PHAS0083 Chap.2 Common Families of Distributions
Common Families of Distributions
Common Families of Distributions gives the building blocks of physics models. It first defines expectation, variance, moments, median and mode, and the preimage rule for a function of a random variable, including the case where the function is not one-to-one. It then works through six families. The uniform describes equally likely values on a set or interval.
The Bernoulli has two outcomes and the binomial counts successes in a fixed number of independent trials, with mean np and variance np(1 - p). The Poisson counts independent events at a constant rate, has mean equal to variance and arises as a limit of the binomial. The exponential is the waiting time between events of that process, written with a scale parameter whose value is the mean.
The normal is fully set by its mean and variance, describes the sum of many small independent effects, and stands in for both the binomial for many trials and the Poisson for a large mean. Recognising which family a physical description implies is a learning objective in its own right, and the chapter repeatedly stresses the conventions that cause errors, such as variance as the second normal argument.
What this chapter covers
- 01
Expectation, variance and moments
- 02
Functions of a random variable
- 03
Uniform and Bernoulli distributions
- 04
Binomial counts and their mode
- 05
Poisson counts and the binomial limit
- 06
Exponential waiting times
- 07
Normal distribution and standardisation
Worked example · free
Standardise a resistor tolerance with the normal family
- 1Standardise both limits: z = (97 - 100)/2 = -1.5 and z = (104 - 100)/2 = 2.0. This allocation is practice weighting, not a UCL marking scheme.
- 2Read the standard normal cdf: Phi(2.0) = 0.9772 and Phi(-1.5) = 0.0668.
- 1Subtract: 0.9772 - 0.0668 = 0.910.
Key terms
- Distribution family
- A functional form indexed by parameters, such as all normal distributions with any mean and variance.
- Binomial distribution
- The distribution of the number of successes in a fixed number of independent trials with equal success probability.
- Scale parameter
- A parameter that stretches a distribution without changing its shape, such as the exponential mean.
Common Families of Distributions FAQ
How do I pick the right family for a word problem?
Look at what the experimental design fixes. A fixed number of independent trials points to the binomial, independent events in a fixed time or space window to the Poisson, the time until the next such event to the exponential, and a quantity built from many small independent effects to the normal. Equal likelihood on a range points to the uniform.
When should I use a Poisson rather than a binomial model?
Use the binomial when there is a fixed number of trials, each with the same success probability. Use the Poisson when events occur independently at a constant rate in a window with no natural upper count, or as an approximation to a binomial with many trials and a small success probability.
Why does the exponential distribution have a scale and a rate form?
The notes write it with a scale equal to the mean lifetime, while many physics texts use a rate equal to one over the lifetime. Both describe the same family; the scale form matches the reference texts and the Python functions used in the course.
How large must the Poisson mean be for a normal approximation?
The notes describe the approximation as already quite accurate once the mean is around 10 or more. For smaller means the Poisson pmf is visibly skewed, with extra probability near zero, so exact Poisson probabilities should be used instead.
Exam move
Build a one-page card for each family with its support, parameter, pmf or pdf, mean, variance and the physical situations that produce it, and derive each mean once from the definition so you can rebuild it under pressure. Then practise recognition: for each word problem, decide what the design fixes, a number of trials, a time window or a waiting time, and name the family before writing anything else.
Drill the cdf manoeuvres that cost marks, especially complements for upper tails and the one-step-below rule for the lower edge of an inclusive discrete range. Keep conventions explicit: Poisson mean as nu, exponential scale as beta, and the normal written with its variance. Finally connect the families by limits, binomial to Poisson and both to normal, and check one limit numerically to see how fast it converges.
Working through Common Families of Distributions in PHAS0083? Sia is AskSia’s AI Statistics tutor — ask any PHAS0083 Common Families of Distributions question and get a clear, step-by-step explanation grounded in how PHAS0083 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.