The main results display is an efficiency scores table. The scores and, if appropriate, the scale efficiency are displayed here. But this is just the start. At the simple click of a button many more useful and easy to interpret displays are available to you. You can sort the scores table in different ways, and include different sections of the data. The distribution graph, shown next, shows the spread of efficiency.
The Score Distribution window presents a quick view of the distribution of scores. This allows you to quickly see whether most units are inefficient, nearly efficient, or well spread between efficient and inefficient. This may tell you either that your units are doing well and can’t improve, or perhaps that you haven’t included input/output variables that will discriminate between them better.
The Frontier Plot shows how DEA works, useful for small models and explaining how Frontier Analyst works.
The Improvement Summary window shows you a graphical display of the possible improvements shown by the analysis. It provides a quick overview of where you should be seeking improvement, and also where there is little to be gained.
The overall display is a pie chart showing the relative percentages of potential improvement for each input/output. This is achieved by adding up the potential improvements for each unit – no weightings are applied.
The summary graphs provide a quick insight into where the greatest efficiency gains can be made. You then use the other graphs to investigate the details. The pie chart shows the total improvements possible. If a ‘slice’ is large, it is worth investigating, but if it is small, there is little to gain by improving that variable.
The alternative potential improvement summary display is a distribution graph. For each input/output, the graph shows the number of units that show a potential improvement in a certain range (0-10, 11-20, 21-30, etc). This display allows you to quickly see how many units can make how much improvement. A large number which can make large improvements is indicative that some effort should be made to improve that particular factor.
The Reference Set Frequency graph shows the number of times an efficient unit appears in other inefficient unit’s reference sets. The higher the number, the more representative of ‘best practice’ a unit is.
X-Y plots allow you to find correlation between two input/output variables to help you choose the best inputs and outputs. If two variables have a high correlation, you may want to remove one of them to increase discrimination.
The Efficiency plot allows you to find correlation between a particular input/output variable and the efficiency scores of each unit.
Remember that all graphs and tables can be easily copied to the clipboard for inclusion in your reports.