Exploring Central Tendency: Mean, Median, and Mode

Every research study-whether it’s a survey on social media habits, an audience rating analysis, or a content analysis of news coverage-generates data. And raw data, no matter how carefully collected, tells you very little on its own. The first step in making sense of it is finding the center: a single number that best represents what’s typical in your dataset. That’s exactly what measures of central tendency do. According to the Australian Bureau of Statistics, a measure of central tendency is a summary measure that attempts to describe a whole dataset with a single value representing the middle or centre of its distribution. The three primary tools for this are the mean, the median, and the mode-each offering a different angle on the same question: what is the typical value here?

Why central tendency matters in research
The mean: the arithmetic average
When the mean misleads: the outlier problem
The median: the middle value
The median’s strength with skewed data
The mode: the most frequent value
Unimodal, bimodal, and multimodal distributions
How the three measures relate to distribution shape
Choosing the right measure: a practical guide
Central tendency in communication research: real applications

Why central tendency matters in research

Before diving into the three measures, it’s worth understanding why they matter at all. The SAGE Encyclopedia of Communication Research Methods puts it well: knowing the center of a distribution gives researchers an important reference point by which to compare data points to one another. Without it, a dataset of even 50 responses is just a list of numbers. With it, patterns emerge, comparisons become possible, and findings become communicable.

Descriptive statistics-of which central tendency is a core component-serve as the starting point for any data analysis. They organize, simplify, and summarize data before more advanced inferential tests are applied. Central tendency measures are part of this foundation. Research published on PubMed confirms that mean, median, and mode are among the most consistently used measures in statistical analysis, crucial for reducing datasets to a single representative value. The challenge, as we’ll see, is choosing the right one for your data.

The mean: the arithmetic average

The mean-commonly called the average-is the most familiar of the three. You calculate it by adding all values in a dataset and dividing by the total number of values. Laerd Statistics describes it as a model of your data: it’s the value that produces the least error when predicting any single data point in the set. An important mathematical property is that the sum of all deviations from the mean always equals zero.

In research notation, the sample mean is written as x̄ (pronounced “x-bar”), while the population mean uses the Greek symbol μ (mu). The formula is straightforward: add all values (Σx) and divide by the number of observations (n).

For example, if you survey five people about their weekly social media usage and get responses of 10, 12, 8, 15, and 20 hours, the mean is (10+12+8+15+20) ÷ 5 = 13 hours. That single number summarizes the group’s typical usage.

When the mean misleads: the outlier problem

The mean’s biggest weakness is its sensitivity to extreme values, called outliers. Consider a workplace where eight employees earn between $12,000 and $18,000 annually, but two senior managers earn $90,000 and $95,000. As Laerd Statistics illustrates, the mean salary for this group comes out to $30,700-a figure that doesn’t accurately represent what most workers actually earn. The two high salaries are pulling the mean upward, distorting the picture.

This is a recurring problem in income, wealth, and audience data research, where a small number of extreme values can make the mean deeply unrepresentative. In such cases, the Australian Bureau of Statistics notes that the mean is more sensitive to outliers than either the median or the mode, and researchers must consider whether it’s the right tool for the job.

The median: the middle value

The median is the middle value when all data points are arranged in ascending or descending order. It divides the distribution in half-50% of observations fall below it, and 50% fall above. A peer-reviewed article in the Journal of Pharmacology and Pharmacotherapeutics describes the median as a “measure of location”-it tells you where the data sits, not what it averages out to.

For an odd number of values, the median is simply the middle score. For an even number of values, you take the mean of the two middle scores. For instance, in a dataset of 10 ordered scores-14, 35, 45, 55, 55, 56, 56, 65, 87, 89-the median falls between the 5th and 6th values: (55 + 56) ÷ 2 = 55.5.

The median’s strength with skewed data

The median’s key advantage is its resistance to outliers. Going back to the salary example: if one retirement-age employee in a dataset retires at 81 instead of 60, the ABS explains that this outlier shifts the mean noticeably upward, but leaves the median unchanged-because the middle value doesn’t move just because one extreme value changed.

Laerd Statistics reinforces this: when data is skewed (i.e., the distribution has a longer tail on one side), the mean gets dragged toward the tail, while the median retains its position as a more accurate marker of the typical value. This is why income distribution data almost always reports the median household income rather than the mean-a handful of billionaires would inflate the mean far beyond what most people actually earn.

The mode: the most frequent value

The mode is the value that appears most often in a dataset. Unlike the mean and median, it requires no calculation-just observation. Scribbr notes that the mode is the only measure of central tendency that can be used with nominal or categorical data-data that can’t be ranked or averaged.

This makes the mode particularly valuable in communication research. If you’re surveying audiences about their preferred news platform-TV, print, online, or radio-you can’t average those categories. But you can identify which one appears most frequently. That’s the mode, and it tells you the most popular choice.

Unimodal, bimodal, and multimodal distributions

A dataset with one clear most-frequent value is called unimodal. When two values tie for the highest frequency, the distribution is bimodal. More than two such values make it multimodal. The NIH-published article on central tendency notes that in a bimodal distribution, the taller peak is called the major mode and the shorter one the minor mode-a useful distinction when analyzing polarized audience data, for instance.

The mode’s limitation is equally important to understand. The ABS points out that in some distributions, the mode may not reflect the centre of the distribution accurately-a dataset where 54 appears three times but most values cluster around 57 will report a mode that’s misleadingly low. And in continuous data, the mode may not exist at all if no value repeats.

How the three measures relate to distribution shape

One of the most important concepts in choosing between these three measures is the shape of the distribution. Research published via PubMed (NIH) explains the relationship clearly: in a perfectly normal (symmetrical, bell-shaped) distribution, the mean, median, and mode are all identical. They converge at the same central point and are equally valid.

In a positively skewed (right-skewed) distribution-where a few very high values extend the right tail-the mean is pulled upward toward that tail, while the mode sits at the hump of the distribution and the median falls between them. The reverse applies in a negatively skewed (left-skewed) distribution.

Ledidi’s statistical resource summarizes the practical takeaway: for skewed distributions, the median is generally the better choice because the mean is influenced by extreme values and may not represent a “typical” value of the dataset. The median, drawn from the middle of the distribution, is not affected by those extremes.

Choosing the right measure: a practical guide

The choice between mean, median, and mode isn’t arbitrary-it depends on your data type and distribution. Laerd Statistics’ FAQ on central tendency outlines the key decision points:

Use the mean when your data is numerical (continuous or discrete), normally distributed, and free of significant outliers. It uses every data point in its calculation, making it the most statistically powerful option when conditions allow.

Use the median when your data is numerical but skewed, or when outliers are present. It’s also the preferred measure for ordinal data-where values have a meaningful order but the intervals between them aren’t equal, such as Likert-scale responses in surveys.

Use the mode when your data is categorical or nominal. It’s the only measure that works with non-numerical data. It’s also useful when you specifically want to identify the most common category, response, or behavior in a dataset-such as the most-watched TV genre in a viewership survey.

Scribbr adds a useful rule of thumb: for normally distributed data, all three measures give you the same answer, so any can be used. For skewed distributions, the median is the safest default because it’s unaffected by extreme values or asymmetric distributions.

Central tendency in communication research: real applications

These aren’t abstract concepts-they directly shape how communication researchers report findings. The SAGE Encyclopedia of Communication Research Methods notes that people use central tendency intuitively every day when judging what’s “normal”-and researchers formalize that intuition into rigorous statistical practice.

In a media consumption survey, the mean might tell you the average number of hours respondents spend watching streaming content per week-useful when the data is roughly normal and no extreme outliers exist. The median becomes essential when reporting on income levels or ad spend data, where a few very high values would distort a simple average. The mode is the go-to when you’re reporting categorical preferences-which social media platform people use most, which news format they prefer, or which time slot draws the most viewers.

Health and demographic researchers similarly flag a common pitfall: using the mean for skewed distributions can produce misleading conclusions, particularly for variables like income, wealth, or any measure where a small group of extreme cases pulls the average far from what most people experience. The same caution applies in communication research when working with engagement metrics, reach figures, or viral content data-where outlier posts or accounts can dramatically inflate the mean.

Understanding when each measure applies-and what it conceals as well as reveals-is what separates competent data interpretation from misleading reporting. A single statistic summarizes hundreds of data points, which is enormously useful. But it’s only reliable when you’ve chosen the right summary for the right kind of data.

What do you think? When a news report cites the “average” income or the “average” screen time, which measure of central tendency do you think they’re using-and does it matter which one they’ve chosen? If a dataset on audience preferences is heavily skewed, what risks does a researcher take by reporting only the mean?

How useful was this post?

Click on a star to rate it!

Average rating 0 / 5. Vote count: 0

No votes so far! Be the first to rate this post.

References

PDF 📄

Table of Contents

Why central tendency matters in research 🔗

The mean: the arithmetic average 🔗

When the mean misleads: the outlier problem 🔗

The median: the middle value 🔗

The median’s strength with skewed data 🔗

The mode: the most frequent value 🔗

Unimodal, bimodal, and multimodal distributions 🔗

How the three measures relate to distribution shape 🔗

Choosing the right measure: a practical guide 🔗

Central tendency in communication research: real applications 🔗