Where Data Comes From and What Goes Wrong

We are learning to collect data from a survey, our own measurements and a published table, to record where each piece came from, and to name the errors each kind of source is prone to.

Before you startAbout 60 minutes · what success looks like · notes for the helper

About 60 minutes. It's fine to stop after the Quick Check and come back another day: the lesson opens where you left it, at the Quick Check.

You'll know you've got it when

I can record the source, method and date of each set of data in a paper log
I can write a survey question that does not steer the answer, and say who I would ask
I can name at least two errors that each kind of source is prone to, and say how I would reduce one of them

For the helperhow to sit beside this lesson, and what this step is for

This step: Spark

Start with the question at the top and chat about it. There is no wrong answer here: it gets your learner thinking about what they already know.

Getting ready: Paper and a pencil, a clear straight-sided container such as a jar or tin with a ruler, and a look at where the household water meter lives (you might lift the lid yourself, since meter boxes can hold spiders and sharp edges). If there is no meter you can reach, a bucket and a watch will do for measuring how many litres a tap delivers in a minute. The survey options need a few people willing to answer, with their permission, and the Table Detective needs a published table from a library, council leaflet or weather summary. Daily readings start today and run for about two weeks alongside the next lessons.
What understanding sounds like: Listen for your young adult saying where a number came from before saying what it means: 'that is from the gauge, read at nine each morning'. When they talk about an error, listen for a cause and a fix, not just 'it might be wrong'.
A likely mix-up, and something that helps: A common belief is that data from a survey is 'what people think, so it is fine' and data from a table is 'official, so it is right'. You might ask, 'Who was asked, how was the question worded, and when was this written?' and see whether those answers are on the page.
If it comes easily: For a stretch, you might ask which of the three sources they would trust most for how much water the household used, and which for what people believe about their own habits, and why those are different questions.
If it's a struggle: If it is a struggle, one way is to work on one source at a time, starting with the measurements because they are the most concrete, and to write the log entry together before looking for errors. Provenance, sample and bias are words worth talking through first, in your home language if that is easier. If asking people questions feels like too much, the measurement and table options need no survey at all.

Step 1 of 6 · Spark

Lesson 1 of 5 · Data You Can Trust: Collecting, Checking and Spotting Trends

Ask the people, ask the meter

For the helper

Start with the question at the top and chat about it. There is no wrong answer here: it gets your learner thinking about what they already know.

Ask twenty households 'Do you waste water?' and nearly all of them will say no. Read the same twenty water meters and the story is rather different. Both sets of numbers are real data, and both were collected honestly, so which one do you believe? Before you read on, decide what the answer depends on. Is it the same for every question you might ask about water? Write your thinking down in two lines, because you will come back to it at the end.

What people say they do, what a meter says they did and what a table says a whole town did are three different kinds of evidence.