Secondary 2 Mean, Median and Mode is the canonical eduKateSingapore guide to measures of central tendency: how to calculate them, how to choose between them, how frequency tables change the calculation, how outliers affect interpretation, and how to reverse a mean problem to recover a missing value. For students searching mean median mode Sec 2, averages, frequency tables, missing mean, weighted mean or statistics, the important question is not merely “which formula?” It is what feature of the data does this summary represent, and when does it become misleading?
The mean uses every value and is sensitive to extreme observations. The median depends on order and is often more robust to outliers. The mode identifies the most frequent value or category. These summaries answer different questions. A high-quality Secondary Mathematics response should calculate accurately and interpret what the result means in the dataset.
This page belongs to the Secondary Mathematics Topic Library. It extends Sec 1 Data Handling and links outward to the broader What Is Statistics? owner for variation, inference and advanced statistical ideas beyond the school-level scope here.
Quick answer: three different centres
- Mean: total of all values ÷ number of values.
- Median: middle value after ordering.
- Mode: most frequently occurring value or category.
There is no rule that one of these is always “best”. Choice depends on the data and the question.
Mean: balance point of the data
For values 4, 7, 8, 11, 15: total = 45, count = 5, mean = 9.
The mean uses every observation. Changing one value changes the mean, even if the middle position remains unchanged.
Median: the ordered middle
For 4, 7, 8, 11, 15, the median is 8.
For an even number of observations, average the two middle values after ordering.
Example: 2, 5, 9, 14 → median = (5+9)/2 = 7.
Mode: the most common
For 3, 4, 4, 5, 5, 5, 7, the mode is 5.
A dataset may have no mode, one mode or more than one mode. For categorical data such as favourite transport type, mode may be useful even when mean is meaningless.
The first comparison: outliers
Data: 5, 6, 6, 7, 30.
Mean = 54/5 = 10.8.
Median = 6.
Mode = 6.
The value 30 pulls the mean upward. Median and mode remain near the cluster of four lower values.
When mean is useful
- numerical data where every value should contribute
- datasets without severe outliers
- total can be meaningfully redistributed conceptually
- later calculations require an arithmetic average
When median may be more representative
- income or property data with extreme high values
- skewed distributions
- ordered data where the middle position is meaningful
- small datasets where one unusual value would distort mean
When mode matters
- most common shoe size
- most selected category
- most frequent score
- categorical data where mean/median may not apply
Mean from a frequency table
If a value appears several times, multiply each value by its frequency before summing.
Mean = Σ(value × frequency) ÷ Σfrequency.
This is not a new average; it is simply a compressed way to total repeated observations.
Worked frequency-table example
Scores 1,2,3,4 have frequencies 2,3,4,1.
Total score = 1×2 + 2×3 + 3×4 + 4×1 = 2+6+12+4 = 24.
Total frequency = 10.
Mean = 24/10 = 2.4.
Median from a frequency table
Do not expand every value unnecessarily. Use cumulative position.
With 10 observations, the median lies between the 5th and 6th values.
Using frequencies 2,3,4,1 for values 1,2,3,4: positions 1–2 are 1, positions 3–5 are 2, positions 6–9 are 3, position 10 is 4. Median = (2+3)/2 = 2.5.
Mode from a frequency table
The mode is the value with the highest frequency. In the same table, value 3 has frequency 4, so mode = 3.
Reverse mean: finding a missing value
If the mean of 6 numbers is 12, the total must be 72.
If five known values total 59, the missing value is 72−59=13.
Reverse-mean problems are total reconstruction problems.
Worked missing-value example
The mean of 8 test scores is 14. Seven scores total 93. Find the eighth.
Required total = 8×14 = 112.
Missing score = 112−93 = 19.
Changing one data value
If one observation increases by 10 in a set of 5 values, the total increases by 10 and mean increases by 10/5=2.
This “change in total ÷ number of observations” idea is faster than recomputing everything.
Adding a new observation
A class has 10 scores with mean 68. A new score of 88 is added.
Old total = 680.
New total = 768.
New count = 11.
New mean = 768/11 ≈ 69.82.
Removing an observation
If an observation is removed, subtract it from the old total and reduce the count by one. The same total-count architecture controls the problem.
Combined mean
Two groups cannot usually be combined by averaging their means directly unless they have the same size.
Group A: 20 students, mean 70 → total 1400.
Group B: 30 students, mean 80 → total 2400.
Combined mean = (1400+2400)/(20+30)=3800/50=76.
The simple mean of 70 and 80 is 75, which would be wrong because the groups are unequal in size.
Weighted mean intuition
Combined mean is a weighted mean: larger groups contribute more weight. Frequency tables use the same idea, with frequency acting as weight.
This is why “average of averages” is dangerous unless the underlying weights are equal.
Mean as equalisation
The mean can be imagined as redistributing the total equally across all observations.
For 2,5,8, the total is 15. Equal redistribution gives 5,5,5. Mean=5.
This interpretation helps explain why mean is tied to the total.
Median and position
Median ignores the exact size of most values; it depends primarily on ordering and middle position. Changing the largest observation from 20 to 200 may not change the median at all.
That is why median behaves differently from mean under outliers.
Comparing two datasets with the same mean
Dataset A: 5,5,5,5,5. Mean=5.
Dataset B: 1,3,5,7,9. Mean=5.
Same mean, very different spread. A measure of centre alone does not describe variability.
This is an important bridge into later statistics.
Choosing the right average in context
Salaries
A small number of very high salaries can pull the mean upward. Median may better describe the middle employee.
Most common product size
Mode can be useful because the practical question is which size occurs most often.
Test performance
Mean can summarise total class performance, but median and distribution may reveal whether a few extremes dominate.
Misleading average claims
“The average is 80” is incomplete unless we know which average, what dataset, how it was collected and whether unusual values matter.
Mathematics should preserve context, not only compute a number.
The central-tendency error taxonomy
- Mean denominator error: divides by wrong count.
- Median order error: finds middle before sorting.
- Even-count error: chooses one middle value instead of averaging two.
- Mode error: picks largest value rather than most frequent.
- Frequency error: ignores repeated counts.
- Combined-mean error: averages group means without weights.
- Interpretation error: uses mean despite a severe outlier without comment.
The four-layer diagnosis
- Calculation: can the student compute each measure?
- Representation: can the student read a frequency table?
- Selection: can the student choose an appropriate measure?
- Interpretation: can the student explain how outliers or sample context affect meaning?
Practice set
Practice 1
Question: Data 4,6,8,10,12: mean
Answer: 8
Practice 2
Question: Data 4,6,8,10,12: median
Answer: 8
Practice 3
Question: Data 2,2,3,4,4,4,7: mode
Answer: 4
Practice 4
Question: Mean of 5 values is 18: total
Answer: 90
Practice 5
Question: Mean of 6 values is 12; five total 58
Answer: missing value 14
Practice 6
Question: Group A n=10 mean 60; Group B n=20 mean 75
Answer: combined mean 70
Transfer task: one dataset, three summaries
Take 3,4,4,5,24. Compute mean, median and mode. Then answer three different questions:
- Which measure uses every observation?
- Which best resists the outlier 24?
- Which identifies the most common observed value?
The exercise teaches selection, not only calculation.
Frequently asked questions
What is the mean?
Total of all observations divided by the number of observations.
Why must data be ordered for the median?
Median is defined by position in the ordered dataset.
Can there be more than one mode?
Yes, if multiple values tie for highest frequency.
Why can’t I average two group means directly?
Unless group sizes are equal, each mean carries a different weight.
Which measure is best with outliers?
Median is often more robust, but the appropriate measure depends on the context and question.
Where does this sit in Atlas?
This is the canonical Sec 2 Mean, Median and Mode owner under Secondary Mathematics Topic Library.
The final Statistics rule
Calculate the centre, then ask whether that centre deserves to represent the data. Mean, median and mode are not rival formulas; they are different summaries with different sensitivities and uses.