The Median is found by counting into the middle of the ordered list of leaves; if there are data points, the median is at the position.
The Mode or modal class is easily identified as the stem with the most leaves, or the specific leaf value that repeats most frequently within a stem.
The Range is calculated by subtracting the smallest value (first leaf of the first stem) from the largest value (last leaf of the last stem).
Quartiles are determined by finding the median of the lower and upper halves of the data set, allowing for the calculation of the Interquartile Range (IQR).
A back-to-back diagram is used to compare two related data sets (e.g., test scores for two different classes) using a single central column of stems.
Leaves for the first group are listed to the right of the stem, while leaves for the second group are listed to the left.
Crucially, the leaves on the left must increase in value as they move away from the central stem (from right to left) to maintain numerical order.
| Feature | Stem & Leaf Diagram | Histogram |
|---|---|---|
| Data Retention | Preserves every individual raw data point. | Groups data into bins; individual values are lost. |
| Data Size | Best for small to medium data sets (e.g., ). | Suitable for very large data sets. |
| Construction | Quick to draw by hand without complex calculations. | Requires calculating frequency density or bin widths. |
Unlike a simple list of numbers, the stem and leaf diagram provides an immediate visual representation of the distribution's skewness and modality.
The Key is Mandatory: Failing to provide a key often results in a loss of marks, as the reader cannot determine the actual magnitude of the data.
Check the Count: Always count the total number of leaves in your final diagram to ensure it matches the number of items in the original data set.
Full Value for Median: When identifying the median, remember to combine the stem and leaf; a common error is providing only the leaf digit as the answer.
Uniform Spacing: Keep the digits in the leaves aligned vertically; if one row has more digits but they are cramped, the visual 'shape' of the data will be misleading.