How to Use Descriptive Statistics in Excel
Once you activate the Data Analysis Toolpak, you can choose a function that will calculate a number of statistics automatically for a sample of data.
Things You'll Need
- Excel 2007 or Excel 2010
Before you analyze a new sample or population of data, it's helpful to examine each variable individually to determine if there's anything unusual about the data you should be aware of during your analysis. The Descriptive Statistics feature in Excel's Analysis ToolPak calculates a number of statistics that help you understand the variability and central tendency of the data.
Enable Analysis ToolPak
Launch Excel. Choose File and then Options.
Select Add-Ins from the list of options in the left column. Locate the Analysis Toolpak entry in the list of add-ins on the right. If it doesn't appear in the list of activated add-ins, select Go... to open an Add-Ins screen.
Check the box next to Analysis ToolPak and choose OK.
After activating the Analysis ToolPak, or if it was already activated, choose OK to continue.
Use Descriptive Statistics
Your objective when using this tool is to calculate descriptive statistics for the data collected for a variable in order to understand that data better. One example might be analyzing the number of seconds it took for a race participant to run one mile.
Open the workbook that contains the data you want to analyze: in this example, seconds per mile. Choose Data from the menu, Data Analysis from the ribbon bar, Descriptive Statistics from the box of selections and then choose OK.
Type the range of cells containing the data for the variable or click the cell selector button, highlight the range with your mouse and click the button again. Select a blank section of the workbook to store the descriptive statistics output and select the kind of information you want to display, such as summary statistics and 95-percent confidence intervals. Choose OK to calculate the statistics.
Use and interpret the descriptive statistics as follows:
- Mean: Average value — here, that would be 716.69 seconds, or almost 12 minutes per mile.
- Standard Error: This is how close your sample values are to the mean. In a normal distribution, almost all values will be within three standard errors of the mean; about 95 percent will be within two standard errors; and about two-thirds will be within one standard error around the mean.
- Median: Middle value — here, 697 seconds, which is about 11.6 minutes per mile.
- Mode: Most common value — 675 seconds, or 11.25 minutes per mile.
- Standard Deviation: An indication of how widely dispersed the numbers are around the mean.
- Kurtosis: Indicates whether the data is peaked or flat when plotted compared to a normal distribution. The kurtosis for a normal distribution is 3.
- Skewness: Indicates whether the data is symmetrical when plotted compared to a normal distribution. The skewness for a normal distribution is 0.
- Range: The difference between the lowest and highest values — 667 seconds, or about 11.1 minutes.
- Minimum: The lowest value in the data set — 399, which is 6.65 minutes per mile.
- Maximum: The highest value in the data set -- 1,066, or 17.77 minutes per mile.
- Sum: The sum of all the data points.
- Count: The number of data points.
- Confidence Level (95%): The interval around the mean where you can be sure with 95-percent probability that the interval contains the actual mean.