Analyze the Graph

 

As you examine the graph, look for areas where the El Niņo-year line extends beyond either the average +1 or average -1 Standard Deviation boundaries. 

Do you notice any problems with this data?

Much as scientists do, we made a subjective decision to use Standard Deviation as our measure of variability. Standard Deviation is one of several appropriate measures to use when the data is normally distributed, as is often the case with temperature data. However, precipitation data are often not normally distributed. For example, during some months, the average -1 Standard Deviation could drop below 0 inches during the year (as is the case for San Diego's precipitation). Obviously, precipitation cannot be measured in negative numbers, so the data makes little sense.

In this scenario, it may be more appropriate to use the upper and lower quartiles as boundaries for the range about the median. You can figure quartiles using the Quartile function in the spreadsheet [=Quartile (array,quart)], where the upper uses the value 3, and the lower quartile uses the value 1. 

Although this problem may complicate matters a bit, it illustrates the point that not all descriptive statistics are appropriate in every situation. The person doing the analysis must consider which functions make the most sense for the analysis.

Now try making a new graph.

Was there any change in the precipitation pattern during El Niņo years?

From observing your graph, do you think the El Niņo event affected precipitation in your selected region?

Procedure 1  |   2  |   3