8/9/2019 e Capacity

1/13

The e-capacity of primary schools: Development of a conceptual model and scaleconstruction from a school improvement perspective

Ruben Vanderlinde * , Johan van BraakDepartment of Educational Studies, Ghent University, Henri Dunantlaan 2, B9000 Ghent, Belgium

a r t i c l e i n f o

Article history:Received 19 June 2009Received in revised form8 January 2010Accepted 17 February 2010

Keywords:ICT integrationConceptual frameworkSchool improvementScale constructionFactor analysis

a b s t r a c t

In the search for factors affecting the use of ICT in educational settings, several authors have presentedholistic conceptual frameworks. In this study, we argue that while these models are valuable sources forconducting qualitative research, they are less useful for quantitative research since few measurementscales have been created. We present an empirically tested conceptual framework to further examine thecomplex process of integrating ICT for instructional purposes. This model has been developed froma school improvement perspective and considers the e-capacity of a school as an overarching concept.E-capacity refers to the schools' ability to create and optimise sustainable school level and teacher levelconditions to bring about effective ICT change. The conditions identi ed are based on a literature reviewin the change and school improvement literature and the ICT integration literature. All conditions havebeen translated into reliable measurement scales. Questionnaire data were collected from a represen-tative teacher sample ( N 471) in 62 primary school in Belgium (Flanders). Exploratory and con r-matory factor analyses were conducted indicating good goodness of t estimates and good internalconsistency.

2010 Elsevier Ltd. All rights reserved.

1. Introduction

One of the central activities in Information and Communications Technology (ICT) research is the investigation of conditions that supportthe integration of ICT into schools ( Hew & Brush, 2007 ). While research has traditionally focused on individual teacher characteristics orconditions at the teacher level, such as individual computer attitudes or genderdifferences, we believe that this ignores the social context inwhich teachers behave. In a recent review study, Hew and Brush (2007) show that the majority of ICT integration research primarily focuseson the role of teacher level variables, and that few studies examine important school level variables that may affect the integration of ICT. Inline with this view, Hermans, Tondeur, van Braak, and Valcke (2008) argue that future research should focus on speci c school conditionsand school culture variables that may explain the use of ICT in classrooms. Other authors argue that ICT integration should be considered asa special case of educational innovation, and that research on ICT integration should build on theories and insights from the educationalchange and school improvement literature (e.g. Tearle, 2004; Watson, 2006 ). Furthermore, Somekh (2007) argues that we should adoptstrategies about successful management of change to enable the use of ICT to support effective teaching and learning. We believe thatresearch on ICT integration should increase its focus on both the role of characteristics of the school organisation or school level conditions,and on ICT integration as a case of educational innovation. The rst research focus implies that ICTresearchers should pay more attention tothe role of school organisational features, the second research challenge means we endorse ICT can have a positive impact on studentlearning, like recently provided by rigorous research evidence (e.g. BECTA, 2007).

Moreover, ICT itself is constantlyand rapidly evolving. Hardware and software are changing very quickly and schools are also confrontedwith new technological developments, such as web 2.0. In certain countries national governments have even administered formal andcompulsory ICT curricula. These curricula have a clear pedagogical foundation and tend to focus more on the use of ICT as a tool for teachingand learning than the development of technical skills. Vanderlinde, van Braak, and Hermans (2009) argue that such ICT curricula causea shift in the policy actions of ICT support (i.e., from a technical rationale with the main focus on funding and resources to a pedagogicalrationale stressing student competencies). This is particularly the case in Flanders, the Dutch-speaking part of Belgium, where the Flemish

* Corresponding author. Tel.: 32 9 264 86 30; fax: 32 9 264 86 88.E-mail address: Ruben.Vanderlinde@UGent.be (R. Vanderlinde).

Contents lists available at ScienceDirect

Computers & Education

j ou rna l homepage : www.e l sev i e r. com/ loca t e / compedu

0360-1315/$ e see front matter 2010 Elsevier Ltd. All rights reserved.

doi:10.1016/j.compedu.2010.02.016

Computers & Education 55 (2010) 541 e 553

mailto:Ruben.Vanderlinde@UGent.behttp://www.sciencedirect.com/science/journal/03601315http://www.elsevier.com/locate/compeduhttp://www.elsevier.com/locate/compeduhttp://www.sciencedirect.com/science/journal/03601315mailto:Ruben.Vanderlinde@UGent.be

8/9/2019 e Capacity

2/13

Government formulated cross-curricular ICT attainment targets for compulsory education in September 2007. In so doing the FlemishGovernment clearly outlines its view of how ICT should be integrated into schools and expects them to put this formal ICT curriculum intopractice (see Vanderlinde et al., 2009 ).

In this study, we present a conceptual framework that we developed from a school improvement perspective. This framework wasdesigned to help identify and measure the factors in uencing both ICT integration and the implementation of ICT curricula. Schoolimprovement is regarded as a strategy for educational change that aims to enhance student outcomes and strengthen the schools'capacity for managing change ( Hopkins, 2001 ). The following key assumptions of the school improvement approach have been outlinedby Harris (2002) : (a) schools have the capacity to improve themselves, (b) school improvement involves cultural change, (c) there areschool level and classroom level conditions for change, and (d) school improvement is concerned with building greater capacity forchange.

2. The need for a new framework: e-capacity

As schools are being given greater autonomy and responsibility for their future, they must now operate as strategic organisations ( vanden Berg, Vandenberghe, & Sleegers, 1999 ). Their capacity to successfully implement educational innovations or policy initiatives ( Geijsel,van den Berg, & Sleegers,1999 ) has become an important issue among researchers and policy makers (see also Stoll,1999 ). Geijsel, Sleegers,Stoel, and Krger (2009) rightly note that the schools' capacity to transform a large-scale reform into an accountable learner-orientedteaching practice has become a major focus in recent research into educational change. However, it is necessary to clearly conceptualise(Stoll,1999 ) and operationalise ( Geijsel et al., 1999 ) this concept of capacity. From a school improvement perspective ( Harris, 2002; Hopkins,2001 ; Stoll,1999 ), the schools' capacity can be generally de ned as the competence of a school to implement educational innovations or tobring about effective change ( Geijsel et al., 1999; Malen & King Rice, 2004 ). A crucial matter in this context is that we assume that schoolshave the capacity to improve themselves while establishing school level and classroom level conditions to support and manage change(Harris, 2002; Hopkins, 2001 ).

In this study, the concept of capacity, as described in the school improvement literature, has been translated and contextualised for thecase of ICT integration. Wespeak about the e-capacity of a school, de ned as the collective competence of a school toimplement ICT in a waythat is a lever for instructional change. From this perspective, e-capacity is concerned with creating and optimising sustainable school leveland teacher level conditions to foster effective change through ICT. The school and teacher level conditions identi ed in this study andpresented in our model (see section 3 ) are based on both a reviewof the change and school improvement literature and the literature on ICTintegration. As a starting point for our own conceptual model, we examined existing conceptual frameworks ( Hew & Brush, 2007; Kozma,2003; Lim, 2002; Tearle, 2004 ), yet as these frameworks do not include measurement scales they were thoroughly revised. Below we outlinethe central ideas within the frameworks studied. See Table 1 for a summary overview of the existing conceptual frameworks.

Kozma (2003) describes the conceptual framework developed by the International Society for Technology in Education. This frameworkoutlines factors that might in uence the use of ICT in the classroom and its impact on educational outcomes. Kozma (2003) speaks about innovative pedagogical practices that use ICT and embeds these practices in a concentric set of contextual levels which include theclassroom (micro level), the school or local community (meso level), and state, national and international entities (macro level). For each

level, Kozma (2003) identi ed actors and factors that mediate change. Factors on the micro level include the classroom organisation, teachercharacteristics, teachers' experience with ICT, and student characteristics. For the meso level, Kozma (2003) identi ed school leaders andparents as possible actors; and the school organisation, ICT infrastructure, technical support, and local culture as possible factors. Nationaland state policies and international trends are identi ed on the macro level. This framework puts emphasis on characteristics of theinnovation (or innovative ICT practices), such as practicality, complexity, clarity, relevance, and need. It also takes into account existingteaching and student practices and the actual use of ICT.

Table 1Brief summary of the existing conceptual frameworks.

Author Theoretical underpinnings Central concept In uencing conditions

Kozma (2003) Comparative education, schoolreform, technology and education,diffusion research, etc.

Innovative pedagogical practicesthat use technology

- Innovation characteristics (e.g. complexity, clarity)- Micro level (e.g. teacher background, classroom size)- Meso level (e.g. leadership, ICT infrastructure)

- Macro level (e.g. policy makers, economic forces)- Outcomes (e.g. teacher competencies)

Tearle (2004) - Management of change- Use of ICT in schools

Use of ICT in teaching - Individuals (e.g. ICT skills, beliefs in ICT)- The ICT implementation process

(e.g. support and training, resource provision)- The whole school (e.g. strong leadership)

Lim (2002) - Sociocultural approach- Activity theory

ICT-based lessons asactivity systems

- Course of study (e.g. curriculum, assessment)- School (e.g. ICT facilities)- Education system (e.g. recruitment and

training of teachers)- Society at large (e.g. publishers)

Hew and Brush(2007)

Technology integrationin K-12 schools

Technology integration forinstructional purposes

- Barriers: resources, institution, subject culture,attitudes and beliefs, knowledge and skills, assessment

- Strategies: vision building, overcoming scarcityof resources, changing attitudes, professionaldevelopment, reconsidering assessment

R. Vanderlinde, J. van Braak / Computers & Education 55 (2010) 541 e 553542

8/9/2019 e Capacity

3/13

The model presented by Tearle (2004) is designed to identify important factors in relation to the implementation of ICT in UK secondaryeducation. The author's whole school approach is characterised by considering ICT integration as a special case for implementing change.Central in this concentric model is the use of ICT in teaching. This practice is in uenced by three layers : the whole school, the ICTimplementation process, and the individual. Whole school characteristics refer to strong school leadership, the school culture, the changeoriented nature of a school, the positive and proactive attitude towards external in uences, and the school's internal processes. The ICTimplementation process, refers to the carefully planned process through which ICT was implemented across the school, matters relating toICT resourcing, and matters relating to support and training. Individual characteristics include a positive attitude to ICT and belief in its useand the importance of ICT knowledge, understanding, and skills.

Somewhat differently, Lim (2002) presents a theoretical framework based on activity theory. One of the central notions in this frame-work is that ICTcannot be studied in isolation, but must be studied within the broader context in which it is situated. From this perspective,the focus should be on the events, activities, contents, and interpersonal processes taking place in the context where ICT is used. The authorproposes to study the activity system as a unit of analysis (i.e.,the ICT-basedlesson) surrounded by different successive levels(i.e., the courseof study, school, education system, and society at large).

In a recent study, Hew and Brush (2007) modelled the barriers and strategies that in uence the integration of ICT for instructionalpurposes into the curriculum. This tentative model contains direct and indirect barriers in uencing ICT integration. Direct barriers include:(a) the teachers' attitudes and beliefs towards using ICT, (b) the teachers' knowledge and skills, (c) the institution (e.g. leadership, schoolplan, etc.) and (d) resources (e.g. availability and access to ICT, support, etc.). In this model, ICT integration is indirectly in uenced by thesubject culture and assessment activities measuring student learning. Interesting in this model is that these authors also present strategiesto overcome the identi ed barriers: having a shared vision on learning and teaching, overcoming the scarcity of resources, changing teacherattitudes and beliefs, and reconsidering assessment activities.

All of the modelsoutlined above are primarily conceptual models highlighting ICT integration from a holistic point of view. ICT is notseenas an isolated phenomenon, but is situated within broader mediating contexts. The greatest weakness of these models is that they do notgive indications or guidelines to further empirically test the factors or conditions presented. While these models can be used for qualitativeresearch designs, such as case studies or in-depth interviews, they are inappropriate for quantitative research designs since measurementscales are not discussed. The present study aims to develop a conceptual framework that leads to the construction of measurement scales.More concretely, research goals are:

To further develop the e-capacity conceptual framework and to identify school level and teacher level conditions in uencing the use of ICT as a lever for instructional change; to identify conditions in the e-capacity model derived from recent insights in the schoolimprovement and ICT integration literature.To translate the identi ed supporting school level and teacher level conditions into items and to construct reliable scales measuring theconditions from the e-capacity framework.

3. Conditions in the e-capacity model

E-capacity refers to the schools' ability to create and optimise sustainable school level and teacher level conditions that can bring abouteffective ICTchange. The identi ed school and teacher level conditions are brought together in the e-capacity model. This model comprisesfour mediating concentric circles; school improvement conditions; ICT related school conditions; ICT related teacher conditions; andteachers' actual use of ICT, surrounding a twofold core element (see Fig. 1). Since teachers play a pivotal role in implementing innovations,the present study will analyse all conditions from a teacher's point of view. This means that all conditions are considered as subjectiveteacher perceptions rather than as objective characteristics (see also Geijsel et al., 1999 ).

The twofold core element of the e-capacity model refers to the transformative nature of ICT ( Selwyn & Brown, 2000; Watson, 2006 ). Itstresses on the one hand the use of ICT as a lever for instructional change, and on the other hand the implementation of ICT curricula. Inother words, the heart of our model takes the new Flemish ICT curriculum (see section 1 ) into account as the current frame of referencestressing the meaning of ICT as a lever for instructional change. In this respect, the twofold core element has one broader focus than theoften used and more narrow variables such as computer or ICT use. The heart of our model is concerned with in uences of ICT on teachingand learning processes (see also Underwood & Dillon, 2004 ). As such, teachers' actual use of ICT was added as an extra layer in the modelinstead of placing it in the centre. Contrary to many other studies, teachers' use of ICT is not seen as a dependent variable. In our model,teachers' actual use of ICT is a process or independent variable leading to different outcomes.

Before discussing the different conditions or layers of the e-capacity model, it is important to stress that this model is situated withina broader societal and political context of national ICT policies and curriculum standards. Kozma (2003) argues that classroom practicesare in uenced by state or national policies in areas such as curriculum and assessment, professional development, and telecommuni-cations. National ICT policies provide schools with resources, such as equipment, network infrastructure, or ICT teachers' professionaldevelopment ( Owston, 2007 ). The rationales behind national ICT polices are also important to take into account. Selwyn and Brown(2000) refer to individual employability and international competitiveness as important policy goals for nation states to promote ICTin education. International ICT policies and the in uence of supranational organisations (e.g. the European Union, the United Nations, theOECD, the World Bank, etc.) are also becoming more important and exert in uence on national and regional educational policies. TheLisbon European Council (March 2000) and the strategic policy framework for the information society (i2010) that was launched by theEuropean Commission in 2005, are important policies for European countries (see also Vanderlinde et al., 2009 ). Another in uentialinternational policy is the ICT competency standards for teachers, which was recently presented by UNESCO (2008) . The objective of thispolicy is to provide guidelines in planning education and training for teachers and prepare them to play an essential role in producingtechnologically capable students. Other factors affecting the conditions presented in the e-capacity model concern the societalsubsystems (e.g. the economic system, the social system and cultural norms) that interact with the educational system in general and ICT

policies in particular.

R. Vanderlinde, J. van Braak / Computers & Education 55 (2010) 541 e 553 543

8/9/2019 e Capacity

4/13

In what follows, the different layers of the e-capacity model will be presented and further analysed through our construction of measurement scales.

3.1. School improvement conditions

The rst layer in the e-capacity model comprises conditions described in the school improvement literature as contributing to theimplementation and realisation of educational change.

3.1.1. LeadershipIn the school improvement literature (e.g. Stoll, 1999 ), leadership is put forward as an important condition for realising educational

change and school improvement. Leadership is understood as an element of the school culture and re ects the extent to which the schoolleader engages in supportive and/or instructional behaviour ( Devos, Bouckenooghe, Engels, & Aelterman, 2007 ). Several studies indicatethat strong leadership is an important condition for successful ICT integration ( Baylor & Ritchie, 2002 ; Granger, Morbey, Lotherington,Owston, & Wideman, 2002; Hadjithoma & Karagiorgi, 2009; Hayes, 2007 ). More speci cally, these studies reveal that school leaders are

in an ideal position to create the conditions of developing a shared vision and policy on ICT, and that they can encourage and supportteachers in the process of ICT integration. Put differently, strong leadership is a necessary precondition to shape ICT policy in schools(Tondeur, Coenders, van Braak, ten Brummelhuis, & Vanderlinde, 2009 ), and in uences relationships amongst the school staff and thegeneral school climate ( Hadjithoma, 2009 ). In this context, Dexter (2008) presents an interesting conceptualisation of ICT related leadershipfunctions, referring to tasks like setting a vision for the use of ICT in education, teacher development and professional community buildingfor teachers to learn about ICT integration, and providing access to supported and managed hardware and software.

3.1.2. Participation in decision making The participation of teachers in the planning and decision making process in schools has been associated with more successful

implementation of educational innovations ( Geijsel, Sleegers, van den Berg, & Kelchtermans, 2001; Harris, 2002 ). It is a crucial condition forfostering the implementation of large innovation programs in schools ( Geijsel et al., 2001 ), and has been identi ed as an important factorwhen integrating ICT ( Bowman, Newman, & Masterson, 2001; Hadjithoma & Karagiorgi, 2009 ). For instance, Hadjithoma and Karagiorgi(2009) argue that this condition is linked with teachers' ownership of the ICT implementation initiative. Recently, in the context of ICTintegration,issues such as data-driven decision making processes have become a point of interest (e.g. Dexter, 2008; Vanderlinde, Hermans,

& van Braak, 2010 ).

Fig. 1. E-capacity model.

R. Vanderlinde, J. van Braak / Computers & Education 55 (2010) 541 e 553544

8/9/2019 e Capacity

5/13

8/9/2019 e Capacity

6/13

3.2.5. ICT infrastructureThelastschool conditionrefers tothe accessand availabilityof an ICTinfrastructure at theschool.Accordingto Tearle (2004) , thiscondition

includes the quantity, type, reliability of computer, access arrangements and location of equipment . While access to hardware is important,a good ICTinfrastructure also concerns the appropriatenessof software. As such,ICT infrastructure is considered as oneof themost notoriousimpediments to ICT integration ( Albirini, 2006; Pelgrum,2001 ). A lack of appropriate material resourcesin theclassroom (i.e., computers andsoftware) can seriously limit what teachers are able to do with ICT ( Mumtaz, 2000 ). It also inhibits learning and causes frustration andresistance among teachers ( Granger et al., 2002 ). Summarizing, Johson and Maddux (2006) argue that this condition is the most favourableone with regard to ICT integration and that hardware, software and connectivity must reach a serious level of sophistication.

3.3. ICT related teacher conditions

Teachers play a pivotal role in implementing educational innovations. Therefore, they are important agents in the concrete imple-mentation of ICT in classroom settings ( Albirini, 2006 ). Teachers have a main responsibility in the process of ICT integration in theirclassrooms (e.g. the organisation of the curriculum with ICT, the organisation of students' ICT work), and different teacher conditions are atstake. In this context, Drent and Meelisen (2008) make a distinction between exogenous factors (non-manipulative like genderand age) andendogenous factors (manipulative). In this study, two endogenous teacher level conditions are put forward; the relevance of knowledge andskills and ways of acquiring them ( Granger et al., 2002 ).

3.3.1. ICT teachers' professional development An important strategy for ICT integration is the professional development of ICT teachers. These activities are designed to improve

teachers' ICT skills and to foster positive attitudes towards ICT ( Galanouli, Murphy, & Gardner, 2004 ). Hew and Brush (2007) describe threeconditions for effective ICT professional development: (1) having a focus on content, (2) giving teachers opportunities for hands-on work,and (3) being consistent with teachers' needs. Similarly, Galanouli et al. (2004) argue that successful ICT professional development needs tobe exible and should re ect the level of ICT competence of the teachers involved. Valcke, Rots, and van Braak (2007) suggest that ICTteacher training should be organised during school hours and use the school as the training location. This implies that ICT training has to belinked with context-speci c questions, needs, or problems. BECTA (2004) argues that training activities need to focus on both pedagogicalaspects and teachers' ICTskills. Professional development in the context of ICT integration is not only about taking part in in-service teachertraining courses, but also about experimenting with ICT before using it in the classroom and keeping informed about ICT ( Albirini, 2006 ).

3.3.2. Teachers' ICT competenceICT competence is an important condition for effective ICT use by teachers ( Hew & Brush, 2007; Pelgrum, 2001 ), and lack of teaching

experience with ICT has been identi ed as an important factor that prevent teachers from integrating ICT ( Mumtaz, 2000 ). Sinnaeve, vanBraak, Tondeur,and Evers (2008) de ne ICTcompetence as a setof knowledge, skills, and attitudesfor an integrated andfunctional useof ICTin an educational context. Similar to Hew and Brush (2007) , these authors further delineate the concept of ICT competence into threecategories: (1) basic ICT knowledge and skills (e.g. howto handle a computer, how to use a spreadsheet, etc.), (2) pedagogical competence touse ICT in the classroom (e.g. having insight into the effects of ICT on teachers' roles and students' motivation and learning), and (3) class

management skills to integrate ICT in the classroom. The last category refers to skills such as how to organise the class effectively so thatstudents have equal opportunities to use ICT or how to respond to students' technical dif culties (see Hew & Brush, 2007 ).

3.4. Teachers' actual use of ICT

Like the model of Kozma (2003) , the e-capacity model takes into account teachers' actual use of ICT. Teachers' ICT use is not placed in thecentreof our model,butis consideredas an extramediatinglayeror processvariable. Inotherwords, teachers'actual useof ICTis consideredasan independent instead of a dependent variable. In previous research ICTuse hasbeen measured andoperationalised in differentways. Someresearchers report the amount of ICT use in the classroom, the time students spend working with ICT, or the use of speci c computerapplications. For instance, Smeets (20 05) measured both the frequency of ICT use during classes, as well as the frequency of use of speci edtypes of ICT applications (e.g. word processing, email, etc.). Baylor and Ritchie (2002) state that many researchers operationalise ICT use interms of a basic dichotomy whereby ICT is either used as the subject of study or as an instructional tool to teach other content. Niederhauserand Stoddart (2001) focus on the use of educational software, making a distinction between skill-based transmission software and open-ended constructivist software. Nowadays, researchers do not consider ICT use as a monolithic process, but emphasise that ICT can be inte-

gratedin many differentwaysin classrooms. In this sense ICTuseis studied as a morecomplexphenomenonreferring todifferenttypes of use.For instance, Tondeur, van Braak, and Valcke (2007) make a distinction between thee types of computer use: (1) the use of ICT as an infor-mationtool,(2) theuseof ICTasa learningtool,and(3) learningbasiccomputerskills. BaylorandRitchie (2002) describenine different typesof ICT use, including the use of ICT for subject-matter content, the use of ICT for collaboration, and the use of ICT for higher order skills .

4. Method

4.1. Procedure and instruments

The second purpose of this study is to develop instruments that measure school and teacher level conditions in uencing the use of ICTfor instructional purposes. To serve this purpose, all conditions from the e-capacity framework were operationalised and translated intoitem statements and a questionnaire was developed. Existing scales were used if available in the research literature, and new items weregenerated for conditions which have not yet been described in terms of measurement in the literature.

For the school improvement conditions, the leadership scales of Hoy and Tarter (1995, 1997) were selected after items were carefully

translated into Dutch and adjusted for appropriateness in the Flemish context. More concretely,

supportive leadership

and

initiating

R. Vanderlinde, J. van Braak / Computers & Education 55 (2010) 541 e 553546

8/9/2019 e Capacity

7/13

structure were chosen as leadership dimensions. The rst dimension is characterised by efforts to motivate teachers by using constructivecriticism and setting an example through hard work. At the same time, the school leader is helpful and genuinely concerned with thepersonal and professional welfare of teachers ( Hoy & Tarter, 1995, 1997 ). Initiating structure is related to task- and achievement-orientedleadership behaviour; the school leader makes his or her attitudes and expectations clear and maintains de nite standards of performance(Hoy & Tarter, 1995,1997 ). For the two other school improvement conditions, Staessens' professional relations among teachers scale, whichmeasures teacher communication and cooperation ( Staessens, 1990; Staessens & Vandenberghe, 1994 ), and Geijsels' participation indecision making scale ( Geijsel et al., 2001, 2009 ) were selected. This last scale measures the extent to which teachers feel that theyparticipate in the schools' decision making process regarding issues of education, innovation, and school improvement.

Based on a comprehensive review of the ICT integration research literature, a number of items were generated and formulated for the ICTschool and teacher level related conditions since no appropriate scales have been described in the literature. Afterwards, these items wereevaluated independently by 11 experts and stakeholders in the eld of ICT integration (researchers, teachers and ICT coordinators) to assesscontent validity, and clarity of the items. The review process resulted in an item bank of 35 items (see Appendix ) with a Likert-type answerformat in order to assess; (1) school related conditions, including the schools' vision of ICT integration, the schools' ICT policy, ICT schoolsupport, ICT coordination, ICT infrastructure, and (2) teacher related conditions, including teachers' ICT competencies and professionaldevelopment activities.

The schools vision of ICT integration items assess the extent to which the school has a clear vision of the place of ICT in education (e.g. The schools vision on the place of ICT in education is acknowledged by the colleagues'). The schools ICT policy items assess the extent towhich the school has a policy and policy plan containing different elements of ICT integration in education (e.g. My school has a welldeveloped ICT policy plan ). The ICT coordination items assess the degree to which ICT integration is coordinated at the school, while the ICT support items assess the extent to which ICTsupport is arranged at the school level. The ICT infrastructure items assess the availabilityand appropriateness of the ICT school and classroom equipment (i.e., hardware, software, and peripheral equipment).

The Teachers ICT competence items assess the degree to which teachers feel competent integrating ICT into their classroom practice.These items include assessment of teachers' perceived technical knowledge and skills, organisational skills, and pedagogical e didacticalskills. Finally, the Teachers ICT professional development items assess the extent to which teachers keep up with developments in the eldof ICT integration, like taking part in in-service teacher trainings.

The examination of teachers' actual use of ICT is based on revised scales of Tondeur et al. (2007) . Three types of ICT use are distinguished:(1) the use of basic ICT skills, (2) ICT as a learning tool (i.e., the use of ICT to support pupils' learning), and (3) ICT as an information tool (i.e.,the use of ICT to select, retrieve, and present information).

4.2. Statistical analysis

Previously validated measurement scales, which measure school improvement conditions and teachers' actual use of ICT, were runthrough a con rmatory factor analysis (CFA) to investigate the structural stability of the scales, and then tested for internal consis-tency by calculating the Cronbach's a (Cronbach, 1951 ). The scale construction of the ICT school and teacher related conditionsinvolved three main stages: (a) exploratory factor analysis (EFA) to identify the number of factors, (b) con rmatory factor analysis(CFA) to examine the stability of the exploratory factor structure, and (c) internal consistency to determine the psychometric quality of the scales.

For the CFA, several t indices were calculated to provide information on the adequacy of the tted model, these included: (a) the c 2 and p-value, (b) the goodness of t index (GFI) and the adjusted goodness of t index (AGFI), (c) the comparative t index (CFI), and (d) the rootmean square error (RMSEA). Following Brown and Cudeck (1993) , values of RMSEA that range between .05 and .08 indicate a fair t. GFI,AGFI, and CFI should be above .90 to indicate an adequate t.

SPSS 15.0 was used for the EFA and reliability analyses and Amos 17.0 ( Arbuckle, 2008 ) was used for the CFA.

4.3. Participants

A representative sample of school teachers in Flanders was selected in order to collect questionnaire data. Teachers were asked to ll ina questionnaire (see Appendix for a presentation of the items) so scales could be constructed in a next step. As an incentive to participate,comprehensive school performance feedback was promised if at least 60% of the teachers participated in the survey. Schools could chosebetween an on-line survey or a paper-and-pencil version of the questionnaire.

Questionnaire data were collected from a sample of 471 primary school teachers from 62 primary schools in Flanders (Belgium). Allparticipants teach in grades 1 e 6 (students' ages 6 e 12) and areevenly distributed acrossall primary schools. The samplewas 78% female; theage ranged between 22 and 61 years with an average age of 38. Teachers reported that they have been using their computer at home for thepast 12 years and in the classroom for the past 8 years. This data set was used for the CFA's on the school improvement conditions and theteachers' use of ICT.

In order to conduct the EFA and CFA of the school and teacher ICT conditions, the sample was randomly divided into two equal groups,matched for gender and age.

5. Results

5.1. Results of the CFA on the conditions for school improvement

Using a CFA on the complete data set ( N 471) we tested the structural stability of the Leadership scales of Hoy and Tarter (1995,1997) ,the Professional relations among teachers scale of Staessens (1990 ) and Staessens and Vandenberghe (1994) , and the Participation in

decision making

scale of Geijsel et al. (2001) .

R. Vanderlinde, J. van Braak / Computers & Education 55 (2010) 541 e 553 547

8/9/2019 e Capacity

8/13

For Hoy and Tarters' scales the results of the CFA indicate signi cant factor loadings for all items (ranging between .70 and .90) onthe two latent leadership factors. The results show a good t between the hypothesised model and the observed data (df 41,c

2 199.1, c 2/df 4.86). The goodness of t estimates were GFI .93 and AGFI .88, CFI .96 and RMSEA .09, indicating fairly good t. The correlation between the two factors is signi cant and very high ( r .85). Cronbach's a for these scales indicates good internalconsistency ( a .94 for the 7-item Supportive leadership scale; a .88 for the 5-item Initiating structure scale).

For the analysis of Staessens (1990 ) and Staessens and Vandenberghe (1994) Professional relations among teachers scale the resultsindicate signi cant factor loadings which range between .55 and .74. One item was deleted to obtain a good t:df 9, c 2 37.4, c 2/df 4.2,GFI .97, AGFI .94, CFI .97, and RMSEA .08. The internal consistency of the 6-item scale is very good with a Cronbach's a of .85.

For the analysis of Geijsel et al.'s (2001) Participation in decision making scale the results indicate signi cant factor loadings whichrange between .48 and .81. Two residuals(e) werecorrelated for one pair of items toobtain a good t:df 4, c 2 19.9, c 2/df 5.0, GFI .98,AGFI .94, CFI .98, and RMSEA .09. The internal consistency of the 5-item scale is good with a Cronbach's a of .81.

5.2. Results of the scale construction of ICT related school conditions (EFA and CFA)

An EFA was performed in order to investigate the underlying structure of the 24 ICT related school items. We used maximum likelihoodEFA based on the assumption that the factors were not correlated (orthogonal rotation). Based on the rst analysis, four items were deleteddue to loadings across factors or low communality value. The second analysis was conducted on the 20 remaining items. The scree testsuggested a four-factor solution. However, as the scree test depends on the reliability of the researchers' judgment, the interpretation of thetest is not always exact ( Tabachnick & Fidell, 2007 ). Therefore, a parallel analysis and Velicer's MAP test were conducted ( O'Connor, 2000 ).Both tests suggested a three-factor solution. The theoretically assumed number of factors was not con rmed in the data structure. In thethird analysis, a three-factor solution was forced, resulting in a model accounting for 59.6% of the common variance. The rst factor was

labelled

ICT school support and coordination

and refers to the degree to which ICT integration is coordinated at the school level and theextent to which ICT support is arranged at the school level. This factor also includes the perceived tasks and roles of the schools' ICTcoordinator. The second factor was labelled Schools ICT vision and policy and comprises (a) the extent to which a school has a clear visionon the place of ICT in education, and (b) the extent to which a school has a policy and policy plan containing different elements of ICTintegration in education. Indeed, Hew and Brush (2007) also combine these components as strategies fostering the integration of ICT intothe curriculum. Furthermore, this factor assesses whether teachers are involved in the process of building an ICT vision and developing anICT policy plan. The third factor is labelled ICT infrastructure and assesses the perceived availability and appropriateness of the ICT schooland classroom equipment (i.e., hardware, software and peripheral equipment).

Next,a CFA was conducted to examine the stability of the three-factorstructure of the ICTschool related items.Basedon the rst CFA, twoitems were removed due to low factor loadings. The nal model comprises 7 items for ICT school support and coordination, 7 items for Schools ICT vision and policy, and 4 items for ICT infrastructure (see Fig. 2). The results show a good t between the hypothesised model

Fig. 2. Results of the CFA: Structure coef

cients for ICT school related items,**

p