Introduction: Making Sense of the Mess

Once you have collected your data, you are often left with a giant pile of numbers or words called raw data. On its own, raw data is hard to understand. In this stage of the Statistical Enquiry Cycle, we learn how to organise, clean, and turn those numbers into clear pictures (diagrams) and summary numbers (measures). This makes it possible to spot patterns and eventually answer your original hypothesis.

Whether you are using a pencil and ruler or a high-tech spreadsheet, the goal is the same: accuracy and clarity.

1. Organising and Cleaning Data

Before you can draw a graph, you must organise your data. This usually involves tabulation (putting data into tables).

Data Cleaning

Data isn't always perfect. Cleaning data involves looking for:

  • Missing responses: What do you do if someone skipped a question on your questionnaire?
  • Errors: Like a human height recorded as \(250\) metres instead of \(250\) cm.
  • Spreadsheet issues: Incorrect formatting that stops a computer from doing calculations.

Grouping Data

If you have too many different values (like 100 different test scores), you might group them into class intervals. Example: Instead of listing every score, you use groups like \(10 \le s < 20\).

Important Tip: Grouping data makes it easier to handle, but you suffer a loss of accuracy because you no longer know the exact original values.

2. Presenting Data Visually

Choosing the right diagram depends on the type of data you have (qualitative or quantitative) and what you want to show.

Common Diagrams (Foundation and Higher)

  • Pictograms: Use symbols to represent frequencies. Always check the key!
  • Pie Charts: Show how a total is split into categories. To find the angle for a section:

    \(\text{Angle} = \frac{\text{Frequency}}{\text{Total Frequency}} \times 360^\circ\)

  • Stem and Leaf Diagrams: Great for showing the shape of the data while keeping the original values.
  • Box Plots: Show the median, quartiles, and range. Excellent for comparing two sets of data.
  • Cumulative Frequency Graphs: Used to estimate the median and quartiles for grouped data.
  • Histograms (Equal Class Width): Similar to bar charts but for continuous data.

Higher Tier Only: Advanced Visualisations

  • Histograms with Unequal Class Widths: You must calculate Frequency Density for the vertical axis.

    \(\text{Frequency Density} = \frac{\text{Frequency}}{\text{Class Width}}\)

  • Comparative Pie Charts: When comparing two different total populations, the area of the circles must be proportional to the total frequency. To find the radius (\(r\)) of a new pie chart:

    \(r_{new} = \sqrt{\frac{\text{Total}_{new} \times (r_{old})^2}{\text{Total}_{old}}}\)

Quick Review: Always label your axes clearly. If a graph looks "squashed" or has a truncated axis (a scale that doesn't start at zero without a break symbol), it might be a misleading diagram!

3. Generating Statistical Measures

Presentation isn't just about pictures; it’s also about finding "summary" numbers that describe the data.

Measures of Central Tendency (Averages)

  • Mean (\(\bar{x}\)): The "total" average. \(\text{Mean} = \frac{\sum fx}{\sum f}\)
  • Median: The middle value when data is in order.
  • Mode: The most common value or modal class.

Measures of Dispersion (Spread)

  • Range: \(\text{Highest value} - \text{Lowest value}\).
  • Interquartile Range (IQR): \(\text{Upper Quartile (UQ)} - \text{Lower Quartile (LQ)}\). This shows the spread of the middle \(50\%\) of the data and ignores extreme outliers.
  • Higher Tier - Standard Deviation: A more sophisticated measure of spread. You will be given the formula:

    \(\sigma = \sqrt{\frac{\sum (x - \bar{x})^2}{n}}\) or \(\sigma = \sqrt{\frac{\sum x^2}{n} - (\frac{\sum x}{n})^2}\)

Higher Tier - Calculating Skew

Skewness tells you if your data is "pushed" to one side. You can identify it by looking at a box plot or by using this formula:

\(\text{Skew} = \frac{3(\text{mean} - \text{median})}{\text{standard deviation}}\)

4. Using Technology

In the modern world, statisticians rarely do this by hand. You should be familiar with how technology helps:

  • Calculators: Use the STAT mode to input data and find the mean and standard deviation automatically. This saves time and prevents calculation errors.
  • Spreadsheets (e.g., Excel/Google Sheets):
    • Use formulas like \(=AVERAGE()\) or \(=SUM()\).
    • Use sorting to find the median or identify outliers.
    • Generate instant charts. If the chart looks wrong, check that you haven't included the "Total" row in your data selection!

Did you know? Computers are great at drawing, but they can't "think." If you give a computer "dirty" data (like letters mixed with numbers), the output will be meaningless. This is called GIGO: Garbage In, Garbage Out!

Common Mistakes to Avoid

  • Forgetting Units: Always state if your mean is in cm, kg, or £.
  • Incorrect Histograms: On Higher tier papers, drawing a histogram with frequency on the y-axis when class widths are unequal is a very common error. Always use frequency density!
  • Miscalculating Angles: Ensure your pie chart angles add up to exactly \(360^\circ\).
  • Scale Errors: When drawing a line graph or scatter diagram, make sure your scale goes up in equal steps (e.g., \(2, 4, 6, 8\)... not \(2, 5, 6, 10\)).

Key Takeaway: Processing and presenting data is about turning a "mess" into a "message." Choose the diagram that best fits your data type, use technology to stay accurate, and always check for errors before you start your analysis.