3.12 Evaluate a stats report 1

Non sampling errors

• are avoidable

• are unavoidable

Sampling errors

• are avoidable

• are unavoidable

Larger samples have larger sampling errors

• True

• False

Non sampling errors are linked to (multiple answers)

• bias

• populations

• samples

• conflicts of interest

Observational studies provide evidence for causal effects

• always

• depending on the type of study

• never

• only if the sample is very large

Experimental studies provide evidence for causal effects

• provided the experimental design is sound

• depends on the type of experiment

• always

• only is the experimental group is very large

We randomly allocate into groups in experiments,

• to minimise the effect of other variables

• to speed the experiment up

• to prevent arguments

• for convenience

An experimental study is one in which two groups are

• compared

• controlled

• experimented on

• challenged

"Before and after" experiments are reliable?

• True

• False

An observational study is defined upon a

• population

• sample

• cohort

• census