Signal vs Noise Examples in Math

Start with the recap, study the fully worked examples, then use the practice problems to check your understanding of Signal vs Noise.

This page combines explanation, solved examples, and follow-up practice so you can move from recognition to confident problem-solving in Math.

Concept Recap

Signal versus noise describes the fundamental challenge of separating meaningful patterns (signal) from random, unpredictable variation (noise) in data — the central task of all statistical analysis.

Is this pattern real or just coincidence? The fundamental question of data analysis.

Read the full concept explanation →

How to Use These Examples

  • Read the first worked example with the solution open so the structure is clear.
  • Try the practice problems before revealing each solution.
  • Use the related concepts and background knowledge badges if you feel stuck.

What to Focus On

Core idea: Signal vs noise is the judgment of separating a meaningful pattern from random fluctuation in the same data.

Common stuck point: The procedure for signal vs noise is the easy part; the trap is calling any difference a real effect. Asking "Is the difference larger than the data's ordinary random fluctuation?" first is what keeps a correct-looking calculation from being attached to the wrong concept.

Sense of Study hint: Ask: Is the difference larger than the data's ordinary random fluctuation?

Worked Examples

Example 1

easy
A teacher tracks class average scores over 6 months: {65,67,64,68,82,85}\{65, 67, 64, 68, 82, 85\}. Identify the noise (random month-to-month variation) and the signal (meaningful trend) in this data.

Answer

Noise: ±3-point fluctuations (months 1–4). Signal: 15-point jump in months 5–6.

First step

1
Months 1–4 scores: {65,67,64,68}\{65, 67, 64, 68\} — fluctuate around ~66; this is noise (random variation)

Full solution

  1. 2
    Months 5–6 scores: {82,85}\{82, 85\} — a sudden jump to ~83
  2. 3
    Signal: the large increase in months 5–6 is a genuine shift in performance, not random noise
  3. 4
    Distinguish: small fluctuations (±3\pm 3 points) are noise; a jump of 15+ points is a signal worth investigating
The signal is the meaningful pattern we care about; noise is random variation. To detect signals, we compare the magnitude of an effect to typical random fluctuation. Effects much larger than typical noise are likely real signals.

Example 2

medium
A polling company surveys 100 people monthly. In January, 48% support Policy X. In February, 51%. Explain whether this 3% change is signal or noise, using standard error.

Example 3

medium
A survey of 400400 voters gives 52%52\% support, with a margin of error of ±5%\pm 5\%. A second survey of 400400 voters gives 48%48\% support. Should we conclude support has dropped?

Example 4

medium
A control-chart rule flags 'any single point more than 3σ3\sigma from the mean'. Mean =50=50, σ=4\sigma=4. A point reads 6363. Should it be flagged?

Example 5

hard
A scientist wants the SD of the mean to drop from 44 to 11. By what factor must she increase the sample size nn?

Example 6

challenge
You test 2020 random independent 'lucky charms' and one shows a statistically significant effect at p<0.05p<0.05. Is this likely a real signal? Explain in one sentence.

Practice Problems

Try these problems on your own first, then open the solution to compare your method.

Example 1

easy
You flip a coin 10 times and get 6 heads. Is this signal (the coin is biased) or noise (random variation)? Calculate the expected range of heads for a fair coin.

Example 2

hard
A radar system detects an object with signal strength 12 units. Background noise has mean 5 and SD 3. Calculate how many SDs the signal is above noise and determine if the object is detectable.

Example 3

easy
A thermometer reads 20.1,19.9,20.0,20.2,19.820.1, 19.9, 20.0, 20.2, 19.8 degrees for the same object. Is the underlying temperature (2020 degrees) the signal or the noise?

Example 4

easy
True or false: random noise can sometimes look like a pattern purely by chance.

Example 5

easy
Five repeated weight readings: 50.0,50.0,50.0,50.0,50.050.0, 50.0, 50.0, 50.0, 50.0 kg. How much noise is present?

Example 6

easy
A signal of 100100 has noise that averages out to 00. What is the average of many noisy readings expected to be?

Example 7

easy
Which is the signal in 'monthly sales rose steadily all year while bouncing up and down week to week'?

Example 8

easy
If you double the number of independent measurements, does the noise in their average increase, decrease, or stay the same?

Example 9

easy
A coin flipped 1010 times gives 77 heads. Is concluding 'this coin is biased' justified by signal here?

Example 10

easy
In the equation reading == signal ++ noise, what does subtracting the estimated signal leave?

Example 11

medium
A dataset of daily temperatures shows a clear seasonal cycle plus random daily wobble of about ±2\pm 2 degrees. A single day reads 33 degrees above the seasonal curve. Signal or noise?

Example 12

medium
Two studies test a drug. Study A: n=20n=20, effect not significant. Study B: n=20000n=20000, the same tiny effect is 'statistically significant.' Is the effect necessarily practically important?

Example 13

medium
A radio receives a 10001000 Hz tone buried in static. Averaging 100100 recordings nearly removes the static but keeps the tone. Why does averaging help?

Example 14

medium
A scatterplot of 200200 points shows a faint upward drift hidden in scatter. With only 88 of those points the drift is invisible. What does this say about sample size and signal detection?

Example 15

medium
Sensor readings: 4.9,5.1,5.0,25.0,5.04.9, 5.1, 5.0, 25.0, 5.0. Which value is most likely noise/error rather than signal, and what is the best signal estimate?

Example 16

medium
A weekly report flags any week whose sales differ from the prior week by more than ±1\pm 1 unit, but normal weekly noise is about ±5\pm 5 units. What is wrong with this alert rule?

Example 17

medium
You measure a length once as 7.37.3 cm. To report the true length with less noise, what should you do, and why?

Example 18

medium
A pattern appears in data, vanishes when the experiment is repeated, then does not return on a third try. Signal or noise?

Example 19

medium
You measure the same resistor 3 times: 99,101,10099, 101, 100 ohms. The labeled value is 100 ohms. Is the ±1\pm 1 variation signal or noise, and what is the best estimate of the true value?

Example 20

challenge
Readings are signal S=10S=10 plus noise with standard deviation σ=4\sigma=4. If you average nn independent readings, the noise SD of the mean is σ/n\sigma/\sqrt{n}. How many readings are needed so the mean's noise SD is at most 11?

Example 21

challenge
A detector with noise SD σ=2\sigma=2 uses the rule 'flag a signal if a reading exceeds 3σ3\sigma above baseline.' Baseline is 00. A reading of 5.55.5 arrives. Does the rule flag it, and is that consistent with calling values within 3σ3\sigma noise?

Example 22

challenge
Two analysts study the same noisy series. One reports a 'trend' from 55 points; the other waits for 500500 points and finds none. Explain, using signal vs noise, who is more likely correct and why.

Example 23

easy
A scoreboard tracks a player's accuracy each game: {70%,71%,69%,72%,70%}\{70\%, 71\%, 69\%, 72\%, 70\%\}. Is the variation here best described as signal or noise?

Example 24

easy
You roll a fair die 6 times and get a 6 twice. Is the 'two sixes' result signal (the die is loaded) or noise (random variation)?

Example 25

easy
True or false: averaging many noisy measurements of the same fixed quantity tends to reduce the noise in the estimate.

Example 26

easy
You read someone's daily step count for one week: {8000,8100,7900,8200,8050,7950,8000}\{8000, 8100, 7900, 8200, 8050, 7950, 8000\}. Does this person have a clear upward trend?

Example 27

easy
You flip a fair coin 100100 times and get 5454 heads. About how many heads would you expect if the coin were fair?

Example 28

medium
A scientist measures a quantity with noise SD σ=6\sigma=6. How many independent measurements must she average so the resulting mean has noise SD at most 11?

Example 29

medium
A website's daily visits look noisy from day to day but rise steadily over six months. Which part is the signal?

Example 30

medium
A class of 2525 scored a mean of 7474 on a quiz; another class of 2525 scored a mean of 7676. If individual-score SD is ≈10\approx 10, is a 22-point gap signal?

Example 31

medium
A new training method raises an athlete's average time by 0.020.02 seconds — but day-to-day times vary by ±0.3\pm 0.3 seconds. Is the change detectable from one workout?

Example 32

medium
Three measurements of the same length read 7.40,7.42,7.387.40, 7.42, 7.38 cm. Estimate the signal (true length) and the noise level.

Example 33

medium
Sensor readings: {10.1,9.9,10.0,10.2,9.8,50.0}\{10.1, 9.9, 10.0, 10.2, 9.8, 50.0\}. Which is most likely a noise spike rather than the signal, and what is the best signal estimate?

Example 34

hard
Signal S=20S=20 plus noise with SD σ=5\sigma=5. After averaging n=25n=25 independent readings, give the SD of the mean.

Example 35

hard
A signal-to-noise ratio (SNR) is defined as S/σS/\sigma. For S=15S=15 and σ=3\sigma=3, find SNR.

Example 36

hard
In year 1 a coach claims a 'breakthrough drill' based on 55 players improving. In year 2, 5050 players try the same drill and the average effect is zero. What probably happened in year 1?

Example 37

hard
Two analysts share the same dataset. Analyst A fits a 9-parameter curve through 10 points; Analyst B fits a straight line. Which is more likely confusing noise for signal?

Example 38

challenge
A weak periodic signal of amplitude 11 sits inside Gaussian noise with σ=4\sigma=4. After averaging nn independent observations of a single time point, you want SNR ≥3\ge 3. Find the minimum nn.
← Back to Signal vs Noise Practice Mode

Background Knowledge

These ideas may be useful before you work through the harder examples.

noisevariability