What Is a Line Chart? How to Read and Present It in Quantitative Research
What Is a Line Chart
You open your data file and see values changing by month, quarter, age group, or measurement occasion. A line chart connects those data points so the reader can quickly see whether the observed variable is increasing, decreasing, stable, or fluctuating.
A line chart usually has a horizontal axis and a vertical axis. The horizontal axis contains time or an established order, such as January through December or measurements taken before and after an intervention. The vertical axis contains the indicator being tracked, such as revenue, a mean score, a usage rate, or the mean value of a scale. Each point on the line represents the value of the variable at a particular position on the horizontal axis.
From a data perspective, a line chart does not create a new statistical test. It turns a table containing pairs of x and y values into a visual display. An upward line therefore shows that the observed value increased between points, but it is not enough to conclude that time caused the increase or that a research hypothesis has been accepted.
If you are looking for background material for the statistics section, you can also consult the Statistics topic and compare the chart with other quantitative analyses.
The Meaning of a Line Chart in Quantitative Research
The main value of a line chart is that it shows a trend. A table with 12 values is difficult to scan in a few seconds, while a line helps you identify the highest month, a period of continuous decline, and a time when an unusual fluctuation occurred.
In a thesis, a line chart is usually suitable in four situations. The first is describing an indicator over time, such as the number of transactions in each month. The second is comparing the same indicator across two or more groups at different points in time. The third is tracking respondents' mean scores before and after a measurement activity. The fourth is presenting ordered sequence data when adjacent points have a meaningful basis for comparison.
A chart can also help you check the data before running a model. A sudden increase may reflect a real change, but it may also result from a data-entry error, incorrect coding, or a group with too few observations. When you find a suspicious point, check the .sav or .csv file, the variable codes, and the number of observations before changing anything. This check is related to identifying outliers, but a high point on a chart is not automatically a statistical outlier.
A line chart does not replace Cronbach's Alpha, EFA, regression, or hypothesis testing. It answers a descriptive question such as “which direction did the value change”, while a statistical test answers a question such as “is the change statistically significant”. You need to present these two questions separately in the results chapter.
What Makes a Line Chart Acceptable
A line chart does not have a single pass threshold like Cronbach's Alpha, KMO, or AVE. You do not evaluate it by whether the line rises, falls, or has an attractive shape. The relevant criteria are data accuracy, the measurement axes, the order of observations, and the chart's ability to communicate information.
The table below gives you a checklist before you place the chart in your thesis. The sources in the table are methodological sources that can be used for the relevant sections, while the chart type itself still depends on the characteristics of your data and your presentation objective.
| Item to check | Level to consider | How to interpret it in the thesis | Source |
|---|---|---|---|
| Number of observations for each point | Must be clearly identified | State whether each point is a total, mean, proportion, or predicted value | (Nguyễn Đình Thọ, 2011) |
| Order on the horizontal axis | A natural order exists | Months, quarters, years, or measurement occasions must be arranged in the correct order | (Field, 2013) |
| Unit on the vertical axis | Consistent | Do not mix revenue, percentages, and mean scores on the same axis without explanation | (Field, 2013) |
| Number of variables in one chart | Sufficient for comparison | If many lines overlap, separate them into charts or sensible groups | (Nguyễn Đình Thọ, 2011) |
| Inferential conclusion | Requires a separate test | The chart describes a trend and does not establish a p-value or causal relationship by itself | (Field, 2013) |
When the vertical axis starts at a value other than 0, the line may look steeper than it really is. This can sometimes be useful for magnifying a narrow range, but you must state the axis limits and avoid giving the reader a misleading impression of the difference. For charts describing proportions, use the same scale when comparing several groups across different figures.
If the chart uses mean scores from a scale, state the sample size and the method used to calculate the mean in the caption or in the paragraph immediately before the figure. A chart may look clear, but without knowing how many observations produced each point, the reader has too little information to assess it. When the vertical axis contains scores from a scale, state how the items were combined.
How to Read a Line Chart in the Output
SPSS can create a line chart through several procedures, depending on how you aggregate the data. For descriptive data, go to Graphs > Chart Builder, select the line-chart group, drag the ordered variable to the horizontal axis, and drag the variable to be described to the vertical axis. In Element Properties, check whether Statistic is set to Mean, Sum, Count, or the proportion that matches your research question.
If you run Analyze > Compare Means > Means, the results table usually gives you the mean for each group. You can use those values to build a chart, but first check that the group order in the table matches the order you want to display. SPSS may sort groups by numeric code or alphabetically rather than by the time sequence you have in mind.
The following is illustrative output, not the result of a real study. Assume that you surveyed the mean purchase intention score across four quarters, with each point calculated from the valid questionnaires collected during that quarter.
| Quarter | Mean of YD | N | Std. Deviation |
|---|---|---|---|
| Quarter 1 | 3.21 | 86 | 0.71 |
| Quarter 2 | 3.38 | 91 | 0.68 |
| Quarter 3 | 3.14 | 88 | 0.76 |
| Quarter 4 | 3.57 | 94 | 0.65 |
Based on the illustrative table, the line will rise from Quarter 1 to Quarter 2, fall in Quarter 3, and rise to its highest level in Quarter 4. You can describe this accurately as follows: “The mean value of YD increased from Quarter 1 to Quarter 2, decreased in Quarter 3, and reached its highest level in Quarter 4”. You should not yet write that Quarter 4 caused higher purchase intention unless your research design and statistical test support that conclusion.
When a chart contains several lines, each line needs a name in the Legend. For example, the horizontal axis may show quarters, with the first line representing new customers and the second representing existing customers. Check the original data table to ensure that both lines were calculated from the same type of indicator. Comparing one Mean line with one Count line produces a misleading conclusion even though the chart itself displays normally.
In the output or chart file, check four points. First, confirm that the axis labels show the correct variable name and unit. Second, check the order of the points. Third, make sure the number of decimal places is consistent. Fourth, confirm that the caption states whether the data are a mean, total, proportion, or predicted value.
How to Handle a Line Chart That Does Not Look Right
If you think the chart “does not look good”, do not edit the figure before checking the data. A jagged line may accurately reflect the pattern in your sample. Flattening the line by deleting a low point or changing a mean only to make the figure look smoother is difficult to justify to your supervisor.
Check Variable Codes and Order
Open Variable View and Data View, then check the value labels, variable type, and code order. If October appears before February because the variable is stored as a string, create a numeric order variable or sort the values according to the correct rule before building the chart.
Check the Number of Observations at Each Point
A point calculated from 12 questionnaires and another point calculated from 180 questionnaires do not have the same level of stability. Add the N column to the descriptive table or state it in the caption. If a group is too small, discuss how to combine groups with your supervisor instead of hiding the number of observations.
Check Unusual Data
Compare the unusual point with the original file, the collection date, and the data-entry procedure. You can consult the guide to outliers to distinguish an extreme observation from an entry error. Remove an observation only when you have a methodological basis and have recorded the reason. Do not remove it simply because it makes the line look unattractive.
Choose Another Chart Type When the Data Structure Does Not Fit
If the horizontal axis contains groups without an inherent order, a bar chart is usually easier to read. If you want to show a distribution, a histogram or boxplot is more suitable. If you want to examine the relationship between two quantitative variables, a scatterplot is more appropriate than a line chart.
Separate a Chart with Too Many Lines
Four or five lines may still be readable, but many intersecting lines make it difficult for the reader to keep track of each group. You can separate the charts by group, add a supplementary table, or retain only the groups directly related to your research question. The presentation should remain consistent with the model, including when the study contains a moderator or moderating effect.
Distinguishing a Line Chart from a Bar Chart
A line chart emphasizes continuity and the direction of change between points. It suits time, repeated measurement order, or a sequence with a meaningful before-and-after relationship. The line connecting two points directs the reader's attention to the movement from one point to the next.
A bar chart emphasizes the separate size of each group. If you are comparing mean scores across faculties, regions, or product types where the order has no natural meaning, bars are usually safer. Connecting those groups with a line may make the reader think that Group A leads into Group B.
A line chart also differs from a scatterplot. A scatterplot uses points to examine the relationship between two quantitative variables, while a line chart uses an ordered axis and generally connects points according to that order. If the horizontal axis is age, income, or a continuous variable that has not been divided into meaningful intervals, connecting points arbitrarily can easily mislead the reader.
You also need to distinguish a line chart from a regression line. A regression line is produced by an estimated model and usually comes with a coefficient, error term, and statistical test. A line connecting monthly means only describes the observed values. The two lines may both slope upward while having completely different statistical meanings.
Common Errors
The first error is using a general title such as “Survey Results Chart”. The title should state which variable is shown, which groups are compared, and which time points are used. For example, “Mean Purchase Intention Score by Quarter” allows the reader to understand the figure before reading the explanation.
The second error is leaving out the axis name or unit. The reader cannot tell whether 3.57 is a Likert score, million dong, or thousand views. The variable name in SPSS may be short, but the name on the figure and in the caption must be clear enough for someone reading the thesis.
The third error is using colors without a legend. If the thesis is printed in black and white, two lines may look almost identical. Combine color with line style, point markers, and Legend, and check the PDF after exporting it.
The fourth error is describing a trend with causal language. The phrase “Quarter 4 increased purchase intention” goes beyond what the descriptive chart provides. Write “the mean value was higher in Quarter 4”, then use an appropriate test or model if you need to assess the difference.
The final error is placing too many charts in Chapter 4 without linking them to the research question. Each figure should have a number, title, data source, and a short interpretive paragraph. If a figure does not help answer the research objective, move it to the appendix or remove it.
Frequently asked questions
What is a line chart and what is it used for?
A line chart displays values along an ordered axis, usually time, by connecting data points. It helps you see increases, decreases, stability, and fluctuations between observed points.
Does a line chart need to start the vertical axis at 0?
There is no single rule for every situation. Starting at 0 shows the overall size of the difference, while a narrower axis can make a small fluctuation easier to see. If the axis does not start at 0, state the axis limits and interpret the chart carefully so that the difference is not exaggerated.
Can a line chart be used for Likert data?
You can use one when presenting the mean score of a scale across time or across a group with a meaningful order. State how the score was calculated, the number of observations, and the measurement unit. If you are only comparing independent groups without an inherent order, a bar chart is usually easier to understand.
If the line rises, is the hypothesis accepted?
You cannot conclude that from the chart alone. An upward line only shows that the descriptive value is higher at a later point. To evaluate a hypothesis, you need the appropriate test, estimated coefficient, p-value, or confidence interval for your research design.
Why is my line chart jagged or showing unusual jumps?
The data may genuinely fluctuate, but the cause may also be an incorrect month order, wrong coding, different numbers of observations across points, or an incorrectly entered value. Check the original data table, the N column, variable labels, and aggregation method before removing or changing any observation.
Can I use Excel instead of SPSS to draw a line chart?
Yes. Excel is suitable for building the figure and adjusting its format, while SPSS is convenient when the data and aggregation procedure are part of the same analysis workflow. Regardless of the tool, retain the source table, calculation method, and caption information so that you can explain the figure when necessary.
Open your data file now and identify the horizontal axis, the vertical-axis indicator, the aggregation method, and the number of observations at each point. Then build a trial chart and compare it with your research question. If you need to run and interpret the analysis using your own .sav or .csv file, see DoThesis M4 data analysis.