Data Representation and Interpretation for the TMUA

Categorical Data Representations

Categorical data, also known as qualitative data, consists of variables that can be sorted into distinct groups or categories, such as eye colour, car makes, or types of university degrees. For the TMUA, you must be proficient in five primary representations of categorical data.

Frequency and Two-Way Tables

A frequency table is the simplest way to summarise categorical data. It lists each category alongside a count of how many times that category occurs in the dataset.

A two-way table is used when we wish to compare two categorical variables simultaneously. For example, if we recorded the favourite subject (Maths, English, Science) and the year group (Year 12, Year 13) of 100 students, the two-way table would show the overlap between these variables. You must be able to calculate marginal totals (the sums of rows and columns) and use them to find missing values.

Example: Two-Way Table Construction

In a group of 50 students, 30 are in Year 12 and 20 are in Year 13. Of the Year 12 students, 12 prefer Mathematics. If 25 students in total prefer Mathematics, find how many Year 13 students prefer other subjects.

1. Calculate Year 13 Maths: $25 - 12 = 13$.
2. Calculate Year 13 Other: Total Year 13 is 20, so $20 - 13 = 7$.

Bar Charts and Pictograms

Bar charts represent the frequency of categories using the height of rectangular bars. Note that for categorical data, bars should be separated by gaps to signify that the data is not continuous. The height of each bar is equal to its frequency.

Pictograms use symbols or icons to represent data. Every pictogram must include a key defining the value of one symbol. For instance, if a symbol represents 10 units, a half-symbol represents 5 units. When interpreting pictograms, always check the key first to avoid scale errors.

Pie Charts

Pie charts show the proportion of a total that each category occupies. The size of each sector is proportional to the frequency of that category. To construct a pie chart, you must calculate the angle for each sector using the formula:

$\text{Sector Angle} = \frac{\text{Frequency of Category}}{\text{Total Frequency}} \times 360^\circ$

To interpret a pie chart without frequencies, you must measure the angle or know the percentage of the total. Remember that pie charts do not show absolute frequencies unless the total is known.

Discrete Numerical Data: Vertical Line Charts

When dealing with ungrouped discrete numerical data (numbers that take specific values, such as the number of siblings or shoe sizes), we use a vertical line chart. Unlike a bar chart, where the categories are words, the x-axis represents numerical values. Lines are used instead of bars because the data points are distinct: there is no such thing as '2.5 siblings', so a bar with width would incorrectly imply a range of values.

Time Series Data: Tables and Line Graphs

Time series data is a sequence of data points recorded at successive time intervals. This data is represented in tables and visualised using line graphs.

In a line graph, time is always plotted on the horizontal axis (x-axis), and the variable being measured is on the vertical axis (y-axis). Individual data points are plotted and then connected by straight lines. These graphs are essential for identifying trends (long-term upward or downward movements) and seasonality (regular patterns that repeat over specific time cycles).

Appropriate Use of Representations

Choosing the right tool depends on the data type and the goal of the analysis:

1. Use Pie Charts to compare proportions of a whole.
2. Use Bar Charts to compare absolute frequencies between categories.
3. Use Two-Way Tables to investigate relationships between two categorical variables.
4. Use Vertical Line Charts for discrete numerical counts.
5. Use Line Graphs to track changes or trends over time.