30. Statistics for Analysts

30.2 The Analysis ToolPak: Load It and Read the Descriptive Statistics Output Line by Line

By Ravindra Bagale · Last updated: 7 October 2026

Example first. Shahrukh needs the same summary for five columns before a meeting. Typing 16 formulas per column is slow. With the Analysis ToolPak he selects the data, ticks Summary statistics, and Excel writes a 14-line table in one click. The table is full of terms like "Standard Error" and "Confidence Level(95.0%)". This lesson loads the ToolPak, runs Descriptive Statistics, and explains every line with the formula that gives the same number.

First look at the data (before → after):

Expected result — ToolPak output (first lines)

File › Options › Add-ins › Manage: Excel Add-ins › Go › tick Analysis ToolPak › OK
Data › Analysis › Data Analysis › Descriptive Statistics
Input Range B1:B21, Labels in first row, Output Range D1, Summary statistics, Confidence Level for Mean 95%

The full output, read line by line

Excel prints more decimals (for example 1.539830616); the table rounds to 4 decimals.

Every number on this page was computed with Excel's documented formulas and re-checked twice: in LibreOffice Calc 25.2 with the same formulas, and in Python with scipy and statsmodels.

The two lines people misread

1. Standard Error is not Standard Deviation. Standard deviation (6.89) describes individual orders. Standard error (1.54) describes the mean: if you took another 20 orders, their average would typically differ from 13.5 by about 1.5 minutes. It shrinks as you collect more data (divide by the square root of the count).
2. Confidence Level(95.0%) is a half-width, not a percentage. It is the "±" part. The 95% confidence interval for the true average delivery time is 13.5 ± 3.22, so 10.28 to 16.72 minutes. It uses the t distribution with n − 1 = 19 degrees of freedom (=T.INV.2T(0.05,19) = 2.0930). The formula =CONFIDENCE.T(0.05,STDEV.S(B2:B21),20) gives the same 3.2229.

Options in the dialog

Several columns at once. Select a block such as B1:F21 (five measures with headers), keep Grouped By: Columns, and the ToolPak writes five summary blocks side by side, which is the fastest way to compare columns.

Microsoft sources (checked 7 October 2026): Use the Analysis ToolPak to perform complex data analysis.

Practice task (try it before you open the answer):

1. Run Descriptive Statistics on the same data with Kth Largest: 2 ticked. What extra line appears and what is its value?
2. Change the data: BLK-5020 becomes 17 minutes. Does the ToolPak output update by itself?
3. What is the new mean and how do you get the new output?