Professional Development in Education Implementing school-based teacher development in Tanzania

This article was downloaded by: [University of York]On: 15 May 2015, At: 09:18Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

Professional Development in EducationPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/rjie20

Implementing school-based teacherdevelopment in TanzaniaFrank Hardmana, Jan Hardmanb, Hillary Dachic, Louise Elliotta,Noel Ihebuzord, Maniza Ntekimd & Audax Tibuhindad

a Institute for Effective Education, University of York, Berrick SaulBuilding, Heslington, York YO10 5DD, UKb Department of Education, University of York, York YO10 5DD, UKc Faculty of Education, University of Dar es Salaam, Dar esSalaam, Tanzaniad Basic Education, UNICEF Tanzania, Dar es Salaam, TanzaniaPublished online: 15 May 2015.

To cite this article: Frank Hardman, Jan Hardman, Hillary Dachi, Louise Elliott, Noel Ihebuzor,Maniza Ntekim & Audax Tibuhinda (2015): Implementing school-based teacher development inTanzania, Professional Development in Education, DOI: 10.1080/19415257.2015.1026453

To link to this article: http://dx.doi.org/10.1080/19415257.2015.1026453

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Implementing school-based teacher development in Tanzania

Frank Hardmana*, Jan Hardmanb, Hillary Dachic, Louise Elliotta, Noel Ihebuzord,Maniza Ntekimd and Audax Tibuhindad

aInstitute for Effective Education, University of York, Berrick Saul Building, Heslington, YorkYO10 5DD, UK; bDepartment of Education, University of York, York YO10 5DD, UK;cFaculty of Education, University of Dar es Salaam, Dar es Salaam, Tanzania; dBasicEducation, UNICEF Tanzania, Dar es Salaam, Tanzania

(Received 1 August 2014; accepted 4 February 2015)

This paper reports on the findings of a pilot school-based professional develop-ment programme for Tanzanian primary school teachers launched in February2011 and evaluated in December 2012 by the Ministry of Education and Voca-tional Training with the support of UNICEF. The study set out to investigate theeffectiveness and efficiency of the pilot programme in changing pedagogicalpractices before it was scaled up nationally. It was found that teachers who hadparticipated in the school-based training showed significant differences in theirpedagogical practices and demonstrated a positive attitude towards their trainingand their pupils, and saw teaching and learning as an interactive, communicativeprocess. Drawing on the findings, the paper explores the challenges and the les-sons learned for scaling up school-based teacher development at the nationallevel in Tanzania and other countries in the east and southern African region.

Keywords: school-based teacher development; teacher education reform;Tanzania; primary education; capacity development; quality education

Introduction

The need to ensure that children in low-income countries receive quality teachingand actually learn as a result of their educational experience has been a runningtheme throughout each of the annual Education For All Global Monitoring Reportssince the establishment of Education for All in Dakar, Senegal in 2000. It points tothe need to strengthen access, quality and equity of provision for all children. How-ever, the latest Education For All Global Monitoring Report estimates that out of atotal world population of 650 million primary-age children, 250 million are notachieving the basic literacy and numeracy skills even though 130 million of themhave spent at least four years in school (UNESCO 2014). In responding to this chal-lenge, the report argues that there is a clear link between pedagogy and learning out-comes, and that training in an effective pedagogy is central to raising achievement.

Although vast numbers of children are still not learning the basics, the 2014Education For All Global Monitoring Report reports that some countries such asTanzania and Kenya have been able to get more children into school and ensure thatonce they are there they learn. For example, in Tanzania between 2000 and 2007,

*Corresponding author. Email: [email protected]

© 2015 International Professional Development Association (IPDA)

Professional Development in Education, 2015http://dx.doi.org/10.1080/19415257.2015.1026453

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

the proportion of children completing primary school has risen from one-half totwo-thirds. At the same time, learning outcomes have also improved with 61% ofpupils at Standard 6 (12/13 years of age) attaining basic levels in reading and 31.5%for mathematics (Southern African Consortium for Measuring Educational Quality2010). However, it is generally agreed that learning outcomes are still too low, par-ticularly for the poorest and most marginalised children, and developing the peda-gogic practices of teachers will be central to improving the quality of education inresource-constrained contexts such as Tanzania.

Tanzania was the first country in the east African region to come very close toachieving universal primary education in the early 1980s under Nyerere’s Africansocialist policy. Education for self-reliance was administered through a village-basedprogramme structured around the notion of ujamaa, or extended family, found intraditional African cultures. School fees were abolished and there was a focus on theestablishment of a more ‘practical’ and ‘agriculturally-oriented’ primary curriculumtaught through Kiswahili as the national language of instruction (Thenjiwe andMulvihill 2009, Wedin 2010). However, in the mid-1980s, school fees were re-intro-duced and by the end of the millennium less than 60% of primary school-aged chil-dren were in schools.

In a new drive to address problems of access and quality, the government abol-ished school fees in 2001, and in 2002 the heavily donor-supported Primary Educa-tion Development Programme was launched. This was followed by the SecondaryEducation Development Programme in 2004, designed to expand secondary educa-tion access by up to 50% by 2010 (Wedgwood 2007). In 2008, given the need toaddress the quality of the rapidly expanding teaching force, the Tanzania Ministry ofEducation and Vocational Training (MoEVT 2008) developed a national continuingprofessional development (CPD) strategy and operational plan based on the TeacherDevelopment and Management Strategy.

Despite major achievements in expanding the primary teaching force, andimproving access and completion rates so that 41% of primary pupils are now goingon to secondary school, it is recognised that by international standards the academicqualifications of those entering primary teaching in Tanzania are low (UNESCO2013). Most students entering teacher training colleges have normally completedfour years of lower secondary school and graduate with an ordinary-level secondaryeducation certificate (Towse et al. 2003, Wedgwood 2007). Data collected in 2007by the Southern African Consortium for Measuring Educational Quality (SACMEQ)as part of a regional study of 15 countries show that around 80% of primary schoolpupils are being taught by teachers with a junior secondary school qualification(known as Grade A teachers) and 16% by teachers with no more than a primaryleaving certificate (known as Grade B teachers) (Southern African Consortium forMeasuring Educational Quality 2010).

The training that student-teachers receive at the initial teacher education (ITE)stage in sub-Saharan Africa is often judged to be of poor quality (Lewin and Stuart2003, O’Sullivan 2010, Akyeampong et al. 2013). In the case of Tanzania, there arecurrently 34 public and 14 private colleges offering a full-time, residential course ofone year or more. The curriculum for teacher education is centrally determined bythe Tanzanian Institute of Education and examined by the National ExaminationCouncil of Tanzania. The teacher education curriculum covers general studies, stud-ies related to students’ intended field of teaching and teaching practice.

2 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

A recent study of three teacher colleges in Tanzania covering a wide regionalspread found that the pedagogical knowledge, skills and attitudes needed to teachprimary subjects effectively through a mixture of theory and practice were rarelybeing practiced (Hardman et al. 2012). The model of teaching the students werebeing presented with was essentially transmission based, delivered through lecturesand stressing a hierarchical learning of knowledge and conventional teacher-frontedclassroom organisation. As in other studies of ITE in sub-Saharan Africa, the studyfound that the linking of college-based learning to its application in the classroomappeared to be the exception rather than the rule. Key dimensions of pedagogic con-tent knowledge (teaching large classes, multi-grade strategies for small schools, lan-guage code switching, constructivist approaches to lesson planning) were thereforelargely ignored in the teaching and curriculum materials reviewed. The poor qualityof teacher education and training often meant that rote and recitation approaches toteaching and learning were the norm.

In the face of these challenges, the government of Tanzania has accepted that inorder to improve the quality of its primary education, there is the need to moveaway from ad hoc CPD provision with little follow-up in the classroom, towards amore long-term sustainable vision of professional development that systemicallyupdates the key competences that teachers require in the classroom, a view sup-ported by international research (Moon 2007, Schwille et al. 2007, Timperley 2008,Avalos 2011, UNESCO 2014). For example, in its most recent review of teachereducation covering 65 countries from around the world, the OECD argued that muchcan be learned from high-performing countries in terms of offering a quality educa-tion for their pupils (OECD 2011). Countries such as Finland, South Korea, Canadaand Cuba place a high value on teacher education at the initial stage and through theprovision of school-based professional learning. In all high-performing educationsystems, teachers have a central role to play in improving educational outcomes, andare also at the centre of the improvement efforts themselves. Such systems are notdriven by top-down reforms but by teachers embracing and leading on reform, tak-ing responsibility as professionals, and thereby developing a wider repertoire ofpedagogic strategies for use in the classroom.

The OECD study also found that the most effective professional developmentprogrammes provide high-quality professional learning that upgrades teacher peda-gogic knowledge and skills over a sustained period of time rather than through dis-jointed one-off courses. In this way, high-performing education systems provideopportunities for teachers to work together on issues of instructional planning, tolearn from one another through mentoring or peer coaching and by conductingresearch on the outcomes of classroom practices to collectively guide curriculum,assessment and professional learning decisions. High-performing education systemsalso benefit from clear and concise profiles of what teachers are expected to knowand be able to do at different stages of their careers so as to guide ITE and CPD,and create a lifelong learning framework for teachers (Darling-Hammond et al.2009).

Launch of the school-based teacher development programme

Building on the Teacher Development and Management Strategy, guidelines for thedevelopment of a school-based teacher development strategy were developed and apilot of the programme was officially launched in February 2011 (MoEVT 2009).

Professional Development in Education 3

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Supported by UNICEF, the MoEVT and the Prime Minister’s Office – RegionalAdministration and Local Government began piloting the new school-based modelin seven district councils from a cross-section of regions in mainland Tanzania witha view to implementing the teacher development strategy nationally.

The six-month school-based programme was designed to be mixed mode: aweek’s residential course delivered by college lecturers followed by supported self-study of modules in mathematics, English and pedagogy. It was intended that teach-ers would experience a range of strategies such as training workshops held within aschool cluster supported by external teacher educators, distance self-study modulesand trained school colleagues to support teachers in class. It was also intended thatthe school-based training would be more cost-effective and pedagogical than tradi-tional, college-based provision because it would involve trainees remaining in theirschools and integrating new learning with their classroom practice, a conclusion sup-ported by a recent systematic review of teacher education in developing countries(Orr et al. 2013).

A key component of the training was the use of distance self-study modules.Three modules were developed over the course of a year using writing teams madeup of both national and international educational consultants (Hardman et al. 2011).A core module in pedagogy and subject-specific modules in English and mathemat-ics were made available to the teachers, with the core and mathematics modulespublished in Kiswahili. The modules emphasised an active teaching model throughthe use of problem-solving and discussion activities, and the promotion of high-quality dialogue and discussion between teachers and pupils in whole class, group-based and one-to-one situations.

In the training modules, dialogue and discussion through, for example, the useof open questions (i.e. allowing for more than one possible answer), probing andbuilding on pupil answers and peer-to-peer discussion were promoted alongside themore traditional drilling, closed questioning and telling, thereby raising cognitiveengagement and understanding. Such an approach was designed to build on thetraditional model of whole class teaching found in many Tanzanian classrooms, butto avoid the simplistic polarisation of pedagogy into ‘teacher-centred’ versus ‘child-centred’ that has characterised much educational discourse in the international donorcommunity (Barrett 2007, Vavrus 2009, Schweisfurth 2013). It was also designed tohelp ensure there was a better balance and blending of local socio-cultural practiceswith internationally informed reforms to teacher education, particularly with regardto adult–child relationships, and to recognise the realities of a resource-poor class-room in which many teachers worked (Abd-Kadir and Hardman 2013). The trainingalso demonstrated to teachers how to plan group work so that it was purposeful,well-structured and appropriate to the learning task, including the training of pupilsin how to work collaboratively and to assign roles and tasks.

Teachers were expected to work through the modules in their own time, sup-ported by weekly school-based study groups and monthly meetings of schools clus-tered together at the council ward level led by CPD coordinators appointed from thedistrict council advisory service and teacher colleges. The training was also designedso that teachers would receive a range of CPD provision – coaching and observationin the classroom at least once a month from a coordinator, inspector or head teachertrained in mentoring teachers – and that the school would become a community ofpractice to help transform pedagogic practice in the classroom (Cordingley et al.2005, Kennedy 2014, Choi Fung Tam 2015). By the end of 2012, a total of 2052

4 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

primary school teachers from 141 schools across the 7 district councils had beentrained through the pilot programme.

Before deciding whether to scale up the school-based teacher development pro-gramme, a process evaluation was commissioned a year after its implementation.Using classroom observation and interviews with key stakeholders, the study set outto investigate the following research questions:

• To what extent had the pilot school-based programme impacted on classroompedagogy?

• To what extent had it met the objectives of the 2009 CPD strategy in terms ofits design, implementation and cost-effectiveness?

• What lessons could be learned for the scale-up of the national CPD strategy?

Research design

To study the impact of the school-based pilot on classroom practices and perceptionsof its effectiveness on improving the quality of learning, a mixed-method approachwas used. The 2012 pilot evaluation built on the 2009 baseline study of classroompractices to allow for a longitudinal comparison. In addition to the baselineevaluation, a comparison group of schools from a district council not participatingin the pilot were selected to act as an additional counterfactual1 (UNICEF 2009).The study made use of lesson observations, systematic analysis of video-recordedlessons and interviews with a range of stakeholders including education officials,head teachers, teachers, parents and pupils. The mixed-method approach allowed forthe interplay of multiple analytic lenses and procedures, and for the lessons to beanalysed at the macro and micro levels. Such methodological triangulation alsoallowed for cross-checking of the reliability and validity of the classroom observa-tion data.

Sample

The baseline sample had included 32 schools from 8 districts of mainlandTanzania (Bagamoyo, Hai, Magu, Makete, Mtwara, Siha, Temeke, Shinyanga) toensure that 8 of the 26 mainland regions of Tanzania were represented. For thepilot evaluation, seven of the eight original district councils were selected for theschool-based CPD pilot to allow for a longitudinal comparison. While Shinyangadistrict council had formed part of the original baseline, it had been implementinga CPD programme supported by Oxfam and so was not included in the follow-upevaluation.

At both stages of the study, a random stratified sampling of four schools wasselected from each district council to ensure an equal balance of schools servingurban and rural contexts. For the 2012 CPD pilot evaluation, in addition to the 28schools in the intervention group, 4 schools from Kwimba, a neighbouring districtof Magu, were selected to act as a comparison group. All teachers participating inboth phases of the study had an ITE qualification and had not taken part in any otherform of CPD training beyond the MoEVT school-based pilot.

Professional Development in Education 5

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Classroom interaction and discourse analysis

Research into effective teaching has identified a range of teaching behaviours inwhole class, group-based and one-to-one teaching that lead to higher learning gains,suggesting there needs to be a focus on classroom interaction and discourse becauseof its centrality to the act of teaching and learning. For example, Hattie’s synthesisof 800 meta-analyses involving more than 50,000 studies related to achievement inschool-aged children in respect of interactive strategies, such as reciprocal teaching,collaborative group work and peer tutoring encouraging student verbalisation andteacher feedback, shows that high-quality classroom talk enhances understanding,accelerates learning and raises learning outcomes (Hattie 2009). Such interactiveapproaches make the learning visible for both teachers and students, allowing forthe monitoring of learning and formative evaluation. It also points to the importanceof investigating what can be observed in the act of teaching (i.e. task, activity, class-room interaction, assessment) as key indicators of quality (Alexander 2008).

While much of the evidence on effective teaching and learning processes hascome from high-income countries, a substantial body of evidence based on observa-tion studies from low and middle-income countries is starting to emerge. A recentreview commissioned by the UK government’s Department for InternationalDevelopment, building on a systematic review of 489 studies and an in-depth studyof 54 empirical studies, concluded that classroom interaction is the pedagogical key(Westbrook et al. 2013). It found that teachers who promoted an interactive peda-gogy also demonstrated a positive attitude towards their training and the students,and saw teaching and learning as an interactive, communicative process. Becausemanaging the quality of classroom interaction is seen as the single most importantfactor in improving the quality of teaching and learning, particularly in resource-poor contexts, investigating the quality of classroom interaction in Tanzanian pri-mary schools was a central focus of both the baseline and pilot evaluation.

Four specific strategies that promoted an interactive pedagogy and visible learn-ing from students were identified: providing feedback; sustained attention and inclu-sion in the classroom; creating a safe environment in which students felt supportedin their learning; and drawing on students’ backgrounds and experiences. From thefour strategies, six effective teaching behaviours were identified: frequent and rele-vant use of visual aids and locally produced learning materials beyond the use ofthe textbook; open and closed questioning; expanding responses; encouraging stu-dent questioning; demonstration and explanation, drawing on sound pedagogicalcontent knowledge; and use of local languages.

To investigate the quality of teacher interaction and discourse and to address thefirst of the research questions, the baseline and pilot evaluation made use of a time-line analysis of live lessons and discourse analysis of a sub-sample of video-recorded lessons to allow for a macro-level and micro-level of lesson analysis. Bothsystematic observation instruments were designed to be comprehensive, manageableand as low-inference as possible in capturing current pedagogical practices inTanzanian primary schools. They built on earlier studies of east African primaryclassrooms and were informed by international pedagogical research focusing onwhat can be observed in the act of teaching (i.e. task, activity, interaction, assess-ment) so as to ensure the observations of classroom processes are as valid and reli-able as is practically possible (Ackers and Hardman 2001, O-saki and Agu 2002,Pontefract and Hardman 2005, Hardman et al. 2009, Wedin 2010).

6 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Both the timeline and discourse analysis drew upon a three-part teachingexchange structure that is central to teacher–pupil interaction (Sinclair and Coulthard1992, Alexander 2001). In its prototypical form, a teaching exchange consists ofthree moves (the IRF structure): an initiation, usually in the form of a teacher ques-tion; a response, in which a pupil, or group of pupils, attempts to answer the ques-tion; and a follow-up move, in which the teacher provides some form of feedback(very often in the form of an evaluation) to the pupil’s response. However, researchinto classroom interaction suggests that the IRF structure can take a variety of formsand functions leading to different levels of pupil participation and engagement, par-ticularly through the use that is made of the follow-up move (Crockett et al. 2009,Mameli and Molinari 2013). By asking pupils to expand on their thinking, to justifyor clarify their opinions or to make connections to their own experiences, the IRFpattern can be said to take on a dialogic function.

Timeline analysis

The timeline analysis schedule required the observer to record the main teachingactivities from a list of 14 prompts at every five-minute interval in the lesson. Build-ing on the research into effective teaching in low-income countries, prompts covereda range of whole class, group-based and individual teaching and learning activities,and captured a repertoire of classroom talk including teacher explanation, rote,recitation and dialogue. In the note section, observers were asked to record activitiesnot covered in the checklist and to provide further contextual information on theactivities ticked in the checklist.

For the baseline study, 236 live lessons were originally analysed using the time-line observation schedule to investigate the structure of lessons and proportion oftime spent on the 14 teaching and learning activities in the teaching of English,Kiswahili, mathematics and science. When comparing the underlying pedagogy usedby teachers, it was found that the majority of lessons used a transmission model ofteaching in which the teacher often used a chalkboard and/or textbook to transmitrecipe knowledge for recall, with no statistically significant differences foundbetween the teaching of English, Kiswahili, mathematics and science at Grades 3and 6, and between urban and rural schools.

For the pilot evaluation, 127 live lessons in English and mathematics wereanalysed using the timeline observation. Because the teaching and learning ofscience and Kiswahili had not been included in the school-based programme, only

Table 1. Lesson observations in English and mathematics by grade for 2009 baseline and2012 evaluation.

Grade Subject

2009 baseline 2012 pilot evaluation

Number ofschools

Number ofobservations

Number ofschools

Number ofobservations

3 English 32 32 32 32Mathematics 32 31

6 English 32 32Mathematics 31 32

Total 127 127

Professional Development in Education 7

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

observation data from the teaching of English and mathematics at Grades 3 and 6from the 2009 baseline and 2012 pilot evaluation were compared. A breakdown ofthe lesson observations in English and mathematics at Grades 3 and 6 conducted forthe 2009 baseline and 2012 evaluation is presented in Table 1.

Discourse analysis

A sub-sample of video-recordings from English lessons at Grades 3 and 6 was sys-tematically analysed to capture teacher–pupil interactions by recording the differenttypes and frequency of discourse moves made by teachers and pupils. For the base-line study, 42 English lessons, half from each grade, were analysed; and for the pilotevaluation it was 32 lessons. The coding system primarily focused on the IRF struc-ture by gathering data on the types of teacher questions (i.e. open, closed, repeatquestions), whether questions were answered (and by whom) and the types of fol-low-up given in response to answers. It also recorded the number of pupil initiationsin the form of questions. Responses were coded according to whether a boy or girlanswered, or whether there was a choral reply. Teacher follow-up to a pupil responsewas coded according to whether it was affirmed, praised or elaborated upon. Havingsystematically coded all of the three-part teaching exchanges using the above cate-gories, the tallies were turned into percentage scores to compare the patterning ofthe teacher–pupil interactions at the macro level across both phases of the study.

Interviews

To address the second and third research questions to access the extent to which theschool-based pilot had met the objectives of the 2009 CPD strategy in terms of itsdesign, implementation and cost-effectiveness, and to explore what could be learnedfor the scale-up of the national strategy, key stakeholders were interviewed atnational and sub-national levels. For the baseline, semi-structured interviews wereconducted with head teachers (n = 32), teachers (n = 109), district officers (n = 8),ward coordinators2 (n = 22) and inspectors (n = 3). The interviews explored theirperceptions of the current state of provision of CPD in the districts. A questionnairesurvey (n = 328) of CPD provision was also conducted in the eight district councils.Both the interviews and survey found that CPD was ad hoc with no follow-up inthe classroom. Overall, just over half the sample of 328 primary teachers had under-taken CPD or upgrading3 in the previous five years, with wide variation in CPDprovision amongst the district councils, ranging from 74% to 5% (UNICEF 2009).

For the pilot evaluation, semi-structured interviews were conducted with headteachers (n = 28), teachers (n = 117) and a focus group of four (two boys/two girls)Standard 6 pupils (n = 28) and focus group interviews with school committees (n = 28)to elicit their opinions on the impact of the pilot on classroom practices and pupillearning and to triangulate the classroom process data. In addition, to evaluateperceptions on the impact of the pilot at the district level, interviews were conductedwith district officers (n = 5), ward coordinators (n = 24), ward inspectors (n = 7) andCPD cluster coordinators (n = 6).

Teacher colleges were also visited to interview principals (n = 4) and collegetutors (n = 6) involved in delivering the school-based pilot, particularly to addresstheir perceptions on the effectiveness and efficiency of the pilot. Semi-structuredinterviews were also conducted with a range of national stakeholders selected from

8 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

the Prime Minister’s Office – Regional Administration and Local Government, theTanzania Institute of Education, the Agency for the Development of Education Man-agement and the Institute of Adult Education and with universities to explore theirperceptions of the effectiveness and efficiency of the pilot (n = 13).

Training of data collectors

For both the baseline and evaluation, 24 and 16 data collectors respectively wereselected and trained over the course of three days. They were made up of representa-tives from the Ministry of Education, teacher colleges and universities. In additionto extensive video analysis, their training included a half-day trialling of theobservation and interview schedules in schools. During the training workshop, inter-rater reliability checks were conducted on the timeline analysis so as to achievereliability ratings of at least 80% amongst the data collectors. Teams of four wereformed and each team was responsible for visiting two districts over a two-week tothree-week period. The baseline data were collected in October 2008 and the pilotevaluation data in October 2011.

Findings

Timeline analysis

In analysing the timeline data from the baseline and pilot evaluation, we counted thenumber of times an activity occurred within a five-minute interval and evenly dis-tributed the percentage of time4 to identify patterns of frequency. Officially, Standard3 and Standard 6 lessons are meant to last 40 minutes so the timeline schedules werecompleted within these time frames. If the lesson ended before the official desig-nated time, observers coded the behaviour as ‘off-task’. Table 2 presents a break-down of the most common teaching and learning activities, as a percentage of thelesson time, during each five-minute interval for the baseline schools. Table 3 pre-sents a breakdown of the most common teaching and learning activities, as a per-centage of the lesson time, during each five-minute interval for the pilot schools.

An independent-samples t-test was used to statistically analyse differences in thetimeline distribution of teaching and learning activities between the baseline and theschool-based trained teachers. It was found that teachers who had participated in thepilot were significantly more likely to involve the pupils in group work and pupildemonstration, and to spend less time on class management issues. There were alsofewer interruptions to their lessons and pupils were more likely to be on-task. It isalso interesting to note that trained teachers were spending more time on ‘classadministration’ (e.g. moving from whole class to group-based activities, giving outteaching and learning materials), suggesting they were using more varied teachingand learning approaches.

The timeline analysis from the 4 comparison schools shows that the distributionof the 14 teaching and learning activities was very similar to the 2009 baselineschools. Table 4 presents the findings of the timeline analysis from the baseline,comparison group and CPD-trained teachers, grouped according to whether theactivity was ‘teacher’ or ‘pupil’ focused. Overall, it was found that teacher-directedactivities (i.e. explaining, question and answer, teacher rote and chorus response,writing on the chalkboard, reading to the class, lesson review, teacher marking, class

Professional Development in Education 9

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Table

2.Percentageof

timespenton

teaching

andlearning

activ

ities

inbaselin

eschools.

Baseline

1–5

6–10

11–1

516

–20

21–2

526

–30

31–35

36–40

Overall

minutes

minutes

minutes

minutes

minutes

minutes

minutes

minutes

A1

21.3

21.4

18.2

15.2

14.5

12.7

13.8

14.1

16.4

A2

13.3

12.9

10.4

4.5

4.3

7.0

1.5

2.8

7.1

A3

17.3

21.4

18.2

13.6

11.6

14.1

12.3

7.0

14.4

A4

6.7

5.7

6.5

4.5

5.8

4.2

3.1

1.4

4.7

A5

4.0

10.0

5.2

6.1

1.4

4.2

3.1

0.0

4.3

A6

1.3

4.3

6.5

10.6

14.5

12.7

16.9

21.1

11.0

A7

4.0

5.7

3.9

6.1

7.2

4.2

1.5

1.4

4.3

A8

2.7

5.7

3.9

3.0

1.4

1.4

0.0

1.4

2.4

A9

0.0

0.0

1.3

3.0

1.4

4.2

1.5

4.2

2.0

A10

1.3

0.0

1.3

1.5

8.7

8.5

13.8

16.9

6.5

A11

13.3

8.6

13.0

21.2

14.5

9.9

13.8

12.7

13.4

A12

12.0

1.4

5.2

3.0

4.3

7.0

4.6

4.2

5.2

A13

1.3

0.0

1.3

1.5

1.4

1.4

1.5

0.0

1.1

A14

1.3

2.9

5.2

6.1

8.7

8.5

12.3

12.7

7.2

Total

100.0

Notes:A1=teacherexplanation/qu

estio

nandansw

er;A2=teacherrote/cho

rusrespon

ses;

A3=teacherwritin

gon

chalkb

oard;A4=teacherreadingto

who

leclass;

A5=pu

pilreadingto

who

leclass;A6=pu

pilsworking

individu

ally;A7=pu

pilsworking

inpairs/grou

ps;A8=pu

pildemon

stratin

gto

class;A9=teacherreview

sles-

sontopic;

A10

=teachermarking

work;

A11

=classmanagem

ent;A12

=classadministration;

A13

=interrup

tionto

lesson

;A14

=pu

pilsoff-task.

10 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Table

3.Percentageof

timespenton

teaching

andlearning

activ

ities

inpilotschools.

Interventio

n1–5

6–10

11–15

16–20

21–2

526–3

031

–35

36–40

Overall

minutes

minutes

minutes

minutes

minutes

minutes

minutes

minutes

A1

18.3

17.7

16.7

17.7

15.2

16.0

16.3

15.0

16.6

A2

13.4

14.3

11.2

9.9

7.7

6.8

5.6

4.2

9.3

A3

16.1

13.0

12.4

11.2

11.7

10.7

9.9

8.7

11.7

A4

4.5

6.1

5.0

3.5

4.0

3.3

3.1

1.9

3.9

A5

5.0

6.3

6.0

6.2

5.5

5.8

3.9

1.9

5.0

A6

6.7

7.3

7.7

7.3

9.1

11.1

13.6

16.7

9.9

A7

2.5

6.6

8.8

9.1

8.8

8.8

7.2

5.9

7.2

A8

4.9

7.7

7.9

7.6

7.7

6.4

6.0

2.8

6.4

A9

5.2

1.6

1.9

2.2

3.5

3.3

2.9

6.4

3.4

A10

0.5

2.1

2.9

4.3

6.2

7.2

9.5

10.4

5.4

A11

7.9

7.1

9.1

9.9

9.7

9.5

10.9

10.6

9.3

A12

14.1

9.6

10.0

8.4

8.2

8.6

8.6

10.4

9.7

A13

0.0

0.0

0.2

0.4

0.4

0.2

0.2

0.2

0.2

A14

0.8

0.7

0.3

2.2

2.4

2.3

2.5

4.9

2.0

Total

100.0

Notes:A1=teacherexplanation/qu

estio

nandansw

er;A2=teacherrote/cho

rusrespon

ses;

A3=teacherwritin

gon

chalkb

oard;A4=teacherreadingto

who

leclass;

A5=pu

pilreadingto

who

leclass;A6=pu

pilsworking

individu

ally;A7=pu

pilsworking

inpairs/grou

ps;A8=pu

pildemon

stratin

gto

class;A9=teacherreview

sles-

sontopic;

A10

=teachermarking

work;

A11

=classmanagem

ent;A12

=classadministration;

A13

=interrup

tionto

lesson

;A14

=pu

pilsoff-task.

Professional Development in Education 11

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

management, administration) took up 69.7%, 69.6% and 69.3% of the lesson timerespectively. However, pupil-focused activities (i.e. pupil reading to the class, pupilsworking individually, paired or group work, pupil demonstration) accounted for22%, 21.9% and 28.5% of the time respectively, reflecting the greater use of groupwork and pupil demonstration by trained teachers. Non-curricular activities (i.e.interruptions, off-task) took up 8.3% and 8.5% of the time in the baseline and com-parison schools compared with 2.2% in the pilot schools. Most of the time off-taskin the baseline and comparison schools occurred in the final 10 minutes of the les-sons, suggesting they were more likely to end early. The timeline findings from thecomparison schools therefore more closely reflect those of the 2009 baseline study(UNICEF 2009).

Analysis of video-recorded lessons

Like other, smaller-scale, studies of the Tanzanian primary classroom, the analysisof the video data from the baseline and pilot evaluation found that most of the class-room talk was teacher-fronted and largely made up of teacher-led explanation,recitation, cued elicitations, chorus responses and use of chalkboard/whiteboard (O-saki and Agu 2002, Barrett 2007, Wedin 2010). However, when the use of IRFmoves in whole class teaching was statistically compared using an independent-samples t-test, the following differences between trained and non-trained teacherswere found to be significant at p < 0.05: teacher informs, teacher questions, cuedelicitations,5 teacher check,6 teacher direct, pupil questions, girl answers, choralanswer, probe, comments, praise and no follow-up.

As in the timeline analysis, the trained teachers were generally more interactivein their teacher questions and participation strategies (cued elicitations and teacherchecks) compared with the baseline and comparison schools. They also used more

Table 4. Timeline findings for baseline, comparison and intervention schools.

Teaching and learning activityBaseline(%)

Comparison(%)

Intervention(%)

Teacher activityA1 = teacher explanation/question andanswer

16.4 16.7 16.6

A2 = teacher rote/chorus responses 7.1 7.6 9.3A3 = teacher writing on chalkboard 14.4 14.6 11.7A4 = teacher reading to whole class 4.7 4.2 3.9A9 = teacher reviews lesson topic 2.0 1.8 3.4A10 = teacher marking 6.5 6.3 5.4A11 = class management 13.4 13.6 9.3A12 = class administration 5.2 4.8 9.7Pupil activityA5 = pupil reading to whole class 4.3 4.7 5.0A6 = pupils working individually 11.0 10.8 9.9A7 = pupils working in pairs/groups 4.3 3.8 7.2A8 = pupil demonstrating to class 2.4 2.6 6.4Non-curricularA13 = interruption 1.1 1.2 0.2A14 = off-task 7.2 7.3 2.0

12 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

teacher explanation, gave more praise in their feedback, encouraged more individualanswers, and probed and offered more comments on answers (i.e. rephrasing, build-ing or elaborating upon an answer). Trained teachers gave more direction mainlydue to the greater use of group/paired work, which necessitated the giving of direc-tions in regrouping the class and setting up activities.

In analysing the use of paired or group work (i.e. when pupils were observedinteracting together and exchanging ideas through exploratory talk), it was foundthat 63% of the trained teachers used some form of peer interaction in their lessoncompared with just 6% of the baseline and comparison group teachers. A similarpicture emerged with the use of open-ended questions: 11% of the questions askedby trained teachers were open compared with 3% asked by non-CPD-trainedteachers.

The video-recording analysis also showed that a greater range of organisationalarrangements were being deployed by trained teachers to meet different educationalgoals. In the 2009 baseline, 98% of classrooms were organised using a traditionalclassroom layout (i.e. desks organised in rows) and this was similar to the compar-ison schools; this compares with 64% of the pilot schools captured in the liveobservations using an alternative classroom layout, often to accommodate the use ofgroup work.

Analysis of interviews

The interviews from the districts and schools where the MoEVT CPD was pilotedexploring stakeholder perceptions of its impact on classroom practices and, wherepossible, the effectiveness and efficiency of its implementation were analysed usinga content analysis approach. Research themes were generated from the data so as tocreate categories into which the interview responses could be assigned.

Teachers

All 117 teachers who had participated in the school-based professional developmentprogramme were asked about their views on the effectiveness of the training. Inresponse to the question about which aspects of the training they had found the mostuseful, almost three-quarters of the teachers stated that they found the study modulesvery useful, with one-third referring specifically to the pedagogy module. Sixty percent of the teachers found the school and cluster group meetings to be very effec-tive, and 40% found the classroom observations to be very useful.

When asked about how the programme had improved their teaching, 70% of theteachers felt that it had improved pupil participation in lessons and broadened theirteaching repertoire through more varied questioning and a greater use of paired andgroup techniques and teaching aids drawn from the local environment. Over one-half of the teachers interviewed also felt that the training had improved lesson plan-ning and helped teachers to identify the learning needs of the pupils they teach.Similarly, in response to the question on how the training had improved pupil learn-ing, over three-quarters of the teachers stated it had increased their active participa-tion and improved their confidence and motivation.

When asked about observations of their teaching as part of the follow-up to thetraining, 70% stated they had been observed at least once. When probed about whohad observed them, one-third of the teachers indicated they had been observed by a

Professional Development in Education 13

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

peer, 24% by a head teacher, 18% by a college tutor and 17% by a ward educationcoordinator. It was interesting to note that in Mtwara, all 21 teachers had beenobserved and college tutors had participated in 16 of the teacher observations, sug-gesting more active involvement of the local teacher college. However, in Hai,Makete, Siha and Temeke none of the teachers had been observed by a collegetutor.

In response to the question about when teachers had undertaken most of theirtraining, over 60% of the teachers interviewed stated that they studied in the eve-nings after school with 30% meeting in study groups. Nearly one-third of the teach-ers said they studied during school hours in breaks and at lunch-time, and a similarnumber stated they studied at weekends.

Overall, the trained teachers were very positive about the training they hadreceived and thought it should be scaled up at the school and national levels toinclude all teachers. They thought more time should be made available for school-based CPD through the allocation of training days and study time during the work-ing week. They also indicated that they wanted the training to be accredited andrewarded through promotion, and for there to be more face-to-face training, clustersupport and follow-up in the classroom. However, 30% of teachers raised concernsabout the timing of feedback on assignments. Twenty per cent of the teachers alsofound the English module more difficult to study than the pedagogy and mathemat-ics modules published in Kiswahili.

Head teachers

Overall, responses from the 28 head teachers were very positive towards the pilotschool-based teacher development programme. There was a general agreement thatthe training had impacted on teacher pedagogy and pupil performance, promotinggreater confidence in teachers and pupils alike, more involvement of pupils in theirlearning, and greater subject matter competence from teachers and pupils. Mentor-ing, cluster groups, observations and reading of the modules by the teachers weredeemed to be the most important features of the programme.

Most head teachers reported the CPD had taken place after school hours (oftentimetabled by the school as part of the school day but after class) but also duringweekends and evenings. Twenty-two of the head teachers reported their teachers hadbeen observed by a visitor, with peer review playing a major role. Some head tea-chers, however, indicated that there was a need for more external help, particularlyfrom college tutors. They also indicated that they would have liked additional train-ing in leadership on the school-based CPD.

In terms of comments on the effectiveness of the programme, the need for aCPD budget and materials for teachers was a constant theme. Other themes focusedon its efficiency, including logistical and infrastructure problems, such as difficultiesfaced by teachers in travelling to cluster meetings due to time, distance and cost, aswell as the need for desks, water, sanitation and housing for teachers. Some moduleswere viewed as being too difficult, particularly with respect to the English module,for which some head teachers requested additional training. Perhaps the most signifi-cant comment from a programmatic point of view was that relating to the need toprovide incentives, such as certificates, additional pay and promotion, for thoseteachers who participated in the school-based CPD.

14 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

School committees

Out of the 28 focus group interviews, 21 of the school committees stated that theywere aware of the school-based pilot and 14 responded that they had been involvedin discussions of the programme. While 26 school committees stated they had beeninvolved in drawing up the school development plan, 21 reported that CPD was notpart of the plan. Most committees therefore only knew about the programme in ageneral sense: aware that it involved teachers working together in the school and atcluster meetings.

While not familiar with the technical detail of how the training was being deliv-ered in their schools, most of the committees (24 out of 28) agreed that the pro-gramme had improved the quality of teaching through improved teacher–pupilrelations and higher levels of commitment from the teachers. Similarly, 26 commit-tees reported the quality of education in their schools had improved, with moreactive pupil participation in lessons and pupils reporting they liked school more.

Overall, many of the committees focused more on budgets and infrastructure intheir discussions and not as much on teacher or school performance. Some commit-tee members would have liked more information on the CPD programme, to havebeen more involved in discussions of its implementation and to have received ori-entation training. They also expressed concerns about the lack of teaching learningresources, poor infrastructure and the need for all teachers to be given CPD training.

Pupils

When asked about the subjects they enjoyed the most, pupils from all eight districtsranked mathematics, English and science as the most popular. Pupils in the pilotschools were also asked about whether they had noticed any changes in the waythey were taught compared with the previous year. There was general agreement thatteaching practices had changed, particularly in the use of group work, teaching andlearning aids and pupil participation in lessons. Many reported that teachers werefriendlier, gave more time to pupils and allowed more pupils questions. They alsoreported that lessons were more varied and that group work was a daily feature oftheir lessons, particularly in mathematics, English and science, supporting the class-room observation finding. This was in contrast to the comparison schools: two ofthe four focus groups reported that group work was rarely used and two reported itwas never used.

Teacher training college principals and tutors

College staff, including principals and college tutors, reported they had beenresponsible for administering the CPD programme, for liaising with the districtcouncils and for mentoring, monitoring and assessing teachers. They were generallypositive about the design of the programme and thought it had improved teachereffectiveness through more interactive forms of teaching and the use of group work.They also thought it was time efficient as teachers could study the modules in theirown workplace, supported by school and cluster-based meetings and follow-up inthe classroom.

The main constraint college CPD coordinators identified was the lack of a budgetto cover travel costs to the schools and clusters, thereby limiting the number of visits

Professional Development in Education 15

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

they could make to mentor, monitor and assess the teachers, and to liaise withdistrict councils. Lack of funds also had an impact on the dispatching of markedportfolios by college tutors back to the clusters and schools.

Overall, in terms of its effectiveness and efficiency, the principals and collegetutors thought the programme should be extended to include all teachers and schoolsubjects, and for there to be academic accreditation of the programmes with officialstudy time being made available to teachers throughout the school year. They alsofelt there was a need for more training of teacher educators including ward educa-tion officers, district CPD and cluster coordinators to improve their capacity in sup-porting and mentoring teachers.

District education officers

Amongst education officials at the district and ward levels there was general agree-ment that the training had improved teacher competence, subject knowledge andconfidence, leading to more interactive teaching and learner-centred approaches.There was a similar level of agreement that the modules, particularly the pedagogymodule, had been effective in training the teachers in their own schools, and that thementoring, observation, school and cluster meetings had also provided a good levelof support. The training was also viewed as being cost-effective because the distancelearning materials ensured teachers were not taken away from their schools, and theycould study in their own time during the school day, in the evenings and atweekends.

When asked about their role in the pilot programme, ward education coordina-tors, district CPD coordinators, cluster coordinators and inspectors indicated thatthey had mainly been involved in the mentoring and observation of teachers, and inmonitoring the programme, attending training sessions, administering the pro-gramme and liaising with the schools, clusters and colleges. School inspectors alsoreported they had noticed improvements in the quality of teacher–pupil interactionin their school visits. In response to the question on the CPD budget, the majority ofeducation officers were not aware of its existence. Not surprisingly, the lack of abudget at the district, ward and school levels was seen as a major challenge foreffectiveness and sustainability of the programme.

There was also general agreement that the programme needed to be scaled upand accredited, and that teachers who took part in CPD needed to be rewarded withhigher pay and promotion. District officers also felt that there was a need for furthertraining, accreditation and career advancement of teacher educators working at theschool level to help build capacity for delivering and monitoring school-based CPD.

National stakeholders

The majority of respondents working at the national level were in favour of theschool-based teacher development programme being gradually scaled up through aphased programme of implementation. They also thought that ‘model districts’ couldbe identified in each of the regions to coach and mentor neighbouring districts,thereby disseminating good practice in terms of cluster coordination, expert teachers,peer tutoring and networking.

It was suggested by teacher educators that the training modules should beadopted by the teacher colleges and used as core or supplementary materials as part

16 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

of ITE. In this way newly qualified teachers would have valuable teaching resourcesto take to their first post and ensure their wider dissemination in schools. It was alsosuggested that the modules should be made freely available in a variety of electronicformats, supported by a website, television and radio broadcasts.

Many of the national stakeholders agreed that school-based teacher developmentwould bring cost benefits, particularly as many of the Teacher Resource Centres setup in the 1990s to support CPD were dysfunctional, mainly due to the withdrawalof funding. Many saw the school cluster as an effective mechanism for supportingcontinuous professional development programmes because of their proximity to theschools and the fact that they were formed and managed by teachers themselves.There was also strong support for devolving ownership of the CPD down to theward, cluster and school levels, and providing incentives for those teachers andteacher educators who took part in the training. They stated that this would, in turn,require devolved powers, responsibilities and resources to local governmentauthorities.

However, it was reported that while district councils were advised to budget forthe school-based programme, there was no government circular to ensure it was allo-cated funds for the 2012/13 fiscal year. Other national stakeholders also commentedon the lack of capacity of staff at the district level, and suggested that otherinstitutions and organisations from the public and private sectors, such as universi-ties and consultancy companies, should be identified and brought in to providetechnical support.

Cost-effectiveness

Available data on the actual costs of implementing the school-based teacher develop-ment programme compared with residential, college-based provision were also gath-ered from UNICEF, the MoEVT and the Prime Minister’s Office – RegionalAdministration and Local Government to measure the cost-effectiveness of the pro-gramme before it was scaled up. By calculating both the development and recurrentcosts of the programme, it was found that the unit cost of training a teacher at theschool and cluster level was US$215 compared with an average cost of $2000 totrain a teacher on a full-time, one-year residential course. The unit cost of theschool-based CPD was also likely to fall as the cluster system was scaled up.

Conclusions

Overall, the findings from the classroom observations and stakeholder interviewsshow that the school-based pilot was well received at the district and school levels,and that it was having a significant impact on teaching and learning practices of theteachers who had received the training. The findings of the pilot suggest that school-based CPD, building on existing systems and structures, and linked to study materi-als, coaching, observation and feedback by more experienced colleagues, helpedteachers to explore their own beliefs and classroom practices and to explore alterna-tive pedagogic approaches (Haßler et al. 2014). The findings also show that workingat the school and cluster levels helped to ensure that teacher education was part of abroader capacity development strategy that supported all actors in the education sys-tem, including, for example, head teachers, district education officers and teacher

Professional Development in Education 17

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

trainers, and that it was cost-effective against other competing demands in aresource-poor environment such as Tanzania.

One of the main lessons to emerge from the evaluation was that a move awayfrom ad hoc provision to a more systematic, longer-term and sustainable approach,where the teacher is much more involved in his or her ongoing professionaldevelopment, working with other teachers at the school level, and where other actorsplay critical supporting roles, will do much to enhance the capacity of Tanzanianprimary teachers to deliver quality education (Avalos 2011, Timperley 2011, Defise2013, Flitton and Warwick 2013). A multi-mode system, including distance learningand teacher development at school and school cluster levels, appears to be the mostcost-effective way for ensuring national CPD coverage (Save the Children 2012, Orret al. 2013).

While the process evaluation findings support the view that teacher educationand professional learning should be largely located in the school environment, itshould also be recognised that such provision requires an investment of time andmoney in building partnerships, collaboration and delegation (Mattson 2006,O’Sullivan 2010, Orr et al. 2013). As the variation in external visits to the schoolsshows, capacity-building and incentives need to be devolved down to thoseresponsible for delivering school-based CPD, with a clear division of roles andresponsibilities between national, regional and district offices, and between head tea-chers, schools and teacher educators. This will ensure that teacher education is partof a broader capacity development strategy that supports all actors in the educationsystem, including head teachers, district education officers and teacher trainers, andthat it is cost-effective against all the other competing demands on the educationbudget (O’Sullivan 2010, Hardman et al. 2012).

It also requires a systemic approach to teacher education reform to establish clearlinks between ITE and CPD and an alignment of policies, plans and institutionalarrangements with regard to curriculum reform, assessment practices and teachermanagement, so that an effective school-based programme can ultimately be imple-mented on a national scale (Tikly 2011, King 2014). In decentralising teacher educa-tion in Tanzania it will also be necessary to encourage transparency about thebudget, to build capacity at all levels of the system and to consult all stakeholderson the distribution of responsibilities, resources and incentives. Putting in place asystematic monitoring and evaluation system with input from stakeholders across alllevels of the education system will help improve accountability, planning and imple-mentation, and assist in knowledge sharing.

One of the limitations of the current evaluation with its focus on classroom pro-cesses was the lack of robust evidence on the impact of the school-based CPD piloton children’s learning. While the general perception amongst stakeholders was thatchildren’s learning was improving as a result of the school-based training, no preand post testing had been conducted during the pilot to study its impact on chil-dren’s learning. There is therefore a need to conduct a rigorous evaluation of thenational scale-up using an experimental or quasi-experimental design with a focuson impact as well as process. This will help determine what works in raising theattainment of children in Tanzanian primary schools, and help understand howschool-based teacher development is being implemented on the ground.

Lessons from the pilot have been incorporated into the design of a school-basedCPD programme launched in 2014 in over 20 of Tanzania’s poorest district councils.Known as the Education Quality Improvement Programme in Tanzania and funded

18 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

by a £60 million grant from the UK’s Department for International Development,the school-based teacher development programme is designed to support improve-ments in the quality of primary education in primary schools in Tanzania over thenext four years. The programme is particularly aimed at girls’ education and theteaching of literacy and numeracy, so that school-based CPD can be scaled up at thenational level with funding from the government and international donor commu-nity. Most importantly, it will help develop the CPD infrastructure at the nationaland sub-national levels in Tanzania to support and sustain school-based teacher pro-fessional development beyond the life of the project (Wells 2014).

This evaluation study does not conclude that school-based teacher developmentalone can address the problems faced by teachers and learners in Tanzanian class-rooms where resources are often scarce and the education system is likely to con-tinue to face many constraints. Rather, it suggests that a holistic approach tobuilding an education system is required that recognises the realities of the contextin which teachers work and emphasises capacity-building and the equitable distribu-tion of resources at national, regional, district and school levels. However, the qual-ity of the teachers’ expertise is essential to raising standards in the country’s primaryschools, and the evaluation of the pilot suggests that systematic school-based CPD,together with management and career structures that result in consistent and high-quality performance by teacher educators and teachers, will contribute to teacherprofessionalism and classroom practice, and will significantly help raise educationalachievement in Tanzania and in other countries in the east and southern Africanregion.

Disclosure statementNo potential conflict of interest was reported by the authors.

FundingThe studies cited in this paper were funded by UNICEF, Tanzania. The authors are gratefulto UNICEF for their support but the views expressed are those of the authors.

Notes1. The evaluation team were only commissioned towards the end of the pilot and so the

opportunity to influence the design of the evaluation using a randomised controlled trialor quasi-experimental design with pre-test and post-test control group design was missed.To strengthen the causal inferences that could be drawn from the study, a comparisongroup of four randomly selected schools from a district council not taking part in thepilot was used in addition to the longitudinal comparison with the baseline data.

2. Ward coordinators act as education advisors and are employed by district councils.3. Because of the low academic qualifications of the vast majority of those entering primary

teaching as Grades A and B, teachers are offered the opportunity to follow in-servicecourses upgrading their subject knowledge so they can gain access to diploma and degreecourses. There is, however, little emphasis on upgrading pedagogical subject knowledge.

4. For example, if there were three activities occurring within a five-minute time slot, eachwould be allocated one-third of the time.

5. A type of questioning exchange designed to encourage pupil participation using a mid-sentence rise in voice intonation to get a response from the pupils during an explanationor following a pupil answer. The answer, usually in the form of a choral answer, takesthe form of a repetition or completion of a phrase or word, initiated by the teacher.

Professional Development in Education 19

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

6. A teacher check on understanding often functions as a ritualised teaching exchange withthe teacher asking the class whether they are following. It was found that teachers rarelyprobe for pupil understanding.

ReferencesAbd-Kadir, J. and Hardman, F., 2013. Reforming teacher education in East Africa: the impor-

tance of socio-cultural context. In: L. Jin and M. Cortazzi, eds. Researching cultures oflearning. London: Palgrave Macmillan, 124–134.

Ackers, J. and Hardman, F., 2001. Classroom interaction in Kenyan primary schools.Compare: a journal of comparative and international education, 31 (2), 245–262.

Akyeampong, K., et al., 2013. Improving teaching and learning of basic maths and readingin Africa: does teacher preparation count? International journal of educationaldevelopment, 33 (3), 272–282.

Alexander, R., 2001. Culture and pedagogy: international comparisons in primary education.Oxford: Blackwell.

Alexander, R., 2008. Education for all, the quality imperative and the problem of pedagogy.London: DFID.

Avalos, B., 2011. Teacher professional development in Teaching and teacher education overten years. Teaching and teacher education, 27 (1), 10–20.

Barrett, A.M., 2007. Beyond the polarization of pedagogy: models of classroom practice inTanzanian primary schools. Comparative education, 43 (2), 273–294.

Cordingley, P., et al., 2005. The impact of collaborative CPD on classroom teaching andlearning. In: Research evidence in education library. London: EPPI Centre, SocialScience Research Unit, Institute of Education, 1–167.

Crockett, M.D., et al., 2009. Exploring discourse-based assessment practice and its role inmathematics professional development. Professional development in education, 35 (4),677–680.

Darling-Hammond, L., et al., 2009. Professional learning in the learning profession: a reporton teacher development in the United States and abroad. Stanford, CA: StanfordUniversity/National Staff Development Council.

Defise, R., 2013. Supporting the implementation of curriculum reform through learningcommunities and communities of practice. Prospects, 43 (4), 473–479.

Flitton, L. and Warwick, P., 2013. From classroom analysis to whole-school professionaldevelopment: promoting talk as a tool for learning across school departments.Professional development in education, 39 (1), 99–121.

Hardman, F., et al., 2009. Changing pedagogical practice in Kenyan primary schools: theimpact of school-based training. Comparative education, 45 (1), 65–86.

Hardman, F., et al., 2011. Developing a systematic approach to teacher education insub-Saharan Africa: emerging lessons from Kenya, Tanzania and Uganda. Compare: ajournal of comparative and international education, 41 (5), 669–684.

Hardman, F., Abd-Kadir, J., and Tibuhinda, A., 2012. Reforming teacher education inTanzania. International journal of educational development, 32 (6), 826–834.

Haßler, B., et al., 2014. School-based professional development in a developing context:lessons learnt from a case study in Zambia. Professional development in education,http://dx.doi.org/10.1080/19415257.2014.938355

Hattie, J., 2009. Visible learning: a synthesis of over 800 meta-analyses relating toachievement. Abingdon: Routledge.

Kennedy, A., 2014. Models of continuing professional development: a framework foranalysis. Professional development in education, 40 (3), 336–351.

King, F., 2014. Evaluating the impact of teacher professional development: anevidence-based framework. Professional development in education, 40 (1), 89–111.

Lewin, K.M. and Stuart, J.S., 2003. Researching teacher education: new perspectives onpractice, performance and policy. Multi-Site Teacher Education Research Project(MUSTER) synthesis report. London: DFID.

Mameli, C. and Molinari, L., 2013. Interactive micro-processes in classroom discourse:turning points and emergent meanings. Research papers in education, 28 (2), 196–211.

20 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Mattson, E., 2006. Field-based models of primary teacher training: case studies of studentsupport systems from sub-Saharan Africa. London: DFID.

MoEVT, 2008. INSET Strategy and operational plan linked to the Teacher Development andManagement Strategy (TDMS), 2008–2013. Dar es Salaam: Ministry of Education andVocational Training.

MoEVT, 2009. In-service education and training strategy for primary school teachers2009–2013. Dar es Salaam: Ministry of Education and Vocational Training.

Moon, B., 2007. School-based teacher development in sub-Saharan Africa: building a newresearch agenda. Curriculum journal, 18 (3), 355–371.

O’Sullivan, M.C., 2010. Educating the teacher educator – a Ugandan case study.International journal of educational development, 30 (4), 377–387.

OECD (Organisation for Economic Cooperation and Development), 2011. Building a highquality teaching profession: lessons from around the world. Paris: OECD Publishing.

Orr, D., et al., 2013. What are the impacts and cost-effectiveness of strategies to improveperformance of untrained and under-trained teachers in the classroom in developingcountries? London: EPPI-Centre, Social Science Research Centre, Institute of Education,University of London.

O-saki, K.M. and Agu, A.O., 2002. A study of classroom interaction in primary schools inthe United Republic of Tanzania. Prospects, 32 (1), 103–116.

Pontefract, C. and Hardman, F., 2005. The discourse of classroom interaction in Kenyanprimary schools. Comparative education, 41 (1), 87–106.

Save the Children, 2012. Review: teacher support and development interventions. London:Save the Children.

Schweisfurth, M., 2013. Learner-centred education in international perspective: whose peda-gogy for whose development? Oxford: Routledge.

Schwille, J., Dembele, M., and Schubert, J., 2007. Global perspectives on teacher learning:improving policy and practice. Paris: UNESCO – International Institute for EducationPlanning.

Sinclair, J. and Coulthard, M., 1992. Towards an analysis of discourse. In: M. Coulthard, ed.Advances in spoken discourse analysis. London: Routledge, 1–34.

Southern African Consortium for Measuring Educational Quality (SACMEQ), 2010. SAC-MEQ III project results [online]. Available from: http://www.sacmeq.org/reports.htm.

Tam Choi Fung, A., 2015. The role of a professional learning community in teacher change:a perspective from beliefs and practices. Teachers and teaching: theory and practice, 21(1), 22–43.

Thenjiwe, M. and Mulvihill, T.M., 2009. Julius Nyerere (1922–1999), an Africanphilosopher, re-envisions teacher education to escape colonialism. New proposals:journal of Marxism and interdisciplinary inquiry, 3 (1), 15–22.

Tikly, L., 2011. Towards a framework for researching the quality of education in low-incomecountries. Comparative education, 47 (1), 1–23.

Timperley, H., 2008. Teacher professional learning and development. Brussels: TheInternational Academy of Education.

Timperley, H., 2011. Realising the power of professional learning. London: Open UniversityPress.

Towse, P., et al., 2003. Non-graduate teacher recruitment and retention: some factors affect-ing teacher effectiveness in Tanzania. Teaching and teacher education, 18 (6), 637–652.

UNESCO, 2013. UIS statistics in brief education (all levels) profile – United Republic ofTanzania. Paris: UNESCO Institute for Statistics.

UNESCO, 2014. EFA global monitoring report 2011: teaching and learning: achievingquality for all. Paris: UNESCO.

UNICEF, 2009. The quality of teaching and learning in Tanzanian primary schools: abaseline study. Dar es Salaam: UNICEF.

Vavrus, F., 2009. The cultural politics of constructivist pedagogies: teacher education reformin the United Republic of Tanzania. International journal of educational development, 29(3), 303–311.

Wedgwood, R., 2007. Education and poverty reduction in Tanzania. International journal ofeducational development, 27 (4), 383–396.

Professional Development in Education 21

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5

Wedin, A., 2010. Classroom interaction: potential or problem? The case of Karagwe.International journal of educational development, 30 (2), 145–150.

Wells, M., 2014. Elements of effective and sustainable professional learning. Professionaldevelopment in education, 40 (3), 488–504.

Westbrook, J., et al., 2013. Pedagogy, curriculum, teaching practices and teacher educationin developing countries. London: EPPI Education Rigorous Literature Review, Instituteof Education.

22 F. Hardman et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f Y

ork]

at 0

9:18

15

May

201

5