The large data set

The large data set

Spec 3.21 · AQA A-Level Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • The large data set is real data published by AQA in advance; know each variable's type, units and how missing readings are shown.
  • Match diagrams to variables: bar chart for categories, histogram or box plot for one quantitative variable, scatter diagram for two.
  • Use a spreadsheet or calculator for subsets and summary statistics: σ=∑x2n−xˉ2\sigma=\sqrt{\frac{\sum x^2}{n}-\bar x^2}.
  • Remove missing-value codes (such as −99-99) and say why; never replace blanks by 0.
  • Compare samples with one measure of location and one of spread, and state limitations.

What the data set is

The large data set is real data published by AQA before the exam. Know each variable's type, units, collection method and how missing readings are shown.

Real data have many records (rows, one per observation) and many variables (columns).

Variables are categorical (labels, such as fuel type) or quantitative: discrete (counts) or continuous (measured, such as temperature).

Exam questions refer to the data set, so use the AQA files and guidance notes to explore it beforehand.

  1. 1

    Categorical

    Labels, such as fuel type

  2. 2

    Discrete

    Quantitative counts

  3. 3

    Continuous

    Quantitative measurements, such as temperature

Types of variable in the large data set.

Variable

  • Categorical
  • One discrete or continuous variable
  • Two quantitative variables

Suitable diagrams

  • Bar chart, pie chart
  • Histogram, box plot, cumulative frequency graph
  • Scatter diagram

Which diagram suits a categorical variable such as fuel type?

Using technology

Use a spreadsheet or statistics package to sort, filter and select a subset, then calculate statistics and draw diagrams. A calculator also works.

Select a subset (for example one location, one year or one fuel type) with sorting and filtering, and use built-in functions for the mean, median, quartiles and standard deviation.

On a calculator, enter the data (or a frequency table) in statistics mode and read off nn, ∑x\sum x, ∑x2\sum x^2, xˉ\bar x, σ\sigma, the median and quartiles.

  1. 1

    Select a subset

    Sort and filter, e.g. one location, year or fuel type

  2. 2

    Calculate

    Mean, median, quartiles and standard deviation

  3. 3

    Draw diagrams

    Choose a diagram that suits the variable

Working with the data using a spreadsheet or calculator.
  • σ\sigma∑x2n−xˉ2\sqrt{\dfrac{\sum x^2}{n}-\bar x^2}

Which formula gives the standard deviation σ\sigma from nn, ∑x\sum x and ∑x2\sum x^2?

Cleaning the data

Remove missing-value codes and impossible values, say why, and do not delete an unusual value without checking it.

Real data are untidy. Look for missing values (blank cells or a code such as −99-99), impossible values, inconsistent units and duplicates.

Missing codes must be removed, not treated as numbers: a −99-99 would drag the mean down and inflate the standard deviation. Replacing blanks by 0 has the same effect, because the 0 adds nothing to the total but counts as an observation.

An unusual but possible value is an outlier: check it, then decide, and say what you did and why.

051015Cleaned (5 values)With −99 included (6 values)Data usedMean
Illustrative data 12, 14, 15, 16, 18. Leaving a missing-value code of −99 in the list drags the mean from 15 down to −4.

Missing values

  • Blank cells or a code such as −99
  • Remove; never use as a number

Impossible values

  • Inconsistent units, duplicates
  • Correct if possible, else remove

Outliers

  • Unusual but possible
  • Check, then decide and explain

A column of temperatures contains −99-99 for missing readings. What should you do before finding the mean?

Outliers and summary statistics

An outlier is defined by a rule the question gives. Compare the median and IQR with the mean and standard deviation to see its effect.

The rule is usually either more than 1.5×1.5\timesIQR above Q3Q_3 or below Q1Q_1, or more than kk standard deviations from the mean.

The median and IQR resist outliers; the mean and standard deviation do not. If the mean is well above the median, the distribution is positively skewed.

05101520253035408.9mean 15.321.725.5

Worked example

A sample has n=40n=40, ∑x=612\sum x=612 and ∑x2=9774\sum x^2=9774. Is 25.5 an outlier by the 2 standard deviation rule?

The mean of a data set is much larger than its median. What does this suggest?

Drawing conclusions

Compare samples with one measure of location and one of spread, in context, and state the limitations of the sample.

Interpret real data in summary or graphical form to investigate a question in context. Compare two samples using one measure of location (median or mean) and one of spread (IQR or standard deviation), and note differences in skew.

A sample may not represent the whole data set. State the limitations: how it was chosen, whether the data cover only some places or times, and whether the conclusion applies more widely.

Correlation does not prove that one variable causes the other.

  1. 1

    Location

    Compare the median or mean

  2. 2

    Spread

    Compare the IQR or standard deviation

  3. 3

    Context

    Explain what the difference means in the question

  4. 4

    Limitations

    How the sample was chosen and whether it is representative

Comparing two samples.
  • Compare two samplesone measure of location and one of spread, in context

Which pair of measures should you use to compare two samples?

Try an exam question

A sample of n=50n=50 readings from the large data set has ∑x=800\sum x=800 and ∑x2=14000\sum x^2=14000. Find the mean and standard deviation, and decide whether a reading of 2828 is an outlier using the 2 standard deviation rule.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Which of the four variables is a continuous quantitative variable?

1 mark, marked against the mark scheme

Revision notes