01 · Learn the idea
Look before you calculate
Read the whole data set before doing anything with it. Notice the largest and smallest values, whether the numbers rise or fall, and whether anything looks out of place.
Patterns are often visible immediately, and seeing one early tells you what to check.
02 · Build understanding
Kinds of pattern
A trend means the values move steadily in one direction, such as temperature rising through the morning. A cycle repeats, such as the Moon's appearance over a month.
A pattern may also be a grouping, where the results fall into two or three clusters rather than spreading evenly.
trend: steady direction · cycle: repeats · cluster: grouped
03 · Build understanding
Values that do not fit
A value far from the rest is an outlier. It may be a mistake in measuring or recording, or it may be real and important.
Investigate it before deciding. Check the notes for that trial, repeat the measurement if you can, and never delete a value simply because it is inconvenient.
04 · Build understanding
Sources of error
Errors come from the instrument, from the person and from the conditions. A thermometer read too soon, a cylinder read from above, a draught from an open window all shift a result.
Naming likely sources of error is part of a good report. A result reported without them claims more certainty than the investigation earned.
Apply and explain
Your turn
Explain what to do before calculating anything with a data set. Describe a trend, a cycle and a cluster. Explain what an outlier is and what to do about one. Then name three sources of error and explain why they belong in a report.
Use paper or explain your thinking aloud before opening the solution.
Check your reasoning
Before calculating, read the whole data set: notice the largest and smallest values, whether the numbers rise or fall, and whether anything looks out of place. Patterns are often visible immediately, and seeing one early tells you what to check. A trend is values moving steadily in one direction, such as temperature rising through the morning. A cycle is a pattern that repeats, such as the Moon's appearance over a month. A cluster is a grouping, where results fall into two or three groups rather than spreading evenly. An outlier is a value far from the rest. It may be a measuring or recording mistake, or it may be real and important, so it should be investigated by checking the notes for that trial and repeating the measurement if possible. It should never be deleted simply because it is inconvenient. Three sources of error are the instrument, such as a thermometer read before it settled; the person, such as reading a graduated cylinder from above; and the conditions, such as a draught from an open window. They belong in a report because a result given without them claims more certainty than the investigation actually earned.
- Reading the data first, with what to look for.
- Trend, cycle and cluster each described.
- Outlier handling, three error sources, and the certainty argument.
Ready to practice?
Try the questions, get feedback, and review your reasoning.
Data and explanations practice →