OCR MEI Further Statistics Minor 2019 June — Question 3 4 marks

Exam BoardOCR MEI
ModuleFurther Statistics Minor (Further Statistics Minor)
Year2019
SessionJune
Marks4
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicCombinations & Selection
TypeCritique of sampling methods
DifficultyEasy -1.8 This is a purely conceptual question about sampling methodology requiring only recall of basic statistical principles (census vs sample, bias in convenience sampling, and sample properties like representativeness). No calculations, mathematical techniques, or problem-solving are involved—just straightforward explanation of standard textbook concepts about sampling methods.
Spec2.01a Population and sample: terminology2.01c Sampling techniques: simple random, opportunity, etc2.01d Select/critique sampling: in context

3 A company has been commissioned to make 50 very expensive titanium components.
A sample of the components needs to be tested to ensure that they are sufficiently strong. However, this is a test to destruction, so the components which are tested can no longer be used.
  1. Explain why it would not be appropriate to use a census in these circumstances. A manager suggests that the first 5 components to be manufactured should be tested.
  2. Explain why this would not be a sensible method of selecting the sample. A statistician advises the manager that the sample selected should be a random sample.
  3. Give two desirable features (other than randomness) that the sample should have.

Question 3:
AnswerMarks Guidance
3(a) Because a census would involve destroying all of
the components.E1
[1]2.3 “Testing all items is expensive”
alone is E0
AnswerMarks
(b)Because as more components are produced a
change might develop in the production process.
It is not a random method.
AnswerMarks Guidance
It will not be representative of all the components.E1
[1]2.2b Only one reason needs to be
givenNB “The sample is too
small” is E0
AnswerMarks
(c)The sample should be:
unbiased
representative of the population
chosen so that components are selected
AnswerMarks
independentlyB1
B1
AnswerMarks Guidance
[2]1.2
1.2Any two reasons (B1 each) NB “The sample needs
to be large” is B0
Question 3:
3 | (a) | Because a census would involve destroying all of
the components. | E1
[1] | 2.3 | “Testing all items is expensive”
alone is E0
(b) | Because as more components are produced a
change might develop in the production process.
It is not a random method.
It will not be representative of all the components. | E1
[1] | 2.2b | Only one reason needs to be
given | NB “The sample is too
small” is E0
(c) | The sample should be:
unbiased
representative of the population
chosen so that components are selected
independently | B1
B1
[2] | 1.2
1.2 | Any two reasons (B1 each) | NB “The sample needs
to be large” is B0
3 A company has been commissioned to make 50 very expensive titanium components.\\
A sample of the components needs to be tested to ensure that they are sufficiently strong. However, this is a test to destruction, so the components which are tested can no longer be used.
\begin{enumerate}[label=(\alph*)]
\item Explain why it would not be appropriate to use a census in these circumstances.

A manager suggests that the first 5 components to be manufactured should be tested.
\item Explain why this would not be a sensible method of selecting the sample.

A statistician advises the manager that the sample selected should be a random sample.
\item Give two desirable features (other than randomness) that the sample should have.
\end{enumerate}

\hfill \mbox{\textit{OCR MEI Further Statistics Minor 2019 Q3 [4]}}