About the Project
Social Science: Human Geography, Environment and Urban Planning
User-generated data, e.g. social media and crowdsourced data, have provided researchers and decision makers with an unprecedented opportunity to monitor and understand society at much higher frequency and granularity. However, these ‘new forms of data’ may have some challenges to be considered alongside traditional data, such as randomised surveys. Their huge sample size and questionable quality (e.g. bias, dependencies, representativeness, and missingness) make inferences from traditional statistical and analytical methods questionable. Adjustments are thus needed to ensure biased and unrepresentative data do not lead to flawed conclusions. This project builds on recent work by incorporating geospatial and social components to two main challenges of using user-generated data in social science, i.e. quality-quantity balance, and missing data imputation.
- This project builds on recent work by incorporating geospatial and social components to:
- understand the effective sample size of “big data”, data with geospatial and social components,
- provide approaches to combine new forms of data with survey data,
- provide mechanisms to impute missing data (Missingness Not At Random (MNAR)) using areal-level effects and identify the functions that links the missingness and the values (e.g. low income) at an aggregated level.
By considering spatiality and connectivity, through network autocorrelation, the proposed work will allow new forms of data to be combined with survey data (Scottish Household Survey, Understanding Society, and National Travel Survey) to make valid inferences about society at a higher frequency and lower cost than has previously been possible.
This project will provide a geospatial and social science-enabled solution to the challenge of “big data paradox” and missing data which will enable wider academic disciplines, including computer science, data science and artificial intelligence, decision makers and policymakers to reliably use user-generated data, and traditional survey data to have a meaningful, realistic and and statistically valid data-driven results.
Applicants must meet the following eligibility criteria:
- 2.1 Honours degree or equivalent
For applicants whose first language is not English, the University sets a minimum English Language proficiency level.
International English Language Testing System (IELTS) Academic module (not General Training) or equivalent English language qualifications:
- 6.5 with no sub-test under 6.0.
- Tests must have been taken within 4 years 5 months of start date. Combined scores from two tests taken within 6 months of each other can be considered.
Please note that all applicants must also meet the ESRC eligibility criteria. ESRC eligibility information can be found here.
For full details and to apply for this studentship, please visit the Scottish Graduate School of Social Science (SGSSS) website here.
Applications will be ranked by a selection panel and applicants will be notified if they have been shortlisted for interview by 9th April 2021. Interviews will take place on 16th April 2021.
• An annual maintenance grant (stipend)
• Fees at the standard institutional home rate
• Students can also draw on a pooled Research Training Support Grant (RTSG)
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.
Big Data Analytics and Mining: investigating and testing distributed formulations of data mining algorithms that are suitable for the MapReduce paradigm and for other distributed computing approaches
University of Reading