30-minute beginner lesson
- 5 minutes: Learn rows, columns, units and storage
- 10 minutes: 1 average practice and number swapping
- 10 minutes: Compare the two methods 2 or compare before and after 3
- 5 minutes: Write one sentence and check results
Statistics for beginners
Start with just five numbers.
Let’s take it one step at a time.
You do not need to memorize formulas. Start with a question, enter your numbers, and learn to explain the result in one sentence.
13 step-by-step lessonsExample data and answers includedPractice separately from your real projects
Start with your question
What do you want to find out?
You do not need to know the tool’s name. Choose the situation closest to yours to find a suitable lesson.
First, distinguish between the two. If you already have numbers, choose the analysis lesson, and if you want to set the conditions for future experiments, choose the DOE lesson. If you measure the same object twice, it is a “paired comparison.”
Getting started with statistical analysis
Your first analysis in six steps
- Open the example first.
Click the green “Practice this example” button in the lesson. A practice screen filled with numbers opens in a new tab. Look back and forth between the Instructions tab and the Practice tab.
- Put one number in each box.
Click the box to correct the number. Instead of putting the units g, seconds, and % in the column, write them in the column name. After correcting the number, click another box to finish entering.
- Choose which columns to view.
Pick two columns if you're comparing two groups, and pick one column if you're only looking at weight. The before and after values for the same person must be in the same row.
- Click “View Results.”
After changing data or settings, click the button again. A disabled report button prevents you from saving an outdated result.
- Read in the following order: size → picture → uncertainty.
First look at how many minutes the difference is and what it looks like in the picture. Then check the confidence interval and p value. You can look up difficult terms in the dictionary below.
- Save the work you want to keep as a file.
Save tasks/plans (JSON) to continue working, and use reports/plans (HTML) to read and share. Don't just rely on your browser's automatic saves to save important data.
One variable per column; one observation per row
| row | Before practice · s | After practice · s |
|---|---|---|
| 1 · Person one | 30 | 27 |
| 2 · Person two | 35 | 31 |
Each row connects one person’s before and after values. Sorting just one column would mix up the people.
Paste a table when you have many values
Click “Paste table” and paste the lesson’s copied data. If the first row contains column names, select “Use the first row as column names”.
No file upload is needed. An empty cell means “not measured”; 0 means the measured value is zero. They are different.
Move the slider and explore
What happens to the mean when one value changes?
Try changing only the first of the five weights. The other four values are fixed at 100, 102, 100, and 100.
Mean 100gsample standard deviation 1.414g
The dots are the weight of each bag, and the vertical line is the average. The larger the standard deviation, the more spread out it is. The numbers in this picture are not saved in the actual work.
DOE selection helper
When you get confused because there are so many names
First, look at the recommendations by purpose and frequency, and use this table when you want to make a more detailed selection. The number of experiments depends on the number of factors, center points, and repetitions.
| name | To put it simply: | When will it be reviewed? | Things to check |
|---|---|---|---|
| Full factorial | All combinations of specified values | When conditions are few and full combinations can be tested | As the number of conditions increases, the number of experiments increases rapidly. |
| Fractional factorial | Some combinations selected with structure | When finding important conditions or combination effects by reducing the number of times | Check resolution and effect overlap. |
| PB | Selecting many candidates through small-impact experiments | Selection stage where there is reason to believe that the interaction is small | Main effects can be mixed with interactions. |
| Orthogonal array | Balanced combination of levels | When planning level 2·3 level selection | Not all interactions are distinct. |
| CCD | Explore curvature with factorial, axial, and center points | When looking at intermediate conditions and combination effects | CCC creates an axis point outside the input range. Separate from CCI/CCF. |
| Box–Behnken | Plan to avoid edges where all conditions are extreme. | Explore curvature with 3 to 5 numerical factors | The results for unobserved edges cannot be determined. |
Currently supported: 2 to 12 numeric factors, separate ranges per design. Categorical, mixture, execution constraints, split-plot design and DOE response analysis and automatic optimization are not supported. View detailed support coverage →
No need to memorize it
Plain-language glossary
Open and read only the words you need. Just read the explanation and return to the practice.
It's okay, check it here
Troubleshooting and common mistakes
time to learn together
Materials directly used in lessons and in-house training
You can do it alone, or you can work in pairs, with one person entering the information and one person explaining the results. All examples are virtual materials for learning purposes only.
60 minutes of quality practice
- 10 minutes: Understanding averages, dispersion, and samples
- 15 minutes: 6 Comparison of three packaging machines
- 15 minutes: 7 specifications and process capabilities
- 15 minutes: Control chart 8 or defective rate 9
- 5 minutes: Organizing conditions to apply to field data
Experiment plan 60 minutes
- 10 minutes: Factor·Level·Y·Iteration explanation
- 20 minutes: Create helicopter planner 11
- 15 minutes: Lesson 12 — compare fractional factorial run counts
- 10 minutes: Lesson 13 — intermediate conditions and curvature
- 5 minutes: Write down your experiment question, scope, and measurement method.
Please schedule separate production and actual measurement times.
Write it down before applying it to your data
- Questions you want to know: ______________________________
- Who/what/when is one line? __________________
- Units and methods of measurement of numbers: ______________________
- Is it a repetition of the same object or different objects? ______
- Features to use and why you chose them: ______________________
- Numbers/pictures to look at first in the results: ____________________
- What cannot be determined based on results alone: __________________
- Check or measure: ___________________
Facilitator Tips Don’t just guess the correct number, but ask questions such as “What are the units?”, “Who did you subtract from whom?”, and “What doesn’t this result tell you?” The key is to practice distinguishing between what is statistically significant and what is practically important.
Once you save HTML, you can read the description and examples offline. The actual tools lab requires an internet connection. For PDF, select “Save as PDF” in the browser print window.
Calculation criteria and learn more
The example numbers from the textbook were compared with the calculation results from the current tool. This short hypothetical is intended for instructional purposes only and does not represent a sufficient sample size for actual process judgment.
ProcessStat Analysis methods/assumptions · DOE Scope and Design Criteria