Math III · S-ID.4

Using Mean and Standard Deviation to Fit Normal Distributions and Estimate Population Percentages

Normal distributions let students model many bell-shaped data sets and estimate percentages using mean, standard deviation, and technology.

Concept Statistics and Probability
Domain Interpreting Categorical and Quantitative Data
Read time 5 minutes

What this learning objective is really asking you to learn

This objective asks students to use mean and standard deviation to fit normal distributions and estimate population percentages with technology. A normal distribution is a bell-shaped, symmetric distribution described by its mean and standard deviation. The mean locates the center. The standard deviation measures typical spread.

Many real data sets are not exactly normal, but some are approximately normal: heights within a similar population, measurement errors, certain test scores, manufacturing variation, and biological measurements under stable conditions. When a distribution is approximately normal, the normal model can estimate what percentage of values fall below, above, or between certain values.

For example, suppose adult heights in a population are approximately normal with mean 170 cm and standard deviation 8 cm. A height of 178 cm is one standard deviation above the mean. A height of 154 cm is two standard deviations below the mean. Using the empirical rule, about 68% of values lie within one standard deviation of the mean, about 95% within two, and about 99.7% within three. Technology can provide more precise percentages.

The objective asks students to use technology because normal percentage calculations often require cumulative areas under a curve. Calculators, spreadsheets, and statistical software can compute these areas. But students still need to understand what the technology is calculating: area under a normal curve represents proportion or probability.

This objective is about model fitting and interpretation. Students must decide whether a normal model is appropriate, identify mean and standard deviation, compute or estimate percentages, and interpret them in context.

Why students should learn this math

Students should learn normal distributions because they are one of the most important models in statistics. Normal models appear in measurement, quality control, psychology, education, biology, manufacturing, polling, and scientific error analysis. Even when raw data are not perfectly normal, normal models often approximate distributions or sampling distributions.

Mean and standard deviation become more meaningful in normal models. The mean is the balance point and center of symmetry. The standard deviation sets the scale. A value's distance from the mean can be measured in standard deviations, which allows comparison across different units and contexts.

For example, a score of 85 on one test may be excellent if the mean is 70 and standard deviation is 5, but ordinary if the mean is 82 and standard deviation is 10. The raw score alone is not enough. Standard deviation gives context.

Normal models also support percentage estimates. If a product dimension is normally distributed with mean 10 mm and standard deviation 0.2 mm, a manufacturer can estimate what percent of parts fall within tolerance. If test scores are approximately normal, a school can estimate percentiles. If measurement errors are normal, scientists can evaluate unusual observations.

The “why” is that normal distributions give a powerful way to connect center, spread, and percentages. They turn raw measurements into interpretable positions within a distribution.

The historical machinery: the bell curve and error

The normal distribution became important through the study of measurement errors and repeated observations. When many small independent sources of variation contribute to a measurement, the resulting errors often form a bell-shaped pattern. This idea was formalized through probability theory and the central limit theorem.

Mathematicians such as De Moivre, Laplace, and Gauss contributed to the development of the normal curve. It is sometimes called the Gaussian distribution because of Gauss's work on errors in astronomy and measurement.

The normal distribution became central because it is mathematically tractable and appears widely. But it is also overused. Not every data set is normal. Income distributions, home prices, waiting times, and many biological or social variables can be skewed. Students should learn normal models as useful tools, not universal truth.

The historical lesson is that normal distributions model certain kinds of variation well, especially symmetric bell-shaped variation and many sampling distributions.

Where this fits in the big map of mathematics

This objective follows inference and report evaluation. It gives students a specific distribution model for estimating percentages.

It connects to mean and standard deviation from earlier statistics objectives.

It connects to z-scores. A z-score measures how many standard deviations a value is from the mean:

\[z = (x - mean) / standard deviation\].

It connects to technology because normal curve areas are usually computed digitally.

It connects to sampling distributions and inference. Normal models often approximate sample statistic behavior.

It connects to real-world quality control, education, measurement, and science.

The big-map role is distribution modeling. Students learn to use a theoretical curve to estimate real percentages.

How to execute the skill technically

Use this routine:

  1. Check whether a normal model is reasonable: roughly symmetric, bell-shaped, no extreme skew or outliers.
  2. Identify mean μ and standard deviation σ.
  3. Convert values to z-scores if useful.
  4. Use technology or empirical rule to find area/proportion.
  5. Interpret the proportion in context.

Example: Test scores are approximately normal with mean 75 and standard deviation 10. Estimate the percent scoring above 90.

Compute z-score:

\[z = (90 - 75)/10 = 1.5\].

Use technology to find area above \(z=1.5\). This is about 0.0668.

So about 6.7% of students score above 90.

Example: Find the percent between 65 and 85.

z-scores:

\[(65-75)/10=-1\].
\[(85-75)/10=1\].

By empirical rule, about 68% of values lie within one standard deviation, so about 68% score between 65 and 85.

Worked example: manufacturing tolerance

A machine fills bottles with amounts approximately normally distributed with mean 500 mL and standard deviation 4 mL. Bottles are acceptable if they contain between 492 mL and 508 mL. Estimate the percent acceptable.

Compute z-scores:

For 492:

\[z=(492-500)/4=-2\].

For 508:

\[z=(508-500)/4=2\].

By the empirical rule, about 95% of bottles fall within two standard deviations. So about 95% are acceptable.

Technology gives a more precise value of about 95.45%.

This example shows how normal models support quality control.

Technology and interpretation

Technology may output an area such as 0.9545. Students must interpret it as about 95.45% of the population or process outcomes under the model. The area under the curve is a proportion. It is not a count unless multiplied by a total number.

If 10,000 bottles are produced, about \(0.9545(10000)=9545\) are expected to be acceptable, assuming the model is accurate.

More normal model examples

Example: Suppose SAT-style scores are approximately normal with mean 500 and standard deviation 100. A score of 650 has z-score

\[z=(650-500)/100=1.5\].

Technology shows that about 93.3% of values are below z=1.5. So a score of 650 is around the 93rd percentile under this model.

Example: A manufacturing process makes rods with lengths approximately normal with mean 20 cm and standard deviation 0.1 cm. Rods must be between 19.8 and 20.2 cm. These are z-scores -2 and 2, so about 95% of rods are within tolerance.

When not to use a normal model

Normal models are inappropriate for strongly skewed data, data with extreme outliers, categorical data, and data bounded in ways that conflict with the bell curve. Household income, home prices, waiting times, and social-media follower counts are often skewed. A normal model may be misleading.

Students should check a histogram or dot plot before fitting a normal model. Does it look roughly symmetric and bell-shaped? Are there strong outliers? Does the context make normal variation plausible?

Technology interpretation

Technology can calculate normal probabilities, but students must know whether they are finding:

  • area to the left of a value;
  • area to the right of a value;
  • area between two values;
  • a value corresponding to a percentile.

The area under the curve is the proportion of the population modeled in that interval. It is not the height of the curve.

Problem Library

Problems in the App From This Objective

174 problems across 12 archetypes in the app.

interpret center and spread.
15 problems Warmup Practice Mixed Review Assessment
Problem 1

Identify mean and standard deviation from normal model heights are normal with mean 68 and standard deviation 3 inches.

Problem 2

Identify mean and standard deviation from normal model N(100,15) where second value is sd.

Problem 3

Identify mean and standard deviation from normal model test scores centered at 75 with typical distance 8.

Problem 4

Identify mean and standard deviation from normal model The weights of adult males follow a normal distribution with an average of 180 pounds and a standard deviation of 20 pounds.

Problem 5

Identify mean and standard deviation from normal model The temperature in July is normally distributed as N(25°C, 3°C), where the second value is the standard deviation.

Problem 6

Identify mean and standard deviation from normal model The commute times to work are normally distributed with an average of 45 minutes and a spread of 10 minutes.

Problem 7

Identify mean and standard deviation from normal model The number of customers per hour follows a normal model with an expected value of 50 and a variability of 7.

Problem 8

Identify mean and standard deviation from normal model A certain population has a characteristic that is normally distributed N(72.5, 5.2).

Problem 9

Identify mean and standard deviation from normal model The lifespan of a certain electronic component is normally distributed with a mean of 5000 hours and a typical deviation of 250 hours.

Problem 10

Identify mean and standard deviation from normal model The amount of soda in a bottle is normally distributed with its center at 355 ml and a standard deviation of 5 ml.

Problem 11

Identify mean and standard deviation from normal model The scores on a standardized test are normally distributed with an average of 500 and a standard deviation of 100.

Problem 12

Identify mean and standard deviation from normal model The daily rainfall in a city is modeled by N(1.5 inches, 0.3 inches).

Problem 13

Identify mean and standard deviation from normal model The error in a measurement device is normally distributed with a mean of 0 and a spread of 0.02.

Problem 14

Identify mean and standard deviation from normal model The delivery times for a package are normally distributed with an average of 3.5 days and a typical distance from the mean of 0.8 days.

Open in simulator
Problem 15

Identify mean and standard deviation from normal model The annual income for a certain profession is normally distributed with an expected value of $75,000 and a standard deviation of $12,000.

mark mean and standard-deviation intervals.
15 problems Warmup Practice Mixed Review Assessment
Problem 16

Sketch normal distribution with mean 50 and standard deviation 10.

Problem 17

Sketch normal distribution with mean 100 and standard deviation 15.

Problem 18

Sketch normal distribution with mean 0 and standard deviation 1.

Problem 19

Sketch normal distribution with mean 20 and standard deviation 5.

Problem 20

Sketch normal distribution with mean 75 and standard deviation 8.

Problem 21

Sketch normal distribution with mean 10 and standard deviation 2.

Problem 22

Sketch normal distribution with mean 500 and standard deviation 50.

Problem 23

Sketch normal distribution with mean 1.5 and standard deviation 0.5.

Problem 24

Sketch normal distribution with mean -10 and standard deviation 3.

Problem 25

Sketch normal distribution with mean 30 and standard deviation 4.

Problem 26

Sketch normal distribution with mean 120 and standard deviation 20.

Problem 27

Sketch normal distribution with mean 0 and standard deviation 0.5.

Problem 28

Sketch normal distribution with mean 250 and standard deviation 25.

Problem 29

Sketch normal distribution with mean 60 and standard deviation 6.

Open in simulator
Problem 30

Sketch normal distribution with mean 1000 and standard deviation 100.

apply 68-95-99.7 rule.
15 problems Warmup Practice Mixed Review Assessment
Problem 31

Use empirical rule for normal model within 1 sd of mean.

Problem 32

Use empirical rule for normal model within 2 sd of mean.

Problem 33

Use empirical rule for normal model within 3 sd of mean.

Problem 34

Use empirical rule for normal model above mean in normal distribution.

Problem 35

Use empirical rule for normal model below mean in normal distribution.

Problem 36

Use empirical rule for normal model below 1 standard deviation from the mean.

Problem 37

Use empirical rule for normal model above 1 standard deviation from the mean.

Problem 38

Use empirical rule for normal model below 2 standard deviations from the mean.

Open in simulator
Problem 39

Use empirical rule for normal model above 2 standard deviations from the mean.

Problem 40

Use empirical rule for normal model below 3 standard deviations from the mean.

Problem 41

Use empirical rule for normal model above 3 standard deviations from the mean.

Problem 42

Use empirical rule for normal model between the mean and 1 standard deviation above the mean.

Problem 43

Use empirical rule for normal model between the mean and 2 standard deviations above the mean.

Problem 44

Use empirical rule for normal model between the mean and 3 standard deviations above the mean.

Problem 45

Use empirical rule for normal model between 1 and 2 standard deviations above the mean.

compute `(x-mean)/sd`.
12 problems Warmup Practice Mixed Review Assessment
Problem 46

Find z-score for value x=86, mean 80, sd 3.

Open in simulator
Problem 47

Find z-score for value x=70, mean 85, sd 5.

Problem 48

Find z-score for value x=112, mean 100, sd 15.

Problem 49

Find z-score for value x, mean mu, sd sigma.

Problem 50

Find z-score for value x=100, mean 90, sd 5.

Problem 51

Find z-score for value x=40, mean 50, sd 2.

Problem 52

Find z-score for value x=75, mean 70, sd 10.

Problem 53

Find z-score for value x=15, mean 20, sd 10.

Problem 54

Find z-score for value x=69, mean 60, sd 4.

Problem 55

Find z-score for value x=35, mean 40, sd 8.

Problem 56

Find z-score for value x=150, mean 150, sd 10.

Problem 57

Find z-score for value x=26.5, mean 20.5, sd 2.5.

state standard deviations from mean.
12 problems Warmup Practice Mixed Review Assessment
Problem 58

Interpret z-score z=2.1 for height.

Problem 59

Interpret z-score z=-0.7 for score.

Problem 60

Interpret z-score z=0.

Problem 61

Interpret z-score z=-3.

Problem 62

Interpret z-score z=1.5 for weight.

Problem 63

Interpret z-score z=-2.5 for test result.

Problem 64

Interpret z-score z=1 for temperature.

Open in simulator
Problem 65

Interpret z-score z=-2 for reaction time.

Problem 66

Interpret z-score z=0.8.

Problem 67

Interpret z-score z=-1.2.

Problem 68

Interpret z-score z=3.1 for rainfall.

Problem 69

Interpret z-score z=0.1 for blood pressure.

use normal CDF output.
15 problems Warmup Practice Mixed Review Assessment
Problem 70

Estimate percentile or area using normal technology for P(X<85) for mean 80 sd 5.

Problem 71

Estimate percentile or area using normal technology for P(X>90) for mean 100 sd 10.

Problem 72

Estimate percentile or area using normal technology for P(70<X<90) for mean 80 sd 5.

Problem 73

Estimate percentile or area using normal technology for P(X<60) for mean 50 sd 10.

Open in simulator
Problem 74

Estimate percentile or area using normal technology for P(X>120) for mean 100 sd 15.

Problem 75

Estimate percentile or area using normal technology for P(150<X<180) for mean 160 sd 10.

Problem 76

Estimate percentile or area using normal technology for P(X<25) for mean 30 sd 2.

Problem 77

Estimate percentile or area using normal technology for P(X>75) for mean 70 sd 5.

Problem 78

Estimate percentile or area using normal technology for P(40<X<50) for mean 45 sd 3.

Problem 79

Estimate percentile or area using normal technology for P(Z<1.5) for mean 0 sd 1.

Problem 80

Estimate percentile or area using normal technology for P(Z>-0.5) for mean 0 sd 1.

Problem 81

Estimate percentile or area using normal technology for P(-1<Z<1) for mean 0 sd 1.

Problem 82

Estimate percentile or area using normal technology for P(X<500) for mean 450 sd 25.

Problem 83

Estimate percentile or area using normal technology for P(X>1100) for mean 1000 sd 50.

Problem 84

Estimate percentile or area using normal technology for P(900<X<1100) for mean 1000 sd 50.

use inverse normal output.
15 problems Warmup Practice Mixed Review Assessment
Problem 85

Find data value from percentile using normal technology 90th percentile, mean 100 sd 15.

Problem 86

Find data value from percentile using normal technology bottom 5 percent, mean 50 sd 8.

Problem 87

Find data value from percentile using normal technology top 10 percent cutoff.

Problem 88

Find data value from percentile using normal technology 75th percentile, mean 60 sd 10.

Problem 89

Find data value from percentile using normal technology bottom 10 percent, mean 200 sd 25.

Problem 90

Find data value from percentile using normal technology top 25 percent cutoff, mean 75 sd 12.

Problem 91

Find data value from percentile using normal technology median value, mean 30 sd 5.

Problem 92

Find data value from percentile using normal technology 99th percentile, mean 1000 sd 50.

Open in simulator
Problem 93

Find data value from percentile using normal technology bottom 1 percent, mean 10 sd 2.

Problem 94

Find data value from percentile using normal technology top 5 percent cutoff, mean 120 sd 10.

Problem 95

Find data value from percentile using normal technology 25th percentile, mean 40 sd 6.

Problem 96

Find data value from percentile using normal technology top 1 percent cutoff.

Problem 97

Find data value from percentile using normal technology bottom 2.5 percent, mean 80 sd 12.

Problem 98

Find data value from percentile using normal technology 95th percentile, mean 500 sd 75.

Problem 99

Find data value from percentile using normal technology top 20 percent cutoff, mean 150 sd 20.

compare z-scores.
15 problems Warmup Practice Mixed Review Assessment
Problem 100

Compare values from different normal distributions score 85 on mean 80 sd 5 vs score 110 on mean 100 sd 20.

Problem 101

Compare values from different normal distributions height 70 with mean 68 sd 3 vs score 90 with mean 75 sd 10.

Problem 102

Compare values from different normal distributions values with z -1 and z 0.2.

Problem 103

Compare values from different normal distributions test score 90 on mean 80 sd 10 vs project score 75 on mean 70 sd 8.

Problem 104

Compare values from different normal distributions weight 180 lbs with mean 170 sd 15 vs height 72 inches with mean 68 sd 2.5.

Problem 105

Compare values from different normal distributions time 55s with mean 60s sd 5s vs distance 105m with mean 100m sd 2m.

Problem 106

Compare values from different normal distributions temperature 20C with mean 25C sd 3C vs pressure 1005hPa with mean 1000hPa sd 2hPa.

Problem 107

Compare values from different normal distributions score 75 on mean 80 sd 5 vs score 50 on mean 65 sd 10.

Problem 108

Compare values from different normal distributions value 100 with mean 100 sd 10 vs value 55 with mean 50 sd 2.

Problem 109

Compare values from different normal distributions value 100 with mean 100 sd 10 vs value 45 with mean 50 sd 2.

Problem 110

Compare values from different normal distributions values with z 1.8 and z 0.9.

Open in simulator
Problem 111

Compare values from different normal distributions values with z -0.5 and z 0.1.

Problem 112

Compare values from different normal distributions values with z -2.3 and z -1.1.

Problem 113

Compare values from different normal distributions student's math score 92 (mean 85, sd 4) vs his reading score 88 (mean 80, sd 5).

Problem 114

Compare values from different normal distributions car's fuel efficiency 35 MPG (mean 30, sd 2) vs its repair cost $400 (mean $500, sd $50).

assess shape and context.
15 problems Warmup Practice Mixed Review Assessment
Problem 115

Decide whether normal model is appropriate for histogram roughly symmetric, single peak, no extreme outliers.

Problem 116

Decide whether normal model is appropriate for strongly right-skewed income data.

Problem 117

Decide whether normal model is appropriate for bounded counts with many zeros.

Problem 118

Decide whether normal model is appropriate for measurement errors around target with symmetric spread.

Problem 119

Decide whether normal model is appropriate for heights of adult males in a large population.

Problem 120

Decide whether normal model is appropriate for number of car accidents per day in a small town.

Problem 121

Decide whether normal model is appropriate for IQ scores from a standardized test.

Open in simulator
Problem 122

Decide whether normal model is appropriate for time until failure for electronic components.

Problem 123

Decide whether normal model is appropriate for distribution of sample means from a large number of samples.

Problem 124

Decide whether normal model is appropriate for daily sales of a niche product with many zero-sale days.

Problem 125

Decide whether normal model is appropriate for weights of newly manufactured bolts with slight random variation.

Problem 126

Decide whether normal model is appropriate for histogram showing two distinct peaks in test scores.

Problem 127

Decide whether normal model is appropriate for uniformly distributed random numbers between 0 and 1.

Problem 128

Decide whether normal model is appropriate for measurement errors in a precise scientific experiment.

Problem 129

Decide whether normal model is appropriate for strongly left-skewed data for age at first marriage.

apply technology/empirical rule and interpret.
15 problems Warmup Practice Mixed Review Assessment
Problem 130

Estimate population percentage in normal distribution for mean 100 sd 10, between 90 and 110.

Problem 131

Estimate population percentage in normal distribution for mean 50 sd 5, above 60.

Problem 132

Estimate population percentage in normal distribution for mean 70 sd 8, below 54.

Problem 133

Estimate population percentage in normal distribution for mean 0 sd 1, between -2 and 2.

Problem 134

Estimate population percentage in normal distribution for mean 100 sd 10, between 70 and 130.

Problem 135

Estimate population percentage in normal distribution for mean 200 sd 20, above 220.

Problem 136

Estimate population percentage in normal distribution for mean 30 sd 3, below 27.

Problem 137

Estimate population percentage in normal distribution for mean 500 sd 50, between 500 and 550.

Problem 138

Estimate population percentage in normal distribution for mean 10 sd 1, between 9 and 10.

Problem 139

Estimate population percentage in normal distribution for mean 60 sd 6, between 66 and 72.

Problem 140

Estimate population percentage in normal distribution for mean 80 sd 4, between 72 and 76.

Problem 141

Estimate population percentage in normal distribution for mean 1000 sd 100, below 700.

Open in simulator
Problem 142

Estimate population percentage in normal distribution for mean 50 sd 2, above 56.

Problem 143

Estimate population percentage in normal distribution for mean 0 sd 1, between -1 and 2.

Problem 144

Estimate population percentage in normal distribution for mean 100 sd 5, between 90 and 105.

describe shift and spread changes.
15 problems Warmup Practice Mixed Review Assessment
Problem 145

Interpret effect of changing normal model parameters mean increases, sd same.

Problem 146

Interpret effect of changing normal model parameters sd increases, mean same.

Problem 147

Interpret effect of changing normal model parameters sd decreases.

Problem 148

Interpret effect of changing normal model parameters mean decreases and sd increases.

Problem 149

Interpret effect of changing normal model parameters the mean value of the data decreases, and the standard deviation is unchanged.

Problem 150

Interpret effect of changing normal model parameters the mean increases while the standard deviation decreases.

Problem 151

Interpret effect of changing normal model parameters both the mean and standard deviation decrease.

Problem 152

Interpret effect of changing normal model parameters both the mean and standard deviation increase.

Problem 153

Interpret effect of changing normal model parameters neither the mean nor the standard deviation change.

Problem 154

Interpret effect of changing normal model parameters the average of the data points increases, and the spread remains constant.

Problem 155

Interpret effect of changing normal model parameters the variability of the data increases, and the mean is stable.

Problem 156

Interpret effect of changing normal model parameters the standard deviation is reduced, keeping the mean fixed.

Problem 157

Interpret effect of changing normal model parameters the average value drops, and the data becomes more dispersed.

Open in simulator
Problem 158

Interpret effect of changing normal model parameters the mean is halved, and the standard deviation is unchanged.

Problem 159

Interpret effect of changing normal model parameters the standard deviation is doubled, with no change to the mean.

catch z-score, tail direction, empirical-rule, and technology mistakes.
15 problems Warmup Practice Mixed Review Assessment
Problem 160

Correct normal-model interpretation error z=-2 means value is 2 units below mean.

Open in simulator
Problem 161

Correct normal-model interpretation error top 5 percent uses invNorm(0.05).

Problem 162

Correct normal-model interpretation error 68-95-99.7 rule used on skewed data.

Problem 163

Correct normal-model interpretation error P(X>90) computed as normalcdf(-infinity,90).

Problem 164

Correct normal-model interpretation error z=1.5 means the value is 1.5 times the mean.

Problem 165

Correct normal-model interpretation error A z-score of 0 means the data value is 0.

Problem 166

Correct normal-model interpretation error P(X < 70) calculated as normalcdf(70, infinity).

Problem 167

Correct normal-model interpretation error P(50 < X < 60) calculated as normalcdf(-infinity, 60).

Problem 168

Correct normal-model interpretation error To find the probability of X < 85, I use invNorm(0.85, mean, std_dev).

Problem 169

Correct normal-model interpretation error To find the 75th percentile, I use normalcdf(-infinity, X, mean, std_dev).

Problem 170

Correct normal-model interpretation error According to the empirical rule, 95% of data falls within 1 standard deviation.

Problem 171

Correct normal-model interpretation error The empirical rule states that 68% of data is within 2 standard deviations.

Problem 172

Correct normal-model interpretation error To find the cutoff for the bottom 10 percent, I use invNorm(0.90).

Problem 173

Correct normal-model interpretation error normalcdf(80) for a normal distribution with mean 70, SD 5.

Problem 174

Correct normal-model interpretation error Z-score for X=80 (mean=70, SD=5) is (80+70)/5.