Quantitative data
Quantitative data is information expressed as numbers that can be counted or measured, such as a rating, a price, a count of visits or a time in seconds. It answers how many, how much and how often, and it can be analyzed statistically.
How it works
Quantitative data comes in two types. Discrete data is counted in whole units: the number of purchases or support tickets. Continuous data is measured on a scale that can be divided finely: time on task, spend, temperature. Rating-scale answers, such as a 1 to 5 satisfaction score, are often treated as quantitative, though strictly they are ordinal.
Sources include surveys with closed questions, analytics, experiments and transaction records. Because the numbers can be summarized with means, medians and percentages, compared across groups and tested for significance, quantitative data suits questions of size, frequency and change over time. Counted values can be averaged as readily as measured ones, as in 2.4 support tickets per customer a year. In a normal distribution the mean and median coincide, while skew pulls the mean toward the longer tail and a few extreme values distort it, so the median suits data with outliers.
A worked example
Checkout speed averages 3.2 out of 5 in a grocery chain's survey of 3,000 shoppers, and 41 percent rate it 1 or 2. Store data adds a continuous measure, median queue time, at 6.5 minutes on weekend afternoons against 2 minutes on weekday mornings.
Related
Automate this in Pulse
Ratings, counts and closed-question answers from a Pulse survey sit in the same study as the interviews that explain them.