Research

Commissioned by Kent County Council, nurtureuk aimed to support the inclusion of pupils with special educational needs and disabilities (SEND) across 300 schools through a whole-school nurturing approach. Working closely with the client, I clarified the research objectives and identified key metrics, including school-wide attendance and exclusion rates, SEND pupil well-being and engagement, and staff feedback on behavioural impacts. I translated these objectives into specific research questions and developed a research proposal, adjusting the design to accommodate the staggered start times of participating schools. Although a controlled trial was not feasible, we used regression models to explore the association between programme participation length and key outcomes. I designed three surveys to gather data to collect data for each objective, and oversaw the design and implementation of qualitative focus groups. I then designed regression models to identify the presence of any associations, and contributed to the writing of the subsequent research reports.

Using the Theory-U framework, I co-designed participatory systems-change workshops to address inequalities in healthcare access and outcomes. These workshops brought together stakeholders including health service commissioners, local government officials, healthcare professionals, community members, and representatives from voluntary and community sectors. I facilitated dialogues that encouraged stakeholders to understand each other’s perspectives, fostering collective reflection and re-evaluation. I guided discussions, helping participants to clarify their objectives and led them through a prototyping process, where they collaboratively shaped and tested new policies and practical solutions. Afterwards, I conducted thematic analysis of the workshop outputs, synthesizing insights on both the specific prototypes and the overall systems-change process. To expand on these findings, I carried out semi-structured interviews with leaders of best practice programmes across the country tackling healthcare inequalities, producing video and written case studies to illustrate key themes. I independently completed the first draft of the national inquiry report, which was subsequently edited and published by London South Bank University.

In my dissertation, I explored methods to assess the readiness and ability of healthcare professionals treating individuals with Myasthenia Gravis to engage in shared decision making. I conducted a Theoretical Domains Framework survey across multiple countries, collecting data from healthcare professionals and patients. I analysed this data using regressions in R to identify the factors influencing professionals’ readiness to participate in shared decision making, as well as the barriers and facilitators involved. Based on my findings, I proposed behavioural change techniques to improve healthcare professionals’ engagement in shared decision making.

Feedback noted how my dissertation stood out for its theoretical grounding, methodological rigor, and practical relevance. With clear objectives, a comprehensive methodology, concise results, and transparent acknowledgment of limitations, markers noted how my paper showcased my reflexivity and offered valuable insights.

In response to an open brief, I developed a nudge aimed at enhancing healthcare access for young people, along with a comprehensive evaluation strategy to assess its effectiveness. I began by dissecting the issue through the Theoretical Domains Framework, which informed my intervention design using the MINDSPACE framework. To address potential risks, I utilized the APEASE criteria for analysis and outlined a quasi-experimental design for evaluating the nudge.

I presented my findings through a recorded poster presentation and submitted a detailed written proposal. Feedback highlighted my thorough understanding of the issue, the clarity of my evaluation strategy, and the logical flow of my writing. Both my poster and proposal were commended for their effective presentation.

In the first project, I designed a classification solution for an object recognition task using classical machine learning methods, including k-Nearest Neighbors and various classifiers, while employing techniques such as cross-validation, dimensionality reduction, and performance evaluation metrics.

In the second project, I focused on predicting cellular composition in histological images, extracting relevant features, and applying regression models like Ordinary Least Squares and Support Vector Regression. I also explored Convolutional Neural Networks to predict total cell counts and individual cell types, analyzing model performance through various metrics and visualizations.

Through both projects, I demonstrated proficiency in data analysis, feature extraction, model optimization, and effective communication of results.

In my proposal “Shame, Need, and Desert: Why People Might Not Accept Food Bank Charity,” I created an experiment to investigate psychological barriers that deter food bank recipients from seeking help. Focusing on shame and perceptions of need and desert, I designed a 3×2 randomized controlled trial to analyze how these factors affect voucher redemption rates.

Feedback highlighted my proposal as excellent, with an exhaustive reference list, a novel and creative idea, and a clean experimental design that effectively isolates the variables of interest. My professor specifically praised the design of the original priming questions as a notable strength, and requested to use the proposal as a best case example.

In my project, I analysed the relationship between deprivation and the prescription frequency of psychoactive drugs across GP practices in London and South East England. I devised and undertook a methodology using regression analysis and classification models. My results highlighted prescribing inequalities and suggests that antipsychotic use in deprived areas warrants further investigation to ensure appropriate treatment.

Feedback praised my detailed introduction, effective use of case studies, and thorough data pre-processing. The hypotheses were clearly stated and addressed with appropriate methods, including clear explanations of the PMCC and classification models. The analysis effectively compared multiple models using various accuracy measures. Conclusions aligned with the analysis, and limitations were well noted. The presentation was commended for its clarity and neat visualizations.

I completed a project using MATLAB to explore psychological theories and models. This included creating graphs to compare Fechner’s Law and Stevens’ Law and analysing how Range Frequency Theory (RFT) affects perceived expensiveness. I calculated the Root Mean Square Deviation (RMSD) between model predictions and data, identifying optimal parameters. I also implemented the Scale-Invariant Memory, Perception, and Learning (SIMPLE) model to study memory recall effects, examining retention intervals and temporal gaps. Additionally, I explored the Schelling segregation model, simulating social dynamics among economists and psychologists in a spatial grid, assessing neighbour tolerance and population distributions.

Feedback said my work demonstrated excellent understanding and engagement, with clean, efficient code and effective use of MATLAB features. Commentary was clear and insightful, particularly in explaining the RMSD calculation and the SIMPLE model. The structure was well-organised, and theoretical explanations added value to your analyses.

During class projects, I employed various statistical analysis methods, including regression analysis, repeated measures ANOVA, and t-tests, to explore relationships between variables and assess subjective outcomes. I used pooling techniques and applied corrections for multiple testing where appropriate. Additionally, I performed follow-up tests to evaluate significant interaction effects.

Feedback praised my ability to identify and resolve data issues, including duplicate rows and typos. My results were well-reported, and I was commended for thorough checks of assumptions related to homogeneity and normality. My figures were noted for their clarity, and my coding was well-structured and commented.

When presented with images of Vincent van Gogh’s paintings, and letters, along with data on people’s emotional responses, I outlined a methodology to explore the relationship between Vincent van Gogh’s use of colour and people’s emotional responses. I drew on existing colour psychology literature and van Gogh’s expressed intentions in his letters in creating my approach. I employed clustering models to identify dominant colors in each image and both frequentist & bayesian regression to interpret the impact of colour on emotional perception. Feedback commended my theory, detailed method, and planned interpretations.

In my project “Risky Choice and Cumulative Prospect Theory (CPT)”, I fitted three versions of CPT to analyse decision-making under risk using data from a study with 30 participants. My approach included parameter estimation, model comparison, correlation analysis, and parameter recovery through data simulation. Feedback highlighted my numerical analysis skills and clear plots.

In my project “Exploring Drivers of Patient Satisfaction When Making a GP Appointment”, I analysed factors influencing patient satisfaction using data from the General Practice Patient Survey. Key drivers identified included ease of contact, receptionist helpfulness, and appointment availability. My analysis employed exploratory data analysis and multivariate linear regression, highlighting the importance of communication and accessibility while suggesting the need for further exploration of appointment offers.

Through this module, I gained foundational knowledge in computing, including programming in Python, statistical methods, and computational mathematics. I learned to effectively use data structures, analyse algorithms and their complexities, and apply logical analysis to solve mathematical problems.