Start with a useful question

Does “usually” mean the same thing to everyone?

You visit a café before work. Some mornings your coffee arrives quickly; on others, you keep checking the clock. You want to answer a simple question: how many minutes does it take to collect an order?

By the end of this lesson, you will have a repeatable measurement rule and know what five observations cannot tell you. No software is needed: paper and a clock will do.

Image unavailable: A white coffee cup on a red table, seen from above.

Define the measurement first

“Waiting for coffee” could start when you enter the café, place an order, or pay. Each definition answers a slightly different question. Throughout this course, we use one rule:

Start the timer when you join the queue. Stop it when you collect your finished drink.

Measure in minutes, for one person ordering one coffee. Note unusual circumstances, such as an espresso machine breaking down. Do not change the starting point just because the queue happens to be long. That might improve the number on paper, but it would spoil the comparison.

Five visits, one dataset

The following lessons use invented teaching data. These numbers do not describe a real café or research study.

Each row is an observation. The letters identify visits; they are not days of the week. We have already arranged the values from smallest to largest. Their chronological order is unknown. That distinction matters: this table does not show that service became slower every day.

Five observations let us calculate a summary of this dataset. They do not justify a promise about tomorrow's wait. The time of day, staffing, and type of order could all change the result.

Try it yourself

Person A starts timing when they join the queue. Person B starts after paying. Can their times be compared as the same measurement without further information? Suggest one improvement.

Compare your answer

No. Person B leaves out the wait to place an order. Agree on the same starting and stopping points, then collect comparable observations. You cannot reliably correct the old values without knowing the omitted time.

Take this forward: define the question, unit, and measurement boundaries before calculating anything. In the next lesson, you will see why the same five numbers can produce two different answers.

Photograph: www.Pixel.la Free Stock Photos, Wikimedia Commons, CC0 1.0.