Measurement analysis · 45 questions

Error-theory drill

Four ideas account for almost the whole measurement-analysis subject area: how a set of observations is summarised, how random error carries through a computation, how observations of unequal quality are combined, and how a result is judged against a tolerance. This drill works them until they are automatic.

The four relations worth memorising

s = √( Σv² / (n − 1) ) sample standard deviation σ_mean = s / √n standard error of the mean E_sum = √( ΣEᵢ² ) propagation through a sum w = 1 / σ² weight from a standard deviation

The n − 1 is not a detail. With n observations of one unknown there are n − 1 degrees of freedom, and dividing by n understates the spread of a short set — which is the set a crew is most likely to have.

Question 1 of 45Measurement analysis · Recall

Which of the following is a systematic error?

Work a full problem

Error-propagation problems