What a line chart shows and what it hides

A line chart connects measured values in time order, which makes it good at showing direction, pace and turning points. It is weaker at showing what happened between measurements, how large the gaps in time were and how much the values varied within each period. Reading a line chart well means taking in what it shows while checking what it leaves out.

What a line chart is good for

A line chart suits a series of values measured repeatedly over time, such as the same quantity recorded every month or every year. The horizontal axis is usually time and the vertical axis shows the value. Because the line keeps the points in sequence, the eye picks up the overall direction faster than it would from a column of numbers. Line charts also let you compare several series over the same period, although more than four or five lines quickly become hard to follow, because they cross and blend together.

A line chart is less suited to categories with no natural order, or to showing how parts add up to a whole at one moment. For those questions, a bar chart or a stacked chart usually works better. Knowing what a chart type is for helps you judge whether a given line chart answers a question it can answer well.

Reading slopes and turning points

The slope of a line segment shows how fast the value is changing between two points. A steep rise means a large change over a short span, a gentle slope means a slow change, and a flat stretch means little or no change. To read slopes reliably, look first at the vertical scale. Steepness depends on the scale as much as on the data. Stretching the vertical axis makes a small change look dramatic, while compressing it makes a large change look modest. Compare charts with similar scales before comparing their slopes.

Turning points are where the direction changes, from rising to falling or the reverse. A peak or trough is a useful signal, because something changed at that moment. Note the date of each turning point and ask what was happening then. Be cautious, though. A single turn in a short series may be ordinary variation rather than a real change of direction, and a sharp-looking turn may be small once you compare it with the full scale.

A line on a logarithmic scale needs a different reading. On such a scale, equal distances on the axis represent equal percentage changes rather than equal amounts, so a straight line means steady percentage growth. Check the axis labels to see whether the scale is logarithmic before reading slopes as absolute change.

Uneven time intervals

A line chart can look evenly paced while the underlying data are not. Many charts place each point at an equal horizontal distance regardless of how much time separates it from the next point. If the data include points for January, February and then April, the chart may show three equally spaced points, which makes the change from February to April look like the change from January to February. The line appears to move at a steady pace even when the time between points is not steady.

To check, look at the horizontal axis. A date axis spaces points in proportion to the time between them, so a gap in the data shows up as a gap in the chart. A category axis places points at equal intervals whatever the time between them. Many tools offer both, and the choice changes how the line looks. When a series has missing periods, some charts connect the points across the gap with a straight segment, which implies a smooth path through a stretch that was never measured. Check whether the chart marks missing data and whether a break in the line matches a break in the data.

What a line can hide

A line implies a value between each pair of measured points, but none of those in-between values were measured. The straight segment is a visual convention, and it can suggest a smooth change when the real path was jagged. For data recorded monthly, the line says nothing about what happened on particular days within the month.

Each point may also be an average, a total or a rate for its period, and the chart often does not say which. An average over a month can hide a sharp spike followed by a quiet stretch. A line that looks steady may conceal wide swings in the underlying data. If a chart uses a moving average, the line has been smoothed, which removes some of the variation, and the chart should say so.

A line also shows the level of one quantity over time, not how its parts contributed. When several lines appear together, each one tells its own story, and the lines do not add up to a total unless the chart says they do. When a line chart supports a claim, a useful final question is whether a table of the same points would tell the same story. If the line seems more convincing than the numbers, the scale or the spacing is probably doing the work.