Home Data Analysis Data processing

Data Processing

Attention: open in a new window. PDFPrintE-mail

© UNICEF/2009Data processing is the process by which quantitative data are analyzed. Data processing can be used to answer the following supply chain questions, among others:

- Duration of each step of the supply chain
- Variability within steps of the supply chain
- Costs of each step of the supply chain

When conducting a supply chain analysis, the following methods are useful in processing data (all are easily created in Microsoft Excel):

Whisker Plots: Whisker plots are a useful way to visually represent the distribution of data in a given data set. As at left, a whisker plot can be used to show the variability at each step of a supply chain. By plotting the median with the smallest observation, second and third quartiles, and the highest observation, whisker plots show where data points are clustered throughout the data set. The more tightly clustered the data, the less variability, and vice versa.

Stacked Time Charts: Stacked time charts are another way to represent the range of each step in the supply chain. Stacked time charts show the minimum, average, and maximum observations as a horizontal bar (see the example at right). Measurements for each step in the supply chain are then stacked in chronological order, and comparison between each step show where the greatest range variability is. The stacked time chart here indicates wide variability across all represented stages of the supply chain, with lower average times in the later ends of the supply chain. One drawback of stacked time charts is that they do not show where clusters of data are concentrated and they do not give an indication of outliers.

Bar Charts:  Bar charts have rectangular bars extending either horizontally or vertically. The length of each bar represents a proportional value of the data being looked at. Bar charts are used to compare groups and frequency of events from a defined data set.

Pie Charts: Pie charts are a useful method for comparing the proportional size of a data set when compared to the whole data set. Pie charts can be an effective way of quickly indicating the relative significance of different data components.

Scatter Plots: A scatter plot displays data as a collection of points, each having the value of one variable determining the position on the horizontal axis and the value of the other variable determining the position on the vertical axis. When large groups of data are processed using a scatter plot, trends in the data can be observed, helping to predict the results for data which have not yet been measured.

Line Graphs: Line graphs are used to compare two variables along different axes, one along the vertical axis and one along the horizontal axis. Line graphs are useful when both variables can be easily determined. For instance, if plotting shipment weights against transport times, shipment weight could be on the horizontal axis and transport time along the vertical axis. By aligning shipment weights from least to greatest, plotting a line graph will indicate if the weight of a shipment has any effect on transport time. Line graphs are useful for looking at scenarios such as this, because the slope of the line can indicate results which have not yet been measured.

 

Strategic Issues in Data Processing

The results determined from data processing are useful in forecasting different elements of the supply chain. For instance, data processing can identify the range of time it takes to transport goods between steps in the supply chain. This information will allow for more accurate forecasting. Or data processing can be used to determine the cost of each step in the supply chain. This information can be used for more accurate forecasting.

In addition to more accurate and reliable forecasting, data processing is an effective tool to determine areas in the supply chain which need improvement. For instance, the results from data processing may identify that sea freight passing through Algeciras as a port-of-transshipment experiences shorter delays than sea freight passing through Tangiers. This information could be used to either improve processing at Tangiers or to strategically route shipments so that they pass through Algeciras.

 

RUTF Case Study: Data Processing