Patricia Chirwa

Patricia Chirwa is a research and data scientist who has consulted for the Commission on Restitution of Land Rights (CRLR) and the South African Labour and Development Research Unit (SALDRU). She has principally worked on the South African Land Restitution Evaluation Study (LRES), developing skills in three main areas: 1) social impact design and implementation, 2) database solution design and management, and 3) data science.

Patricia holds a masters degree in economics from the University of Cape Town and is currently working towards her PhD in economics. Her proposed study applies social network analysis and machine learning techniques to assess the degree of fragmentation and conflict risk and cooperation failure in communities that have received land through South Africa's land restitution programme. This work employs proprietary data captured using a relational database system designed and developed by Patricia as a solution to the CRLR’s large-scale administrative data challenges.

11/2019-12/2022

South African Labour and Development Research Unit

Project: Land Restitution Evaluation Study 

Role: Research and Data Scientist/ Junior Project Lead

Responsibilities: • Supported the lead in the design and implementation of an impact evaluation study on the South African land restitution programme • Constructed quantitative and qualitative models and frameworks around financial compensation and land restoration claims • Supported the development and programming of multi-system survey tools to collect social network, behavioural, and cognitive game task data • Developed training guides and facilitated field team training sessions • Lead the drafting and preparation of research reports and presentations   •  Performed preprocessing (ETL) on administrative and survey data (including spatial data) •  Applied cluster and diffusion algorithms and unstructured learning tools with python (including packages such as pandas, numpy, scikit-learn, etc.) and visualisation and reporting (including google data studio and jupyter notebook). • Provided data analysis and visualisation inputs for steering committee and stakeholders on project progress. • Provided technical and strategic advice on land restitution business processes and implementation.  

Skills:   Impact Assessment, Monitoring and Evaluation, Econometrics, Economic Modeling, Cluster Analysis, Regression Analysis, Behavioural Economics, Project Management, Economic Research, Data Science, Solution Design, Communication, Reporting and Presentation.

Results: Developed a network analysis model to predict land restitution success or failure by assessing the degree of historical fragmentation and cooperation in communities restored through land. While preliminary results are promising, I am currently investigating the use of machine learning algorithms like neural network models and random forest models to improve the model fit.

01/2023-