Statistics: Line Graphs and Time Series Plots

Understanding Statistics: Line Graphs and Time Series Plots

A plotted line can suggest that every value between two observations is known. Often, it is not. If a school records attendance once each week, the straight segment between two dots is only a visual connection.

Daily attendance may have moved up and down in between. This matters most when data are collected at wide intervals. A line is most reasonable when the quantity changes smoothly, such as temperature measured every minute.

For separate events, such as the number of goals scored in each match, a bar chart may communicate the data more honestly. Always notice how often measurements were taken and whether missing dates could hide important changes.

The steepness of a line describes the rate of change. A rising line means the measured value increased over the chosen interval. A falling line means it decreased.

Steeper sections show faster change, but the units on both axes must be checked before comparing steepness. A climb of ten metres in one second differs greatly from a climb of ten metres in one hour. The rate is found by dividing the change in the measured value by the change in time.

Students meet this idea in speed graphs, cooling experiments, population records, electricity use, and bank balances. A flat section can mean no change, though it can sometimes mean that the measuring tool was not sensitive enough to detect small changes.

Real time data are rarely perfectly smooth. Short term variation is called noise. It can come from natural randomness, measurement error, or one unusual event.

Daily temperature changes because of clouds and wind, even when the season is getting warmer overall. Sales at a shop may rise every weekend. Rainfall may follow wet and dry seasons.

These repeated patterns are called seasonal or cyclical patterns. Looking at several years helps separate a long term trend from a regular yearly pattern.

A moving average can help reveal the underlying direction by averaging nearby values. It reduces random bumps, but it can hide sudden real changes, so the original data should still be inspected.

A sudden jump or drop deserves careful attention. It might show a genuine event, such as a new law, a storm, a machine failure, or a change in public behaviour. It might instead be caused by a changed survey method, a new sensor, or an error when data were entered.

Read titles, units, source notes, and dates before drawing conclusions. Comparisons need equal time intervals and a sensible vertical scale. A narrow vertical range can make a small change look dramatic, while a very wide range can hide a meaningful shift.

Finally, two lines moving together do not prove that one causes the other. A third factor may affect both. Good graph reading means describing the visible pattern first, then checking whether the evidence supports any explanation.