2016/17 Taught Postgraduate Module Catalogue

EDUC5064M Statistical modelling in educational research

30 creditsClass Size: 30

Module manager: Dr Matt Homer
Email: m.s.homer@leeds.ac.uk

Taught: Semester 2 (Jan to Jun) View Timetable

Year running 2016/17

Pre-requisite qualifications

Some introductory statistical background (eg a basic knowledge of hypothesis testinc). Some basic knowledge of statistical software (eg SPSS or Stata).

This module is approved as an Elective

Module summary

This module provides an introduction to advanced statistical modelling approaches through the application to real educational datasets of appropriate software tools (e.g. SPSS). You will learn how to choose the appropriate statistical model based on the nature of the data you have available and the research questions you are interested in. You will discover how to modify and improve statistical models in an iterative process as the analysis proceeds. You will also consider how to evaluate both the substantive findings coming out of the analysis, but also the 'quality' of the model(s) themselves. There will also be emphasis on you being able to meaningfully interpret what the models are, and are not, telling you, and on careful and clear writing up of your findings.Whilst the focus will be very much on applications of statistical methods to real data, you will also consider the wider issues concerning quantitative research, and research more generally – how reliable is the data? What are the main issues with using secondary data in educational research? What exactly can we infer from our analysis? What assumptions have we made? How can the literature inform what we have found? What other studies have considered similar issues? What methods have these used? Are their better methods that could prove fruitful? What does a p-value tell us and what are effect sizes? And so on.The module is taught through a mixture of formal teaching, hands-on practical classes using computers, class-based discussion and presentation of work, and a range of online activities. This module will provide an excellent grounding in sophisticated statistical approaches to quantitative data analysis.

Objectives

The aims of this module are to enable students to:

- Investigate a variety of approaches to statistical modelling and to understand the different contexts where each is most appropriately applied.
- Use appropriate software to construct statistical models using secondary data
- Assess the quality of the models so constructed and develop them to improve their explanatory power
- Report the findings of such analyses in a clear non-technical manner to a range of audiences

Learning outcomes
On completion of this module, students will be able to apply appropriate software tools such as SPSS and Stata to real world datasets in order to:

- Choose appropriate statistical modelling approaches to the data in hand and the research questions being investigated.
- Evaluate the 'quality' of the model(s) using standard methods – e.g. proportions of variance explained, patterns in the residuals.
- Interpret the outcomes from such models and write up findings in a clear, non-technical way, appropriate to the educated lay-person. This will include a focus on effect sizes rather than statistical significance.
- Understand the benefits and limitations of the modelling approaches being employed, and the challenges of secondary data analysis.

Syllabus

1. Introduction to probability
The main distributions used to underpin statistical modelling: Binomial, Multinomial, Poisson and Normal Bayes theorem and the odds ratio
2. Regression models
Simple and multiple ordinary least squares
General linear models including repeated measures
Multi-level models for continuous variables
3. Models for categorical data
Logistic regression
Log linear analysis
The use of the odds ratio
4. Latent variable approaches
Exploratory factor analysis
Confirmatory factor analysis using AMOS
5. An introduction to more advanced techniques
Structural Equation modelling
Item Response Theory
Rasch models

Teaching methods

Delivery type	Number	Length hours	Student hours
Practical	12	2.00	24.00
Tutorial	6	2.00	12.00
Private study hours			264.00
Total Contact hours			36.00
Total hours (100hr per 10 credits)			300.00

Private study

In order to prepare for the sessions, students will have to read and comment on academic and non-academic articles (both substantive and methodological), explore and model appropriate datasets and write-up findings. They will also be required to read additional material about specific statistical techniques (e.g. video and other online resources) and how to carry them out in SPSS and other statistical software as appropriate.

Opportunities for Formative Feedback

In every session, students will be set particular modelling tasks using real data and SPSS and other software to answer research questions. This work will be formatively assessed by the lecturer monitoring students as they work through them. Additional tasks will be assigned for work outside the class (for exampling reading and summarising key articles on benefits and limitations of particular methods) and these will be followed up in the subsequent sessions through discussion and/or other approaches (e.g. informal presentations). Each week there will be on online quiz for students to work on out of class – this will be used as a stimulus for discussion during the following session.

Methods of assessment

Coursework

Assessment type	Notes	% of formal assessment
Essay	6000 word or equivalent	100.00
Total percentage (Assessment Coursework)		100.00

Normally resits will be assessed by the same methodology as the first attempt, unless otherwise stated

Reading list

The reading list is available from the Library website

Last updated: 23/02/2017

Disclaimer

Browse Other Catalogues

• Undergraduate module catalogue
• Taught Postgraduate module catalogue
• Undergraduate programme catalogue
• Taught Postgraduate programme catalogue

Errors, omissions, failed links etc should be notified to the Catalogue Team.PROD