Collecting and Organising Data

Data is simply information. Every time you record a score, a measurement, a colour or a count, you are collecting data.

But data on its own is not very useful. The point of collecting it is to answer a question — a statistical question — such as "What is the most popular sport in our class?" or "How long do students take to get to school?"

A good investigation always begins with the question. Once you know what you are trying to find out, you can decide what to collect and how to collect it.

The data you collect yourself is called primary data. Data that someone else has already collected — from a website or a book — is secondary data. Both are useful, but primary data lets you answer your own exact question.

Data comes in two big families, and spotting which is which helps you choose the right chart later.

Categorical data describes something using words or groups — categories. Favourite colour, type of pet, country of birth and eye colour are all categorical. You cannot do arithmetic with categorical data; "blue + green" makes no sense.

Numerical data is made of numbers — things you can count or measure. Number of brothers, height in centimetres, test score and shoe size are all numerical. You can do arithmetic with numerical data, such as finding a mean.

A quick test: ask "Could I find an average of this?" If the answer makes sense, it is numerical. If not, it is categorical.

Numerical data splits again into two types — and the difference is whether the data is counted or measured.

Discrete data can only take certain separate values, usually whole numbers. You count it. The number of goals scored, the number of pets, the number of students — these jump in whole steps. You cannot score 2.5 goals.

Continuous data can take any value within a range. You measure it. Height, mass, temperature and time are continuous. A height could be 152 cm, 152.4 cm, or 152.43 cm — there are no gaps, only the limits of your measuring tool.

The simple rule: if you count it, it is discrete; if you measure it, it is continuous.

When you collect data, you need to record it quickly and without mistakes. A tally chart is perfect for this.

Each time a result comes up, you draw a stroke. To make counting easy, every fifth stroke is drawn across the previous four, making a clear "gate" of five.

Counting the strokes for each row gives the frequency — how many times that result happened. Writing the frequencies into a table gives a frequency table, which is the tidy version ready for drawing charts or finding averages.

The total of all the frequencies should equal the number of items you collected — a quick way to check nothing was missed.

Collecting good data takes a little planning. Once you have your question, think through these steps:

• Decide what to record. What single piece of information answers your question?
• Choose how to collect it. A survey, a questionnaire, a measurement or an observation?
• Decide who or what to include. This group is called your sample.
• Make it fair. Everyone in your sample should have an equal chance of being chosen, and your questions should not push people towards an answer.

A questionnaire is a common tool. Good questions are clear, specific and have non-overlapping answer options.

If your sample is too small or unfair — only asking your friends, for example — your results will not represent everyone, and your answer to the question will be misleading.

Collecting and organising data is the first step of every statistics task — everything else builds on it:

• Data representation — a frequency table is exactly what you need to draw a bar chart, pictogram or pie chart.
• Averages and range — once your data is in a frequency table, you can find the mean, median, mode and range.
• Probability — listing all the possible outcomes of an event is a kind of organising, just like a tally chart.
• In science and geography, you will plan fair experiments and surveys, recording results in tables exactly as you do here.

Whenever you see a neat chart or a confident statistic, remember it began with someone asking a clear question and organising their data carefully — just as you now know how to do.