The first rule of data analysis is to plot your data. Minitab’s Graphical Summary provides both graphical (a histogram with an overlaid normal curve) and numeric summaries of the raw data and key statistics (mean, median, standard deviation, maximum value, minimum value, and sample size).

Answers the questions:

- What do the sample data look like? Are the data symmetric, skewed, or reasonably normal?
- What are the estimates of the process mean and standard deviation? Are they different from expectations?
- How good are these estimates?

When to Use | Purpose |
---|---|

Pre-project | Assess your process and verify the process is producing output significantly different from expectations, validating the need for an improvement project. |

Mid-project | Characterize any input or output. |

End of project | Verify the process is producing expected output after implementing improvements. |

Single column of continuous data with an optional column of categorical data to be used as a By, or grouping, variable.

- In a Minitab worksheet, enter the raw data into a single column.
- Optionally, you can use a categorical (By) variable entering the values into a second column.

- Use confidence limits for the standard deviation to evaluate if the variance is equal to a standard or benchmark.
- To do a 1-variance analysis with a 1-sided alternate hypothesis (greater than or less than), double the alpha (for example, for a 5% alpha-level, use a confidence level of 90% which provides 5% on each side).
- To use the graphical summary method for testing normality, you must have a sample size of 50 or more data points.
- If you have discrete numeric data from which you can obtain every equally spaced value and you have measured at least 10 possible values, you can evaluate these data as if they are continuous.