What is Dr. Zubair Khalid's research focus?

Dr. Zubair Khalid specializes in molecular virology, mRNA vaccine development, and computational biology, with a focus on avian pathogens like IBDV and Avian Reovirus.

Where is Dr. Zubair Khalid currently working?

Dr. Zubair Khalid is a Postdoctoral Research Associate at the University of Maryland (UMD), specifically within the Department of Animal and Avian Sciences.

How to Find Mode: Mean, Median, Mode Guide

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-20

The mode is the value that appears most often in a dataset. The mean is the arithmetic average, found by adding every value and dividing by the number of values, while the median is the middle value once the data are placed in order.

Those three sentences carry most of what a student needs for an exam, but they hide the part that actually matters in research: each measure answers a different question, and choosing the wrong one can make a real effect look like noise or make noise look like an effect. A single unusually large value can drag a mean far away from the typical case while leaving the median untouched. A mode can be the most useful number in the room when you are counting categories, and almost meaningless when you are measuring something continuous. This guide walks through how to find the mode, how it relates to the mean and median, and when each measure earns its place in a results section.

What Mean, Median, and Mode Actually Mean

Mean, median, and mode are the three standard measures of central tendency, a term that simply means a single number used to summarize where the center of a dataset sits. Each one defines "center" differently, so each one behaves differently when the data are skewed, when outliers appear, or when the values are categories rather than quantities.

Mode meaning in math

The mode is the most frequent value in a dataset. That is the whole definition. If you sort a list and count how many times each value occurs, the value with the highest count is the mode. A dataset can be:

Unimodal: exactly one value occurs most often.
Bimodal: two values tie for the highest frequency.
Multimodal: three or more values tie for the highest frequency.
No mode: every value occurs exactly once, so no value is more frequent than any other.

That last case trips up a lot of students. If a dataset is 4, 8, 15, 16, 23, 42, there is no mode. It is not zero, and it is not the average of the list. The correct answer is that the dataset has no mode.

Mean meaning in math

The mean, also called the arithmetic average, is the sum of all values divided by the count of values. For a dataset of n values, you add them all and divide by n. The mean uses every data point, which makes it informative but also makes it sensitive to extreme values. One outlier can shift the mean substantially.

Median meaning in math

The median is the middle value when the data are arranged in order from smallest to largest. If the dataset has an odd number of values, the median is the single middle value. If the dataset has an even number of values, the median is the average of the two middle values. Because the median depends only on position, not on magnitude, extreme values do not move it much. That property is called robustness.

Worked Example: One Dataset, Three Answers

Take the dataset:

2, 3, 3, 5, 7, 7, 7, 9

There are eight values, so n = 8.

Mode. Count the frequency of each value:

Value	Frequency
2	1
3	2
5	1
7	3
9	1

The value 7 appears three times, more than any other value. The mode is 7. This dataset is unimodal.

Median. The data are already in ascending order. With an even count of eight, the median is the average of the fourth and fifth values. The fourth value is 5 and the fifth value is 7, so the median is (5 + 7) / 2 = 6.

Mean. Add every value: 2 + 3 + 3 + 5 + 7 + 7 + 7 + 9 = 43. Divide by 8: 43 / 8 = 5.375.

So for this one dataset, the mode is 7, the median is 6, and the mean is 5.375. Three legitimate summaries of the same eight numbers, and they disagree by more than 1.5 units. That gap is not a mistake. It is the entire reason statisticians distinguish between these measures.

Now change one value. Replace the 9 with 100. The mode stays 7, the median stays 6, but the mean jumps to 43 / 8 + (100 - 9) / 8 = 16.75. A single extreme observation moved the mean by more than 11 units while the mode and median did not budge. This is the outlier sensitivity that makes the mean a poor summary for skewed data such as income, wait times, or viral load measurements.

How to Find the Mode Step by Step

Finding the mode is mechanical once you have a clean dataset. The steps below work for any list of values, whether they are numbers, categories, or binned ranges.

List every observation. Write out the full dataset, including repeats.
Sort the list from smallest to largest. Sorting is not strictly required, but it makes counting far easier and prevents missed values.
Tally the frequency of each distinct value. A frequency table with two columns (value, count) is the cleanest format.
Identify the highest frequency. Scan the count column and find the largest number.
Report every value tied at that frequency. If one value has the top count, the dataset is unimodal. If two tie, it is bimodal, and you report both. If three or more tie, it is multimodal.
Check the no-mode case. If the highest frequency is 1, meaning every value appears exactly once, the dataset has no mode.

A categorical example

Modes shine with categorical data. Suppose a clinic records blood types for ten patients: A, O, O, B, A, O, AB, O, A, B. The frequencies are O = 4, A = 3, B = 2, AB = 1. The mode is O. You cannot compute a mean or median for blood type, because the categories have no numerical order or magnitude. The mode is the only measure of central tendency available.

A bimodal example

Consider test scores: 70, 75, 75, 80, 85, 85, 90. The value 75 appears twice and 85 appears twice. Both share the highest frequency, so the dataset is bimodal with modes 75 and 85. Reporting only one of them would misrepresent the data. Bimodality often signals that two distinct groups are mixed in one sample, which is a finding worth investigating rather than smoothing over.

A multimodal example

Values 1, 1, 2, 2, 3, 3, 4 produce three values tied at frequency 2. The dataset is multimodal with modes 1, 2, and 3. In practice, three or more modes in a small dataset usually means the sample is too small or too heterogeneous for the mode to be a useful summary.

Comparison Table: When to Use Each Measure

Measure	Definition	Sensitivity to outliers	Works with categories	Best used when
Mean	Sum of values divided by count	High	No	Data are roughly symmetric with no extreme values, and you need a mathematically tractable summary
Median	Middle value of the ordered data	Low (robust)	No	Data are skewed or contain outliers, or you want the typical value
Mode	Most frequent value	None	Yes	Data are categorical, or you want the most common value or the peak of a distribution

The mean is the foundation of most parametric statistics, so it appears constantly in the literature even when the median would describe the data better. The median is the standard summary for skewed biomedical measurements such as anticholinergic burden scores, where a small number of patients with very high exposure would distort a mean [1]. The mode is the standard summary for nominal categories and for identifying the most common price, response, or rating.

Mode in Real Research Practice

The mode is not a classroom curiosity. It appears in published studies whenever the most frequent value is the quantity of interest.

A multicenter study of digital pathology adoption surveyed 54 of 66 pathologists about the utility of the transition to screen-based diagnosis. The median utility grade was 9 out of 10 at both sites, the mode was 10, and the mean was 8.5 [2]. Reporting all three gave readers the full picture: the typical pathologist rated the change very highly (median 9), the single most common rating was the maximum (mode 10), and the average was pulled slightly lower by a small number of less enthusiastic respondents (mean 8.5). No single measure told the whole story.

A study of tobacco pricing at a US Army installation calculated the mode, mean, and median price of each product to compare on-post and off-post retailers [3]. The mode was essential here because the Department of Defense policy required on-installation retailers to set prices equal to the most common community price. That is a policy written around the mode, not the mean. If the most common price of a pack of cigarettes off-post was $7.00, the on-post retailer was expected to match $7.00, not the average of all nearby prices.

A cross-sectional study of anticholinergic burden in older adults compared twenty established scales and seven experimental scales built using different mathematical functions, including maximum, mode, median, and mean, to see which best correlated with cognitive and functional impairment [1]. This is a direct example of researchers treating the choice of central tendency measure as an empirical question rather than a default.

How the Mode Is Used With Continuous Data

The mode has a well-known limitation: for truly continuous data, every value may be unique, so no mode exists. Measure the heights of ten people to the nearest millimeter and you will likely get ten distinct values. The mode is undefined.

The standard fix is binning, also called grouping. You divide the continuous range into intervals (bins) and count how many observations fall into each bin. The bin with the highest count is the modal class, and its midpoint is often reported as the mode. A histogram makes this visual: the tallest bar marks the modal bin.

Binning introduces a choice. Change the bin width and the modal class can change. A dataset with a bin width of 5 units might show one peak, while a bin width of 2 units might reveal two. This is why the mode is rarely reported for continuous variables unless the bins are pre-specified by convention or by the measurement instrument itself. When you see a mode reported for continuous data, check how the data were grouped.

For discrete data with a limited range, such as a 1-to-10 rating scale, the mode is perfectly well defined and often the most intuitive summary. The digital pathology survey is a good example: ratings on a 10-point scale are discrete, so the mode of 10 is meaningful and easy to interpret [2].

Mean, Median, and Mode in Study Design

The choice of summary measure should be made before data collection, not after, because it affects sample size calculations, the statistical tests you can run, and how you present results.

The mean pairs naturally with the standard deviation and with parametric tests that assume roughly normal distributions. The median pairs with the interquartile range and with nonparametric tests. The mode pairs with frequency tables and with categorical comparisons such as chi-square tests.

In clinical research, the median absolute error is a standard accuracy metric precisely because it is robust to outliers. A study comparing six intraocular lens calculation formulas in 212 eyes reported both mean absolute error and median absolute error for each formula, and the Barrett Universal II formula showed the lowest values on both metrics across most subgroups [4]. Reporting both let readers see that the formula's advantage was not driven by a handful of unusual eyes.

A porcine study of myocardial cold ischemia compared in-vivo and ex-vivo simulation methods and reported median values for physiological measurements, a common choice when animal-to-animal variability is high and sample sizes are small [5]. With eight pigs, a single atypical animal could dominate a mean.

The general rule that emerges from the literature is straightforward. Use the mean when the distribution is symmetric and you need a value that supports further arithmetic. Use the median when the distribution is skewed, when outliers are present, or when you want the typical case. Use the mode when the data are categorical, when the most common value is the policy or clinical target, or when you are describing the peak of a distribution.

Common Mistakes and Limitations

Reporting "no mode" as zero. A dataset with all unique values has no mode. Zero is a number, not an absence of a mode. These are different statements.

Reporting only one mode when two exist. If two values tie for the highest frequency, both must be reported. Silently choosing one is a data integrity problem.

Using the mode for continuous data without binning. Ungrouped continuous measurements usually produce all-unique values and therefore no mode. Bin the data first, and state the bin width.

Confusing the mode with the maximum. The mode is the most frequent value, not the largest value. In the dataset 2, 3, 3, 5, 7, 7, 7, 9, the maximum is 9 but the mode is 7.

Assuming the mean is always the best summary. For skewed data such as income, hospital length of stay, or antibody titers, the mean can sit far from any actual observation. The median is usually the better description.

Forgetting that the median requires ordering. The median is defined by position in the sorted list. Computing it from unsorted data without sorting is the most common arithmetic error in introductory statistics.

Treating the three measures as interchangeable. They answer different questions. A results section that reports only the mean for heavily skewed data is incomplete, and a reviewer will often ask for the median as well.

Ignoring the effect of sample size on the mode. In small samples, the mode is unstable. Add one observation and the mode can change entirely. The mode becomes more reliable as the sample grows, but it remains the least stable of the three measures for quantitative data.

Quick Review

The mode is the most frequent value. A dataset can be unimodal, bimodal, multimodal, or have no mode at all.
The mean is the sum divided by the count. It uses every value and is sensitive to outliers.
The median is the middle value of the ordered data. It is robust to outliers.
In the dataset 2, 3, 3, 5, 7, 7, 7, 9, the mode is 7, the median is 6, and the mean is 5.375.
The mode is the only measure of central tendency that works for categorical data.
For continuous data, bin the values first and report the modal class and the bin width.
Choose the measure before you collect data, and report it alongside a matching spread measure (standard deviation with the mean, interquartile range with the median).

Frequently Asked Questions

What is the mode if all values are different?

There is no mode. If every value in the dataset occurs exactly once, no value is more frequent than any other, so the dataset has no mode. Do not report zero or the mean as a substitute.

Can a dataset have more than one mode?

Yes. If two values tie for the highest frequency, the dataset is bimodal and both values are reported. If three or more tie, it is multimodal. Ties are common in small datasets and in data measured on coarse scales.

How do I calculate mean, median, and mode for the same dataset?

Sort the data, count frequencies to find the mode, find the middle position to get the median, and sum then divide by the count to get the mean. For 2, 3, 3, 5, 7, 7, 7, 9, the mode is 7, the median is 6, and the mean is 5.375.

Why is the median used instead of the mean for skewed data?

The median depends only on the position of values in the ordered list, so extreme observations do not move it. The mean uses every value's magnitude, so one very large or very small number can pull it far from the typical case.

Is the mode useful for continuous measurements?

Rarely, unless the data are grouped into bins first. Ungrouped continuous measurements usually produce all-unique values, which means no mode exists. When data are binned, the bin with the highest count is the modal class and its midpoint is often reported.

What is the difference between mode and median?

The mode is the most frequent value and works for categories. The median is the middle value of the ordered data and requires values that can be ranked. A dataset can have no mode but always has a median.