The principal and teachers of a small independent South African school noticed a significant gap in the implementation of an effective Intermediate Phase (IP) reading curriculum. The principal decided to establish a Community of Inquiry (CoI) at the school to focus on this problem, and to research the process. The main purpose of this research was to investigate the affordances and constraints of the establishment of a CoI for the professional development of teachers as an instructional leadership intervention by the principal. The sub-questions that emerged from the main research question were: what would be the affordances and constraints of the principal establishing, facilitating and researching the CoI; what would be to the benefit of the teachers (and their learners) of establishing such a community for professional development and reading instruction in the IP; and, what processes are entailed in establishing an in-school CoI? To what degree, if any, would the CoI be a generating space to answer the research questions and aims? The literature review for the study explored the means of professional development available to educators and principals, especially Professional Learning Communities and their more focused interventions – Communities of Inquiry, and that professional development needs to target four levels – the teacher, their teaching, the community and collegiality at the school, and the principal as a developmental leader, in order for there to be an improvement in students’ learning. The concerning South African context with regard to reading in the IP was outlined, as well as that in the school. The South African curriculum IP learning outcomes and assessment standards for reading were critically explored against the backdrop of international curricula. The process of learning to read was traced. Hindrances to effective reading were also explored. The qualitative research design was an applied case study. Grounded Theory methods were used to reduce the data from the transcripts of CoI sessions. The findings of the research were that there are many affordances to establishing an in-school CoI, and that these benefits far outweigh the constraints. The CoI provided a germinating locus in which participants could begin to address the problems related to reading and reading instruction in the IP. It enabled the principal to develop as an instructional leader, and the teachers to develop as professionals and reading instructors. In hindsight, this project was a vital one, but considerably ambitious, difficult to implement, and perhaps even constrained in the sense of the principal establishing, facilitating and researching this process herself. However, without the principal driving, facilitating and researching the CoI, the question emerges whether such an intervention would have been established. Sometimes external insistences from authority can provide valuable impetus for change at schools, as long as these are sensitively handled and one has the support and trust of participants. The actual acceleration in learning that the participants experienced was extensive, and, sometimes it is only the principal that can play this role – especially in newly established communities. Much more research and support for principals in becoming instructional leaders and in establishing CoIs at their schools is required in the South African context.

Our research deals with the problem of devising models for fitting non- identical dependent Bernoulli variables and using these models to predict fu- ture Bernoulli trials.We focus on modelling and predicting random Bernoulli response variables which meet all of the following conditions: 1. Each observed as well as future response corresponds to a Bernoulli trial 2. The trials are non-identical, having possibly different probabilities of occurrence 3. The trials are mutually correlated, with an underlying complex trial cluster correlation structure. Also allowing for the possible partitioning of trials within clusters into groups. Within cluster - group level correlation is reflected in the correlation structure. 4. The probability of occurrence and correlation structure for both ob- served and future trials can depend on a set of observed covariates. A number of proposed approaches meeting some of the above conditions are present in the current literature. Our research expands on existing statistical and machine learning methods. We propose three extensions to existing models that make use of the above conditions. Each proposed method brings specific advantages for dealing with correlated binary data. The proposed models allow for within cluster trial grouping to be reflected in the correlation structure. We partition sets of trials into groups either explicitly estimated or implicitly inferred. Explicit groups arise from the determination of common covariates; inferred groups arise via imposing mixture models. The main motivation of our research is in modelling and further understanding the potential of introducing binary trial group level correlations. In a number of applications, it can be beneficial to use models that allow for these types of trial groupings, both for improved predictions and better understanding of behavior of trials. The first model extension builds on the Multivariate Probit model. This model makes use of covariates and other information from former trials to determine explicit trial groupings and predict the occurrence of future trials. We call this the Explicit Groups model. The second model extension uses mixtures of univariate Probit models. This model predicts the occurrence of current trials using estimators of pa- rameters supporting mixture models for the observed trials. We call this the Inferred Groups model. Our third methods extends on a gradient descent based boosting algorithm which allows for correlation of binary outcomes called WL2Boost. We refer to our extension of this algorithm as GWL2Boost. Bernoulli trials are divided into observed and future trials; with all trials having associated known covariate information. We apply our methodology to the problem of predicting the set and total number of passengers who will not show up on commercial flights using covariate information and past passenger data. The models and algorithms are evaluated with regards to their capac- ity to predict future Bernoulli responses. We compare the models proposed against a set of competing existing models and algorithms using available air- line passenger no-show data. We show that our proposed algorithm extension GWL2Boost outperforms top existing algorithms and models that assume in- dependence of binary outcomes in various prediction metrics. / Statistics