You enter a mean, a standard deviation and a boundary into a calculator. A probability appears, but you are unsure whether it answers the question. Did the question ask for values below the boundary, above it or between two boundaries?
A sketch can settle that interpretation before numerical work begins. Mark the mean, locate the relevant values and shade the requested region. Then standardise or use an appropriate calculator function with the correct parameters.
This S1 normal distribution guide uses original examples to connect the event, the graph and the calculation. Check your examination-year syllabus and permitted resources alongside your practice. For structured subject learning, view AS Level Math S1 at Abeel Eschool.
Identify the variable and its units
Suppose X represents the mass of a manufactured item in grams. Under an illustrative model, X is normally distributed with mean 80 grams and standard deviation 4 grams.
The mean locates the centre of the distribution. Standard deviation describes its scale or spread. A boundary such as 84 grams is one standard deviation above the mean; 72 grams is two standard deviations below it.
Write what X measures before using notation. This helps distinguish the original measurement from a standardised value. A z-value is a relative position, not another mass in grams.
Also distinguish a model from a statement about every possible real-world measurement. The normal model is continuous and theoretically unbounded; its appropriateness depends on the context and assumptions given in the question.
Read variance and standard deviation carefully
Normal notation often states a mean and a variance, commonly written N(μ, σ²). If X ~ N(80,16), the standard deviation is four because the variance is sixteen.
Using sixteen as the standard deviation would make the distribution appear much wider than intended. The calculator can still return a plausible-looking decimal, so this mistake may not be obvious from the output alone.
Before calculating, write three labelled values: mean, variance and standard deviation. If one is not needed immediately, it still helps clarify the notation you are using.
Always follow the convention explicitly used by the question, textbook or function. Some calculator interfaces ask directly for standard deviation even when the written distribution notation provides variance.
Standardising measures distance from the mean
For a normal variable with mean μ and positive standard deviation σ, standardisation uses Z = (X−μ)/σ. The resulting Z has a standard normal distribution with mean zero and standard deviation one.
In the mass example, X = 84 gives z = (84−80)/4 = 1. X = 72 gives z = (72−80)/4 = −2.
The subtraction measures distance from the mean. Dividing by standard deviation expresses that distance on a common scale. Values above the mean have positive z-values, and values below it have negative z-values.
A sign check is therefore useful. If a boundary below the mean produces a positive z-value, inspect the order of subtraction before continuing.
Match a shaded region to a probability
Suppose the question asks for P(X < 84). The relevant region lies to the left of 84, which is above the mean. Standardising gives P(Z < 1), approximately 0.8413.
This answer should exceed one half because the boundary is above the centre and includes the whole lower half plus an additional region. That check supports the calculation.
If the question asks for P(X > 84), the requested region is the upper tail. Its probability is approximately 1 − 0.8413 = 0.1587.
The same boundary creates two complementary questions. A correct lower-tail calculation does not automatically answer an upper-tail question, even though both use z = 1.
Find the area between two boundaries
For P(76 < X < 84), the standardised boundaries are −1 and 1. The required probability is P(Z < 1) − P(Z < −1), approximately 0.8413 − 0.1587 = 0.6826.
Draw both boundaries before calculating. The region is centred on the mean in this example, but not every interval is symmetric. Do not assume equal tails unless the boundaries are equally far from the mean.
For a non-symmetric example, P(80 < X < 88) becomes P(0 < Z < 2), approximately 0.9772 − 0.5 = 0.4772.
Keep the interpretation consistent with the function or table you use. If the resource reports an area between zero and z rather than a lower-tail area, the arithmetic must reflect that convention.
Understand the role of symmetry
The normal distribution is symmetric about its mean. For the standard normal variable, P(Z < −a) equals P(Z > a) for a positive a.
This gives a useful check for paired boundaries. In the mass example, P(X < 76) equals P(X > 84), since both boundaries lie four grams from the mean on opposite sides.
Symmetry does not mean that every positive and negative boundary in the original units can be paired by simply changing a sign. The centre is μ, not necessarily zero. Reflect a value across the mean.
For example, the reflection of 72 around a mean of 80 is 88, not −72. Standardising first makes the symmetry easier to see.
Treat inverse problems as a location question
An inverse question gives an area and asks for a boundary. Suppose P(X < k) = 0.95 for X with mean 80 and standard deviation 4.
The boundary must be above the mean because ninety-five per cent lies below it. The corresponding lower-tail standard normal value is approximately z = 1.6449.
Rearranging gives k = μ + zσ = 80 + 1.6449 × 4, approximately 86.58 grams. State the boundary in the original units.
If your calculator returns a negative z-value for this lower-tail area, check whether its input expects the upper tail or a different convention. A sketch of the ninety-five per cent region should make the intended side clear.
Convert upper-tail language before using an inverse function
Suppose the largest ten per cent of items exceed a boundary c. That means P(X > c) = 0.10, so P(X < c) = 0.90 for this continuous model.
An inverse function using lower-tail probability should therefore receive 0.90, not 0.10. The resulting z-value is positive, about 1.2816, and c is approximately 85.13 grams.
Entering 0.10 in a lower-tail inverse function would give a boundary below the mean, corresponding to the smallest ten per cent. That answers a different question.
Write the converted probability explicitly before the calculator step. The phrase “top ten per cent” is an interpretation task first and a numerical task second.
Use two probability statements to find unknown parameters
Some questions give two boundaries and their cumulative probabilities, then ask for the mean and standard deviation. Convert each probability into its corresponding standard normal z-value.
For example, suppose P(X < 50) is approximately 0.1587 and P(X < 70) is approximately 0.8413. These correspond approximately to z = −1 and z = 1.
The equations are (50−μ)/σ = −1 and (70−μ)/σ = 1. Equivalently, 50 = μ−σ and 70 = μ+σ. Adding gives μ = 60, and substituting gives σ = 10.
Use adequate precision if the given probabilities do not correspond to simple rounded z-values. The method is the same: two standardisation relationships provide two equations for the unknown parameters.
Do not add a point mass to a continuous model
For a continuous normal variable, the probability of exactly one numerical value is zero. Consequently, P(X < a) and P(X ≤ a) are equal in the model.
That differs from an integer-valued distribution such as binomial, where including a particular count can change a probability. Keep the distinction between continuous measurement and discrete count clear.
A reported measurement may be rounded in practice, which can make a statement about a displayed value correspond to an interval of underlying measurements. Follow the wording and level of modelling required by the question.
Do not import a discrete correction or another approximation procedure automatically. This article concerns a variable already modelled as normal; other approximations have their own conditions and methods.
Check calculator entries against a simple manual case
For the example with mean 80 and standard deviation 4, the probability below 80 should be one half. Use this as a simple test of parameter order and function behaviour.
Next, compare the probability below 84 with the expected standard-normal result near 0.8413. A dramatically different value suggests a parameter, boundary or function-entry error.
Keep enough written working to show the event and parameters. A bare decimal does not reveal whether you used the right tail or accidentally entered variance as standard deviation.
Use the permitted functions and resources for your examination. Practise on the calculator you will actually use, rather than relying on another model’s sequence of buttons.
Build revision around the interpretation step
Take a mixed set of questions and sketch every requested region without calculating. Label whether each is lower tail, upper tail, interval or inverse boundary. Then add the standardised boundaries.
After calculating, compare each answer with the sketch. A tail above a positive z-value should have probability below one half. A wide central interval should have more area than a narrow interval with the same centre.
Record the earliest error in your log: parameter reading, boundary translation, tail choice, inverse-area conversion or arithmetic. Target that step on a fresh question rather than simply copying the worked solution.
Abeel’s Cambridge past-paper guide can support the practice routine, while the AS/A2 planning guide helps organise it around other subjects.
Explore Abeel Eschool and its course catalogue for current options. Use admissions and FAQs to ask about relevant lessons, including the topic and examination component you are preparing.
Frequently asked questions
What does standardising do?
It expresses a value’s distance from the mean in standard-deviation units. For a normal variable, Z = (X−μ)/σ has the standard normal distribution.
Should I enter variance or standard deviation?
Read the calculator function’s required input. Written normal notation often supplies variance, while the function may ask for standard deviation.
How do I know which tail to use?
Translate the wording into an inequality and shade the region. Below a boundary is a lower tail; above it is an upper tail.
Why does a top-ten-per-cent inverse problem use 0.90?
A lower-tail inverse function needs the area below the boundary. If ten per cent is above it, ninety per cent is below it.
Can a normal variable equal one exact value with positive probability?
A continuous normal model assigns zero probability to an exact point. Probabilities over intervals can be positive.
Have a question about this topic?
Send your query to Abeel Eschool and our admissions team will get back to you.