Unit information: Advanced Quantitative Modelling Techniques in Education in 2013/14

Unit name	Advanced Quantitative Modelling Techniques in Education
Unit code	EDUCM5509
Credit points	10
Level of study	M/7
Teaching block(s)	Academic Year (weeks 1 - 52)
Unit director	Professor. Thomas
Open unit status	Not open
Pre-requisites	None
Co-requisites	None
School/department	School of Education
Faculty	Faculty of Social Sciences and Law

Description including Unit Aims

This unit is designed to build upon and extend the knowledge and skills developed in mandatory quantitative research methods units (Statistics in Education and Multivariate Statistical Methods in Education). In this unit students will be introduced to larger data sets and more advanced statistical modelling techniques using analytical tools (Mlwin software) for longitudinal and multivariate data, linear and multiple regression and multi-level modelling. This will be undertaken through a series of practical exercises drawn from research projects carried out within the School and from other sources (e.g.DfES data archives). Students will be encouraged to understand the complex and hierarchical nature of educational settings (e.g. pupils nested within classrooms, departments, schools and LEAs), how contrasting pictures can be derived from different manipulations of the data and the implications this has for particular types of research data gathering and analysis.

Aims:

• To provide students with a statistical understanding of the complex nature of educational data, and, more specifically, of the modelling of educational outcomes such as examination and assessment results, in relation to a variety of explanatory factors comprising for example, educational processes, inputs and context.

• To provide students with an understanding of when complex quantitative modelling methods are appropriate and how these can contribute to a more robust/powerful evidence base in educational research.

• To provide students with the knowledge and skills to apply basic multilevel modelling techniques to larger educational datasets using the Mlwin computer package and to interpret basic statistical output in relation to specific research questions.

Intended Learning Outcomes

Students will be able to:

• Understand, from a statistical viewpoint, the complex/multilevel nature of educational data and how educational outcomes can be modelled using quantitative approaches.

• Understand when complex quantitative modelling methods are appropriate.

• Demonstrate a working knowledge of basic multilevel modelling techniques (using Mlwin computer software) applied to larger educational datasets and an understanding of how to interpret basic statistical output in relation to specific research questions.

Teaching Information

Students will carry out set exercises using a prepared dataset followed by group discussion of the results. The tutor will provide explanations of the theoretical rationale underlying different multilevel models verbally and through printed material and worksheets. Documentation and datasets will be utilised, as appropriate, from ‘state of the art’ training resource websites such as http://tramss.data-archive.ac.uk/documentation/MLwiN/, http://www.mlwin.com/

The needs of a wide range of students, including those with disabilities, international students and those from ethnic minority backgrounds have been considered. It is not anticipated that the teaching and assessment methods used will cause disadvantage to any person taking the unit. The Graduate School of Education is happy to address individual support requests as necessary.

Assessment Information

Using Mlwin students will carry out a basic OLS multiple regression and/or multilevel analysis involving a dataset obtained from an appropriate data archive (e.g. ESRC, LEA, DfES) and will be asked to present their findings briefly to the class group. The analysis will be written up formally in the form of 2,000 word (maximum) report comprising a short account of the research background and research questions, and followed by a critical interpretation of the results.

Reading and References

Goldstein, H (1997) Methods in School Effectiveness Research, School Effectiveness & School Improvement, 8, (4): 369-395

Kreft, IGG and De Leeuw, J (1998) Introducing Multilevel Modelling. London: Sage.

Rasbash et al (2000) A user’s guide to Mlwin. London: Institute of Education.

Snijders T and Bosker R (1999) Multilevel Analysis. London: Sage.