Quantitative Data Graphs

Quantitative data consists of numerical measurements that can be ordered and compared mathematically — things like heights, test scores, temperatures, or reaction times. Before calculating any statistics, a data analyst's first job is to visualise the distribution: What values are typical? Are values spread out or clustered? Is the shape symmetric, skewed, or unusual? Three foundational graphical tools — dotplots, histograms, and stemplots — each reveal these features in a slightly different way, and choosing the right one depends on the size of your dataset and the level of detail you need. A distribution's shape is described by its modality (how many peaks, or modes, it has), its symmetry or skewness, and the presence of outliers — individual values that fall far from the bulk of the data. A symmetric, bell-shaped distribution has roughly equal tails on both sides. A right-skewed (positively skewed) distribution has a long tail pulling toward larger values, while a left-skewed distribution has a long tail toward smaller values. Identifying these features from a graph is a core skill in statistics because they influence which summary statistics are most appropriate. This article walks through the construction and interpretation of all three graph types, illustrates their strengths and weaknesses with a real worked example, and connects them to the broader goal of exploring data before drawing conclusions.

Key takeaways
  • Dotplots — One Dot per Data Value
  • Histograms — Frequency by Interval
  • Stemplots (Stem-and-Leaf Plots) — The Hybrid Approach
  • Describing a Distribution: SOCS

Dotplots — One Dot per Data Value

A dotplot places each individual data value as a dot above a number line. Identical values are stacked vertically. Dotplots are ideal for small datasets (roughly n ≤ 30) because every single observation remains visible. To draw one: (1) Draw a horizontal axis scaled to cover the full range of data. (2) For each data value, plot one dot directly above its position; stack dots for repeated values. (3) Label the axis and give the graph a title. Reading a dotplot: gaps reveal possible outliers; clusters show where values concentrate; the overall spread from leftmost to rightmost dot is the range.

Histograms — Frequency by Interval

A histogram divides the data's range into equal-width intervals called bins (or classes) and draws a bar whose height represents the count (frequency) or relative frequency (proportion) of values in each bin. Unlike a bar chart for categorical data, histogram bars touch each other because the data is continuous. Construction steps: (1) Choose 5–15 bins; a common rule is to use approximately √n bins. (2) Find the bin width: width = (max − min) / number of bins, rounded to a convenient number. (3) Count how many data values fall in each bin (a value on a boundary goes into the higher bin by convention). (4) Draw bars of appropriate height; bars must be adjacent. The area of each bar is proportional to frequency, so if you use relative frequency on the y-axis, all bar areas sum to 1. Histograms are best for larger datasets (n > 30) where individual values matter less than the overall shape.

Stemplots (Stem-and-Leaf Plots) — The Hybrid Approach

A stemplot preserves the actual data values while still showing the distribution's shape — giving it a unique advantage over histograms. Each data value is split into a stem (leading digit(s)) and a leaf (the final digit). Construction: (1) List all stems in order in a vertical column. (2) For each data value, write its leaf next to the appropriate stem. (3) Leaves are written in order from smallest to largest. (4) Include a key, e.g. '3 | 7 means 37'. A back-to-back stemplot uses the same stem column with one dataset's leaves growing left and another's growing right — perfect for comparing two groups. Stemplots work best for datasets of size roughly 10–50.

Describing a Distribution: SOCS

Regardless of which graph you build, always describe a quantitative distribution using SOCS: Shape (symmetric, skewed left, skewed right, unimodal, bimodal), Outliers (any unusual values?), Center (roughly where is the middle of the data?), and Spread (how variable are the values?). For example: 'The distribution of exam scores is roughly symmetric and unimodal with a center near 74 points. The scores range from 48 to 98, a spread of 50 points, with one potential low outlier at 48.' This structured description turns a visual impression into a precise, communicable summary.

The Key Equation

Bin Width=max⁡(x)−min⁡(x)k\text{Bin Width} = \frac{\max(x) - \min(x)}{k}
  • \max(x) — Maximum value (same units as data): The largest observed data value in the dataset.
  • \min(x) — Minimum value (same units as data): The smallest observed data value in the dataset.
  • k — Number of bins (dimensionless): The chosen number of equal-width intervals; a common choice is k ≈ √n where n is the sample size.

Worked Example

Problem. A teacher records the following 15 quiz scores (out of 20) for her class: 12, 14, 15, 15, 16, 17, 17, 17, 18, 18, 19, 19, 19, 20, 9. (a) Construct a stemplot. (b) Determine an appropriate bin width for a histogram using k = 4 bins. (c) Describe the distribution using SOCS.

Bin width=20−94≈3,Median=17,Range=11\text{Bin width} = \frac{20-9}{4} \approx 3, \quad \text{Median} = 17, \quad \text{Range} = 11

Answer. Stemplot shows a left-skewed, unimodal distribution. Bin width ≈ 3 gives 4 histogram bars. The distribution is left-skewed with center near 17, spread of 11, and one outlier at 9.

Figure
Diagram
Quiz Scores: Dotplot, Histogram, and StemplotDotplot (n=15 scores)91214151617181920Histogram (bin width = 3)[9,12)[12,15)[15,18)[18,21)0246FreqScoreStemplot0 | 91 | 2 4 5 5 6 7 7 7 8 8 9 9 92 | 0Key: 1|7 = score of 17
Origins · 1 minJohn Tukey and how this was discovered
JT
John Tukey
1915–2000 · American

John Wilder Tukey was a mathematician and statistician at Princeton University and Bell Labs who essentially invented the field of Exploratory Data Analysis (EDA). He introduced the stemplot (stem-and-leaf plot) and the boxplot as practical tools for visualising data before formal modelling. His landmark 1977 book 'Exploratory Data Analysis' changed how statisticians approach datasets by emphasising looking at data graphically first. Tukey coined the word 'bit' (short for binary digit) in 1947 and also invented the term 'software' in a 1958 paper — two of the most fundamental words in computing — making him as influential in computer science as in statistics.

  • Measures of Center (Mean and Median) — After visualising a distribution with a graph, the next step is to quantify its center. The shape revealed by a histogram or stemplot tells you whether the mean or median is a more appropriate measure.
  • Measures of Spread (Range, IQR, Standard Deviation) — The spread visible in dotplots and histograms is formally measured by the range, interquartile range, and standard deviation — concepts that build directly on graphical exploration.
  • Categorical Data Graphs (Bar Charts and Pie Charts) — Understanding the contrast between graphs for quantitative data (histograms, dotplots) and categorical data (bar charts, pie charts) is essential for choosing the correct display.
  • Boxplots and Outliers — Boxplots are another key graphical tool for quantitative data that make outliers and the five-number summary explicit, complementing the detail provided by histograms and stemplots.