Data synthesis

Objective. Connect probability, estimation and model criticism.

1. Learn
2. Worked example
3. Practice
4. Feedback
5. Continue

Learn: the key idea

Data synthesis connects probability, estimation and model criticism into one workflow: choose a probability model for the data, estimate its parameters (commonly using maximum likelihood or the method of moments), then critically assess whether the fitted model is reasonable. Criticism includes checking residuals, comparing observed and expected frequencies, and questioning whether assumptions such as independence are justified by how the data were collected. For example, fitting a Poisson model to call-centre arrivals, estimating λ as the sample mean, then checking if the variance of the data is close to λ (as a Poisson model requires) tests the model's validity. A good estimate is only useful alongside honest criticism of the model.

Worked example

In a data synthesis exercise, a sample of 10 values from a Poisson model has mean 4.2; use maximum likelihood estimation to estimate λ.

1 $For Poisson data, the MLE of λ equals the sample mean$
2 $State the estimator: λ̂ = x̄$
3 $Substitute the given sample mean$
4 $λ̂ = 4.2$

Answer: λ̂ = 4.2

Concept mastery

Based on your past attempts at this lesson — the question types to practise again come first.

Complete a practice set to see your concept breakdown here.

Practice

Practice happens on its own screen, one question at a time. Answers stay hidden until you submit yours, and you can stop and pick up at the same question later.

Start practising

Common mistake. Fitting a model and reporting the parameter estimate without ever checking whether the model's assumptions actually fit the data; an estimate from a poorly fitting model can be precise but meaningless.

Your progress

Tick each step as you finish it. Your exact reading position is saved automatically, so Continue drops you back on the same line.

0 of 5 steps complete

Your place saves automatically as you read.

My dashboard