What factors control the extent of random sampling error

However, what we have just described is random error, and there is an important difference from bias. There is no longer any way that we could predict whether error would take the form of over- or under-estimation. Indeed, if we carried out such a survey repeatedly, we would probably suffer very similar numbers of over- and under-estimates. [Pg.39]

Any given sample has an equal chance of under- or over-estimating the population mean value. [Pg.39]

Even the best designed experiments are subject to random error. [Pg.39]

3 The rest of this book is concerned with random error only - not bias [Pg.39]

For the rest of this book we will be concerned solely with the problem of random error. Bias just should not be there - good experimental design should see it off. However, random error is ever with us and the science of statistics is there to allow us to draw conclusions despite its presence. Statistics will not make random error disappear, it just allows us to live with it. [Pg.39]

Big Chemical Encyclopedia

Chemical substances, components, reactions, process design ...