7SSGN110 Chap.2 Visual Exploration and Data Communication
Visual Exploration and Data Communication
Make the geometry answer the question
Visual exploration has an analytical job: expose structure that a single summary can hide. A histogram describes the shape of one quantitative variable, grouped boxplots compare distributions, a scatterplot preserves paired measurements and a line plot represents ordered change. Select among them from the research question and variable roles rather than from visual fashion.
The first graph should make the later method easier to justify or easier to reject.
Use axes as part of the evidence
An axis label names the environmental quantity and its unit, not merely the software column. Tick precision should reflect measurement precision, and the scale must make important variation visible without manufacturing drama.
A restricted vertical range can be appropriate when bounds are explicit; a logarithmic scale needs a scientific reason and interpretable ticks. Time axes must retain genuine gaps instead of joining missing periods as though observations were continuous.
Inspect data quality before polishing
Look for skew, clusters, plateaus, detection limits, isolated points, changing spread and nonlinear relationships.
An unusual value might be an event, a transcription error or evidence from another process. Verify it against field notes and instrument limits before deciding whether any exclusion is defensible. Transparency and jitter can reveal overlapping points, but they alter only the display.
Calculations must continue to use the original recorded coordinates.
Distinguish variability from uncertainty
Error bars are meaningful only when the caption says what they represent. Standard deviation describes dispersion among observations; standard error describes precision of an estimated mean; a confidence interval applies a stated coverage procedure.
Identical-looking bars can therefore answer different questions. Include group sample sizes and avoid comparing bar lengths when denominators or error definitions change across panels.
Write captions that constrain interpretation
A useful caption identifies the variables, units, groups, place or period and the visible pattern.
It should not convert sequence into causation or imply that one chart tested every assumption. Tables serve exact values; figures serve position, direction and shape. Generate both from the same prepared dataset and saved script.
A reviewer should be able to connect the graphic with the numerical result and regenerate it without manual editing after export.
Compare panels on a common basis
When several groups or periods share a figure, use compatible scales unless a clearly labelled difference is essential. Keep colour meaning stable, place legends near the evidence and label panels in the order discussed.
Small multiples can separate overlapping patterns while preserving comparison. Check accessibility without relying on colour alone.
Finally, read the chart without its surrounding paragraph: axes, legend and caption should still reveal what was measured, how observations were grouped and which visible feature matters.
For revision, redraw one figure from a plain data description and justify every axis, scale and caption phrase. Then choose a deliberately unsuitable display and explain the information it would hide.
This contrast makes graph selection a reasoned analytical decision rather than a memorised mapping of variable types.
What this chapter covers
- 01
Histogram
- 02
Boxplot
- 03
Scatterplot
- 04
Line plot
- 05
Error bars
- 06
Captions and tables
Worked example · free
Repair a crowded water-level plot
- 2Recover the environmental quantity, unit, site and period.
- 2Use ordered time with readable tick spacing.
- 2Write a factual caption that identifies the visible pattern.
Key terms
- Histogram
- A Histogram displays the frequency distribution of a quantitative variable across bins.
- Boxplot
- A Boxplot summarises a distribution through its median, quartiles and whiskers.
- Error Bar
- An Error Bar displays a named quantity such as standard deviation, standard error or a confidence interval.
Visual Exploration and Data Communication FAQ
Which graph should I use first?
Choose from the variable roles and question. Use a histogram for one continuous distribution, a boxplot for grouped distributions, a scatterplot for paired quantitative variables and a line plot when order through time is part of the evidence.
What should an axis label contain?
Name the environmental quantity and its unit rather than the software column. Choose a scale and precision that reveal the relevant pattern, and explain transformations or jitter so the reader can distinguish display choices from changes to the analysed data.
What belongs in a figure caption?
Identify the variables, groups, place or period represented, then describe the visible feature the reader should notice. Keep causal explanation in the main interpretation unless the design supplies the evidence needed to support that stronger claim.
Assessment move
For each dataset, sketch four candidate displays and reject three with a reason based on variable role or order. Audit the chosen visual for units, readable precision, missing periods, overplotting and a caption that describes evidence without overselling mechanism.
Working through Visual Exploration and Data Communication in 7SSGN110? Sia is AskSia’s AI Statistics tutor — ask any 7SSGN110 Visual Exploration and Data Communication question and get a clear, step-by-step explanation grounded in how 7SSGN110 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.