1 Data Know Best: Tracking Learner Achievement and Assessing the Impact of the Old Mutual Flagship Program in Nine Project Schools in Motheo District, Free State Province, South Africa. Zotha Zungu A field practicum report submitted in partial fulfilment o f the requirements in respect of the Master of Sustainable Development Degree at the University of Florida in Gainesville, Florida, USA. December 2017 Supervisory Committee: Dr. Sarah McKune (Chair) Dr. Glenn D. Israel (Member)
2 ACKNOWLEDGEMENTS I thank God for the opportunity, strength and perseverance; my family and friends for their prayers and support. Thank you to New Leaders Foundation and Old Mutual Education Flagship Program teams, as well as The University of the Free State Schools Partn ership Program for which this study would not have been possible. To my supervisory committee members Dr. Sarah McKune and Dr. Glenn Israel for their wisdom, patience and guidance. To the University Masters of Sustainable Development Practice Program Dr. Glenn Galloway for academic support. To Dr. Andy Noss for his kindness; always going above and beyond his duties to advise, support and assist. To the MDP Cohor t 6, Class of 2017 for making my academic journey worthwhile, and to Bertha Kitching, Director of Strategic Planning, Policy and Research and her team at the Free State Department of Education for her swift efforts and support in making sure I complied wit h the necessary ethical requirements in conducting the study. Finally, to me: ometimes it is the people no one can imagine anything of; who do Do the unimaginable!
3 T ABLE OF C ONTENTS Acknowledgem 2 4 6 7 7 8 1 1 2.4 11 2.5 Data 3 .............................. . ..... . ............... ....1 4 1 4 1 8 2 3 2 9 4 5 5 6 7 2 2 3 Educat 6 4 6 5
4 Abstract The term big data made waves and became widespread by 2011 through technology firms (Gandomi, 2015 ). Organizations across the spectrum have become increasingly consumed with how data they own or have access to can potentially revolutionize their operations or increase their customer share. Governmental departments have also become interested in how the y can use data they generate and own to improve their systems and services. It is against this backdrop that a study was conducted focusing on producing data for the purposes of improving learner achievement and governance in schools in South Africa. The o bjective of the study was to implement an evaluation tool (Data Driven Districts Tracker) in nine schools that are recipients of the Old Mutual Education Flagship Program (OMEFP) in Motheo District in the Free State Province. Six metrics district leade rship, school leadership, subject leadership, educator skills, learner results and school governance and community involvement were used as a baseline assessment of the OMEFP. The tracker, a survey instrument with questions specific to each group of educ ators, subject heads of department, deputy principals and principals, was the primary data collection method. A second method the Data Driven Districts Dashboard (DDD) was also used. An online system that extracts school data from a national data system ca lled the South African School Management Administration System (SA SAMS), individualizes it and presents it in visuals with a less complex interface for schools to access and use to inform decision making. The tracker and DDD results form a progress report mechanism to funding and participant stakeholders and informs them of the scores in learner achievement and how educators and school management teams fare in various categories aligned to the six metrics. The results also
5 highlight possible deficiencies a nd challenges that exist in the schools that necessitate urgent interventions and mechanisms focused on the improvement of instruction, skills or governance.
6 Introduction The United Nations Secretary General Mr. Ban Ki Moon describes the 17 Sustainable Dev contract between the world and its people; a to do list and a blue print for success that purposes reduce and eradicate poverty, inequality, and climate change amongst other global challenges, as well as improve the quality of education over the next 15 years (United Nations, 2015) The field practicum is based on goal number 4 of the 17 SDGs, a continua tion of Goal 2 of the Millennium Development Goals (MDGs), which focused on achieving universal primary education. Targets 4A and 4A1 of the SDGs explain access to resources and infrastructure as a component in reaching goal 4, and target 4c looks at the i ncrease of qualified teachers as the key to achieving all the SDG 4 targets. The field practicum aligns with target 4c which requires governments and institutions to make concerted efforts towards closing the equity gap in the provision of quality educatio n; a gap that is evident in the inadequate and uneven distribution of professionally trained teachers and lack of management support and capacity, particularly in disadvantaged areas. Educator empowerment, motivation, professional training, support within well resourced efficient and effectively governed systems, and adequate recruitment and instrumental in the delivery of quality education (UNESCO, 2016).
7 Chapter 1 Background and Context 1.1 Country Context South Africa presently in its 23 rd year of democracy has enjoyed and suffered significant changes. The country suffered segregation and inequality under the rule of apartheid which dates back from 1948 until 19 94. This period saw quality human services being provided to an elite white minority and industrial labor and poor human services delegated to the black majority (McKeerver, 2016). These inherited racist systems and policies that continue to plague the Sou th African government and remedying their impact is an ongoing challenge. With a population of 55,908,865 million people having increased from 40, 558 495 million in 1994 (World Bank Data 2017) and 5.8 million people unemployed in late 2016 (Statistics Sou th Africa Anual Report, 2016/17), the country remains one of the most economically unequal countries globally with a Gini Coefficient of .65 (World Bank 2011). During the first quarter of 2017 the South African economy moved into recession with a contract ion of 0.7% following a contraction of 0.3% in the fourth quarter of 2016 (Statssa, 2017). Negative economic growth adversely affected employment across the board and specifically within the construction, non formal agricultural, manufacturing and communi ty services sectors. Unemployment increased by 1.2% to reach 27.7% during the first quarter of 2017 (Statistics South Africa Labor Force Survey 2017). As all human services strain in delivering services to the population, education is not exempt under a sh rinking economy and increasing budget deficits. Challenges in resource allocation, skills development and lack of infrastructure development plague the department of education which is held accountable for the low learner achievements rates and the overall inferior quality of education. In the next section I discuss the
8 Department of Basic Education and its mandate as a precursor to the study and analysis of the field practicum. 1.2 Basic Education in South Africa The core functions of the Department of Bas ic Education in South Africa are to; 1. Improve the quality of teaching and learning 2. Undertake regular assessments in order to track changes in learner performance 3. Improve early childhood development 4. Provide human resource capacity; right skills at the right time (The South African Department of Basic Education Service Delivery Improvement Plan 2015 2018). South Africa has 12.3 million learners, 386 600 teachers and 26 292 schools. High schools account for 6000 of the total and primary schools th e remaining 20 292 (The Eastern Cape Department of Education 2017). The lack of resources leading to infrastructure deficiencies and lack of student services characterize many schools. For example, around 27 per cent of public schools do not have running w ater, 78 per cent are without libraries and 78 per cent do not have computers (UNICEF South Africa 2014). The ratio of teacher to learner in public schools, 32.6 to 1 in comparison to private schools that have one teacher for every 17.5 learners (The Easte rn Cape Department of Education 2017), creates a burden for educators in adequately teaching and dedicating time and attention to each learner. Added to these challenges, a study conducted by the Annual National Assessment on numeracy and literacy skills f or foundation to intermediate phases (grades 1 6) revealed an achievement of partial level of learning (depending on the subject and grade) of 30 47 percent. Other findings indicated that the quality of teaching was poor, further contributing to poor learn er performance (UNICEF South Africa 2017).
9 In a 2015 survey by Statistics South Africa on the reasons given by persons aged 7 18 years for not attending an educational institution, 20.3 percent of respondents gave poor academic performance as a primary re ason (Statistics South Africa 2015). Figure 1 shows the bachelor pass percentage for all provinces in South Africa for the period 2016. The very low national pass rate of 26% confirms the need for interventions such as the Old Mutual Flagship Project, the objectives and activities of which will be explained later in further detail. Figure 1 : Bachelor Pass Percentage in South Africa for 2016 Source: Gauteng Department of Basic Education MEC Presentation: Education Lekgotla 2017 Figure 2 compares 2015 wit h 2016 grade 12 pass rates among all provinces in South Africa, including the field practicum province of the Free State which achieved a pass rate above 80% for both 2015 and 2016.
10 The Eastern Cape Province performed the lowest at 55% and 59% for both y ears followed by Kwa Zulu Natal at 60% and 66%. The poor results are due to a lack of resources, skills, capacity and management particularly in lower grades. These problems are magnified in the senior grades and particularly in grade 12 results which are critical and essential in applying and getting into tertiary education, getting employment and ultimately improving living conditions. Figure 3 provides a closer look at the field practicum project area. The Free State Province, a predominantly rural area with vast agricultural land and a 35.5% unemployment rate (Statssa Quarterly Labour Force Survey: Quarter 1, 2017), also bears a number of educational challenges, which include low pass rates in the key subjects of Mathematics, Physical Science, English an d well as the support of mathematics and science educators by school heads of department (Digital Classroom 2014). Low quality is characterized by lack of adequa tely skilled educators who do not have support and adequate resources to deliver the curriculum. The Free State also experiences decreased percentages of learners opting for pure Mathematics (as opposed to
11 Mathematical Literacy) and Physical Science (Digit al Classroom, 2014). It is against this backdrop that the OMEFP was implemented in the Thaba Nchu and Botshabelo rural areas. Figure 3. Free State Province highlighting Motheo District, Bloemfontein. Source: http://www.businessinsa.com/free state/ 1.3 The Old Mutual Flagship Project (OMEFP) The Old Mutual Education Flagship Project is an initiative by Old Mutual, an international financial institution operating in South Africa. It has a social development arm, the Old Mutual Foundation, which has a range of focus areas that target marginalized citizens in the rural and peri urban areas. The focus areas are: enterprise development, skills capacity building, education, staff volunteerism, and vul nerable communities (Old Mutual Foundation 2017). For 2013 the initiative supports basic education interventions focused on improving school functioning, which affects learner achievement. This is proposed to be done by, among other
12 things, building capacity within school leadership and governance; strengthening district level leadership; improving and supporting teaching skills through educator capacity building within Mathematics and Science subjects; curriculum and pedagogical support to educators; resource and infrastructure support; encouraging the establishment and maintenance of Communities of Practice (CoPs) amongst principal educator community for best pr actice (Old Mutual Foundation 2017). The program specifically focused on curriculum support, school management and leadership has produced positive results in areas like Limpopo, Gauteng and the Eastern Cape where outcomes include improved learner scores a nd achievement ( Old Mutual Foundation 2017) The project includes the stakeholders whose roles and contributions are listed below. Stakeholder Interests and Contributions Figure 4 summarizes individual stakeholder contributions and expectations in the OM EFP. The text in black outlines inputs of each stakeholder, and the blue text highlights project expectations in terms of outcomes. The Old Mutual Board of Trustees represents the funder and is the strategic guide for the project, whose interests are in t he provision of quality education and project impact. The University of the Free re quality data, capacity building and improved bachelor pass percentage. New Leaders Foundation is the implementing organization responsible for monitoring and evaluation and the Data Driven Districts Dashboard project. Their interests are improved learne r achievement specifically in Math and Science subjects, quality data, and improved bachelor pass percentages. School Districts and School Management Teams are the participants and respondents in the study who
13 teach learners and manage the schools. Their i nterests are improved learner achievements, school management, skills development, data and project outcomes. Figure 4: OMEFP Stakeholder Inputs and Expectations. 1.4 Host Organization New Leaders Foundation (NLF) is a not for profit organization based i n Parktown Johannesburg that focuses on transforming education through leadership development and empowerment (New Leaders Foundation, 2017). Involved primarily with the distribution of the Data Driven Districts Dashboard project funded by the Michael and Susan Dell Foundation, it facilitates the roll out and training of management teams of the department of basic education in utilizing the online data system for better quality data. The impact the organization has achieved is commendable with direct impact in 86 districts and 900 circuits, and indirect impact throughout the Department of Education, with lower levels reaching 9 million learners (New Leaders Foundation, 2017).
14 The NLF became a partner in the OMEFP when the project was rolled out in Jane Furse Limpopo province in January 2016. The organization was tasked with implementing the flagship Data Driven Districts (DDD) program and associated DDD dashboard in 12 project schools (New Leaders Foundation, 2017). In June 2016, the OMEFP leadership tasked NLF to lead an M&E process in the Eastern Cape area of King Williams Town where it would identify the various quantitative metrics/indicators that OMEFP should be tracking to assess its impact in the following six outcome targets: district leadership, scho ol leadership, subject leadership, educator skills, learner results and school governance and community involvement. The task gave rise to a draft Excel based quantitative tool (DDD tracker) which was commissioned to be implemented in all the OMEFP project areas. Project areas included the Motheo District where the study was conducted. The OMEFP in the Free State is known as the Schools Partnership Program (SPP) housed at the University of the Free State (UFS). A major project that was launched in 2011 (th at includes many other funding stakeholders) with an overall goal of improving academic achievement in key subject areas (Math, English, Science, Accounting) using curriculum mentoring for educators. The NLF and UFS SPP are partners in the implementation o f the OMEFP and the monitoring and evaluation process. To understand the impact of data based decision making in education projects globally, we review similar programs carried out and their outcomes. 1.5 Data Based Decision Making in International Educat ion Interventions have been carried out in Brazil (Minas Gerais), Pakistan (Punjab), and New York, and in each case data have been used to improve education outcomes. In Minas Gerais, 18,000 schools faced challenges in reading levels: in 2003, only 49% of eight year olds were reading at
15 the recommended level, and 31% were reading poorly (The World Bank, 2003). A targeted strategy including continuous monitoring based on robust gathering and use of data, scripted literacy instruction materials, standardize d teaching techniques, and teacher training and coaching led to an impressive turn around. By the year 2010, 86% of eight year olds were reading at the recommended proficiency level, and the portion of those scoring poorly in reading had been reduced to 6%. Results were informed by rigorous data collection and analysis. Data have been used to effect changes in the outcomes and improve literacy (The World Bank 2003). In New York, Grow Network's print and web bases school data reports aim to link data resul ts with learning and improved learner achievement (Brunner et, al. 2009). The system is designed to organize and provide raw data on test scores and state performance levels in line with the New York State standards. The system organizes raw data for diff erent users (i.e. teacher, administrator, student, parent), providing specific data for each as they mean different things for different audiences. Data are interpreted into information that informs audience decisions related to instruction, learner and cu rriculum development and support. A survey was done along with interviews to understand how the tool was used by educators across New York and how it informed their decision making in their classrooms. The results showed that 37% of teachers used the Grow Reports monthly, and 32% of the teachers reporting using the reports three to six times throughout the year (Brunner et. al, 2009). This can be interpreted as finding the system useful for educator decision making. The tool also facilitated different resul ts for different audiences, i.e. better planning for educators, diverse needs of diverse learners, more discussions about student learning and shaped professional development for educators. Administrators and school heads could identify areas of need and d irect resources and attention where it needed. The Grow reports proved to be an
16 effective navigational tool for educators and a critical tool for administrators and executives for accountability purposes (Brunner et, al. 2009). In 2011, the government of P unjab launched a program targeted at improving access, quality, and infrastructure challenges that faced more than 60 000 schools in the rural areas. Here again the strategy included data driven decision making, low tech highly innovative approaches, clear targets and detailed context specific implementation plans. The program had remarkable results in just 18 months: one million additional primary age children enrolled in school, and teacher absenteeism was reduced by 35% (The Department of Education South Africa, 2014). The DDD dashboard is similar to projects implemented in the United States namely, the K 12 Assessment reporting system which allows users (teachers, administrators) to track student performance to effect data driven instruction. Ribant (20 13) highlights the benefits of a data driven approach; students are having problems with learning achievement, create a strategy for helping them and then monitor student progress throughout the year. This approach really pays off for students and teachers because the objective data starts relevant conversations about what The data driven decision making model has been used in several American institutions of higher education where it has been instrumental in identifying strategies and influencing improvements in retention (Picciano, 2011). Schifter, et al. (2014) discuss a data driven decision making project in th e education sector, initiated by the National Science Foundation. The project process aimed at helping middle school science teachers uses diverse data from a lts showing promise for a model of training teachers to use data from the dashboard and data driven decision making principles, to identify science misunderstandings, and to use the data to design lesson options to address those r et, al. 2014, p.419).
17 dashboard. The dashboard is designed to serve administrative, researcher, teacher, and student needs at the same time. Teachers track student pe rformance using the dashboard to make the relevant learning interventions in class. In the following section we discuss the theory of change and conceptual framework of the OMEFP. The model in Figure 5 summarizes the program process and how the different stakeholders are involved in the expected outcomes of improved learner achievement and quality education.
18 Chapter 2 Theory of Change Figure 5 outlines what each stakeholder contributes and how the project aims to achieve its outcome. On the right hand si de are stakeholders that contribute to inputs and outputs whilst on the left are beneficiaries (educators and school management teams) who are responsible for outcomes. The inputs and outputs on the right hand side of the model should lead to decision maki ng that includes findings from the extracted and reported data taken from the tracker, curriculum mentoring results and data from the DDD dashboard. The expectation is that the inputs and outputs contribute to an overall change in learner achievement and q uality education. Figure 5: Field Practicum Process Model
19 Figure 6: Old Mutual Flagship Program Conceptual Model
20 Figure 6 presents the conceptual model of the OMEFP project and what it aims to achieve in reaching improved learner achievement an d quality education. Poor learner achievement, the problem to be addressed (depicted in the orange square), is caused by a plethora of contributing factors represented by the blue rectangle immediately below, for example, governance and administration defi ciencies to poverty; uninvolved parents and communities; limited educator skills. In addition, poor school data quality and the lack of technological infrastructure to support data management affect decision making. Even political instability can ostracize some schools and limit resources and skills. To manage these factors the interventions in the green squares stand out as inputs that would address the existing challenge: support for educator and school management; capacity building for administration sta ff in input of data and managing the development of school infrastructure can be supported by external structures including corporations and foundation by allocating necessary resources to educators and learners (books, apparatus). Curriculum support for educators as done by the OMEFP/UFS SPP project where educators receive curriculum advice and resources from mentors who are subject specialist; socio economic support for learners to mitigate challenges they face at home related to poverty, psychological disturbances experienced through violence and absent parents, as well lack of access to services like social services and health care. Parents also need to be involved in the learning process and school programs that affect their students and this can be done through more effective outreach to communities by the schools. To understand the flow of these inputs and interventions and how they contribute to improved outcome s, it is important to understand the theory of change for the Data Driven Districts
21 Dashboard which is an input in the conceptual framework and allows for data to be accessible to schools for them to visualize and use the data to inform decision making. T heory of Change: Data Driven Districts Dashboard Figure 7 illustrates of the theory of change for the Data Driven Districts Dashboard to show how the project contributes to data based decision making and learner outcomes. The DDD tool makes data available to schools for school administration staff and key users to better understand how the school is performing. The tool facilitates access to data and offers superior quality data (as opposed to the national South African School Administration and Management System) focused on each school which then allows each school to target challenge areas more specifically. School administration is then trained to use the tool, to trouble shoot and produce superior data that will inform targeted interventions and better d ecision making towards improved learner outcomes and effective data management.
22 Figure 7: Databased Districts Dashboard Theory of Change Source : OMEFP Evaluation Report (PDG) 2015
23 Chapter 3 Methods A paper based survey method was used to colle ct data. Three surveys were designed by the host organization (NLF) and funder Old Mutual Foundation with specific questions (see appendix). The surveys were completed in pen and administered over 3 days, on the 27 th 28 th and 29 th June 2017. They were adm inistered to the three groups of participants (principals and deputy principals, educators and heads of department) with specific questions for each group. The hers as well as heads of department for Math, English and Science subjects only in the project schools. The decision to design separate surveys for the school management teams (principals and heads of department) and educators was made to encourage objecti vity and accurate measurement in participant responses and to provide broader insight. heading. A Likert scale was used for each response (answers of either strongl y agree, slightly agree, agree, disagree, or strongly disagree). The principal and deputy principal questionnaires included questions on how the district supports the school, community involvement, and school governance. The educator questionnaire consiste d of questions relating to staff relationships, school infrastructure, school governance, and community involvement while the head of department questionnaire contained questions relating to subject and district subject support, staff relationship, school infrastructure, school governance, and community involvement. Samples of the survey instruments are found in the appendix. Participant schools were selected based on their socio economic and geographical location (i.e. schools in underprivileged rural are as) in Motheo District (Thaba Nchu and Botshabelo) and on
24 their inferior performance/scores in Math, English and Science. The distance from Bloemfontein (central business district) to Thaba Nchu rural township was approximately 71km and a further 17km from Thaba Nchu to Botshabelo. Schools in each area were not more than 5 km apart, which did not take more than 20 minutes considering the adverse conditions of the roads (potholes and flooding due to underserviced roads and storm water drains, and in other ar eas lack of infrastructure). In one school that was in the deep rural area, distance travelled was a further 60km (1 hour 30 minutes) outside of Thaba Nchu. (Figure 8). Figure 8: Map showing the distance between project schools Source: https://g oo.gl/maps/fA5GyuhGmwR2
25 Surveys and survey instructions were given to principals of each school to administer to the participants. Because the responses were not confidential, this might have led to socially desirable responses by some individuals. We were unable to personally administer the process due to the limited time as schools were closing in the same week. In 8 of the schools, completed surveys were collected the following day. In one remaining school they were collected on the same day due to t Figure 9 represents respondents by group and school. School names have been anonymized and replaced with MT codes. The number of respondents that were counted were; 144 educat ors, 15 principals and deputy principals, and 29 heads of department. A total of 188 respondents completed the surveys. Figure 9: Respondents Per School Source: New Leaders Foundation Figure 10 presents the survey response rate by group in each project school. The overall response rate was 90% with the majority of low response rates seen in educators, the lowest reported in
26 HOD responses and the highest in principal responses. One of the reasons for the low educator response rate was due to educators not being present at the time surveys were being completed. In all of the schools, some educators would not complete the surveys. A possible reason could be because the surveys were administered by the principal, which may have resulted in respondents being u nsure about confidentiality. Figure 10: Survey Response Rate by Group and Per School Source: New Leaders Foundation Limitations Limitations that were experienced included: not being able to personally administer the surveys and explain the questions to each respondent which would allow them to ask questions and request clarity where necessary. The use of one field data collection method was another limitation, more information could be gathered using multiple methods over an extended period
27 for in de pth assessment. The main reason for not conducting follow up qualitative methods was due to having limited time as schools were closing on the same week. Double barreled compound questions in a Likert Scale were also another challenge. In this instance res pondents are put in a position where they must decide which question to answer, with a limiting response scale. This confuses the respondent and produces questionable data. The survey was also difficult to read as it was clustered to save space, to fit as many questions as possible. This may be overwhelming to the respondent, further affecting how the respondent answers the questions and how much time they dedicate to each question. The reliability and quality of the data produced may then be in question. C hallenges Challenges experienced whilst conducting the field practicum included delays in being granted ethical clearance. This precluded additional planned interviews and diminished survey completion rates. The survey responses may have not been impartial due to participants being aware that the survey was from a funding partner and stakeholder. This could create courtesy bias in that school principals may think any future support or funding Old Mutual would provide would be compromised if questions are an swered in a disapproving manner. conviction on the program impact in his school. He communicated along with his deputy that they did not believe the program was effe ctive. He accepted the surveys but returned half of them stating that he knew that educators would not complete them. The fluctuating scores in significant enough. For the senior phase there are decreases instead of improvements according to the data in this study.
28 Another challenge was accessing raw data from the University of the Free compare with results from the DDD dashboard. Regarding data from the dashboard, the main challenges were the system being down most of the time and apparent discrepancies and gaps in scores due to non completion of data input from schools. This brings question to data reliability. Causes for the non completion of data inputs included on going training and capacity building as well as general delays from school administrators and technical support. Analysis Analysis of the quantitative data from the surveys is presented using descriptive measures in graphical format wit h percentages. Raw survey data from each survey document were uploaded into a google survey document which was then transferred into an Excel spreadsheet and presented as descriptive statistics in simple graphical format. All categories present in the surv ey document were represented in the Excel document. Scores from out of 5 were allocated to each category and a score of 3 and above considered as acceptable Data for subject achievement was sourced from the DDD dashboard analyzed by the NLF analytics tea m and data for the bachelor passes and grade 12 passes was sourced from the UFS SPP p r o j e c t report for 2017 w h i c h w a s m a d e a v a i l a b l e b y t h e U F S S P P t e a m
29 Chapter 4 Results Out of 238 surveys that were administered to the participants, 188 (79%) surveys were returned Figure 11 depicts the overall average score from the project schools for categories relating to the six indicators, with a score out of 5. The results show the lowest score, 2.98, on parental involvement in school activities and the highest score, 4.33, for circuit office support. Educ ator support and resource availability scored 3.46, which is a very important indicator as this directly affects how and what learners are taught. Overall Scores Figure 11: Overall Indicator Score for All Participant Schools Source: New Leaders Foundati on Availability of Material Figure 12 looks at the availability of sufficient practical materials, specifically for physical and natural sciences in the school laboratories. These responses are from heads of department (HOD), and educators. Educators of 5 schools (TM 01, 02, 03, 04, 06, and 07) scored this
30 category below 3, meaning there were insufficient practical material available in their schools. For their part, heads of department (HODs) for only 3 schools (TM 02, 06, and 07) scored this category b elow 3. The highest score was from HODs in TM 05 with a score of 4. 7 5 and the lowest was 1.00 from HODs in TM 07. Educator results disagree with HOD results on the availability of practical material at some schools therefore this outcome must be explored further to understand probable causes and to further investigate the divergence in responses from both groups of respondents. A score below 3 means schools are under resourced which affects the way in which classes like natural science are taught. If appar atus like Bunsen burners, acids, and Figure 12: HOD and Educator Scores for Availability of Su bject Resource Material Source: New Leaders Foundation tripod stands are not available, educators struggle to conduct experiments, and learners face challenges in understanding lessons that require these experiments.
31 Problem Solving Processes Figure 13 s hows problem solving process scores, the ability of the district office to solve problems that affect school and learner achievement. Examples of problems would be lack of support in terms of the curriculum, subject/educator development and resources affec ting the school. Results show score ranges of 3 to 5 with all schools achieving a score of 3 and above, meaning there is a favorable relationship between HODs and the District. This should contribute to improved management in schools and improved learner a chievement. The survey asked how the district supports HODs and whether they can log/report issues successfully with the district. Figure 13: Problem Solving Processes Score Source: New Leaders Foundation Educator Subject Resource The educator subjec t resource score considers the availability of teacher resources for subjects that support the learning process for educators and students, for example, textbooks. The
32 unavailability of resources for educators is known to adversely impact learner achieveme nt. When teachers lack resources and support, they are unable to provide maximum support and direction to students. Scores ranged from 2.65 to 4.12 (figure 14). Figure 14: Educator Subject Resource and Support Source: New Leaders Foundation HOD Sta ff Relationships The HOD Staff relationship score considers how well these two groups work together in the school environment to deliver lessons and improve learner achievement (figure 15). The lowest scoring school with 2,89 reflects communication deficie ncies between the two participants which negatively affects student learning for subjects concerned. A healthy collaborative relationship seen in the TM 07 score of 4. 1 7, must exist between educators and subject heads of department to deliver maximum qual ity instruction for each subject.
33 F igure 15: Head of Department Staff Relationship Score Source: New Leaders Foundation fr om them, all schools responded with an average score of 4 which indicates a favorable level of support being provided by HODs to the educators in all subjects (figure 16). Another observation is that MT 04 and MT 05 did not respond about their perceptions on the Math HOD. Further investigation must be made to ascertain the reasons for not answering the question. There are various possibilities.
34 Figure 16: Educator Perceptions of HOD Scores Source: New Leaders Foundation School Governing Body Effecti veness The school governing body (SGB) effectiveness score considers the structure of the SGB; whether it is functional, effectively follows and communicates the strategic direction of the school and legally complies with the education department requirem ents ( F igure 17). TM01 received the lowest score on all 3 sub categories whil e TM02 received the highest on all 3. TM 04 received the lowest score on strategic direction (note: questions from this category relate to whether SGBs provide strategic leade rship without interfering in school management ) With an overall low score for TM 01, this result could be indicative of the failure of some structures of the SGB.
35 Figure 17: School Governing Body Effectiveness Score Source: New Leaders Foundation Scho ol Parental Involvement The school parental involvement score communicates the level in which parents are involved in TM01, 02 and 09 score quite low, between 1.33 and 2.67. The highest score TM07 at 3.50. For parents to understand the challenges schools and learners face, it is important for them to become involved in the school through parent teacher meetings, school governing body meetings (parent representatives ), fundraising events, and even volunteering for diverse needs the school may have. Parent absence may be due to lack of interest, unwelcoming school environment, inaccessibility of schools because of their distance from learner homes or parents who feel i nadequate to participate due to being uneducated themselves.
36 Figure 18: School Parental Involvement Score Source: New Leaders Foundation Comparison of Indicators: All Schools Figure 19 compares overall survey results across all nine project schools. T he lowest scoring school was MT06 and the highest, MT 07. Most schools scored below 3 for School Parental Involvement. For HOD Staff Relationships questions that were answered were in relation to: communication on mission and vision of the school an d the opportunity to share ideas. All schools scored above 3. School Parental Involvement questions related to: whether parents attend meetings, check homework and whether they are involved in extra school activities. The scores were low in all schools, 4 scored below 3. Educator Subject Resource support saw scores of above 3 for all schools. District Support questions related to: visits from the circuit manager 3 times a year, whether the visits were informative and constructive visits and how frequent and open communication is. In Problem Solving Processes questions related to whether there is clear
37 communication whether issues are logged/reported/attended to in agreed time frame and whether there is a clear process of reporting issues. All schools scored above 3. In Educator Perceptions on HODs (Math and Science) questions related to HOD experience capability, skills and support. All scores were above 3. It would be beneficial going forward to analyze the differences in scores and consider how schoo ls can learn from one another; those that have not scored well learning from those that have scored well. This exercise promotes self evaluation and sharing of best practices between the schools, further supporting improvement of school management/governan ce and learner achievement. Figure 19: Comparison of Indicators Across Schools
38 Data Analysis from SPP Reports and DDD: Learner Outcomes 2015 2017 The DDD dashboard provided data on both OMEFP schools and NON OMEFP schools and comparisons were made based on the OMEFP/SPP mentoring interventions. The scores focus on F urther E ducation and T raining (FET) phases (Grades 7 9) and senior phases (Grades 10 12). Non OMEFP school or control school were selected based on the location (schools in d isadvantaged areas), size (number of students in the school) and quintile. Quintile refers to a category South African schools are allocated by the national education department. This system of categories was introduced in 1998 to improve equity in educati on, since poverty is a barrier to schooling in South Africa. community in which the school is located. The ranking is based on the average level of income, the une mployment rate and level of education within the community each of which are given a specific weighting that is determined by the Department of Education. Schools falling in the bottom 20% of this ranking (i.e. the poorest schools) are classified as being Quintile 1 schools. The distribution of schools according to quintiles for 2015 is represented in the figure below. Note that the names of the provinces hav e been abbreviated on the horizontal axis ( i.e. EC Eastern Cape, FS Free State, GT Gauteng, KZ Kwa Zulu Natal, LP Limpopo, MP Mpumalanga, NC Northern Cape, NW North West, WC Western Cape ) In F igure 19 we observe that the Free State province has most o f its schools in the lowest quintile, this being the highest among the provinces. The Western Cape Province is observed as having the highest number of schools in the highest quintile, with Gauteng following behind. With the Free State province having the most schools in under resourced and disadvantaged areas, this contributes to the learner achievement challenges faced by the schools in the area and particularly the project schools.
39 Figure 20: Distribution of South African Schools by Province Acc ording to Quintiles for 2016. Source: (Van Wyk, 2015) Bachelor Passes A greater percent is seen in bachelor passes for project schools compared to control schools for most years (Figure 21) A slight increase to 15.8% in 2012 around the time the intervention was first implemented, to a notable 33.9% in 2016 However, there is a notable decrease in 2014/15. What is important to note is that these are a different cohort of students for each year and the changes may be attributed to t his. The r esults suggest that project schools are making progress, indicated by pass rates in the project schools in four out of six years.
40 Bachelor Passes: Project Schools vs Control Schools Figure 21 : Bachelor Pass Percentages in Project Scho ols and Control Schools from 2011 to 2016. Source: The University of Free State Schools Partnership Program Annual Report 2017 Bachelor Passes: Ex Model C Schools vs Project Schools Figure 21 compares percentage passes between project schools and ex mod el c schools (dark blue and dark green) in Thaba Nchu and Botshabelo for the period 2014 2016. The term ex infrastructure and resources. According to the results project scho ols have improved in 2016 compar ed with the historically resourced and better organized ex model c schools. After 1994 and in the wake of democracy, racially segregated model c schools were scrapped, and all gh they were still with fewer inferior quality resources under the Department of Basic Education. However, this meant ex model c schools continued to benefit as they had already been ahead compared to blacks only schools. Thus, the
41 increase in percentage p asses for the project schools compared to the ex model c schools is very important because the schools achieve with less resources. A decrease from 86.6 % to 82.8% passes is observed in 2015, thereafter 2016 experiences recovery to 89.1%. What must be noted is that the cohort of students is different and thus the variations may be attributed to this, another factor to consider and discussed by the mentorship team was a recently implemented law of progressing learners. Grade 12 Passes: Ex Model C vs Pro ject Schools Figure 22: A Grade 12 Pass Percentage in Ex Model C Schools and Project Schools (2014 2016). Source: The University of Free State Schools Partnership Program Annual Report 2017 The system of progressing learners launched in June 2015 dictat es that learners in the FET phase must not be retained for longer than 4 years; they must be promoted if they fail a phase more than once (The Department of Basic Education, 2015). Learners who are progressed in an
42 automatic fashion not on academic perfo rmance are not well equipped and carry through their learning deficiencies into the next class. This impacts educators who are prepared for learners that are academically capable and have been promoted through merit. Educators may be burdened with dedica ting more time and attention to progressed learners whilst trying to continue with the curriculum as scheduled. A result would be students who do not grasp the curriculum contributing to the decrease in pass rates. Senior and FET Phases: Math and Science Scores The DDD dashboard also provided information on specific subject performance and how project schools fared from quarter 2 in 2015 to quarter 2 in 2017; compared to non project/control schools. The pass rates are disaggregated by phase. Figure 22 and 23 show the pass rate for Mathematics and Physical Science in the FET (Grades 10 12) and senior phases. M athematics Score s Figure 23: Mathematics Pass Rate for Project and Control Schools by Senior (Grade 7 9) and FET (Grade 10 12) Phases for 201 5 2017. Source: Free State Data Driven Districts Dashboard 2017.
43 In Figure 23 the pass rates for mathematics are presented, still considering that the cohorts are different; slight improvements from Q2 to Q4 are observed in the senior phase, with the co ntrol schools scoring better than project schools. The opposite happens in the FET phase with project schools showing slight increases compared to control schools, with Q 2 of 2017 improving though not significantly enough compared to control schools. The reason for senior phase scoring better than FET phases needs to be investigated as well as the fluctuations and considerable decreases in Quarters 2 and 4 (2015) and the spike in Quarter 4 (2016) in the senior phase. Overall the results are not encouraging and further investigation needs to be made, also considering the reliability of the DDD dashboard. A comparison with data from the University of the Free State would assist in understanding the data a bit more. Natural Science Scores Figure 24: Natu ral Science Pass Rates for Project and Control Schools by Senior (Grade 7 9) and FET (Grade 10 12) Phases for 2015 2017. Source: Free State Data Driven Districts Dashboard 2017.
44 Figure 24 shows FET and Senior phase physical and natural science pass per centages for OMEFP and control schools for quarters 2 and 4 between 2015 and 2017. Again, the cohorts may be different and thus may partly explain the variability in scores. An increase is seen in the OMEFP FET phases betwee n quarter 2 of 2016 and quarter 2 of 2017. A decrease is observed in quarter 4 of 2016 whereas in the control schools there is a steady increase from quarter 2 (2016) to quarter 2 (2017). Once again, the senior phase control schools fare better than projec t schools even though the scores do not seem high and the increases may not be significant. As mentioned for the mathematics scores more investigation needs to be made because results do not support the notion that the OMEFP is making a tangible improve ment in natural science. Also, it is important to know what changes or increases are considered significant by the OMEFP and partners.
45 Chapter 5 Discussion and Recommendations Learner Achievement From the evidence provided in this study, we can deduce that overall the OMEFP has contributed to learner outcomes with slight improvements in senior phase though with surprising decreases in the FET phase in math and science The inconsistency in scores could be a result of several fact ors; possibly changes in the cohort of students, changes in subject teachers, expected variances each year or changes in policies. There needs to be further investigation and comparison conducted using data from the DDD dashboard and data produced by the U niversity of the Free State to clearly ascertain the reasons for the stark differences/variation. DDD Tracker According to the tracker/survey results, school governance and management in most schools is efficient however, since the survey was a baseline s tudy, future studies must be carried out to further monitor how the project schools are performing in the specified areas and to compare results and further engage with the data to present insightful results. um development and mentorship would seem to have positively contributed to equipping educators for improved instruction in the classroom. This was also confirmed by mentors in an informal discussion. It would have been even more insightful to use some form of qualitative method like focus groups or interviews with educators that receive the mentorship, to engage them on their views on issues mentioned in the survey and how the program has contributed to their teaching skills, curriculum delivery and learner achievement.
46 Sustainability and ascertaining the impacts of the OMEFP. It also hinges on beneficiaries seeing the benefit of the program and actively taking own ership of it by effectively incorporating the changes in daily instruction and school management so that external interventions are no longer regularly needed. Mentors, in their daily interaction with educators must aim to permanently transfer the skills a nd resources they are providing to the educators to facilitate capacity building that will allow the beneficiaries to independently utilize the skills. Contributing Variables The OMEFP focusses on school governance and curriculum development only, there a re other factors that influence learner outcomes, that must be investigated. Adopting a holistic approach to delivering interventions is critical to the improvement of learner achievement. The focus areas of the OMEFP areas are key in providing quality edu cation however, there may be other expansion of focus in collaboration with service partners is suggested. Below are some focus areas and contributing factors tha t could be incorporated in the expansion. Extra Lessons Learners may require extra lessons to better grasp the concepts and to have more time to remark that som e educators were providing extra time to their students and that they believed improvements could be attributed to their dedication.
47 Learner Socio Economic Status Socio economic factors of learners therefore must be considered and incorporated in the impl ementation of the program and in the data collection. In many rural areas socio economic challenges contribute to poor learner performance. These factors must not be ignored, and interventions that support means of alleviating socio economic challenges mus t be implemented to rule out critical factors contributing to poor achievement. In a meta analysis conducted between 1990 and 2000 in schools around North America, results presented a medium to strong relation between learner achievement and socio economic status (SES) (Sirin, 2005). Components of SES that were discussed, in addition to parent income, were: parental education levels, occupation, and availability of resources at home. Location of schools was also identified as closely relating to SES with no table differences according to where schools were located. This component of SES is seen in how the quintile system in South Africa tries to mitigate SES challenges in schools. Learner Access to Social Capital The issue of social capital is a key factor t o consider particularly because the area in which the schools are located is classified as a previously disadvantaged area with a history of segregation. th an ingredient in the success of a collective working towards any specific goal for the common good. In an area like Motheo district where communities experience high inequality, unemployment, poverty and where racial and religious boundaries still thrive, the possibility of social capital formation is in question. Bradshaw (2009) list some of the challenges that affect
48 the formation of social capital as: lack of service delivery, fault lines of race, polarized opinions on salient issues, lack of interracial contact, land reform crime and emigration. How does the lack of social capital affect education and learner achievement? Since the formation of social capital depends of the presence of certain resources and systems that lack where project schools are situated, this negatively affects any possibility of learners leveraging the presence of social capital in their learning journey. Israel Beaulieu and Hartless (200 1 ) explores the issue of social capital in educational achievement and how learner achievement is little investment and support in improving le arner achievement in a social capital poor community/society, the cycle of an ailing and weak social capital is reinvented further negatively affecting growth and progress in terms of economic benefits and well being. Learners who go to schools in areas th at have a strong social capital have proven to achieve better. There is an apparent need to investigate whether lack/absence of social capital a s a factor in Motheo, to further focus on key factors that directly link to poor learner achievement being expe rienced; whether leaners experience shared values, good will, social interaction across hierarchies, sympathy and mutual understanding in their communities and families as these structures contribute to their learning journey. Resources In a discussion wit h a subject mentor it was mentioned that some of the challenges that contribute to the low scores relate to the unavailability of apparatus and supplies for experiments and that at times these would be borrowed from neighboring schools (as some had even ex pired). The results from the survey show each school scoring differently in this area, some requiring much support and others scoring favorably. Educators in most of the schools scored lower,
49 supporting the hypothesis that there is a need for resources. Wh at is of interest is the stark difference in educator and HOD scores with the former scoring notably higher. This contradiction may be a contributing cause for the unavailability of resources. This section requires more research to ascertain the reason be hind the scores and whether schools are resourced or not; proposing solution where necessary. Schools in need of much support may explore other avenues to supplement the lack of resources, i.e. they can devise fundraising strategies and seek individual don ations from neighboring businesses and community members. Hannover Research in their 2013 report on Alternative Revenue Generation for School Districts, discuss fundraising strategies and revenue alternatives schools can explore to support their programs a nd supplement lack of resources. One of the examples cited is a program in Broward County in Florida, United States where schools use existing resources that are underutilized like kitchen facilities to provide catering services or lease the resources for a fee. Another example They may also foster relationships with local tertiary institution science departments and lobby equipped laboratories. Bias in Educator Learner Allocation Another possible contributing factor to consider is bias in the allocation of effective educators to as must be considered as it shapes the differences in Hammond (2000) discusses this apparent bias and the use of the Tennessee Va lue Added Assessment System, which presents differential teacher effectiveness as a strong determinant of student learning.
50 Parental Involvement In the tracker results for parental involvement, four out of five schools scored below three. Low parental o r community involvement is a challenge the schools must consider. Having the support and involvement of parents encourages collaboration between educators and parents and educators and learners (Topor et al. 2010). Topor further supports this statement by asserting that teacher parent relationships strengthen learner teacher relationships, which influence learning and ultimately performance of learners. Therefore, schools must create platforms where all parents and the community are encouraged to become reg ularly involved to foster unity in improving learner achievements and finding solutions to challenges the schools face. The question of why communities are not involved in schools, if schools create a platform for community/parent involvement is also impor tant. A probable reason for non involvement would have little or no education. With 5% recorded in the Free State as having no education at all and 17.2% with on ly primary school education (Statssa Quarterly Labour Force Survey: Quarter 1, 2017), Motheo and Botshabelo being in the rural areas fall within these statistics. This may then cause parents to shy away from being in the school environment as it makes the m feel inadequate; being in a school to participate or address issues may create a certain level of discomfort. Kohl, et, al. (2000) support this assertion. Also, schools may not possess a welcoming environment for parents, as a study conducted by Felix, e t, al. (2008:108) found in three schools in the Eastern Cape province of South Africa. Parents may feel belittled and not able to make valuable contributions. These are some examples of the barriers to parental involvement. Davis Kean highlights parent edu
51 Kean, 2005). A concerted effort by schools needs to be made in involving parents and being sensitive to their backgrounds. Overall, the OMEFP shows pr omising results and may be scaled to neighboring schools and other provinces experiencing low learner achievement rates. The use of the DDD and the OMEP tracker supports the move to utilizing big data as an integral part of decision making in the education system. In the case of Motheo it has allowed stakeholders to understand where they are in their objectives of improving learner achievement and delivering quality education. The use of data provides a roadmap that guides stakeholders, directs them to gaps that exist and highlights present successes being achieved. It is a tool for encouraging present efforts and supporting decisions for the future. The contentious policy of progressing learners who fail to meet academic requirements for a pass, results in learners moving through grades without the relevant knowledge and skills that prepares them for the respective grades and beyond. This can create complacency in both educators and learners with the latter knowing they will not fail more than once and the f ormer relying on progression as a pass. The system is responsible for over populated classrooms, strain on educators receiving those progressed learners as well as delays in curriculum completion due to the slower students who are not equipped for the next level. The progression system can only be remedied through policy review however, learner centered initiatives that support the challenged learners through tutoring can improve performance. An example is the Ikamva Youth Peer to Peer Program which operat es in two provinces in South Africa, with 1400 students and tutor student ratio of 5:1. The program focuses on delivering after school tutoring programs to learners; learners bring questions they struggle with for various subjects to the group sessions, an d the tutors and students help solve the problems together. Its
52 success has seen 80 100% of its students passing grade 12, 60% of whom access and pursue tertiary education Spaull, N (2015). DDD Dashboard Results The dashboard results suggest slight incre ases in Math and Science scores, however, a need for continued interventions that focus on the improvement of math and science scores is evident, particularly in the senior phases, which have low scores and low improvement rates. The dashboard experienced periodic technical problems during the study and these may be mitigated by using superior data available from the University of the Free owns and manages data that is of better quality, separately from that produced by the DD D dashboard. There were some restrictions in accessing the data from the SPP; however, restrictions are understood as management and precautionary measures. More collaboration and access are necessary in the creation of dependable, quality data that inform focused decision making and assists in moving processes and interventions forward. Data Collection Methods Used How data is collected also informs the quality of data being produced and the instructional decisions being made (Bongiorno, D, 2011). The sur vey instrument did not allow for respondents to elaborate on their responses. Going forward a more flexible response scale that factors in participant written responses and elaboration would provide detailed information and better quality data. Cross Sca le and Cross Disciplinary Implications: Sustainable Development Goals (SDGs) The provision of quality education is linked to the advancement of most SDGs. Providing quality education is linked to increases in opportunities for skilled labor which contribut es to poverty
53 reduction (SDG1). It is also linked to making informed decisions about health and nutrition which contributes to the reduction of vulnerability to health risks such as HIV, TB, Malaria, Diabetes etc. Populations that are educated are a ware of health risks and can use accessible information to make the necessary changes; education is also effective in the reduction of hunger (SDG 2 and 3) through having more employment opportunities and access to information for skills such as farming an d other vocational skills. Education is a significant driver of economic progress and allows those who have access and take advantage of it, to acquire decent, better paying jobs making them contributors to the economy (SDG8) as consumers and creators of b usinesses/industry. It is also effective in reducing gender and income inequality (SDG 5 &10). There has been a global movement to educate girl s based on how women drive economies and how much they sustain households even without education. Affo rding them opportunities allows them to further contribute to economic sustainability and household stability. Education is also instrumental in creating an innovative society that creatively finds solutions to infrastructure deficiencies and support growt h in the different economic industries (SDG9). My project focused on assessing learner achievement and providing a baseline analysis of the management and governance of schools as both these are contributing factors in realizing quality education. Resear ch has found that many poor performing schools lie within communities that are poorly educated or not educated at all. These communities are found to have high unemployment rates, poor infrastructure and human services and a high rate of health issues. The lack of quality and functional systems then reinvents the wheel keeping these communities poor, uneducated and with fewer opportunities.
54 Links Between Education and the Labor Market n: Poverty Trap supports the discussion above on cross scale considerations, with graphical representation of the link existing between education, labor and/access to opportunities and how al most 80% of the population in South Africa exists between the s emi skilled and unskilled sections of the labor market. Various pieces of literature (e.g., Hanushek & Wmann 2008) allude to the notion that the quality and level of education determines job and economic opportunities individuals ac cess/receive Spaull (2015). Better quality of education and the completion of secondary school then leads to reduction in poverty, and ultimately improvement in individual well being. Improvement in the quality of education can be considered a contributor to and instrument for poverty alleviation ( Lewin 2007). In the image below (read from bottom left to right going up), the acqua color represents low quality education and how it leads to individuals going into low productivity jobs as well as vocat ional training. The lack of quality and completion of basic level education creates the green tier in the model, of unskilled and semi skilled individuals. Only 15% who have access to higher quality education become skilled professionals with high producti vity jobs and higher incomes. considerable number of the population living in disadvantaged areas. This affect social capital and access to quality education for t he majority. The improvement of education is a critical challenge South Africa faces and as mentioned in our discussion there are many factors that need to be addressed. There needs to be a holistic approach that aims to address all factors that directly affect the provision of quality education and thus learner achievement. The OMEFP focusses on just one aspect and working together
55 with other stakeholders in other disciplines with better support the initiative and yield noteworthy results. Fi gure 25: Links Between Education and the Labor Market Source: N Spaull (2015) Schooling in South Africa: How Low Quality Education Becomes a Poverty Trap.
56 Conclusion The Old Mutual Education Flagship Program has positively contributed to capacity bu ilding and skills transfer in the nine project schools in Motheo. In an informal discussion with the mentors they noted that they believed their work was welcomed and appreciated by educators and school management teams even outside the sessions conducted in the schools. The resources were incorporated into daily instruction by educators (according to the mentors), and Whatsapp groups were created for further communication and support outside of the mentorship sessions. Regarding the use of the DDD tracker in collecting data on school management and governance issues, the OMEFP has created a mechanism and platform for the schools to reflect and revisit the state of governance and management. Data collected is key in unlocking some of the challenges affectin g learner achievement. Making the data available is a crucial step in addressing the factors that negatively impact the running of the school, which directly affects learner achievement. The use of big data to inform decision making is the central idea in the study, and the OMEFP tracker with the information it presents will allow stakeholders to engage on ways to improve and strengthen systems for the attainment of quality education and the improvement of learner outcome. Though the study is a baseline exe rcise, the data can now be used to further engage on the apparent challenges and successes the schools are enjoying, as well as using the findings to address challenges experienced in learner achievement.
57 References Brunner, C. (2009). Linking Data a nd Learning: The Grow Network Study. Journal of Education for Students Placed at Risk 241 267. Bongiorno, D. (2011). Using Student Achievement Data to Support Instructional Deci s i on Making. Retrieved from National Association of Elementary School Principals : http://www.naesp.org/sites/default/files/Student%20Achievement_blue.pdf Data Driven Districts Dashboard (2017). Data Driven Districts Dashboard. Retrieved from www.dbedashboard.co.za: https://www.dbedashboard.co.za/Account/LogOn?ReturnUrl=%2f Dell Founda tion, (2017, August 8). Data Driven Districts Retrieved from www.msdf.org: https://www.msdf.org/initiatives/data driven districts/ Dell Foundation, (2016). Success by Numbers Report. Retrieved from www.msdf.org: https://www.msdf.org/other reports/success by numbers/ Digital Classroom, (2014). Motheo District Digital Classr o om. Retrieved from www.digitalclassroom.co.za: http://digitalclassroom.co.za/digitalclassroom/latestnews/20 news free state/71 free state motheo district bloemfontein Eastern Cape Departm ent of Education, (2017). Scholar Information. Retrieved from www.ecdoe.gov.za: http://www.ecdoe.gov.za/scholar information/schools
58 Felix N., Dornbrack J., & Scheckle, E. (2008). Parents, homework and socio economic class: discourse of deficit and disadv Practice and Critique, 7(2), 99 112. Gandomi, H. (2015). Beyond the Hype: Big Data Concepts, Mthods and Analytics. International Journal of Information Management 137 144. Kohl, G. O., Lengua, L. J., & M cMahon, R. J. (2000). Parent involvement in school conceptualizing multiple dimensions and their relations with family and demographic risk factors. Journal of School Psychology, 38(6), 501 523. Hanushek E & Wmann L (2008) The role of cognitive skills in economic development. Journal of Economic Literature, 46: 607 668. Social Capital on Educational Achievement. Rural Sociology 66: 43 68. Lewin KM (2007) Transiti ons and Equity. Improving Access, Equity and Transitions in Education: Creating a Research Agenda. Project report. Falmer, UK: Consortium for Research on Educational Access, Transitions and Equity (CREATE). P. 2. McKeerver, M. (2016). Educational Inequalit y in Apartheid South Africa. American Behavioral Scientist 114 131. Mourshed, M. (2010). How the Worlds Most Improved Schools Systems Keep Getting Better. France: McKinsey. Retrieved from www.mckinseysociety.com: http://mckinseyonsociety.com/downloads/rep orts/Education/How the Worlds Most Improved School Systems Keep Getting Better_Download version_Final.pdf
59 Murray, M (2016). Does Poor Quality Schooling and/or Teacher Quality Hurt Black South African Students Enrolling for a Degree at the University of Kwa Zulu Natal? PLOS ONE New Leaders Foundation (2017). New Leaders Foundation. Retrieved from www.newleaders.co.za: http://www.newleaders.co.za/ New Schools, (2007). Achieving with Data: How High Performing Schools Use Data to Improve Instruction for Element ary Students. Retrieved from www.newschools.org: http://www.newschools.org/files/AchievingWithData.pd Old Mutual Foundation, (2017). Do Great Things. Retrieved from www.dogreatthings.co.za: http://dogreatthings.co.za/foundation/about old mutual education f lagship project/ SA SAMS, (2017). The South African Schools Management System. Retrieved from www.samsa.co.za: http://sasams.co.za Schifter, C.C. (2014). Data Driven Decision Making: Facilitating Teacher Use of Student Data to Inform Student Instruction. Contemporary Issues in Technology and Teacher Education Sirin, S. R. (2005). Socio economic Status and Academic Achievement: A Meta Analytic Review of Research. Review of Education Research 417 453. Spaull N (2015) Schooling in South Africa: How low qual ity education becomes a poverty trap. In: De Lannoy A, Swartz S, Lake L, et al. (eds) South African Child Gauge 2015, Cape Statssa, (2017). Annual Report 2016/17. Pretoria: Statistics South Africa. Statssa (20 17). Quarterly Labour Force Survey: Quarter 1 2017. Pretoria: Statistics South Africa.
60 Statssa (2017, August 1). The South African Economy Shrinks by 0.7% Retrieved from www.statssa.gov.za: http://www.statssa.gov.za/?p=9989 Statssa (2017). South Africa: E ducation. Retrieved from www.unicef.org: Statssa. (2016). www.statssa.gov.za. Retrieved from Statistics South Africa: http://www.statssa.gov.za/publications/P0211/P02111stQuarter2016.pdf The Department of Basic Education, S. A. (2015). Guidelines for the I mplementation of Promotion and Progression Requirements for Grades 10 12. Pretoria: The Department of Basic Education, South Africa. Retrieved from Association of Christian Schools International. The Department of Basic Education (2017). Districts. Retriev ed from www.dbe.gov.za: http://www.education.gov.za/Informationfor/EducationDistricts.aspx The Department of Basic Education, (2013). Education Statistics. Retrieved from www.education.gov.za: http://www.education.gov.za/Portals/0/Documents/Publications/Ed ucation%20Statistic% 202013.pdf The National Association of Elementary School Principals, (2011). Success by Numbers. Retrieved from www.naesp.org: http://www.naesp.org/sites/default/files/Student%20Achievement_blue.pdf Topor, David R., Susan P. Keaner, Te Journal of Prevention & Intervention in the Community 38 (3): 183 197. UNESCO. (2016). Unpacking Sustainable Development Goa l 4: Education 2030. UNESCO.
61 United Nations (2015, December 30). United Nations Sustainable Development Goals Retrieved from www.un.org: www.un.org/sustainabledevelopment/ UNICEF, (2017). Overview: Education and Adolescent Development, South Africa. http s://www.unicef.org/southafrica/education_344.html University of the Free State, (2017). The Schools Partnership Project Report. Bloemfontein: University of the Free State. World Bank (2003, September 22). Br M i nas Gerais Basic Education. Retrieved from Th e World Bank Group: http://lnweb90.worldbank.org/oed/oeddoclib.nsf/24cc3bb1f94ae11c85256808006a0046/8 525682e0068603785256cd700528196?OpenDocument World Bank (1994, April 5). Brazil State of Minas Gerais : Basic Education Quality Improvement (Pro Qualidade ) Project Retrieved from www.worldbank.org: http://documents.worldbank.org/curated/en/612311468231287584/pdf/multi0page.pdf Wyk, C. V. (2015). An overview of Education data in South Africa: An Inventory Approach Cape Town: University of Stellenbosch.
62 APPENDIX 1. Field Practicum Gantt Chart
65 4. Sample of Heads of Department Survey
1 GRANT OF PERMISSIONS In reference to the following title(s): Zotha Zungu. D a t a K n o w B e s t : Tracking Learner Achievement and Assessing the Impact of the Old Mutual Flagship Program in Nine Project Schools in Motheo District, Free State Province, South Africa. Gainesville, FL: University of Florida, December 2017. I, __Zotha Zungu__, as copyright holder or licensee with the authority to grant copyright permissions for the afore mentioned title(s), hereby authorize the University of Florida, acting on behalf of t he Board of Trustees of the University of Florida, to digitize, distribute, and archive the title(s) for nonprofit, educational purposes via the Internet or successive technologies. This is a non exclusive grant of permissions for on line and off line use for an indefinite term. Off line uses shall be consistent either, for educational uses, with the terms of U.S. copyright legislation's "fair use" provisions or, by the University of Florida, with the maintenance and preservation of an archival copy. Digit ization allows the University of Florida to generate image and text based versions as appropriate and to provide and enhance access using search software. This grant of permissions prohibits use of the digitized versions for commercial use or profit. __ _______________ Signature of Copyright Holder __Zotha Zungu___________________________ Printed or Typed Name of Copyright Holder _12/07/2017______________ Date of Signature Attention: Digital Services / Digital Library Center Smathers Libraries Univers ity of Florida P.O. Box 117003 Gainesville, FL32611 7003 P: 352.273.2900 DLC@uflib.ufl.edu