IB Maths: Applications and Interpretation SL
4.1 Sampling, data quality and outliers
Total 27 marks
Name
Class
Date
- 1A council wants to find out how its residents travel to work. A researcher stands outside the main railway station one morning and questions the first 80 people who walk past. The council has 12 000 residents on its electoral register.(a)Which sampling technique has the researcher used?[1 mark]
- ASimple random sampling
- BConvenience sampling
- CSystematic sampling
- DQuota sampling
Show the mark scheme
Mark scheme
Answer: B(b)Which statement best explains why the sample may be biased?[1 mark]- AThe sample size of 80 is too small to calculate a mean
- BCommuting times are continuous data, which cannot be sampled
- CResidents who do not use the station, such as those who drive or cycle, have no chance of being selected
- DQuestioning the first 80 people gives every resident an equal chance of selection
Show the mark scheme
Mark scheme
Answer: C(c)Describe how the council could take a simple random sample of 80 residents from the electoral register.[2 marks]Show the mark scheme
Mark scheme
B1: number the residents to (or list every resident)
B1: use a random number generator (GDC) to choose different numbers in this range and select those residents
Accept: names drawn from a hat or any method giving every resident an equal chance
Do not accept: choosing every 150th resident (systematic)Total for question 1: 4 marks
- 2A sports club has 600 members: 240 juniors, 210 adults and 150 seniors. The club takes a stratified sample of 80 members for a survey about its facilities.(a)How many juniors should be in the sample?[1 mark]
- A32
- B40
- C24
- D16
Show the mark scheme
Mark scheme
Answer: A(b)How many adults should be in the sample?[1 mark]- A21
- B27
- C35
- D28
Show the mark scheme
Mark scheme
Answer: D(c)Explain why a stratified sample is more suitable than a simple random sample for this survey.[2 marks]Show the mark scheme
Mark scheme
R1: the members fall into distinct groups (juniors, adults, seniors) whose opinions or facility use may differ
R1: stratified sampling guarantees that each group is represented in proportion to its size, whereas a simple random sample could by chance under-represent a group
Accept: reference to the sample being more representative of the whole membershipTotal for question 2: 4 marks
- 3Eleven students record the time, in minutes, they take to finish a puzzle: 12, 14, 15, 15, 17, 18, 19, 21, 22, 24, 41. The student who took 41 minutes says she forgot to stop her timer. Use your GDC to find quartiles.(a)Find the interquartile range (IQR) of the times.[3 marks]
Show the mark scheme
Mark scheme
A1: lower quartile
A1: upper quartile
A1: minutes
Accept: follow-through from their quartiles
Do not accept: the range(b)Show that is an outlier, and comment on whether it should be kept in the data.[4 marks]Show the mark scheme
Mark scheme
M1: upper boundary
A1:
R1: , so is an outlier
Accept: follow-through from their IQR in (a)
R1: the value is a recording error (timer left running), so it should be corrected or removed rather than used in calculationsTotal for question 3: 7 marks
- 4A company has 2400 employees listed in a database. The HR manager takes a systematic sample of 60 employees for a questionnaire about hours worked. Of the 60 questionnaires sent out, 52 are returned. One returned questionnaire states 400 hours worked last week.(a)The first employee selected is number in the list.[6 marks]
(i) Find the sampling interval.
(ii) Find the position in the list of the fifth employee selected.
(iii) Find the position in the list of the last employee selected.Show the mark scheme
Mark scheme
M1: (i)
A1: (i)
M1: (ii)
A1: (ii)
M1: (iii)
A1: (iii)
Accept: follow-through from their interval(b)(i) Explain why the 8 missing questionnaires may bias the results.[6 marks]
(ii) Suggest one way of dealing with the missing data.
(iii) Explain how the reply of 400 hours should be treated.Show the mark scheme
Mark scheme
R1: (i) employees who did not reply may differ from those who did, e.g. those working the longest hours have least time to reply
R1: (i) so the sample no longer represents all employees and average hours may be underestimated
R1: (ii) follow up the non-respondents, or select replacements at random from the remaining employees
R1: (ii) any second valid detail, e.g. state the response rate and treat the results with caution
R1: (iii) there are only hours in a week, so is impossible (a recording error)
R1: (iii) check with the respondent and correct it, or remove it from the dataTotal for question 4: 12 marks
End of questions