Approaches to quantitative analysis on student performance Dr Diego Bunge and Dr Daniel Uribe
Contents Institutional student data Issues with key variables Ethnicity Socioeconomic background Entry scores Cohort vs year by year Cohort analysis and final outcome Logistic regressions Profiles
Institutional student data UCAS data Data at Programme level DLHE data LEO data Data at Module level Bursaries and Household Income data SITS
Issues with key variables Ethnicity Socioeconomic background Entry scores
Ethnicity Ethnicity Ethnic categories cultural heritage / identity social or cultural characteristics historical experience White Asian Black Arab Chinese Mixed Other BME
Socioeconomic background Higher managerial & professional occupations Lower managerial & professional occupations Intermediate occupations Small employers & own account workers Lower supervisory & technical occupations Semi-routine occupations Routine occupations No answer Parental occupation (NS-SEC) Higher education Not higher education Information refused or no answer Parental education Zero income Up to £13,500 £13,501 to £22,500 £22,501 to £31,000 £31,001 to £44,106 Over £44,106 Non means tested-refused Not reported Bursary recipients Household income
Entry scores Tariff points (UCAS) Best 3 A levels Other Main issue spread scores Best 3 A levels Main problem exclusion of cases Other e.g. what grades do we accept students on?
degree classification Cohort vs year by year Year by year analysis 2014/15 2013/14 Institutions -e.g. the Higher Education Statistics Agency (HESA) Academic studies e.g. McNabb et al, 2002; Connor et al, 2004; Naylor and Smith, 2004; Broecke & Nicholls, 2007 Looks at degree classification
Cohort vs year by year Looks at final outcome Cohort analysis 2015/16 2012/13 entrants 2015/16 3-4 years Looks at final outcome
Cohort vs year by year Year by year analysis Cohort analysis
Year by year and progression Progression of new entrants 2012/13 entrants 2013/14 entrants Compare progression rates Progressed Didn’t progress Withdrew 2013/14 2014/15
Analysis of 2011/12 cohort Data source: SITS In 2011/12, 1,674 students started a programme at Faculty X. After 3 to 4 years, 1,393 (83%) obtained a degree: 262 achieved a 1st and 845 a 2.1; 199 scored 2.2, 45 a 3rd or a pass and 42 other qualifications. On the other hand, 204 withdrew or failed and 77 remained with chances to complete a degree. Data source: SITS Note: Total numbers do not include students who changed school or those who still have chances to obtain a degree.
Logistic regression results Factors increasing the odds of Higher entry scores, state-funded school, living at parental home High entry scores, being female, having attended state funded schools Zero income reported, Clearing, living at parental home Data source: SITS Note: Total numbers do not include students who changed school or those who still have chances to obtain a degree.
Logistic regression results Factors decreasing the odds of BME, Zero income reported Zero Income reported, Clearing Higher entry scores, being female, state-funded school Data source: SITS Note: Total numbers do not include students who changed school or those who still have chances to obtain a degree.
Profiles H&SS S&E
Diego Bunge d.bunge@qmul.ac.uk and Daniel Uribe d.uribe@qmul.ac.uk Thank you Diego Bunge d.bunge@qmul.ac.uk and Daniel Uribe d.uribe@qmul.ac.uk