VCE Applied Computing (Year 11) · Unit 1 — Data Analytics · 16m read
Data Visualisation and Analysis
Ever wonder why Netflix shows you graphs instead of spreadsheets? Data visualisation is about turning raw numbers into pictures—charts, graphs, and plots—that reveal patterns your eyes can actually spot. You'll learn which visuals work best for different types of data and how to read them like a detective, uncovering trends and insights that matter. This skill is gold for VCE statistics, business studies, and any career where you need to make decisions based on real information.
Learning objective: Select appropriate visualisations for different data types and interpret patterns in data.
By the end of this lesson you can:
- I can select the most appropriate chart type for a given data type and purpose.
- I can calculate and interpret mean, median, mode, range, and standard deviation.
- I can identify patterns, trends, correlations, and outliers in data.
Visualisation is a torch on data. The right chart lights up patterns instantly.
Visualisation selection and statistical analysis
### Choosing the right chart
| Data type | Purpose | Chart type |
|-----------|---------|-----------|
| Categorical | Compare quantities | Bar chart, column chart |
| Categorical | Show proportions of a whole | Pie chart, stacked bar |
| Continuous | Show distribution | Histogram, box plot |
| Time series | Show change over time | Line graph |
| Two numerical variables | Show relationship | Scatter plot |
| Hierarchical | Show composition | Treemap |
| Geographical | Show spatial patterns | Choropleth map |
### Descriptive statistics
Measures of central tendency:
- *Mean:* $\bar{x} = \dfrac{\sum x}{n}$. Affected by outliers.
- *Median:* Middle value when ordered. Robust to outliers.
- *Mode:* Most frequent value. Useful for categorical data.
When to use which:
- Skewed data (income, house prices): use median.
- Symmetric data (heights, exam scores): mean and median similar.
Measures of spread:
- *Range:* Max − Min. Simple but sensitive to outliers.
- *Standard deviation ($\sigma$):* Average distance from the mean. Large $\sigma$ = more spread.
$$\sigma = \sqrt{\frac{\sum (x - \bar{x})^2}{n}}$$
### Interpreting patterns
Trend: Direction of change over time (increasing, decreasing, stable, seasonal).
Correlation: Relationship between two variables.
- *Positive:* Both increase together.
- *Negative:* One increases as the other decreases.
- *No correlation:* No pattern.
- Use correlation coefficient $r$ (−1 to +1).
Correlation ≠ causation: Two variables moving together doesn't mean one causes the other.
Outlier: A data point far from the main distribution. May indicate: data entry error, measurement error, or a genuine unusual case worth investigating.
### Data visualisation principles
- Purpose: What question does this chart answer?
- Clarity: Can the reader understand it quickly?
- Accuracy: Does the chart represent the data truthfully? (watch for truncated axes, misleading scales)
- Accessibility: Colour choices should work for colour-blind users.
AARI TIP: Ask Aari to help you choose the right chart for a specific dataset and purpose.
Statistical Formulas
- Mean:
\bar{x} = \dfrac{\sum x_i}{n}(Affected by outliers) - Standard deviation:
\sigma = \sqrt{\dfrac{\sum (x_i - \bar{x})^2}{n}}(Average spread from mean) - Correlation coefficient:
-1 \leq r \leq 1(+1=perfect positive; 0=none; −1=perfect negative)
Key ideas at a glance
- Choose chart for data type: Bar, line, scatter.
- Label axes and units: Clarity.
- Avoid misleading scales: Start at zero.
- Tell a story: Highlight insight.
Worked example (Medium)
A dataset shows monthly sales figures over 2 years. What chart type and statistics would you use to analyse it?
- Identify data type. Monthly sales over time = time series data. The key question is: how do sales change over time?
- Chart selection. A line graph is ideal — shows trends and seasonal patterns over time. A bar chart for each month could also work to compare individual months.
- Central tendency. Mean monthly sales gives average performance. Median more appropriate if some months are unusually high (outliers).
- Spread. Standard deviation shows how variable sales are month-to-month. High SD = unpredictable; low SD = stable.
- Interpretation. Look for: overall trend (growing? declining?), seasonal patterns (Christmas peak?), outliers (unusual months requiring explanation).
Answer: Line graph for trends; mean/median for central tendency; SD for variability; interpret trends, seasonality, and outliers.
Worked example (Medium)
You have weekly canteen sales and want to identify weekly patterns. Choose the most appropriate chart and justify.
- Time-series data. Sales by week — a line graph is best for showing trend.
- Add seasonality. Overlay a 4-week moving average to smooth noise.
- Avoid. A pie chart hides time order; a histogram loses the date axis.
Answer: For temporal patterns use a line chart with a smoothing layer — pie/histogram are wrong tools.
Common mistakes
- Using a pie chart for more than 5-6 categories Pie charts become unreadable with many small slices. Humans are poor at comparing angles. Fix: Use pie charts only for 2-5 categories showing proportions. For more categories, use a horizontal bar chart — lengths are easier to compare than angles.
- Using the mean when data is skewed The mean is pulled toward outliers, misrepresenting the "typical" value in skewed datasets like income or house prices. Fix: Report the median for skewed data. Check the distribution shape with a histogram before choosing a measure of central tendency.
- Claiming causation from a scatter plot correlation Students see a strong pattern on a scatter plot and conclude that one variable causes the other. Fix: Always state: "correlation does not imply causation." Identify possible confounding variables. Causation requires controlled experiments and elimination of alternative explanations.
Data storytelling — from numbers to narrative
Raw data and statistics are powerful but often fail to persuade or inform unless they are presented as a narrative. Data storytelling combines data, visualisation, and narrative to communicate findings effectively.
A good data story has three components:
Context: Why does this data matter? What question are we trying to answer? What is the background the audience needs? Without context, even accurate charts are meaningless.
Insight: What does the data reveal? The insight is not the data itself — it is the interpretation. "Sales increased 15% in Q3" is data. "The marketing campaign launched in July drove a 15% sales increase, concentrated among 18-25 year olds" is insight.
Recommendation: What should happen next? Good data analysis leads to action. "Given the campaign's success with younger demographics, we recommend increasing social media ad spend by 20%."
Visualisation principles for storytelling include: highlighting the key finding (use colour to draw attention to the important data point); simplifying (remove gridlines, unnecessary labels, and decorative elements); sequencing (present charts in a logical order that builds understanding); and annotating (add text directly to charts to explain what the audience should notice).
For Unit 1 Applied Computing, data storytelling is relevant to your School-based Assessment — presenting your analysis clearly and persuasively is as important as performing the analysis correctly.
Practice questions
For comparing the distribution of exam scores across a class, the most appropriate chart is:
- Pie chart
- Line graph
- Histogram
- Scatter plot
(1 marks)
Explain why "correlation does not imply causation" using an example.
(3 marks)
Summary
- Chart selection: bar (compare categories), line (trends over time), histogram (distribution), scatter (correlation).
- Mean: average (affected by outliers). Median: middle value (robust). Mode: most frequent.
- Skewed data → use median. Symmetric → mean and median similar.
- Standard deviation: spread from mean. Large SD = variable data.
- Correlation ≠ causation. Confounding variables can cause spurious correlations.