Topic: Data Envelopment Analysis
Data Envelopment Analysis Coursework
“The module will be assessed 100% by an individual coursework, in which the students will analyse a case study and present their findings in a report.”
This document specifies the requirements you should follow in selecting an appropriate case study and how you should report your findings.
1 Selecting a Case Study
If you are wishing to obtain a mark of 60%+ (ie a first or 2i) you should select your own case study. However if your aspirations are lower and you are happy with a mark of 40% to 60% you are advised to consider the example case study.
If selecting your own case study, a self assessment task has been created where you can post the details about the case study you have are thinking of selecting.
The intention of this task is to allow you to receive feedback on the viability of your proposed case study and will enable me to provide advice as to issues you may need to consider and alternative sources of data.
In selecting a Case study you should consider the following:
1. Availability of suitable data
The data should be publicly available (ideally in electronic form) and ideally should have comparable information for more than one year. Data can come from multiple sources (see examples above). There will need to be data on both the resources used (inputs such as Number and/or cost of staff) and successful production (outputs for example revenue and or number of products sold)
2. Sufficient decision making units for comparisons
What constitutes a sufficient number will depend on the number of inputs and outputs in the DEA models you are using. For the purposes of the coursework you may have a small number of Decision Making Units (DMUs) as long as you consider their performance over several time periods (eg, 15-25 DMUs with 5 to10 years of annual data) or a larger dataset over a shorter period (eg 200 DMUs over two years of annual data).You should have more than one year’s data.
3. Personal interest
You will be spending a considerable amount of time analyzing the data and writing up your report so pick an area that interests you!
4. Ability to demonstrate understanding of the whole DEA module
Pick something which is quite rich in alternative data, so that you can compare alternative models of performance; perhaps where league tables or some other measure of performance is already used so you can compare you results with these. Having data for (mostly) the same DMUs over several years allows you to look at changes over time. You should consider some other groupings of the DMUs (say rural or urban, private or public ownership)
So do not jump to make a selection too quickly, first do some research to find out what data might be available. But also do not spend all your time choosing, you are best making your choice and sticking too it as long as data is available.
You are best writing the report as you go along, that way you can use the surgery sessions to get feedback on your progress. Some tips on where to start are available in blackboard.
There is also a second self and peer assessment task to ensure you have located some appropriate data and have looked at some simple DEA models and obtained feedback to ensure you are on the right track.
2 The Form and Content of the Report
The report that you hand in should be a report on the efficiency of the operations of your case study organization(s) for one of its stakeholders. You should not assume that the stakeholder understands the Data Envelopment Analysis technique and so will need to explain it to the reader (but should include technical details in an appendix). The following is intended as a possible outline for your report it also gives a guideline as to the number of words expected, some variation from this guideline is expected depending on your actual case study.
1. General introduction (250-750 words).
Here you introduce the operations you are studying and what good performance might mean for them. If there is already a method of performance measurement used you should discuss it.
2. Introduction to the Data Envelopment Analysis technique (500-1500 words).