30  Descriptive Analytics

Descriptive analytics is the foundation of data analysis, focusing on summarizing historical data to extract meaningful insights (John W. Tukey, 1977). It helps organizations understand past trends, identify patterns, and support decision-making through statistical and visual techniques. Unlike predictive or prescriptive analytics, which look ahead, descriptive analytics primarily answers “what happened” based on past data — last season’s yields, this month’s rainfall, the price trend at the local mandi.

30.1 Overview of Descriptive Analytics

Descriptive analytics involves processing raw data and presenting it in a meaningful way through summary statistics, visualizations, and structured reports. In agribusiness, it’s the everyday work of turning a season’s field records into numbers a cooperative board, a lender, or an extension officer can act on.

30.1.1 Key Characteristics

  • Summarizes past data to provide insight into historical performance.
  • Uses statistical measures to describe a dataset’s characteristics.
  • Employs data visualization to represent trends and patterns effectively (see Data Visualization in R).
  • Forms the basis for the predictive and prescriptive analytics covered later in this book.

30.1.2 Examples of Descriptive Analytics in Action

  • Agribusiness: Summarizing a cooperative’s average yield per farm across a season.
  • Agronomy: Tracking how rainfall has varied across a growing season.
  • Market analysis: Describing the typical mandi price for a commodity, and how much it fluctuates week to week.
  • Farm input supply: Reporting the most commonly purchased fertilizer type across a district.

30.2 Measures of Central Tendency

Measures of central tendency are statistical metrics that summarize the center point or typical value of a dataset. They’re crucial in data analysis because they provide a simple, single-number summary of a sample. The three main measures are the mean, median, and mode — each offering a different lens on the same dataset.

Central tendency measures identify the value data points cluster around, offering insight into a dataset’s overall behavior.

  • Mean: the arithmetic average, calculated by summing all observations and dividing by their count.
  • Median: the middle value in an ordered dataset, splitting it into two equal halves.
  • Mode: the most frequently occurring value(s) in a dataset.

30.2.1 Mean

The mean, often called the average, is calculated by adding all the numbers in a dataset and dividing by the count of those numbers. It’s the most common measure of central tendency.

  • Formula: \(\text{Mean} = \frac{\sum_{i=1}^{n} x_i}{n}\), where \(x_i\) represents each value in the dataset and \(n\) is the number of values.
  • Sensitive to outliers: the mean is influenced by unusually high or low values, which can skew the result.
  • Used for: interval and ratio levels of measurement.

Example

Consider the yield (tons/ha) recorded across five farms in a cooperative: 3.2, 3.8, 2.9, 4.1, 3.2.

To calculate the mean:

  1. Sum all the values: \(3.2 + 3.8 + 2.9 + 4.1 + 3.2 = 17.2\).
  2. Divide by the number of farms: \(17.2 / 5 = 3.44\).

So the mean yield is 3.44 tons/ha.

Calculation in R

Application

The mean is the natural summary for a cooperative’s average yield, average rainfall, or average procurement price. It provides a quick snapshot of overall performance — but it can be misleading if a few farms performed exceptionally well or badly compared to the rest.

30.2.2 Median

The median is the middle value in a dataset once the values are sorted. With an even number of observations, the median is the average of the two middle values.

  • Represents: the 50th percentile of the dataset.
  • Not sensitive to outliers: unlike the mean, the median isn’t pulled around by outliers, making it a better summary for skewed distributions.
  • Used for: ordinal, interval, and ratio levels of measurement.

Example

Using the same five yields — 3.2, 3.8, 2.9, 4.1, 3.2 — sort them first: 2.9, 3.2, 3.2, 3.8, 4.1.

The median is the middle value: 3.2 tons/ha.

If a sixth farm is added with a yield of 3.6, the sorted set becomes 2.9, 3.2, 3.2, 3.6, 3.8, 4.1. With six values, the median is the average of the two middle ones: \((3.2 + 3.6) / 2 = 3.4\).

Calculation in R

Application

The median is valuable when a dataset has a few extreme values — a district’s median farm income is often more representative than the mean, which a handful of very large or very small landholdings can pull in either direction.

30.2.3 Mode

The mode is the value that appears most frequently in a dataset. A dataset may have one mode (unimodal), more than one mode (bimodal or multimodal), or no mode at all if every value is unique.

  • Useful for: identifying the most common category or value in a dataset.
  • Can be applied to: nominal, ordinal, interval, and ratio data — it’s the only measure of central tendency that works with nominal (category) data.
  • Limitations: the mode may not represent the dataset well when there are many unique values or several competing modes.

Example 1 (unimodal)

In the yields 3.2, 3.8, 2.9, 4.1, 3.2, the mode is 3.2 tons/ha — it appears more often than any other value.

Example 2 (bimodal)

In a different cooperative’s yields — 2.5, 2.5, 3.0, 3.5, 3.5, 4.0 — the dataset is bimodal: both 2.5 and 3.5 appear twice.

Example 3 (no mode)

If every farm’s yield is unique — 2.5, 3.0, 3.5, 4.0, 4.5 — there is no mode, since no value repeats.

Calculation in R

table(farm_yield) builds a frequency count of each value, which.max(...) finds the most frequent one, names(...) extracts it, and as.numeric(...) converts it back from a character label to a number.

Application

The mode is used to identify the most commonly grown crop variety in a district, the most frequently used irrigation type among a cooperative’s members, or the most common soil type across a set of surveyed plots — all cases where the “typical” value is a category rather than something to average.

30.2.4 When to Use Each Measure

  • Mean: ideal when there are no major outliers and every value matters — for example, average monthly rainfall across a season.
  • Median: best for skewed data or when outliers are present, such as farm income, where a handful of very large landholdings can pull the mean upward.
  • Mode: useful for categorical data or to find the most common value — for example, the most common crop variety planted across a district.

30.2.5 Choosing the Right Measure

  • Symmetrical distributions: the mean is typically preferred, since it takes every value into account.
  • Skewed distributions: the median is better when outliers are present, since it isn’t pulled by extreme values.
  • Categorical data: the mode is the only sensible choice when identifying the most common category.

30.2.6 Importance

Each measure of central tendency offers a different insight. The mean gives a mathematical average, the median gives a midpoint unaffected by outliers, and the mode reveals the most frequent value. Choosing the right one depends on the nature of the data, its distribution, and the question being asked — understanding all three sharpens the ability to summarize and act on agricultural data.

Summary

Concept Description
Overview of Descriptive Analytics
Key Characteristics Descriptive analytics summarizes past data using statistics and visualization, and underpins predictive and prescriptive analytics
Examples of Descriptive Analytics in Action Summarizing average yield, tracking seasonal rainfall, describing typical mandi prices, and reporting the most common fertilizer type are all descriptive analytics
Measures of Central Tendency
Mean The arithmetic average of a dataset; sensitive to outliers
Median The middle value of a sorted dataset; not sensitive to outliers
Mode The most frequently occurring value; the only central tendency measure suited to categorical data
When to Use Each Measure Choose the mean for symmetric data, the median for skewed data or outliers, and the mode for categorical data