# Line Graph: Definition, Examples and When to Use One A line graph is a chart that shows how a numerical value changes across an ordered sequence, usually time, by plotting data points on two axes and connecting them with straight line segments. The x-axis carries the ordered variable and the y-axis carries the measured value, so the shape of the line shows direction, rate and pattern of change. Line graphs matter because many biological questions are questions about change. Bacterial growth, patient temperature, reported disease cases, [gene expression](/blog/guides/gene-expression) after a stimulus and pollutant concentration across a season all become easier to read when paired values are plotted in sequence. A well-built line graph turns a table into a pattern the eye can follow. A careless one can make a small shift look dramatic or a noisy series look precise. ## Line Graph at a Glance | Feature | Line graph | Line plot | Bar graph | |--|--|--|--| | Main question | How does a value change across an ordered sequence? | How often does each value occur? | How do separate categories compare? | | x-axis | Ordered numeric values or time | A single number line | Categories or groups | | Marks | Points joined by line segments in x order | Marks stacked above values on a number line | Separate bars | | Best use | Trends, rates, growth curves, repeated measures | Distribution of a small numeric data set | Counts, means and categorical comparisons | | Main risk | Clutter with many series and false precision between points | Confused with a connected line chart | Ordering categories can suggest a trend that does not exist | These charts answer different questions. A line graph asks how a value changes. A line plot asks how often values occur. A bar graph asks how categories compare. The confusion begins when software uses the label "line plot" for any connected line chart. ## What Is a Line Graph? A line graph is a two-axis chart in which each observation is a point defined by an x value and a y value. The points are joined in x order with straight segments. The x-axis is horizontal and the y-axis is vertical. The x variable must be ordered. Time is the common example, but dose, temperature, age, distance and concentration can also work when they have a meaningful order. The y variable is numerical and measured in defined units. A line graph is appropriate when the question concerns direction, rate or pattern across the x variable. Line graphs make it easy to read values and to see increases and decreases from the slopes of the lines [1]. That strength depends on a clear scale and honest point placement. ### Anatomy of a Line Graph A complete line graph has several parts: 1. **Title**: a short statement of what the graph shows. 2. **x-axis**: the ordered variable, with its unit. 3. **y-axis**: the measured variable, with its unit. 4. **Scale and tick marks**: the mapping from data units to physical distance. 5. **Data points**: the measured x-y pairs. 6. **Line segments**: straight connections between adjacent points in x order. 7. **Legend**: a key that maps colors or line styles to groups. 8. **Annotations**: labels for events, thresholds or missing periods. A scale is the rule that turns a value such as 37.5 degrees Celsius into a position on the page. Tick marks are the labeled reference points along an axis. A legend is needed when more than one series appears. A gap in measurement should remain a gap, not a straight line that hides missing data. ### What the Line Segments Mean The segments are interpolations between measured points. They do not prove that a measurement exists between two points. A steep segment means y changes quickly per unit x. A flat segment means y did not change over that interval. A bend or change in slope can mark the start of a growth phase, a plateau, a threshold or a response to treatment. A line graph is not a causal claim. A rising line shows association with x, not proof that x caused the rise. A steep slope is not the same as a strong correlation. A double line graph does not permit unlike units to share one y-axis without clear labeling. ## Line Graph Examples Line graph examples appear across laboratory science, clinical monitoring and public health. ### Example 1: Bacterial Growth Curve Suppose a teaching lab measures optical density at 600 nm every hour. The readings are 0.05 at 0 hours, 0.08 at 1 hour, 0.18 at 2 hours, 0.42 at 3 hours and 0.85 at 4 hours. Time goes on the x-axis and optical density on the y-axis. The line is nearly flat at first, then steepens. The slope between 2 and 3 hours is 0.24 optical [density units](/blog/research-skills/density-units-explained-si-cgs-and-conversions) per hour. Optical density is a proxy for biomass, not a direct cell count, so the graph supports a growth trend without proving a specific cell number. ### Example 2: Clinical or Veterinary Monitoring A clinician may plot body temperature, body weight or blood glucose at scheduled visits. The line shows direction and response over time. A rising fever curve, a stable weight or a sudden drop becomes visible quickly. Missing visits appear as gaps. For an individual patient, the graph is one part of the record and not a diagnosis by itself. ### Example 3: Public Health Dashboard A Covid-19 tracker for India used graphs for daily trends and state-wise comparisons, alongside tables and maps [2]. Daily case counts over time are a natural line graph because the x variable is an ordered date sequence and the goal is to see acceleration, peaks and declines. ### Example 4: Linked Environmental Series Pollen dispersal can relate to the previous summer's temperatures, so the two series are separated by a time lag [3]. A line graph of each series can show lagged peaks. A combined heatmap and line graph view gave better readability and comparability of values than a line graph alone in that work [3]. The example matters because not every useful line graph has x and y measured at the same moment. ### Example 5: Double Line Graph A hypothetical study compares tumor volume in a vehicle group and a treated group at weeks 0, 2, 4 and 6. A double line graph places both series on shared axes with one line per group. The legend identifies the groups, and the two slopes support comparison. If the groups use different units, separate panels are safer than a dual y-axis that creates false crossings. ## How to Make a Line Graph The steps are simple, but each one affects whether the graph is honest. 1. **State the question.** Write the question in one sentence. "Does optical density rise faster at 37 degrees Celsius than at 30 degrees Celsius?" identifies the ordered variable and the outcome. 2. **Choose x and y.** Put the ordered variable on the x-axis and the measured value on the y-axis. Record units before plotting. 3. **Build a clean data table.** Use one row per x value and one column per series. Keep the raw values and note missing measurements. 4. **Set the axis ranges.** Match the scale to the data and the question. A line graph may start above zero when the question concerns small changes, but the truncated axis must be labeled and must not exaggerate the change. 5. **Place points at true x positions.** Uneven time intervals should be spaced by actual time, not forced into equal spacing. A reading at day 1 and day 30 should be far apart. 6. **Connect points in x order.** Do not connect unordered categories. If the x variable is not ordered, a scatter plot or bar graph is usually better. 7. **Label the axes and add a title.** Include the quantity and the unit, such as "Time (hours)" and "Optical density at 600 nm." 8. **Add a legend for multiple series.** Use direct labels when possible. Keep colors distinguishable and use line styles as a second cue. 9. **Review the graph before sharing it.** Check the scale, missing data, line count and caption. Remove decorative effects that do not carry information. When values span several orders of magnitude, a standard linear axis can compress small values into a flat line. A log scale or an adapted log-line chart can make the pattern readable, and the axis label should say that a log scale is used [4]. A line graph with a log axis is still a line graph. It changes the spacing of values, not the meaning of the line. ## Reading Line Graphs Reading line graphs starts with the frame, not the line. Read the title, both axes, the units and the scale. Check whether the y-axis starts at zero. A truncated axis can be honest for small changes, but it can also make a minor shift look steep. Next, separate measured points from interpolated segments. The points are observations. The segments are visual connections. A line that passes through a gap is an assumption, not a measurement. A flat segment means no observed change over that interval, not proof that nothing happened. Read direction first, then rate. A line that rises from left to right shows an increase. A line that falls shows a decrease. The slope between two points is the change in y divided by the change in x. In the growth example, a rise from 0.18 to 0.42 over one hour gives 0.24 optical density units per hour. The unit belongs in the interpretation. Look for change points. A sudden bend can mark a shift in growth rate, a treatment effect or a sensor problem. Compare series only when they share units and a sensible baseline. For large detailed sets with many series, a summarized line graph can reduce the data into representative data, analytical highlights and a data envelope [5]. That design helps a reader see the main pattern without pretending that every line can be read at once. ## Line Graph vs Bar Graph The line graph vs bar graph choice follows from the x variable. A bar graph uses separate bars to compare magnitudes across categories. The spaces between bars signal that the categories are independent. Species counts, treatment groups at one time point and survey response categories are natural bar graph subjects. A line graph uses connected points to show change across an ordered sequence. Repeated measures over time, dose-response curves and growth curves are natural line graph subjects. If the x categories have no order, a line implies a path that does not exist. If time has only a few discrete periods and the goal is a simple comparison, a bar graph can work. If the goal is a trend across many periods, a line graph usually communicates more. The two charts can both display time, but they encode different ideas. Bars emphasize separate magnitudes. Lines emphasize continuity and rate. A reader who sees a line expects adjacent x values to be related in sequence. ## Double Line Graph and Multiple Series A double line graph is a line graph with two data series plotted on shared axes, one line per series. It is useful for comparing two treatments, two genotypes, two sites or two time periods. A legend maps each color or line style to its group. Direct labels at the end of each line can reduce the work of matching a legend. A double line graph works best when the series share units and a common scale. If the units differ, two stacked panels are often clearer than a dual y-axis. A dual axis can make unrelated series appear to cross at a meaningful moment when the crossing is only an artifact of scale. More lines do not always add more insight. Line graphs lose visibility as the number of lines increases [1]. Combined heatmap and line graph designs help by clustering series and showing selected lines next to a heatmap [1]. A later design extended that approach to data with time differences and reported better readability and comparability of values than a line graph alone [3]. On small displays, a space-filling line graph preserves key line graph properties while freeing space for extra information [6]. These designs respond to a real limit, not a cosmetic preference. ## Line Plot vs Line Graph A line plot is not the same as a line graph. A line plot, sometimes called a dot plot, uses a single number line. Each observation is a mark above its value. Repeated values are stacked. The data set 1, 2, 2, 3, 3, 3 produces one mark above 1, two above 2 and three above 3. The result shows frequency, spread and outliers. A line graph uses paired x-y observations and connects them in x order. It has a measured y variable and an ordered x variable. A line plot has no separate y-axis measured variable. Software sometimes uses "line plot" for a connected line chart, but in statistics the terms describe different displays. Calling every line chart a line plot creates confusion in reports and classrooms. ## When to Use a Line Graph Use a line graph when the x variable is ordered and the question concerns change. Good cases include time series, dose-response data with ordered doses, growth curves, temperature ramps and repeated clinical measurements. Use a line graph when the points can be connected in a meaningful sequence and when a reader needs to see direction, rate or pattern. Choose another chart when the categories are unordered. A bar graph is better for separate groups. A histogram or box plot is better for a distribution. A line plot is better for a small set of repeated numeric values. A scatter plot is better for the relationship between two variables when the x variable is not an ordered sequence. Automatic selection between line graph and scatter plot depends on whether the data express trends over time [7]. Chart choice is contextual. A visualization recommender can use the state of a monitored system and the preferences of the end user to suggest a dashboard layout [8]. The basic rule remains simple. Use a line graph when x is ordered and the line has a real sequence to trace. Avoid it when the line would invent an order. Irregular sampling is another boundary. When measurements arrive at uneven intervals, a line graph can still work if points are placed at true x positions. Highly irregular data may need multi-scale visualization to preserve timing and avoid false smoothness [9]. A user should not treat evenly drawn segments as equal time steps when the underlying schedule was uneven. ## How Researchers Test Line Graph Designs Visualization researchers evaluate line graphs with defined tasks. Common tasks include identification, discrimination, estimation and trend detection [4]. Identification asks a reader to find a value. Discrimination asks which series is higher. Estimation asks for a magnitude or rate. Trend detection asks for direction and change over time. Small multiple time series displays are used for monitoring current values and historical trends. In one evaluation, integrated size encoding on the current value improved response time by 28 percent for tasks involving large current values and showed minimal interference with trend lookup [10]. A summarized line graph reduced large detailed multiple time series into representative data, analytical highlights and a data envelope and achieved the best overall result against line graphs, band graphs, stream graphs and horizon graphs on four representative tasks [5]. These studies show why graph design is not only a matter of taste. A line graph can be accurate and still be hard to use. Labels, scale, line count and the task faced by the reader all affect speed and accuracy. ## Common Mistakes and Limitations Several mistakes appear again and again in student reports and professional dashboards. - **Connecting unordered categories.** A line through "control, low dose, high dose" may be acceptable if dose is treated as ordered. A line through "liver, kidney, brain, heart" is not. - **Hiding a truncated axis.** A y-axis that starts at 95 percent can make a 1 percent change look huge. Label the axis and explain why it starts where it does. - **Drawing uneven intervals as equal.** If samples were taken at days 0, 1, 2 and 60, the last interval is much longer. Place points by true x value. - **Using too many lines.** Line graphs lose visibility as lines increase [1]. Large-scale time series can become overplotted, which is why interactive systems use density fields and representative lines [11]. - **Treating interpolation as observation.** The line between two points is a visual bridge. It is not evidence of a value at an unmeasured x position. - **Ignoring missing data.** A gap should look like a gap. A continuous line across a gap can mislead. - **Confusing a line graph with a line plot.** The two answer different questions. - **Using a line graph for an unordered relationship.** A scatter plot is often better when the x variable is not a sequence [7]. - **Ignoring large value ranges.** Values spanning several orders of magnitude may need a log scale or a specialized chart [4]. - **Forgetting the reader.** A graph that is clear to its maker can be opaque to a busy reader. Line graphs also have limits. They are poor at showing parts of a whole and weak at showing distributions. They can hide overlap between series. They can suggest continuity where sampling was sparse. When many time series are present, alternatives such as calendar views, spiral charts, theme rivers, dynamic views and parallel coordinates may fit the task better [12]. For an individual patient or animal, a clinician or veterinarian should interpret a trend alongside the full record and not from the graph alone. ## Quick Review 1. A line graph plots paired x-y observations and connects them in x order. 2. The x variable must be ordered. Time is common, but dose and temperature can also work. 3. The line segments are interpolations, not new measurements. 4. Slope shows rate of change and direction. 5. A bar graph compares separate categories. A line graph shows change across an ordered sequence. 6. A double line graph plots two series on shared axes and needs a clear legend. 7. A line plot is a one-axis frequency display, not a connected x-y chart. ## Frequently Asked Questions ### What is the difference between a line graph and a line plot? A line graph plots paired x-y observations and connects them in x order. A line plot uses a single number line with stacked marks to show how often values occur. ### When should I use a line graph instead of a bar graph? Use a line graph when the x variable is ordered and the goal is to show change or trend. Use a bar graph when the goal is to compare separate categories. ### Can a line graph show more than one data series? Yes. A double line graph plots two series on shared axes with a legend. More than a few lines can become cluttered, so separate panels may be better. ### How do I choose the scale for a line graph? Match the scale to the data range and the question. Label units clearly, show zero when it matters and state when a log scale is used. ### What does a flat line mean on a line graph? A flat segment means y did not change over that x interval. It does not prove that nothing happened, especially when measurements are sparse or insensitive. ### Are line graphs good for large value ranges? They can be, but a standard linear axis may compress small values. A log scale or a specialized log-line chart can help when values span several orders of magnitude. ## Sources 1. [Visualization of Time Series Data Using Clustered Heatmaps and Line Graphs](https://doi.org/10.1145/3678698.3678705) 2. [Covid-19 Tracker: A data visualization tool for time series data of pandemic in India](https://doi.org/10.1109/CSDE53843.2021.9718425) 3. [Visualization of Time Series Data with Time Differences Using Clustered Heatmaps and Line Graphs](https://doi.org/10.1145/3801619.3801668) 4. [Reclaiming the Horizon: Novel Visualization Designs for Time-Series Data with Large Value Ranges](https://pubmed.ncbi.nlm.nih.gov/37871083/) 5. [Bridging the Data Analysis Communication Gap Utilizing a Three-Component Summarized Line Graph](https://doi.org/10.1111/cgf.13696) 6. [SF-LG: Space-Filling Line Graphs for Visualizing Interrelated Time-series Data on Smartwatches](https://doi.org/10.1145/3447526.3472040) 7. [Line Graph or Scatter Plot? Automatic Selection of Methods for Visualizing Trends in Time Series](https://pubmed.ncbi.nlm.nih.gov/28092562/) 8. [VisCARS: Knowledge Graph-Based Context-Aware Recommender System for Time-Series Data Visualization and Monitoring Dashboards](https://pubmed.ncbi.nlm.nih.gov/38869997/) 9. [Stroscope: Multi-Scale Visualization of Irregularly Measured Time-Series Data](https://doi.org/10.1109/TVCG.2013.2297933) 10. [The Now and Then: Integrating Current and Historical Data in Small Multiple Time Series Visualization.](https://pubmed.ncbi.nlm.nih.gov/42853720/) 11. [KD-Box: Line-segment-based KD-tree for Interactive Exploration of Large-scale Time-Series Data](https://pubmed.ncbi.nlm.nih.gov/34587082/) 12. [A Survey of Time Series Data Visualization Research](https://doi.org/10.1088/1757-899X/782/2/022013) ## Related Articles - [How to Make a Line Chart in Excel (Step by Step)](/blog/data-analysis/how-to-make-line-chart-excel) - [Bar Graph: Definition, Types and How to Read and Make One](/blog/data-analysis/bar-graph-definition-types-examples) - [How to Make a Scatter Plot with a Trendline, Equation and R-squared](/blog/research-skills/how-to-make-a-scatter-plot) - [Stationarity Definition: What It Means in Time Series](/blog/data-analysis/stationarity-definition-time-series) - [What Is ARIMA? Time Series Model Explained](/blog/data-analysis/arima-time-series-model)