AQMeN - Applied Quantitative Methods Network

Registration for this event has now closed.  Candidates will be notified whether or not they have a place no later than 25th February.

Presented by Tony Coxon (University of Cardiff & University of Edinburgh) and Phil Hawkins (SERAC, Cardiff)

The course will take place at the University of Edinburgh.

Course Content:
This course will explore the use of multidimensional scaling and its practical application in the social sciences.  The main focus will be on the basic distance model (aka "smallest space analysis") and on the interpretation of configurations. It will then generalize to deal with two-mode models (Unfolding, Vector, Correspondence Analysis) and then to three-way models including Individual Differences Scaling. 

Analyses will be made using SPSS,  NewMDSX (a not-for-profit package devoted to a wide range of MDS programs) and PERMAP (an interactive free program).

Lectures will be combined with practical sessions in order to provide hands-on experience. A "data surgery" will be offered on days 2 and 3, for pre-arranged individual consultation.

Click here to view a draft programme

Who should attend?
This course is intended for people working in social sciences who wish to use geometrical data analysis to obtain visual/graphical representation of their data and gain insight into the data's underlying structure and meaning.

Why attend?
MDS is a much under-used resource, primarily used for teh descriptive analysis of many kinds of data.  It is a family of models, each one dealing with a particular sort of data, a given level of measurement (data-transformation) and a specific type of model.  MDS has affinities with Factor Analysis, Principal Components Analysis, Clustering and Conjoint/ANOVA Analysis, but generally provides more wide-ranging options and is subject to less demnding requirements.  Potential advantages include:

• Use of a wide range of levels of measurement data that can be analysed using different quantitative assumptions about data. In particular, MDS routinely uses both ordinal (“non-metric”) and linear (“metric”) assumptions. 
• A choice between distance models and vector (scalar product/factor) models, each involving very different assumptions about the composition of underlying variables. Although distance models are usually Euclidean (and therefore free of fixed dimensions); other variants do allow for fixed dimensions.
• Perhaps most importantly, MDS has been generalized to deal specifically with a wide range of direct data (e.g. pair-comparisons, profiles, ratings, sortings, triads ). But in addition MDS accepts a very wide range of measures of association, correlation etc, and MDS programs are rarely tied to any particular measure.
• Although data input usually refers to aggregate data, other programs allow for the analysis of sets of individual data, such as Likert ratings or pair-comparisons and represent the individuals separately in the model and analysis.

Course objectives

• To introduce participants to the use of multidimensional scaling.
• To understand the rationale for, and form of, the main MDS models.
• To choose appropriate programs for the analysis of their data, using appropriate software (NewMDSX, SPSS and PERMAP), and to provide hands-on experience of their use.
• To introduce techniques used to interpret the results of MDS analyses.

Learning Outcomes

• Participants will be able to understand the principles and assumptions underlying MDS models for analysing different types of data, in different formats and different levels of aggregation, and recognise the conditions under which the use of each model or program is appropriate.
• Participants will know how to recognise when a program has reached an optimal solution, what may legitimately be done to the solution configuration and what differences are significant rather than a consequence of model assumptions.
• Participants will be able to compare and contrast similarities and differences in solutions from different  programs.
• Participants will understand the principles of  good practice and be able to use them in reporting their results. 

Pre-requisite criteria
The course is pitched at an intermediate level. Participants should be familiar with basic statistical methods, including exploring data numerically and graphically. A basic knowledge and understanding of descriptive and inferential statistics, and of simple multivariate procedures (e.g. Principal Components; Multiple Regression) will be assumed. All participants must be competent in the use of a PC and have some experience of SPSS.  Participants need to indicate what they wish to achieve from the course, and how they plan to use the techniques that they learn. 

Application Process
There are 15 places available on this course.  Places will be allocated following a process of application and priority will be given to students and academic staff in Scotland who meet the selection criteria, and for which the course is free to attend.

Non-academics and academics from outside Scotland may apply and will be placed on a waiting list to be allocated any remaining places.  There will be a charge for non-academics and academics outside Scotland as follows:

• Academic students and staff in Scotland - FREE
• Students and Academic staff outside Scotland - £150 (£50 per day)
• Public and Voluntary Sector - £225 (£75 per day)
• Private sector  £300 - (£100 per day)

To apply for the course:

1. Sign up for the event using the link below AND
2. Complete the application form and return to events@aqmen.ac.uk by 5pm on Wednesday 16th February.

Download the application form

Successful applicants will be notified by 5pm on Friday 25th February.

PhD Expenses
PhD students studying in Scotland are eligible to claim for travel, accomodation and subsistence during the event.  Please note that any claims must be discussed with Heather Thomson or Angie Dickson prior to the event.  failure to do this may result in us being unable to reimburse you.  AQMeN will only reimburse travel to/ from an address in Scotland as provided on the application form.    Accommodation will be arranged by AQMeN on a bed and breakfast basis.  Evening meals may be claimed up to the value of £15 excluding alcohol.  Original travel tickets and fully itemised receipts MUST be provided.  Credit card receipts and Email booking confirmations will not be accepted by our auditors.