Loehlin, J.C., Perloff, R., Sternberg, R.J. and Urbina, S. (1996), “Intelligence: knowns andunknowns”, American Psychologist, Vol. 51, pp. 77-101.

Office of Educational Research and Improvement (1988), Creating Responsible and ResponsiveAccountability Systems, US Department of Education, Washington, DC.

Ostroff, C. (1992), “The relationship between satisfaction, attitudes, and performance: anorganizational level analysis”, Journal of Applied Psychology, Vol. 77 No. 6, pp. 963-74.

Ostroff, C. (1993), “Comparing correlations based on individual-level and aggregated data”,Journal of Applied Psychology, Vol. 78, pp. 569-82.

Ryan, A.M., Schmit, M.J. and Johnson, R. (1996), “Attitudes and effectiveness: examiningrelations at an organizational level”, Personnel Psychology, Vol. 49.

Scientific Software International (2000), HLM 5: Hierarchical Linear and Nonlinear Modeling,Scientific Software International, Chicago, IL.

Schneider, B. (1990), “The climate for service: an application of the climate construct”, inSchneider, B. (Ed.), Organizational Climate and Culture, Jossey-Bass, San Francisco, CA,pp. 383-412.

Schneider, B., White, S.S. and Paul, M.C. (1998), “Linking service climate and customerperceptions of service quality: tests of a causal model”, Journal of Applied Psychology,Vol. 83, pp. 150-63.

Shouse, R.C. (1996), “Academic press and sense of community: conflict, congruence, andimplications for student achievement”, Social Psychology of Education, Vol. 1, pp. 47-68.

Sergiovanni, T.J. (1995), The Principalship: A Reflective Practice Perspective, 3rd ed., Allyn andBacon, Boston, MA.

Sobel, M.E. (1982), “Asymptotic intervals for indirect effects in structural equation models”, inLeinhart, S. (Ed.), Sociological Methodology 1982, Jossey-Bass, San Francisco, CA,pp. 290-312, available at: http://quantrm2.psy.ohio-state.edu/kris/sobel/sobel.htm

Spector, P.E. (1997), Job Satisfaction: Application, Assessment, Cause, and Consequences, Sage,Thousand Oaks, CA.

Steele, C.M. (1997), “A threat in the air: how stereotypes shape intellectual identities andperformance of women and African Americans”, American Psychologist, Vol. 52,pp. 613-29.

Tornow, W.W. and Wiley, J.W. (1991), “Service quality and management practices: a look atemployee attitudes, customer satisfaction and bottom-line consequences”, HumanResource Planning, Vol. 14, pp. 105-16.

Waldman, D.A., Bass, B.M. and Einstein, W.O. (1987), “Effort, performance and transformationalleadership in industrial and military settings”, Journal of Occupational Psychology, Vol. 60,pp. 177-86.

Further reading

Avolio, B.J. and Bass, B.M. (1987), “Charisma and beyond”, in Hunt, J.G. (Ed.), EmergingLeadership Vistas, Lexington, Boston, MA, pp. 29-49.

Guthrie, J.P. (2001), “High involvement work practices, turnover, and productivity: evidence fromNew Zealand”, Academy of Management Journal, Vol. 44, pp. 180-90.

Hater, J.J. and Bass, B.M. (1988), “Superiors’ evaluations and subordinates’ perceptions oftransformational and transactional leadership”, Journal of Applied Psychology, Vol. 73No. 4, pp. 695-702.

Principaltransformational

leadership

355

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 26: 09578230410534667

Huselid, M.A. (1995), “The impact of human resources management practices and turnover,productivity, and corporate financial performance”, Academy of Management Journal,Vol. 38, pp. 635-72.

Jantzi, D. and Leithwood, K. (1996), “Towards an explanation of variation in teachers’perceptions of transformational school leadership”, Educational Administration Quarterly,Vol. 32 No. 4, pp. 512-38.

Pitner, N. (1988), “The study of administrator effects and effectiveness”, in Boyan, N. (Ed.),Handbook of Research in Educational Administration, Longman, New York, NY,pp. 99-122.

Scott, C. and Dinham, S. (1999), “The occupational motivation, satisfaction and health of Englishschool teachers”, Educational Psychology, Vol. 19 No. 3, pp. 287-309.

Scott, C. and Dinham, S. (2003), “The development of scales to measure teacher and schoolexecutive occupational satisfaction”, Journal of Educational Administration, Vol. 41 No. 1,pp. 74-86.

Appendix

Study Intercorrelations

Scales/variables M SD 1 2 3 4 5

Charisma/Inspiration 3.03 0.28 1.00Individualized consideration 2.87 0.33 0.84** 1.00Intellectual stimulation 3.09 0.37 0.85** 0.83** 1.00Staff job satisfaction 3.14 0.27 0.78** 0.76** 0.87** 1.00School staff turnover 11.73 7.36 20.31** 20.24** 20.37** 20.41** 1.00School achievement progress 20.551 8.48 0.30** 0.21* 0.28** 20.36** 20.03

Notes: N ¼ 117 schools; *p , 0.05 and **p , 0.01 two-tailed

Table AI.Intercorrelationsamong study scalesand variables

JEA42,3

356

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 27: 09578230410534667

This article has been cited by:

1. Kostas Dimopoulos, Katerina Dalkavouki, Vasilis Koulaidis. 2014. Job realities of primary school principalsin Greece: similarities and variations in a highly centralized system. International Journal of Leadership inEducation 1-28. [CrossRef]

2. Maria Eliophotou Menon. 2014. The relationship between transformational leadership, perceived leadereffectiveness and teachers’ job satisfaction. Journal of Educational Administration 52:4, 509-528. [Abstract][Full Text] [PDF]

3. Bolanle Ogungbamila, Emmanuel A Fayankinnu. 2014. Some psycho-social factors fostering workplacecommitment among head teachers in Nigeria. International Journal of Research Studies in Psychology 3:4. .[CrossRef]

4. Jiafang Lu, Xinhui Jiang, Huen Yu, Dongyu Li. 2014. Building collaborative structures for teachers’autonomy and self-efficacy: the mediating role of participative management and learning culture. SchoolEffectiveness and School Improvement 1-18. [CrossRef]

5. Khalid Arar. 2014. Deputy-principals in Arab schools in Israel: an era of reform. International Journal ofEducational Management 28:1, 96-113. [Abstract] [Full Text] [PDF]

6. Fabio Alivernini. 2013. An exploration of the gap between highest and lowest ability readers across 20countries. Educational Studies 39:4, 399-417. [CrossRef]

7. Herman H.M. Tse, Xu Huang, Wing Lam. 2013. Why does transformational leadership matter foremployee turnover? A multi-foci social exchange perspective. The Leadership Quarterly 24:5, 763-776.[CrossRef]

8. Ibrahim Duyar, Sedat Gumus, Mehmet Sukru Bellibas. 2013. Multilevel analysis of teacher work attitudes.International Journal of Educational Management 27:7, 700-719. [Abstract] [Full Text] [PDF]

9. Robin Stanley Snell, Zhang Yi, Almaz M.K. Chak. 2013. Representational predicaments for employees:their impact on perceptions of supervisors' individualized consideration and on employee job satisfaction.The International Journal of Human Resource Management 24:8, 1646-1670. [CrossRef]

10. Robin Stanley Snell, May Mei-ling Wong, Almaz Man-kuen Chak, Sandy Suk-kwan Hui. 2013.Representational predicaments at work: How they are experienced and why they may happen. Asia PacificJournal of Management 30:1, 251-279. [CrossRef]

11. Anecia Robyn, Ronel Du Preez. 2013. Intention to quit amongst Generation Y academics in highereducation. SA Journal of Industrial Psychology 39:1. . [CrossRef]

12. Ali Ibrahim, Shaikah Al-Taneiji. 2013. Principal leadership style, school performance, and principaleffectiveness in Dubai schools. International Journal of Research Studies in Education 2:1. . [CrossRef]

13. Ashley Oleszewski, Alan Shoho, Bruce Barnett. 2012. The development of assistant principals: a literaturereview. Journal of Educational Administration 50:3, 264-286. [Abstract] [Full Text] [PDF]

14. Ori Eyal, Guy Roth. 2011. Principals' leadership and teachers' motivation. Journal of EducationalAdministration 49:3, 256-275. [Abstract] [Full Text] [PDF]

15. María Paola Sevilla. 2011. Liderazgo directivo y resultados de los estudiantes: evidencia a partir de laAsignación de Desempeño Colectivo. Un análisis del período 2005-2008. Pensamiento Educativo: Revistade Investigación Educacional Latinoamericana 48:1, 1-14. [CrossRef]

16. Janelle E. Wells, Jon Welty Peachey. 2011. Turnover intentions. Team Performance Management: AnInternational Journal 17:1/2, 23-40. [Abstract] [Full Text] [PDF]

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)

Page 28: 09578230410534667

17. Zaidatol Akmaliah Lope Pihie, Amir Sadeghi, Habibah Elias. 2011. Analysis of Head of DepartmentsLeadership Styles: Implication for Improving Research University Management Practices. Procedia -Social and Behavioral Sciences 29, 1081-1090. [CrossRef]

18. Daniella Smith. 2010. Making the case for the leadership role of school librarians in technologyintegration. Library Hi Tech 28:4, 617-631. [Abstract] [Full Text] [PDF]

19. Kara S. Finnigan. 2010. Principal Leadership and Teacher Motivation under High-Stakes AccountabilityPolicies. Leadership and Policy in Schools 9:2, 161-189. [CrossRef]

20. DANIEL S. WHITMAN, DAVID L. VAN ROOY, CHOCKALINGAM VISWESVARAN. 2010.SATISFACTION, CITIZENSHIP BEHAVIORS, AND PERFORMANCE IN WORK UNITS: AMETA-ANALYSIS OF COLLECTIVE CONSTRUCT RELATIONS. Personnel Psychology 63:1,41-81. [CrossRef]

21. Aamir Ali Chughtai, Finian Buckley. 2009. Linking trust in the principal to school outcomes.International Journal of Educational Management 23:7, 574-589. [Abstract] [Full Text] [PDF]

22. Paula Kwan. 2009. Beginning teachers' perceptions of school human resource practices. Asia Pacific Journalof Education 29:3, 373-386. [CrossRef]

23. John Patrick, Gemma Scrase, Afia Ahmed, Michal Tombs. 2009. Effectiveness of instructor behavioursand their relationship to leadership. Journal of Occupational and Organizational Psychology 82:3, 491-509.[CrossRef]

24. John Chi‐Kin Lee, Paula Kwan, Allan Walker. 2009. Vice‐principalship: their responsibility roles andcareer aspirations. International Journal of Leadership in Education 12:2, 187-207. [CrossRef]

25. Paula Kwan. 2009. Vice‐principals' dilemma – career advancement or harmonious working relationship.International Journal of Educational Management 23:3, 203-216. [Abstract] [Full Text] [PDF]

26. Paula Kwan. 2009. The vice‐principal experience as a preparation for the principalship. Journal ofEducational Administration 47:2, 191-205. [Abstract] [Full Text] [PDF]

27. Nadine Engels, Gwendoline Hotton, Geert Devos, Dave Bouckenooghe, Antonia Aelterman. 2008.Principals in schools with a positive school culture. Educational Studies 34:3, 159-174. [CrossRef]

28. Chien-Wen TsaiLeadership style and employee's job satisfaction in international tourist hotels 293-332.[Abstract] [Full Text] [PDF] [PDF]

29. Kenneth Leithwood, Doris Jantzi. 2005. A Review of Transformational School Leadership Research 1996–2005. Leadership and Policy in Schools 4:3, 177-199. [CrossRef]

Dow

nloa

ded

by U

NIV

ER

SIT

I U

TA

RA

MA

LA

YSI

A A

t 03:

40 1

8 N

ovem

ber

2014

(PT

)