What is data visualisation and why use it?
Data visualisation means turning data into a picture: a chart, graph or map. Our eyes find the biggest bar or a rising line much faster than we can read a table.
A chart helps you to compare, to see a trend, to spot an outlier (an unusual value), and to tell others what you found. It is one step in data analysis: collect → clean → analyse → visualise → conclude.
Choosing the right chart
- Bar chart: compare separate groups (fruits, cities, subjects). Bars must start at 0.
- Pie chart: show parts of one whole (how 100% is shared). Use it only for a few slices.
- Line chart: show change over time (temperature by month, sales by year).
- Scatter plot: show the link between two number variables (height and weight, hours and marks). Each dot is one item.
- Histogram: show how a number is spread out, using touching bars for ranges (marks 0–10, 10–20 …).
- Table: best when people need exact values.
Ask: What is my question? Compare, part of a whole, change over time, or relationship? The question picks the chart.
Parts of a good chart
Every chart needs:
- A clear title that says what and when.
- Axis labels with units (°C, hours, ₹, kg).
- A sensible scale with equal steps.
- A legend if colours mean something.
- The source of the data.
Keep it simple: no 3D effects that hide values, not too many colours, and colours that people with colour blindness can tell apart.
Reading charts and spotting patterns
When you read a chart, look for:
- Biggest and smallest values.
- Trend: going up, going down, or steady.
- Correlation in a scatter plot: positive (dots rise), negative (dots fall) or none (a cloud).
- Outliers: points far away from the rest.
Remember: correlation is not causation. Ice-cream sales and sunburn both rise in summer, but ice cream does not cause sunburn. The hot weather causes both.
Misleading charts
Charts can trick the eye, by mistake or on purpose:
- Cut axis: a bar chart whose axis starts at 50 instead of 0 makes a small difference look huge.
- Unequal steps on an axis.
- Pie slices that do not add to 100%.
- 3D or tilted pies that make front slices look bigger.
- Cherry-picking: showing only the months that fit the story.
Always check the axis, the units and the source before you believe a chart.
Making charts with software
Spreadsheets (like LibreOffice Calc, Google Sheets or Excel) make a chart in a few clicks: type the data in columns, select it, choose Insert → Chart, then pick the type and add a title and axis labels. Programmers use libraries such as Python's matplotlib: plt.bar(fruits, counts) draws a bar chart. Whatever the tool, you still choose the chart type and check it is honest.
Try it: ask 10 friends or family members their favourite fruit. Make a tally, then draw a bar chart on paper and a pie chart (angle = count ÷ total × 360°).
Key formulas and definitions
- Pie slice angle = (value ÷ total) × 360°
- Percentage = (value ÷ total) × 100%
- Bar height ∝ value (axis from 0)
- Compare → bar · Part of whole → pie · Over time → line · Two variables → scatter · Spread → histogram
Worked examples
1. A survey: Mango 12, Apple 8, Banana 6, Orange 4, Grapes 3. Find the pie angle for Mango.
Total = 33. Angle = 12/33 × 360° ≈ 131°. Percentage = 12/33 × 100 ≈ 36%.
2. Which chart would you use to show how a city's rainfall changes month by month?
A line chart (or a bar chart by month), because it shows change over time. The months go on the x-axis in order.
3. A bar chart shows Team A = 52 and Team B = 55 with the axis starting at 50. B's bar looks 2.5 times taller. Is that fair?
No. Above 50, A shows 2 and B shows 5, so B looks 2.5 times taller. The real ratio is 55/52 ≈ 1.06, only 6% more. The axis should start at 0.
4. In a scatter plot of age of car vs price, the dots go down from left to right. What does this show?
A negative correlation: older cars tend to have lower prices.
5. A pie chart has slices of 40%, 35% and 30%. What is wrong?
They add to 105%. Slices of one whole must add to 100%, so the data or the chart is wrong.
Common mistakes
- Using a pie chart for change over time. Use a line chart for time.
- Starting a bar chart's axis above 0, which makes small differences look big.
- Forgetting the title, axis labels or units, so nobody knows what the numbers mean.
- Thinking a correlation in a scatter plot proves that one thing causes the other.