Rss Ordinary Cert Module1 Print

EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY
(formerly the Examinations of the Institute of Statisticians)
ORDINARY CERTIFICATE IN STATISTICS, 2001
Paper I
Time Allowed: Three Hours
There is no restriction on the number of questions that a candidate may attempt,

nor on the order in which they are attempted. Candidates are not required
to answer all the questions: they should answer as many as they can.
The number of marks allotted to each question or part-question is shown in brackets.

The total for the whole paper is 100.
A pass may be obtained by scoring at least 50 marks.
Graph paper and Official tables are provided.
Candidates may use silent, cordless, non-programmable electronic calculators.
Where a calculator is used the method of calculation should be stated in full.
This examination paper consists of 8 printed pages. This front cover is page 1. The reverse of the front
cover, which is intentionally left blank, is page 2. Question 1 starts on page 3.
There are 8 questions altogether in the paper.

SS 2001
2
1. In England and Wales, the government carries out a longitudinal study called 2. You are employed in the marketing department of a national newspaper "The
the Youth Cohort Study. This work is designed to study young people from National Daily" that publishes Monday to Saturday. Your paper is running a
school leaving age (16) and tracks the routes they subsequently follow into prize draw. In return for supplying data about themselves and their reading
further and higher education and the labour market, documenting the training habits, readers aged 18 or over will be entered into a draw to win a valuable
and qualifications they receive. prize. You have been asked to design the form to be used in the newspaper as
an entry to the draw.
For one cohort, self-completion questionnaires were issued at ages 16, 17, 18
and 23. The percentage responses were: The form has to elicit the following information:
Age % response Name, address including postcode (or zip-code), telephone number,
16 77 email address, age group, on which days the reader usually buys "The
17 76 National Daily", which other national daily newspapers are bought
18 76 regularly during the week by the reader and which national Sunday
23 62 newspapers are bought at least once a month.
At the second and subsequent ages the figure was given as a percentage of
those who responded in the previous survey. Design a form that could be used for this purpose. Marks will be given for
clarity of layout and ease of use both by the readers and by those processing
(i) Define the terms longitudinal study and cohort. the completed forms.
(2) (12)
(ii) Explain how the initial choice of the cohort was, or could have been,
made. You should suggest a suitable sampling unit, sampling frame
and sampling design.
(4)
(iii) Calculate the percentage of the original sample responding at each age
correct to the nearest whole number. Comment on, and give a possible
reason for, the figures. 3. An opinion research company has been commissioned to conduct a survey of
(5) 2,000 voters in a town to estimate support for the candidates in a forthcoming
parliamentary election. There are 80,000 persons listed on the electoral roll.
(iv) List three groups of people who might be interested in the results of They are divided geographically into six wards that have distinct
such a survey. characteristics in terms of economic well-being. The company is considering
(3) three ways of selecting the sample of potential voters:
simple random sampling
stratified random sampling (stratified by ward)
quota sampling (quotas for each ward).
(i) Give two advantages and two disadvantages for each of these three
proposals.
(12)
(ii) It is nearly a year since the electoral roll was compiled. Explain what
consequences this may have for the survey.
(4)
3
Turn over 4
4. A local radio station carries out regular polls of its listeners on items of current 5. Explain what is meant by the term sampling fraction.
interest. In one such poll listeners were asked to telephone the station and just (1)
answer yes or no to the following question:
A population has three strata A, B and C. The numbers in the three strata and
Do you think dogs should be allowed in public places only if on the lead? the standard deviations of a variable of interest are given in the table.
The poll was carried out between 8 am and 9 am one morning. At 8.30 am the Number in stratum Standard deviation
announcer said that the percentage yes vote was 63%. When the poll closed at A 1000 10
9 am he announced that the percentage yes vote was 52%. B 4000 5
C 5000 2
(i) What additional number should have been announced at the end of the
poll in order to assess how accurate this percentage is? Explain briefly You are required to draw a sample of size 400 from the population.
why this number is needed.
(3) (i) Calculate the required numbers in the sample from each of the strata
using a uniform sampling fraction.
(ii) List three problems associated with this method of polling and suggest (4)
why each problem might cause misleading conclusions to be drawn.
(6) (ii) Calculate the required numbers in the sample from each of the strata
using sampling fractions proportional to the standard deviation in each
(iii) If respondents could have been asked one question about themselves stratum.
when they telephoned, suggest a suitable question which might be (6)
relevant to their response, and explain how conclusions from the survey
could have been extended using the extra information. (iii) Say, with reasons, which of the two sampling methods you would
(3) expect to lead to a better estimate of the mean of the variable.
(2)
5
Turn over 6
6. It has been suggested that shoppers are more likely to stay longer in a shop and 7. (i) List three types of non-response which may arise in an interview survey
spend more money if there is an agreeable aroma such as that of newly baked to be conducted at the homes of pre-selected interviewees.
bread, freshly ground coffee or a floral perfume. A researcher is testing this (3)
hypothesis in a small gift shop. She has permission to observe customers in
the shop on two days. On one day there will be no artificially induced scent (ii) Explain why non-response may lead to bias in the survey results.
but on the second day a floral perfume will pervade the store. (2)
The researcher needs to know how long a customer spends in the store, how (iii) Outline why it may be necessary to use randomly chosen substitutes in
much inspection of goods takes place before a decision is made as to whether the case of non-response and explain why using such substitutes does
or not to purchase, and the value of any goods purchased. There are normally not necessarily overcome the problems of non-response bias.
several customers in the shop at any one time but she cannot observe them all. (4)
Advise the researcher on how to carry out her observations. Your answer
should include the selection of customers for observation and the recording of (iv) Outline two other ways of trying to minimise non-response and one
the behaviour on a suitably designed form, together with a mention of any method of adjusting for non-response bias.
practical difficulties you foresee in the data collection process. (6)
(10)
8. A bank is redesigning its customer records system and intends to set up a new
database to help it to identify customers who may be interested in new
financial products. The database will contain data about the following
customer attributes:
Name and address including postcode
Home telephone number
Number and type of accounts (e.g. savings, current, loan) held

with the bank.
Design a database for this purpose; your answer should list the fields, giving
appropriate field names, together with field types and suitable widths.
(8)
7
Turn over 8
Read the following three paragraphs before answering any of questions 1, 2, 3
In some parts of England, local government consists of two levels. There are county
councils responsible for large areas, and within these there are borough councils, each
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY of which covers a small section of the county area. Borough councils are responsible
(formerly the Examinations of the Institute of Statisticians) for some public services.
One such borough contains 43 342 households and 103 456 residents. It is partly rural
and partly suburban, with two main township centres. It publishes a "newsletter" two
or three times a year, and this is distributed free to each household in the borough.
In June 2000, the newsletter contained a questionnaire about highway services (roads,
street signs and nameplates, covered shelters for passengers waiting for buses, etc) for
which the borough council is responsible. People were asked to return the
questionnaire to the council offices by post, or to telephone its Highway Services Unit
ORDINARY CERTIFICATE IN STATISTICS, 2002 with their response. A total of 52 people did so, and the results appeared in the April
2001 newsletter.
Paper I Questions 1, 2, 3 concern the responses to this survey.

1. A borough councillor has questioned whether the results can be taken as a
good representation of the views of residents in the borough.
Explain the concept of bias in surveys, and discuss ways in which the results
of this survey may be biased.
Candidates may attempt all the questions.
(8)
The total for the whole paper is 100. 2. Reliability of results means that if the survey were to be repeated in exactly the
same way, similar results could be expected.
Discuss whether the results of this survey are likely to have reliability in this
sense.
(5)
3. For this borough, explain how you would carry out a survey of residents'
opinions by each of the following methods. State the advantages and
Candidates may use silent, cordless, non-programmable electronic calculators. disadvantages of each method.
(i) A simple random sample using the register of voters and a postal
survey. (8)
(ii) A stratified random sample using the register of voters and personal
interviewers. (8)
1
(iii) A random sample selected from the local telephone directory,
interviewed by telephone. (8)
This examination paper consists of 4 pages, each printed on one side only.
This front cover is page 1.
Question 1 starts on page 2.
******************************************************
There are 8 questions altogether in the paper. 2

RSS 2002 Turn over
4. A survey is to be conducted to estimate the average daily time it takes for the 6. An education authority undertakes a survey of 700 parents, to determine their
residents of an area to travel to and from work. A pilot survey has been carried views on the availability of out-of-school facilities (such as sports and other
out, with the area divided into rural and urban parts. Data from the pilot leisure activities) and holiday projects for children in the area. A computer
survey were as follows. data-analysis program is available to enter and analyse the responses to the
survey questionnaire.
Stratum, Population stratum size, S.D. (estimate),
i Ni si (i) Explain:
Rural, i = 1 12 000 11
Urban, i = 2 80 000 8 (a) how the questionnaire should be designed so that data may
easily be transferred to the computer;
The sampling costs, ci, are thought to be about 18 per response in the rural
stratum and 5 per response in the urban stratum. (b) how the computer database should be set up;
(i) Using proportional allocation, the sample sizes ni in each stratum are (c) how the accuracy of the data input might be checked;
in the same ratio as the stratum sizes Ni in the population. Find the
numbers to be sampled from each stratum by this method if the total (d) how missing observations might be dealt with.
sample size is to be 500. Find also the cost of taking this sample. (11)
(5)
(ii) Discuss briefly the advantages and disadvantages of computer analysis
(ii) An alternative optimum allocation method minimises the variance of compared with analysis by hand of the data from this survey.
the estimate, and in order to obtain this ni must be taken proportional to (4)
N i si / ci . Also the total budget is 3600, of which 500 will be
required for fixed costs. Using the results from the pilot survey, find
the sample sizes n1, n2 needed by this method. 7. The UK Index of Retail Prices is used as a measure of price inflation, and
(9) other countries have similar measures. For the Index of Retail Prices or a
similar index of your own choice (you should say which index you are
(iii) Do you consider the optimum method sufficiently better than using):
proportional allocation to make it worth using for this survey? Justify
your answer briefly. (i) explain the way in which the general categories of expenditure are
(3) grouped, and how weights are allocated to them;
(ii) discuss the problems involved in collecting data on prices;
(iii) mention any difficulties in drawing conclusions from such an index

when considering expenditure in different types of household for
example those with children, those in various types of work, the
elderly.
5. Discuss reasons (other than variance estimation as used in question 4) for (12)
carrying out a pilot survey.
(7)
8. For EACH of the two methods quota sampling and cluster sampling, give an
example where the method is suitable. State clearly what populations are
being studied, what sampling frame is being used, and why the method is
better than alternatives.
(12)
3
Turn over 4
1. Describe the method of data collection, and the types of data collected, in
either the UK decennial population census or some other large national or
official population survey known to you. Who might use such data, and for
what purposes?
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY (15)
Questions 2 to 8 all relate to the following scenario. The county council for a
region of the United Kingdom has 3000 staff in a town-centre complex in a large
city. The central canteen for this complex at present has a main self-service area
for hot meals, and a salad bar. The canteen is not open to the general public but is
intended for council staff, who pay prices that are slightly subsidised by the
council. The head of the canteen service intends to carry out a survey.

2. (i) Briefly explain the concept of a target population for a survey.
(3)
Paper I
(ii) Suggest two possible target populations that the canteen head could
Time Allowed: Three Hours adopt for this survey, relating these to possible aims for the survey.
(6)
The following methods for conducting the survey have been suggested:
Candidates may attempt all the questions. A. Leave paper questionnaires near to the pay desks in the canteen with an
invitation to anyone to complete one.
The total for the whole paper is 100. B. Arrange for one of the canteen staff to conduct personal interviews with
a random sample of employees in their workplaces.
C. Send an email questionnaire to all staff (each member of staff has a
work email address).
Graph paper and Official tables are provided. D. Set up a display in the main entrance foyer of the central council
building, with questionnaires available on it for anyone passing to pick
up.
Candidates may use silent, cordless, non-programmable electronic calculators. E. Send out a questionnaire with a prepaid reply envelope to a stratified
sample of staff.
3. (i) Briefly explain the concepts of selection bias and response rate and
why they may cause problems in the interpretation of survey results.
1 (4)
(ii) Discuss how these concepts might relate to the above suggestions A to
This examination paper consists of 4 pages, each printed on one side only. E, noting the advantages and disadvantages of each suggestion.
This front cover is page 1. (12)

RSS 2003 Turn over
4. The strata referred to in method E are the four council departments listed: 7. A suitable questionnaire is to be designed for use in the survey.
Department Number of staff (i) It is thought that some staff, who have specialised diets for religious,
Education 1100 health or other reasons, feel unable to utilise the canteen. Draft one
Social Services 900 open question and one closed question that concern this issue,
Chief Executive's 320 explaining what your questions are trying to achieve and why you have
Environment and Resources 680 worded them as you have.
(6)
(i) Using simple proportional allocation, how many staff should be (ii) Explain the general difference between open and closed questions, and
sampled from each department to give a total sample of 450? discuss the relative advantages of each.
(4) (6)
(ii) Why might stratified random sampling be an improvement on simple (iii) Explain how the wording of a question may lead to biased responses.
random sampling of the entire population? Give an example of a possible question for this survey that is worded
(4) in a way likely to produce a bias, and give a rewording of the question
that you believe should avoid such bias.
(8)
5. (i) Discuss any other factors (such as cost, speed of completion or "need
for people to feel consulted") that might enter into the decision on the
sampling method to be used for this survey. 8. Explain the following in the context of this survey.
(4)
(i) The advantages and disadvantages of computer analysis using a data-
(ii) The head of the canteen service has heard that a multi-stage survey analysis package (such as SPSS or some other package known to you)
can be cheaper. Discuss situations in which this may be true, and compared with analysis of the data by hand.
whether it is likely to be relevant in the present instance. (4)
(4)
(ii) How the questionnaire might be formatted to enable easy data entry to
(iii) Assume that the aims of the survey are both to improve the service to the computer. [Optical Character Recognition is NOT available.]
existing customers and to extend the customer base. State, giving (3)
reasons, your recommendation for a sampling method to be used to
meet these aims. (iii) How the data analysis package might be used to help check the
(4) consistency and accuracy of the data entered. Give two possible
examples.
(5)
6. (i) Briefly explain the difference between longitudinal and cross-sectional

surveys, and how the distinction might apply in this proposed survey.
(4)
(ii) Explain how the canteen might use a panel for a longitudinal attitude
survey, and what the advantages and disadvantages of such an
approach might be.
(4)
3
Turn over 4
1. What are the advantages and disadvantages of quota sampling, as contrasted with
random sampling?
(10)
2. A survey about leisure activities is planned in a community with about 3000

households. A sample of households is to be taken and it is desirable that responses
are obtained from households of different sizes and types.
(i) A list of addresses for delivery of mail is available. Some addresses might
contain several households. Describe three potential problems in using a list of
ORDINARY CERTIFICATE IN STATISTICS, 2004 addresses to obtain a sample of households and suggest how these might be
overcome.
Paper I (6)
Time Allowed: Three Hours (ii) One method of obtaining a sample is to ask interviewers to collect information,
with the choice of household left to the interviewer. What instructions would
you give interviewers to help ensure that the sample obtained is representative
of households of different sizes and types? There is no need to repeat any
suggestions you made in part (i) as regards obtaining a sample of households
Candidates may attempt all the questions. from a list of addresses.
(6)
3. A pharmacist wishes to keep details of regular customers on a database to help ensure

that customers do not take medication which might have adverse effects. Design a
Graph paper and Official tables are provided. database suitable for this purpose. It should contain data about the following customer
attributes:
Name and address, including postcode/zipcode if there is one
Candidates may use silent, cordless, non-programmable electronic calculators. A contact telephone number
Name of doctor
Where a calculator is used the method of calculation should be stated in full. Up to two allergies of the customer
Medication prescribed with dates customer started and finished taking each
one.
Your answer should list the fields, giving appropriate field names, together with field
types and suitable widths.
1 OC Paper I 2004 (12)

SS 2004 Turn over
4. A survey is to be carried out among employees who work in a large city, to study 7. A survey of farms classified into three size groups (Small, Medium and Large) is to be
various aspects of their travel to work. A sample of employees will be asked to undertaken in a particular region. The cost of sampling each farm is the same,
complete a questionnaire. Information on where they work, how long it takes to travel regardless of size, and the budget is sufficient for sampling 40 farms. The main
from home to work, how far they travel, what method(s) of transport they usually use, objective of the survey is the estimation of the area under crops. The table shows the
how much (each part of) the journey costs, and whether there are any disadvantages in number of farms in each size group, and values of the standard deviation of the area (in
their present method(s) of travel will be sought. hectares) under crops obtained from a sample survey undertaken last year.
Draft a short questionnaire to obtain this information, using multiple choice questions Size group Number of farms SD of area (last year)
where appropriate. Small 203 6.4
(12) Medium 115 11.6
Large 82 27.3
Suggest three other questions that might be useful for someone seeking to improve the
experience of employees in their travel to work. (i) Find the numbers of farms to be sampled from each size group if the sample
(3) sizes are to be in the same ratio as the group sizes in the population of farms.
(4)
(ii) Calculate the numbers of farms to be sampled from each size group if the
sample sizes are proportional to "population size times the standard deviation"
in each group.
5. In dietary studies, subjects are sometimes asked to keep a diary for a limited period of (6)
time in which they record what foods and drinks they have consumed, and when.
Outline the advantages and disadvantages of collecting information using a diary rather (iii) Comment on the methods of sampling described in parts (i) and (ii), drawing
than relying solely on a questionnaire. attention to any advantages and disadvantages.
(10) (4)
8. A library has N books (where N is known), arranged in bookcases, each of which has
6. Discuss the issues involved in advising someone who plans to do a survey on what several shelves. A sample of about n books is to be taken in order to estimate the
size of sample they should take. Assume that the method of sampling is to be simple average age of books.
random.
(9) (i) Describe in detail how each of the following might be implemented:
(a) systematic sampling;

(4)
(b) cluster sampling.

(4)
Outline the advantages and disadvantages of each of these methods in this

application.
(5)
(ii) Would stratified sampling be a suitable method in this application? Justify

your answer.
(5)
3
Turn over 4
Questions 1 to 3 relate to the following situation.
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY A country is divided into 50 administrative districts. In 2006, the Education
(formerly the Examinations of the Institute of Statisticians) Ministry plans to undertake a sample survey into the careers and career plans of
those who left secondary school during 2003, and wishes to survey about 10% of
the relevant population.
The following information will be available about each school in the country: size
(number of pupils on 31 January 2006); district; names and contact details of
students who left the school in 2003, arranged in order of leaving date.
1. Your answers to this question should relate to the proposed survey.

ORDINARY CERTIFICATE IN STATISTICS, 2005 (i) Suggest two different two-stage sampling methods of selecting a suitable
sample. Each of these methods should use a different combination of simple
Paper I random, stratified, systematic and cluster sampling, and each of these four
methods of sampling should be used once only. You may assume that
Time Allowed: Three Hours sampling frames suitable for your purpose exist.
(6)
(ii) For each of the sampling methods that you have suggested, comment briefly on
any advantages and disadvantages it has for this survey.
(6)
The number of marks allotted to each question or part-question is shown in brackets. 2. Draft a short self-completion questionnaire to obtain the following information from
the members of the sample.
Month, year and age when secondary school was left.
For each of 2004 and 2005, the occupation and the district in which this was
undertaken. (If a respondent had more than one occupation during the year,
you should direct him or her to give information about the main occupation,
Graph paper and Official tables are provided. and the dates spent in it. Note that being a student is to be regarded as an
occupation; a respondent should be asked to specify the main subject studied.)
Long-term career plans.
(12)
Candidates may use calculators in accordance with the regulations published in 3. The information which is to be obtained from the respondents, as detailed in question
the Society's "Guide to Examinations" (document Ex1). 2, is to be stored in electronic form, together with contact details of the respondents.
Indicating which, if any, responses are to be coded:
either design a database suitable for this purpose. Your answer should list the fields,
giving appropriate field names, together with field types and suitable widths;
1 OC Paper I 2005 or design a spreadsheet suitable for this purpose. Your answer should list the
variables, giving appropriate variable names, together with their types and
This examination paper consists of 4 pages, each printed on one side only. suitable cell widths.
(12)

SS 2005 Turn over
6. Explain what is meant by interviewer bias. Give examples to show how this
Questions 4 to 8 relate to the following proposed design for a survey whose main might occur in the collection of prices in this survey, and suggest how these
objective is to determine the variation of prices of fruit and vegetables over time biases might be reduced.
and in different regions of a small country where a large proportion of the (10)
population experience rural poverty.
The country is divided into 7 regions, 4 mountainous and 3 coastal. One 7. A study of various aspects relating to a widely eaten staple vegetable was done
mountainous region and one coastal have been selected as being typical. The in another region of the country last year. In the table below, costs and prices
survey is to be of a number of markets taking place in rural areas in those two are measured in the local currency of the country. The table shows the
regions. The markets are considered to belong to one of two strata: large markets standard deviations, si, of the price per kilo of this vegetable, as found in the
and small markets. A sample of markets is to be taken from each stratum in each previous study. It also shows the numbers Ni of large and small markets in the
region. Local representatives at each market are to be asked for background coastal region to be used in the proposed survey, and the expected costs ci (in
information on production and supply. the local currency) of sampling one market of each type.
Each market in the sample is to be visited once a month over a period of four Stratum Number of markets SD of price of Cost of sampling
months. At each visit, prices of a selection of fruit and vegetables on sale that day (Ni) vegetable (si) one market (ci)
are to be collected by interviews with a number of merchants. The merchants Large markets 27 0.05 15
present in a market and the produce on sale are likely to change from day to day. Small markets 53 0.07 12
(i) Calculate the required number of markets in the sample from each of
the strata using a uniform sampling fraction to achieve an overall
4. (i) Suggest two advantages and two disadvantages of using a non-random method, sample of about 8 markets. What is the total cost of sampling markets
such as that described here, to select two regions. using these sample sizes?
(6) (4)
(ii) Discuss the difficulties, in the proposed survey, of (ii) The optimal allocation method of choosing sample sizes ni to estimate
a mean minimises the variance of the estimate and, in order to obtain
(a) selecting merchants to interview, this, the ni must be taken proportional to Ni si/ (ci). Calculate the
required number of markets in the sample from each stratum using this
(b) deciding for which fruit and vegetables prices should be obtained.
method if the total cost of sampling markets in this region is to be no
In each case, suggest how these difficulties might be addressed. more than 105 in the local currency.
(10) (8)
(iii) Which of the two methods of calculating sample sizes used in parts (i)
and (ii) do you prefer here, and why?
5. (i) Background information, about the production and supply of produce for sale, (4)
times of openings and numbers of merchants who have stalls, is required for
each market. Local representatives are to be approached to obtain this. What
advice would you give as regards to choosing suitable local representatives, and 8. The information to be collected in the survey is to be used to obtain a price
in collecting the information needed from them? index for fruit and vegetables. By making reference to an official index of
(6) prices with which you are familiar, or otherwise, suggest how this might be
done. Your answer should indicate what decisions need to be made in defining
(ii) Comment briefly on the difficulties of obtaining accurate details, such as last the index in detail, whether other information needs to be collected in order to
year's prices of products, in a survey undertaken this year. calculate it, and how it might be updated over time.
(4) (12)
3
Turn over 4
Questions 1 to 3 carry 41 marks in total, and relate to the following situation.
Farmers in a developing country are to be interviewed as part of a project
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY investigating the control of a pest XX which attacks a particular food crop. Farms
(formerly the Examinations of the Institute of Statisticians) are small and family run, and some are a considerable distance from the nearest
village. No sample frames or official data are available for the population of the
country, and there are no reliable large-scale maps, but basic geographical and
agricultural data are available on the 50 villages in which public meetings, to
generate awareness of the project, are to be held. Fifteen of these villages are to be
chosen before the public meetings to provide a sample of farmers to interview.
The numbers of farmers required in each of these villages are fixed beforehand.
Two methods, A and B, are proposed for selecting the farmers from the 15 chosen
villages. In method A, a random sample of farmers is to be selected from those
ORDINARY CERTIFICATE IN STATISTICS, 2006 attending the public meetings in the selected villages, and the farmers selected are
to be interviewed at the end of the meeting. In method B, interviewers are to travel
Paper I round the selected villages and nearby areas and to interview farmers at systematic
samples of farms.
1. (i) What general criteria should be considered when selecting the 15 villages in
which farmers are to be interviewed?
(3)
Candidates may attempt all the questions. (ii) Discuss any advantages and disadvantages of method A. Where possible
The number of marks allotted to each question or part-question is shown in brackets. suggest ways in which the disadvantages might be overcome.
(7)
(ii) Discuss any advantages and disadvantages of method B. Where possible
A pass may be obtained by scoring at least 50 marks. suggest ways in which the disadvantages might be overcome.
(6)
Graph paper and Official tables are provided. 2. Draft a questionnaire, which will be completed by an interviewer, to obtain the
following information from farmers. Your questionnaire should include detailed
instructions to the interviewer.
Name, age and highest educational level of all members of the household,
including the farmer
Candidates may use calculators in accordance with the regulations published in Relationship of every member of the household to the farmer
the Society's "Guide to Examinations" (document Ex1). The number of hours a week that each member of the household, including the
farmer, works on the farm at the time of year when the interview is held
The types of livestock owned by the farmer, and the numbers of each type
The crops grown by the farmer and the area under each crop
1 OC Paper I 2006 Whether the farmer thinks the risk of pest XX is high, medium, or low
What measures the farmer takes to control pest XX
This examination paper consists of 5 pages, each printed on one side only. (13)

RSS 2006 Turn over
3. The information obtained from farmers, as detailed in question 2, is to be stored in

electronic form.
(i) It is thought that farmers' households have between one and five members Questions 5 to 8 carry 45 marks in total, and relate to the following situation.
(including the farmer). What problems might occur in storing the information
relating to the different household members in electronic form as a result of A survey research organisation has been commissioned to undertake a study of
this variation in household sizes? how people in a particular industrialised country use their time. The country is
(4) divided into several regions, each of which is divided into a large number of
administrative districts, with each district classified as urban or rural. For each
(ii) The information relating to livestock, crops, pest risk and pest control is to be district, data on the total number of adults, the numbers in full-time employment,
stored in a separate file, in order to produce tables and diagrams for each and the numbers in part-time employment, are available for the previous year.
village. Explain how the responses to the questionnaire will help determine
what variables are required. State the type of each variable, and, where
appropriate, indicate the coding system to be used.
(8) 5. It is required to make comparisons of time use for adults in the three groups of full-
time employment, part-time employment and the remainder (neither in full-time nor
part-time employment), and by whether they live in rural or urban districts.
Question 4 carries 14 marks in total.
(i) Describe how each of stratified sampling and cluster sampling might be
4. A psychologist is studying the behaviour of children while they attend nursery school. implemented to choose a sample of administrative districts from which a
One of the main objectives of the study is the estimation of the mean number of sample of adults will be interviewed face to face, so that these comparisons can
minutes per hour that children interact with one another. The psychologist plans to be made. (Consideration of choice of sample size is not required.)
observe individual children for one hour each in three schools and to record the (5)
number of minutes, m, in the hour that the child under observation interacts with other
children. The cost of observing a child is the same in each school, and the budget is (ii) Which of the two methods of part (i) do you prefer here, and why?
sufficient to observe 20 children altogether. The table shows the number of children (3)
at each school, and the standard deviations of m obtained from a preliminary study in
these schools.
SD of m 6. Two ways of selecting a sample of adults to be interviewed about their use of time are
Nursery school Number of children
(preliminary study) under consideration.
1 25 10.2
2 30 7.7 Simple random sampling from a register of voters that was compiled a year
3 37 8.9 ago and which shows names and addresses
Quota sampling
(i) How many children should be sampled from each nursery school, if the sample
sizes are to be in approximately the same ratio as the numbers of children in Describe the non-sampling errors that might occur in each of these two selection
the three schools? methods in this application.
(4) (8)
(ii) How many children should be sampled from each school if the sample sizes
are to be approximately proportional to "population size times the standard
deviation of m" in each school?
(6)
(iii) How might you choose between the two methods of sampling described in
parts (i) and (ii) here?
(4)
3 4
Turn over Turn over
7. A sample of adults is to be asked to record their activities for each 10-minute period
during one week. For this they will be given a diary in which these 10-minute periods
are listed for every day. Respondents are asked to record both the primary activity EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY
and, where applicable, a secondary activity. (For example, the primary activity might (formerly the Examinations of the Institute of Statisticians)
be driving, with a secondary activity of listening to the vehicle's radio.)
(i) What are the advantages of collecting time-use data by a diary, over collecting
this information by a questionnaire?
(5)
(ii) What are the disadvantages of collecting time-use data by a diary, as opposed
to collecting this information by a questionnaire?
(7)
Paper I
8. The time use of children aged between 5 and 16 is also to be studied.
(i) It has been suggested that, in the case of younger children, an observational
study would be a suitable method of collecting information. Discuss the
practicalities of this method here, and whether or not reliable results are likely
to be obtained.
(6)
(ii) The firm conducting the survey decided to publish a questionnaire for older
children on the web. The number of marks allotted to each question or part-question is shown in brackets.
(a) Suggest and comment on two ways in which a sample of children
might be obtained. A pass may be obtained by scoring at least 50 marks.
(6)
(b) Explain why it is particularly important to pilot the questionnaire in

this case. Graph paper and Official tables are provided.
(5)
Candidates may use calculators in accordance with the regulations published in

the Society's "Guide to Examinations" (document Ex1).
1 OC Paper I 2007

5 ©RSS 2007
1. A holiday tour operator wishes to obtain feedback from those of its customers who 3. The manager of a hotel has been asked to find out what people staying there think
have been on one or other of its guided package tour holidays. Design a self- about its facilities. The hotel is used mainly for group bookings made by several
completion questionnaire, to be given to each holidaymaker, to obtain the following travel firms, but also by independent travellers. Groups travelling with different firms
information from respondents. [You should try to use "closed" questions where might arrive on different days, but those in a given group register on their arrival at
possible.] approximately the same time as one another. The manager is considering two ways of
choosing a sample of people to interview.
Which particular holiday they took
Taking a quota sample of residents during a particular week
Their sex and age group Taking a systematic sample of those who arrived during a particular week from
the registration records. These records are in order of registration.
How satisfied they were with the accommodation
(i) Write a clear list of instructions for the manager to use in order to carry out
How satisfied they were with the programme of excursions each of these methods. [You may assume that the manager has been told how
large the samples should be.]
How satisfied they were with the tour leader (6)
Whether they would recommend the holiday to others (ii) Outline any potential advantages and disadvantages each method has for
obtaining a sample of people staying at the hotel and for achieving a desired
What they liked best about the holiday number of interviews.
(8)
What they liked least about the holiday.
(12)
4. An administrator at a large hospital wishes to carry out a survey of the in-patients (i.e.
those who stayed at least one night) who are in the hospital on the night of 31 May
2007. Patients are allocated to beds in one of three types of ward, Observation,
Surgical Care, or Intensive Care. Two methods of choosing a sample of patients, who
2. In the context of question 1, a similar questionnaire contains an open-ended question will be asked to complete a short questionnaire, have been proposed.
asking respondents what kinds of holiday activities they like best. Some respondents
will name just one kind of activity, but others will name several. Many respondents (A) Take a simple random sample of wards, and then take a simple random sample
will name the same kinds of activity as others, but not necessarily using the same of patients from the selected wards.
terms. (B) Considering the wards of each type to be a single group of patients, take a
simple random sample of patients from each of the three groups.
(i) By giving examples of the kinds of responses that might be given to
this question, discuss what decisions are needed when coding the
responses and how the responses will help determine the coding (i) Give the technical name of each sampling method.
system. (2)
(7)
(ii) For each sampling method (A) and (B), state the conditions which would help
(ii) Explain how responses to this question in this open-ended form can ensure that the method works well.
help you reword the question in a closed form for use at a later date. (3)
(2)
(iii) For each sampling method, state one advantage and one disadvantage of using
it for this survey.
(4)
2 3
Turn over Turn over
5. There are three wards for patients requiring intensive nursing care in a particular 7. A careers advisor at a university is interested in finding out details relating to the
hospital. Wards A and B are in the main hospital building, and ward C is at a remote occupations of the university's graduates one, five, and ten years after graduation.
site. Samples of patients are to be taken from each ward and interviewed about State the advantages and disadvantages of collecting this information by
various aspects of their stay. The table below shows the number of patients (Ni) in
each ward, and the estimated cost (ci) in dollars of interviewing a patient in that ward. (i) a longitudinal study of a sample of last year's graduates,
It also shows the standard deviation (si) of the length of time (in minutes) that patients (7)
waited to see a nurse on admission to the ward, obtained from a survey of patients who
were in that ward in May last year (none of these patients is still in the hospital). (ii) sample surveys of those who graduated one, five, and ten years ago.
(6)
Ward Number of patients Cost of sampling one patient SD of length

(Ni) (ci) of time waited (si)
A 50 5 1.81
B 25 5 3.23
C 10 10 2.18
(i) Using a uniform sampling fraction, calculate the required number of patients to 8. (i) Explain why non-response is a problem in social surveys.
sample from each of the wards to achieve an overall sample of about 36 (6)
patients. What is the total cost of sampling patients using these sample sizes?
(4) (ii) List five reasons for non-response in a postal survey. By addressing these
reasons, suggest ways in which non-response might be reduced.
(ii) The optimum allocation method of choosing sample sizes ni to estimate a mean (8)
minimises the variance of the estimator, and, in order to obtain this, the ni must
be taken proportional to Ni si / ci. Calculate the required number of patients in
the sample from each ward using this method if the total cost of sampling
patients is to be no more than 200 dollars.
(9)
(iii) Which of the two methods of estimating sample sizes used in parts (i) and (ii)
do you think is better for selecting the sample of patients, and why?
(4)
6. (i) Present arguments in favour of conducting interviews over the telephone

instead of face to face.
(6)
(ii) Why might telephone interviews be less successful in some circumstances than
face to face interviews?
(6)
4
Turn over 5
1. A professional organisation wishes to discover its members' opinions about its

EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY monthly newsletter. Suppose the organisation decides to mail a self-completion
questionnaire to a random sample of members. Draft such a questionnaire, with
instructions on how to complete it, and a suitable covering letter. Questions are to be
asked on the following topics.
Year of joining
The member's sex and current employment status (e.g. student, full-time
employee, etc.)
Whether the member reads the whole newsletter
ORDINARY CERTIFICATE IN STATISTICS, 2008 What the member does with the newsletter when he or she has finished with it
What the member thinks of the special features
Paper I
Whether advertisements should be included
Whether changes are needed to the frequency of publication and, if so, what
changes
General satisfaction with the newsletter
Your questionnaire should include at least one closed question, at least one open
Candidates may attempt all the questions. question, and at least one question to which the response is to be given on a rating
scale.
The number of marks allotted to each question or part-question is shown in brackets. (13)
2. The organisation referred to in question 1 has approximately five thousand members.
Instead of mailing the questionnaire to a random sample of them, it considers
including the questionnaire as an insert in the newsletter, sent monthly to all members.
Graph paper and Official tables are provided. Discuss the advantages and disadvantages of these two methods.
(12)
3. A professional organisation has three grades of member: student, ordinary, and

Candidates may use calculators in accordance with the regulations published in retired. A spreadsheet gives an alphabetical list of members, along with the years they
the Society's "Guide to Examinations" (document Ex1). joined the organisation and their current membership grades. Two ways of taking a
systematic sample of members are under consideration: the first uses the alphabetical
list, the second uses the list re-ordered by year of joining.
(i) Describe, with examples, the potential advantages and disadvantages of each
of the two possibilities.
(8)
1 OC Paper I 2008
(ii) Would there be any advantage in taking a systematic sample with the list
This examination paper consists of 4 pages, each printed on one side only. ordered by grade of membership? Justify your answer.
This front cover is page 1. (3)

©RSS 2008 2
1 Turn over
4. The professional organisation referred to in question 3 is interested in monitoring the 6. The marketing director of a chain of supermarkets would like to set up an interview
annual amount, Y euros, spent by members on conference attendance. The table below survey of a sample of customers. The supermarkets are located in several different
shows the number of members (Ni) in each grade and the standard deviations (si) of Y regions, and are classified into three types according to size. A sample of
obtained in a survey two years ago. The cost of obtaining information from a member supermarkets is to be taken, and samples of customers will be drawn from these
is the same for all grades, and there is sufficient money in the budget to sample up to supermarkets.
95 members.
(i) Suggest how cluster sampling could be used to choose the sample of
Membership Number of members (Ni ) SD (si ) of Y from previous study supermarkets. Discuss any potential advantages and disadvantages of this
grade method for the proposed survey.
Student 141 226.21 (5)
Ordinary 782 550.12
Retired 239 303.60 (ii) Suggest how stratified sampling could be used to choose the sample of
supermarkets. Discuss any potential advantages and disadvantages of this
(i) How many members should be sampled from each grade if the sample sizes method for the proposed survey.
are to be in approximately the same ratio as the numbers of members in the (6)
three grades?
(4)
(ii) How many members should be sampled from each grade if the sample sizes
are to be approximately proportional to Ni si in each grade? 7. Customers in the supermarkets referred to in question 6 are to be interviewed in store.
(6) The interviewers are provided with a questionnaire, and are told how many customers
to select. After they select the customers, the interviews will take place immediately.
(iii) Why might the method described in part (i) be preferable to that of part (ii)?
(4) (i) What advice would you give the interviewers as regards selecting customers to
interview?
(7)
(ii) What advice would you give the interviewers as regards the conduct of the
interview?
(6)
5. (i) Distinguish carefully between a pilot survey, a sample survey and a census,
giving reasons why each might be done.
(9)
(ii) Give reasons why results from a census might differ from the true values in the 8. Discuss how an observational study in supermarkets might be undertaken to find out
population. Wherever possible, illustrate your answer with examples from a what kinds of customers buy organic vegetables and fruit, and what particular organic
census with which you are familiar. vegetables and fruit they buy. Outline the advantages and disadvantages that this
(5) method has in this context over the use of a questionnaire.
(12)
3
Turn over 4
1. A country is divided into several regions, each of which is further divided into a large
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY number of administrative districts, with each district classified as urban or rural. Lists
of adults living in private households six months previously are available in each
district. These lists are arranged by address.
(i) Suggest two different two-stage sampling methods of selecting a sample of

adults from the available lists. Each of these methods should use a different
combination of simple random, stratified, systematic and cluster sampling, and
each of these four methods of sampling should be used once only.
(6)
ORDINARY CERTIFICATE IN STATISTICS, 2009 (ii) For each of the sampling methods you have suggested, comment briefly on any
benefits and drawbacks it might have in the situation under consideration.
Paper I (6)
Time Allowed: Three Hours (iii) Suggest how the sample of adults in private households might be
supplemented to obtain a sample of all adults. State any potential problems
with the procedure you suggest.
(4)

2. A national survey is being planned in which adults living in private households will be
A pass may be obtained by scoring at least 50 marks. interviewed in order to obtain their perceptions of crime. Design a questionnaire,
which should include detailed instructions to the interviewer, to obtain the information
listed below. You may assume that potential respondents have been chosen at random
from a list giving their names and addresses and have been sent a letter telling them
Graph paper and Official tables are provided. about the survey.
Age group, sex, marital status and occupation of respondents

Whether they think crime is a problem in their neighbourhood
Whether they think the level of crime in their neighbourhood has
Candidates may use calculators in accordance with the regulations published in increased in the last five years
Whether they think the level of crime in the country as a whole has
increased in the last five years
Whether they think violent crimes are on the increase in the country as
a whole
Their views on violent crimes committed by those aged under 16 years
of age
1 OC Paper I 2009 (13)

©RSS 2009 2
1 Turn over
3. A survey is being planned to find out whether people have been victims of crime and 6. All surveys are subject to several sources of error. Describe the different sources of
to obtain details of any crimes involved. Discuss whether it is better to collect error that might arise in an interview survey of a sample of cruise passengers selected
information by means of a telephone interview or by a self-completion questionnaire. at random from passenger lists.
(10) (12)
7. Explain the difference between questions in open format, closed format with a single
4. Describe answer, and closed format with multiple answers. Illustrate your answer by drafting
one question in each format for a survey of cruise passengers. One question should
(i) the advantages, ask them how many cruises they have gone on previously, one should ask them what
newspapers they read at least once a week, and one should ask them if they hope to go
(ii) the disadvantages,
on another cruise within the next two years.
of quota sampling as compared with simple random sampling. (10)
(6 marks for each of part (i) and part (ii))
8. The owners of a chain of newsagents wish to use sample data to monitor changes in
the amount of money taken from customers each month from the sale of magazines
5. The marketing manager of a travel firm wants to carry out a survey of the 750 and newspapers.
customers who had by 31 December booked to go on a cruise during the following
year. A stratified sampling scheme is to be used in which there are three strata defined (i) What data do you think would be suitable for this purpose, and what decisions
by the type of cabin booked: budget booked by 377 customers, standard booked by would be needed concerning the data to collect and when to collect such data?
303, and de luxe booked by 70 customers. No customer booked to go on more than (4)
one cruise. An overall sample size of about 150 is required.
(ii) Suggest how the data collected might be summarised, and how summaries
One of the most important quantities to be estimated is the mean annual income of might be updated over time.
customers who book to go on cruises. Published results of a survey of passengers on (3)
cruises last year give £8850, £13005 and £4695 as estimated standard deviations of
annual income for all passengers in budget, standard and de luxe cabins respectively. (iii) Devise a sampling scheme which would enable suitable information to be
collected for the summary you have suggested in (ii).
(i) How many customers booking each type of cabin should be sampled, if the (6)
sample sizes are to be in approximately the same ratio as the numbers of
customers booking each type of cabin?
(4)
(ii) How many customers booking each type of cabin should be sampled if the
sample sizes are to be approximately proportional to "population size times the
standard deviation of annual income" in each group of customers?
(6)
(iii) Comment on whether there are likely to be any problems with the methods of
parts (i) and (ii) in this survey.
(4)
3
Turn over 4
1. (i) Briefly explain the concepts of selection bias and response rate and why they
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY might give rise to errors or difficulties in the interpretation of survey results.
(4)
(ii) A chain of self-service restaurants has outlets throughout a country. A survey

of customers' opinions of the restaurants is planned and the following two
methods for conducting the survey in any outlet have been suggested.
A. Leave paper questionnaires near to the pay desks with an invitation to

customers to complete one.
ORDINARY CERTIFICATE IN STATISTICS, 2010 B. Arrange for a member of staff to interview customers at a random
selection of tables as determined in the office.
Paper I
Discuss how the concepts of selection bias and response rate might relate to
Time Allowed: Three Hours the methods A and B above, noting the advantages and disadvantages of each
suggestion.
(8)

2. The chain of restaurants described in question 1 is in a country that is divided into
eight regions, each containing both large and small urban areas. There are between
ten and fifteen outlets in every region and these outlets are in both small and large
Graph paper and Official tables are provided. urban areas. Small urban areas have only one outlet, but large urban areas have
between two and five outlets.
(i) Devise a multi-stage sampling method to select outlets in which customers are
to be surveyed. State the technical method used at each stage of the scheme.
(4)
the Society's "Guide to Examinations" (document Ex1). (ii) There are between 6 and 15 tables at outlets. Describe two methods for
selecting tables at which customers will be interviewed, commenting on the
sample size.
(6)
1 OC Paper I 2010

©RSS 2010 2
1 Turn over
3. A survey is being planned to investigate stress at work among teachers in colleges 4. There are three colleges, A, B and C for students aged 16–19 in a city. The table
where the students are aged between 16 and 19. Design a self-completion below shows the number of teachers (Ni) in each college, the estimated cost (ci) in £ of
questionnaire, with instructions on how to complete it, and a suitable covering letter, travelling to and interviewing a teacher at each college, and the standard deviation (si)
to obtain the following information. of the number of years teachers have been at the college.
The teacher's sex, age group, subject taught and position held in the college Number of Cost of sampling one teacher SD of years at college
College teachers ci si
What arrangements there are for flexible working Ni
A 307 5 7.5
The extent to which the teacher B 200 10 2.8
feels that the working hours are excessively long C 103 7 5.3
feels that his or her work makes a valuable contribution to society
(i) Using a uniform sampling fraction, calculate the required number of teachers
feels valued at work to sample from each of the colleges to achieve an overall sample of about 150
feels stressed at work teachers. What is the total cost of sampling teachers using these sample sizes?
feels the pay is adequate (4)
How likely the teacher is to leave the sector during the next year (ii) Among several objectives, it is required to estimate the mean number of years
teachers have spent at the colleges. The optimum allocation method of
choosing sample sizes ni when estimating a mean minimises the variance of
Your questionnaire should include at least one closed question, at least one open the estimate, and, in order to achieve this minimum, the ni must be taken
question, and at least one question to which the response is to be given on a rating proportional to Nisi/ ci. Calculate the number of teachers required in the
scale. Indicate which questions are of each type. sample from each college using this method if the total cost of sampling
(13) teachers is to be no more than £1050.
(10)
(iii) Discuss, with reasons, which of the two methods of calculating sample sizes
used in parts (i) and (ii) you prefer here.
(4)
5. A researcher is interested in finding out whether the views of teachers concerning

stress at work at a large college are different one, two and five years after joining the
college. State the advantages and disadvantages of collecting this information by
means of
(i) a longitudinal study of a sample of teachers who joined the college last year,
(7)
(ii) sample surveys of teachers who joined the college one, two and five years ago.
(6)
3 4
Turn over Turn over
6. Explain what is meant by a pilot survey and discuss in detail reasons for doing one.
(11) EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY
7. A survey is being planned in a community with about 5000 households. A sample of

households is to be taken and it is desirable that responses be obtained from
households of different sizes and types.
(i) A list of addresses for delivery of mail is available. Some addresses might ORDINARY CERTIFICATE IN STATISTICS, 2011
contain several households. Describe three potential problems in using a list of
addresses to obtain a sample of households and suggest how these problems
might be overcome.
(6) Module 1 : Collection and compilation of data
(ii) One method of obtaining a sample is to ask interviewers to collect information, Time allowed: Three Hours
with the choice of household left to the interviewer. What instructions would
you give interviewers to help ensure that the sample obtained is representative
of households of different sizes and types? There is no need to repeat any
suggestions you made in part (i) as regards obtaining a sample of households Candidates may attempt all the questions.
from a list of addresses.
(5) The number of marks allotted to each question or part-question is shown in brackets.
8. It has been suggested that shoppers are more likely to buy goods that are placed at eye
height than those which are low down on the shelves. To investigate this suggestion, Graph paper and Official tables are provided.
the owner of a small supermarket has decided to place the cheaper brands of tinned
goods at eye height for two weeks with the more expensive goods at the lowest
position, and in the following two weeks to reverse these positions. A researcher has
been asked to observe shoppers' behaviour at these shelves. What advice would you
give on what to observe, how to select shoppers to observe, and how to record the
information? Mention any practical problems you foresee in the collection of data. Candidates may use calculators in accordance with the regulations published in
(12) the Society's "Guide to Examinations" (document Ex1).
1 OC Module 1 2011
This examination paper consists of 5 printed pages, each printed on one side only.

© RSS 2011
5
1. Describe a price index in a country of your choice, for instance the Retail Prices Index 3. An estimate is required of the mean value of a variable defined on a large population.
in Great Britain.
The population contains two strata in proportions p : (1 – p), where 0 < p < 1.
Your answer should include comments on the following.
(i) Explain what is meant by strata.
• How often it is published (3)
• How the data are obtained
• How the index is constructed (ii) Give an example of such a population.
(2)
• How and why the constitution of the index is changed over time
(9) A random sample of size n is to be drawn, comprising n1 from stratum 1 and n2 from
stratum 2, where n1 + n2 = n. For each stratum, the sample means, x1 and x2 , and the
Give three ways in which the index is used by different groups for different purposes.
sample standard deviations, s1 and s2, will be computed.
(6)
The formula for the estimate of the population mean is x px1 (1 p ) x2 .
An estimate for the standard deviation of x (the standard error of the mean) is
p 2 s12 (1 p ) 2 s22
se .
n1 n2
2. You are interested in the mean value of a measure in a given population. The
population is large and you decide to estimate this mean using a simple random (iii) Supposing that n1 = 20, n2 = 10, p 2
3
, x1 5 , x2 8 , s1 = 1 and s2 = 3,
sample from the population. compute x and se.
(4)
(i) What would you compute to estimate the population mean?
(1) (iv) Supposing that n1 = 12, n2 = 18, p 2
, x1 5 , x2 8 , s1 = 1 and s2 = 3,
3
compute x and se.
(ii) Explain in detail what is meant by bias in sampling.
(4)
(2)
(v) Given the evidence of your computations, should strata sample sizes be chosen
(iii) Why are the sample members chosen randomly?
in proportion to strata sizes in this example? Justify your answer.
(1)
(2)
(iv) Explain what is meant by sampling accuracy and how it is measured.
(3)
(v) How can sampling accuracy be improved?

(1)
(vi) Describe the circumstances in which the following sampling methods might be
used, and say what advantage they each have over simple random sampling.
(a) Stratified random sampling.
(b) Cluster sampling.

(7)
2 3
Turn over Turn over
4. The management of a package tour company wishes to gather information from 6. A researcher is interested in studying the side effects of a particular brand of pain
customers who have completed their holidays. You are to advise on how this may best killer. She can question individuals who use this brand about their experiences, or she
be achieved. can set up a clinical trial to measure any observable side effects.
You should specify the questions to be asked and the forms of the answers that are Describe what is involved in following each of these approaches. Compare and
required, but you need not produce a design for how the questionnaire might appear. contrast them.
(15)
Your answer should include the following.
• Your recommendation regarding the mechanism by which the data should be

collected
(3)
• No more than 5 questions to determine appropriate personal details

(3)
7. Identify a potential source of bias error in each of the following proposed sampling
methodologies. You should also specify the type of sampling that is involved in each
• No more than 5 questions to determine appropriate details about the holiday
case.
(3)
(i) To investigate energy consumption, a researcher sends out interviewers to 25%
• No more than 10 questions to determine customer experiences and level of of the houses in a neighbourhood. He selects the sample by choosing every
satisfaction with appropriate aspects of the holiday fourth house listed on the electoral register.
(4)
(ii) A market research company employs interviewers to conduct telephone
• How you might persuade customers to complete questionnaires, and an surveys. Each interviewer selects residential numbers at random to ring during
explanation of why this is important working hours until that interviewer's quota of respondents have agreed to be
(2) interviewed.
Why might it not be necessary to collect personal and holiday details? What would be (iii) Employees of a large corporation are e-mailed invitations to take part in job
the advantage of not collecting these details? satisfaction surveys.
(2)
(iv) Machines in a factory are chosen at random on two separate days and the
amount of wear in a specific component is measured. Components are
replaced when wear reaches a certain level but some components reach that
level more quickly than others. The intention is to estimate the average rate of
wear.
(15)
5. A restaurant in an Alpine ski resort has an outdoor seating area on the slopes,
separated from the restaurant by a service road. To improve efficiency, the waiters
and waitresses are equipped with hand-held electronic devices with wireless
capability. Staff enter customer orders on their handsets, and the orders are instantly
transmitted to the kitchen.
Apart from shortening the time between a customer placing an order and receiving his
or her meal, the handsets provide electronic sales data that could help the business of
the restaurant in various ways. Give four areas of the business in which such data
might be useful, and describe briefly how the data could be used in each area.
(8)
4 5
Turn over
1. Answer the following questions in the context of a real survey of which you are aware.
(i) Briefly describe the survey in terms of its purpose, the target population, the
sampling procedure, and the methods of collecting and processing the data.
(5)
(ii) When carrying out a survey there are many sources of error. Identify the
possible sources of error in your chosen survey and describe the strategies used
to try to minimise them. State, giving reasons, whether you think these
strategies were adequate and suggest any modifications.
(9)

The manager of a hotel wants to use a survey to find out clients' opinions about the hotel.
Questions 2 5 relate to this survey.
Module 1 : Collection and compilation of data
Time allowed: Three Hours

2. The manager needs to know the following.
The age and sex of the client
The client's reason for needing hotel accommodation
The client's satisfaction with the booking procedure, the room and the meals
The number of marks allotted to each question or part-question is shown in brackets. What the client liked best about the stay
What the client liked least about the stay
Any other comments that the client may have.
Using closed questions where possible, design a self-completion questionnaire to elicit
this information.
(12)
3. Use examples from the questionnaire you designed in question 2 to illustrate your
answers to the following.
Candidates may use calculators in accordance with the regulations published in (i) Explain why it is generally easier to code the answers from closed questions
the Society's "Guide to Examinations" (document Ex1). than from open questions.
(6)
(ii) The hotel manager ran the survey in 2011. In his questionnaire he used several
open questions, but found it very difficult to analyse these. He decides that he
will repeat the survey in 2013 in order to compare results for the two years.
Explain how he could replace open questions with closed questions where
1 OC Module 1 2012 possible. Discuss the advantages and disadvantages of what you recommend.
(6)
This examination paper consists of 4 printed pages.

© RSS 2012 2
4. The manager is undecided about ways of administering the questionnaire. In 7. A health organisation wants to run a survey to determine what foods mothers are
particular he is unsure whether to feeding to infants less than one year old.
leave questionnaires in the hotel rooms or post them to clients' home addresses
Discuss the practical problems of obtaining this information in a country with which
give the questionnaire to all clients who book in during a 2-month period, or to you are familiar and how they could be overcome for different types of location, such
a proportion of clients who book in during a full year. as a large town or a rural area.
(8)
Discuss the advantages and disadvantages of these approaches.
(6)
8. An organisation wants to gather information on its employees, 90% of whom are male.
It has chosen to use interviews, which will be quite long, and will be conducted during
working hours. Some of the questions are sensitive, and people can decline to answer
5. It is suggested that people often do not like self-completion questionnaires and so
any subset of them.
clients should be interviewed, either face-to-face or by telephone.
The average salary of the women is lower than that of the men, so it is more costly to
Discuss the advantages and disadvantages of self-completion questionnaires, face-to- interview the men. It is estimated that an interview with a male employee will cost
face interviews and interviews by telephone for this particular survey. What would £50 on average, and that for a female employee £30. The total budget for the
your advice be in this case? interviews is £1000. The organisation has been advised to use stratified sampling.
(6)
(i) Explain what is meant by stratified sampling and why it might be important
here.
(3)
6. (i) Two market research companies have run similar surveys on comparable
populations. They used the same questionnaire and cluster sampling. (ii) How many male and female employees could be interviewed if a uniform
sampling fraction is used?
Company A sent out 5000 questionnaires and received 1000 responses. (2)
Company B sent out 1500 questionnaires and received 750 responses.
Company A claims that their survey is better because they have a larger (iii) It is suggested that a uniform sampling fraction might not be the most
number of responses thereby enabling them to calculate better estimates from appropriate method, and that instead the following formula (in the usual
the survey data. notation) should be used:
ni Si
Comment on whether you think this claim is justified, giving your reasons. i 1, 2.
Ni Ci
(3)
(ii) A researcher undertaking a postal survey has been advised that she needs 100 (a) Explain what is meant by each of the terms in the formula. (2)
responses in order to estimate mean values with the required precision.
(b) No one in the organisation knows the values of Si. What advice would
(a) Explain what is meant by precision, and why you might expect it to be you give? (4)
lower if fewer than 100 responses were obtained. (3)
(c) Assuming that S1 = S2 , use the given formula to calculate the number of
(b) The researcher believes that the response rate to the questionnaire will male and female employees that could be interviewed. (4)
be about 60%. She does not know whether to send out 167
questionnaires in order to get about 100 back, or to send out fewer (iv) Further advice is that, if the mean values of some of the answers for men and
questionnaires, followed by reminders to people who do not respond to women are to be compared, there should be approximately similar numbers of
the first questionnaire. each in the sample. How many men and women could be interviewed using
this rule?
Discuss the relative merits of these two approaches. (5) (2)
(c) Discuss briefly other strategies that the researcher could use to try to (v) Overall what would your advice be about the numbers of men and women to
minimise non-response to the questionnaire. (6) interview in this study?
(8)
3 4
1. (i) Explain the difference between a census and a sample survey.
(2)
(ii) In some countries it is a legal requirement to carry out a census, in order to

determine the demographics and working patterns of all residents. Choose a
country with which you are familiar and discuss the following aspects of such a
census in that country.
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY (a) How the population could be identified.
(b) How each person could be contacted.
ORDINARY CERTIFICATE IN STATISTICS, 2013 (c) Examples of types of people who might be difficult to contact, or who
may not be able to take part, and how these problems could be handled.
MODULE 1 : Collection and compilation of data (d) Examples of possible biases in the responses, and how these could be
minimised.
Time allowed: Three Hours (17)

2. A database contains 100 records with identifiers A1–A20, B1–B40, C1–C10 and
The number of marks allotted to each question or part-question is shown in brackets. D1–D30.
The total for the whole paper is 100. You are required to draw a stratified sample. This should comprise 5 records
randomly selected from each stratum, where the letter prefixes (A–D) identify the four
A pass may be obtained by scoring at least 50 marks. strata.
Use the random numbers below to draw a suitable sample. You must provide
sufficient detail to ensure that your procedure can be fully understood.
92 17 63 21 58 64 70 73 31 41
09 08 63 23 26 66 85 20 26 60
56 84 65 72 51 65 78 74 06 61
06 02 78 90 90 63 98 83 47 84
Candidates may use calculators in accordance with the regulations published in 11 02 46 24 06 39 73 47 98 87
the Society's "Guide to Examinations" (document Ex1). 00 60 52 89 65 05 95 16 49 66
57 15 62 63 50 60 27 95 64 31
70 30 11 74 31 41 04 94 60 18
63 02 79 09 44 01 91 75 88 50
95 99 09 98 74 47 71 77 46 93
(8)
1 OC Module 1 2013


© RSS 2013 2
3. A survey is to be carried out on makers of organic cheese in England. The aim is to 5. The following extract appeared in a report of a survey carried out in India.
assess their training needs in relation to their knowledge of food health and safety
laws. Cheese makers are a very busy group of people who start work early in the "The sampling was carried out as follows. In the first stage, 2000 clusters were chosen
morning and often work right through the day. They work in controlled environments, from 3 strata: 400 from urban areas, 200 from the tea estate sector and 1400 from rural
widely scattered throughout England, to which the general public are not normally areas. The second stage involved systematic selection of 10 houses from each cluster.
allowed access and which they cannot leave for long. You have been given the names, In this way we assured that the results of the survey were representative of the whole
addresses and telephone numbers of the people in the sample, together with the country."
questions to be asked. Some questions are closed and some are open. You have a
fixed budget and a short timescale. (i) Define the terms clusters and strata.
(4)
(i) Outline the advantages and disadvantages of each of the following methods for
obtaining the required information in this survey. (ii) What other information would you require before you could comment on how
representative the results are likely to be?
(a) Interview at a visit. (7)
(b) Postal questionnaire.
(c) Telephone interview.

(18) 6. A sports club offers a gym, a swimming pool, aerobics and a café. It is possible to buy
annual membership. The manager wants to survey a sample of 100 annual members
(ii) State which method you would recommend. For your chosen method provide to find out their views of the club. He has designed the following questionnaire.
details of how the cheese makers would be contacted and how the information
would be acquired. Justify your decisions. XX Sports Club
(7)
Dear Member, please complete this short questionnaire so that we can improve
facilities. Put it in the box at reception. Thank you for your time.
What is your date of birth? ....................................

4. An audit has been commissioned in a hospital. The aim is to estimate the frequency of
occurrence of a particular adverse event for the whole hospital over a 12-month Are you male or female? ........................................
period. The people carrying out the audit have been advised that they do not need to
visit every ward, but that they can select wards and also months in which to carry out For how many years have you had annual membership at this club? ...........................
the audit.
Which facilities do you use most? ................................................................................
(i) Identify the type of sampling used, giving reasons for your answer.
(3) Do you consider annual membership value for money and will you renew it? ............
(ii) Discuss whether you think this is an appropriate method of sampling for this
audit. (i) Identify problems with this questionnaire.
(8) (10)
(ii) Re-design the questionnaire to obtain similar information, but addressing the
problems you identified.
(5)
3 4
7. (i) Briefly describe the problems caused by missing data in questionnaires
returned in a sample survey.
(2)
(ii) A questionnaire about library usage contains the following two questions.
A. Currently the library is open late on three days. Which suits you
best? Please tick only one answer.
Monday
Wednesday
Thursday
No preference
B. Which of the following facilities do you use? Tick all that

apply.
Photocopier
Computers BLANK PAGE
Reading room
DVDs
(a) Suggest a reason for including the answer option 'No preference' in
question A.
(b) Explain why it would be impossible to identify all missing data for
question B, and re-design the question to overcome this problem.
(c) Describe how answers to the two questions could be coded in a

statistical worksheet, paying attention to how you would code missing
values. Justify your decisions.
(9)
5 6
BLANK PAGE BLANK PAGE
7 8
1. (i) Define strata in the context of sampling.
(2)
A database comprises 1000 records of three different types with the prefix A, B or C.
The records are labelled as follows.
A1 A500
B1 B100
C1 C400
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY A sample of 100 records is required.
(ii) Explain why a simple random sample is not appropriate in this situation.
(2)
(iii) The following sample is drawn.
MODULE 1 : Collection and compilation of data A10 A20 A30 A40 A470 A480 A490 A500 (sample of size 50)
B10 B20 B30 B40 B80 B90 B100 (sample of size 10)
Time allowed: Three hours
C10 C20 C30 C40 C370 C380 C390 C400 (sample of size 40)
How does this sample appear to have been obtained? Describe any problems
Candidates may attempt all the questions. with this method.
(3)
(iv) What would be your advice about the best way to select an appropriate sample
The total for the whole paper is 100. of 100 records?
(3)

1 OC Module 1 2014


© RSS 2014 2
2. In a survey of dog owner attitudes and their experiences in managing diabetic dogs, 3. (i) Distinguish between open and closed questions, stating the advantages and
questionnaires were distributed to a sample of 36 veterinary clinics. At each clinic, the disadvantages of each method of obtaining data in a survey.
questionnaire was given to all dog owners whose dogs were undergoing treatment for (7)
diabetes.
The manager at a garage wants to obtain views of customers who have had their cars
(i) Which method of sampling was used in this study? Discuss the merits of this serviced at the garage. He wants to ask about the following.
approach for this survey.
(5) The facilities for arranging the service, communicating with the garage
staff, and paying for the work
(ii) Some owners had more than one dog undergoing treatment for diabetes, and The quality of the work carried out
these owners received one questionnaire for each dog. Discuss the
implications for the results of the survey if some owners completed several Any other comments the customer may have
questionnaires.
(2) (ii) Justifying all your decisions, design a questionnaire that could be used for this
purpose, and describe how you would administer this questionnaire.
(iii) 93 questionnaires were returned. What extra information would you need in (10)
order to be able to assess whether or not this was a good response rate?
(3)
(iv) The developers of the questionnaire tried to ensure that it was straightforward
to read and complete. Describe the features of the questionnaire which you 4. People designing sample surveys often ask the question "How many replies do I need
think would be important in order for it to be straightforward to read and for my sample to be representative?"
complete.
(8) (i) Explain what is meant by representative and why it is desirable for a sample
to be representative.
(v) The questions were modified following responses from a pilot study of ten dog (2)
owners.
(ii) Outline the advantages of receiving a large rather than a small number of
(a) What is the purpose of piloting in a sample survey? responses in a sample survey.
(1)
(b) Contrast the following two methods of piloting the questionnaire.
(iii) Draft an answer to the original question.
Asking dog owners to complete it and then asking them how (3)
straightforward it was
Talking to owners as they complete the questionnaire
(c) What other aspects of this survey would you consider important to
pilot? Justify your answer. 5. A researcher wants to carry out a postal survey of people who have made use of a
(9) counselling service. Some of the questions are of a sensitive nature, and so the
researcher believes that it would be best to keep the responses totally anonymous, thus
ensuring that the identities of respondents cannot be linked to any of their answers.
Comment on whether you think this is sensible or not.
(4)
3 4
6. (i) Distinguish between cross-sectional and longitudinal surveys. Give an 7. Based on experience in her clinic, a doctor believes that women who report eating a lot
example of a question that could be investigated by a longitudinal survey but of a certain foodstuff during pregnancy tend to have shorter labours than women who
not by a cross-sectional survey. eat little or none of it. The doctor is proposing doing a more formal investigation into
(3) this phenomenon. She plans routinely to ask all pregnant women attending her clinic
about their consumption of this foodstuff and then to relate this to the details of their
(ii) Describe the main problems caused by dropout in longitudinal studies. subsequent labours. Her objective is to prove that eating it tends to reduce length of
(2) labour.
(iii) A tutor wants to investigate how her students' understanding develops over the (i) Is this proposed investigation observational or experimental? Justify your
three years of the course that she teaches. She has developed a test paper of 50 answer.
questions assessing key knowledge. She gave the test to all her students at the (2)
end of 2011 and collected in the answers. The average mark for the 43
students in their first year was 55%, the average mark for the 35 students in (ii) What advice would you give the doctor about her ability to meet her objective?
their second year was 60%, and the average mark for the 20 students in their (4)
third year was 65%. The tutor claims that these results show how students'
scores tend to increase over the three years of the course.
(a) Explain why this is a cross-sectional study.

8. (i) Distinguish between accuracy and precision when estimating features of a
(b) Explain why the tutor's conclusion is incorrect and why such a population from a sample.
deduction cannot be made from a cross-sectional study. (2)
(3)
(ii) Researchers often want to estimate the proportion of people who agree with a
(iv) The tutor decides to use a longitudinal design and studies a single cohort of simple statement. Respondents are required to answer either 'yes' or 'no'. For
students on her course. In 2011 the average mark for the 43 students was 55%. each of the following scenarios discuss, with reasons, the likely accuracy and
Of these students, 40 completed the second year, with an average mark of 52% precision of the results.
in 2012. Of these students, 12 completed the third year and their average mark
in 2013 was 70%. The students had no access to the questions between (a) The question is posted on an open-access website. The target
examinations. population is English-speakers world-wide. A total of 3000 responses
is obtained.
What conclusions can the tutor draw from these results? What other
information would be helpful in interpreting the results? (b) An anonymised postal survey is carried out of members of a small
(4) society. Of the 20 members, 18 reply.
(c) From a society of 1200 members, 100 attend a meeting and indicate
their view by a show of hands.
(9)
9. A company has a large number of employees. All employees have work email
addresses and easy access to computers, but a sizeable minority of them are not
required to use computers as part of their job. The company routinely distributes
email messages that are of interest to employees but not essential to their work. It is
known that some employees never read these emails. The management wants to carry
out a survey of the staff to establish how employees could be motivated to read emails
regularly. They have a very low budget for this activity. What advice would you give
them about how to do this?
(7)
5 6
7 8
1. (i) Briefly explain the difference between a census and a sample survey.
(2)
(ii) Briefly describe three ways in which a sampling frame might be inadequate.
What major problem might arise from having an inadequate sampling frame?
(4)
(iii) Briefly describe a sample survey in a country of your choice. Identify the
sampling frame used and describe two steps that could be taken to maximise its
adequacy.
(5)
MODULE 1 : Collection and compilation of data

2. A village in a remote country consists of 36 houses built around a central area. A
Time allowed: Three hours
researcher wants to survey people in this village by carrying out six lengthy face-to-
face interviews. In order to do this she wants to select six of the 36 houses at random.
Someone has suggested the 'spin the bottle' method. This involves standing at the
centre of the village and spinning a bottle on the sandy ground. The house whose
direction is nearest to the direction in which the neck of the bottle points is selected,
provided that it is not already in the sample.
(i) Describe a village layout where this method might give approximately random
sampling.
(1)
(ii) Describe a village layout where it would be difficult to obtain a random sample
using this method.
(1)
(iii) Explain what is meant by a table of random numbers in the range 00 to 99.
(2)
(iv) Describe a more scientific method of obtaining a simple random sample of six
of the 36 houses in this village, and explain why your method would work
equally well for each of your village layouts described in parts (i) and (ii).
(7)
1 OC Module 1 2015


© RSS 2015 2
3. A team of university researchers is investigating the experiences and opinions of 4. A manager of a restaurant wants to survey a sample of customers after they have had a
nurses working in nursing homes for elderly people. They specifically want to meal. He has decided to use a paper-based questionnaire to be given out with a free
compare responses across three areas of the United Kingdom: England and Wales, coffee after the meal and to be handed in at the reception desk.
Scotland and Northern Ireland. This is because there are variations in the regulations
relevant to their research questions between these three areas. The research is to be (i) One of the questions he wants to ask is:
led by three universities, one in each geographical area. There are many more nursing How satisfied were you with the range of choice on the menu?
homes for elderly people in England and Wales than in the other two areas.
Would this question, as worded, be open or closed? Discuss the advantages
The researchers have decided to select 30 nursing homes in each of the three areas. and disadvantages of open versus closed questions in such a questionnaire.
These are to be near to the participating universities and known to the researchers. It (5)
is believed that this will increase the response rate of the selected participants in the
research. Within each selected nursing home, three nurses will be chosen at random (ii) As well as the range of choice, the manager also wants to ask about the
from the nursing staff at the home. following.
The quality of the service.
(i) Explain why the proposed survey uses stratified sampling with the three areas
of the United Kingdom as strata. The acceptability of the environment.
(1) The value for money.
(ii) Within each stratum, what sampling method has been used to select the homes? Design a suitable questionnaire for this purpose. Restrict your questions to
Comment on and criticise the method used. these four issues, and choose open and closed questions carefully, justifying
(2) your decisions.
(8)
(iii) Comment on the claim that the method of selecting nursing homes is likely to
increase the response rate.
(3)
(iv) The researchers claim that they will get a random sample of nurses working in
nursing homes in the United Kingdom. Comment on this claim.
(5)
(v) Comment on the appropriateness of aiming for equal numbers of nurses from
each of the three areas.
(3)
3 4
5. As part of a sample survey a researcher has collected the following data on 8 6. (i) Explain the key difference between an experimental study and an observational
individuals. study.
(2)
Age Sex Number of Age of Marital status
(years) M=male children oldest child M=married or cohabiting (ii) Why is an experimental study generally preferred to an observational study
F=female (years) S=single when trying to establish which of two treatments is the more effective?
D=no longer with partner (5)
12 M 0 No answer S
25 F 2 4 M (iii) A researcher wants to investigate the safety of a procedure that is carried out on
21 F 1 0 No answer women whose babies get 'stuck' and who are therefore in distress during labour
10 F 2 6 D in childbirth. In particular he wants to compare the outcomes for both mother
26 M 3 6 M and baby from this procedure with those for another, more common, procedure
31 M 1 12 M that is used in similar situations. He has listed the outcomes of interest. He
14 G 0 0 S intends to collect data over the next 12 months from mothers who attend a
27 F 0 3 D particular large maternity hospital and who experience one of the two
procedures during routine care.
(i) Explain the differences between missing values and errors in a dataset like this (a) Is this an experimental or observational study? Justify your answer.
and describe the problems each can cause. (1)
(6)
(b) Describe any variables, other than outcome variables, that you think
(ii) Identify all the problems in these data. would be important in this study, explaining why you believe them to
(7) be important.
(4)
(iii) Explain briefly one way in which you could set about resolving the problems.
(1) (c) Describe any problems you would anticipate in collecting the data.
(4)
(iv) Assuming that the problems cannot be resolved, explain how you would code
the data ready for entry into a computer-based statistical package. Pay (d) Briefly state the steps that would be necessary in processing the data
particular attention to coding of errors and missing values, justifying your ready for analysis.
decisions. (4)
(6)
7. A survey is to be carried out on the alcohol consumption of full-time students

attending a particular university. Many of the questions are sensitive in nature. All
such students have an email account and ready access to a computer which they are
required to use in their courses. Each student's term-time address is recorded centrally
by the university. It has been decided that it would not be appropriate to carry out
face-to-face interviews or phone interviews for this population.
(i) Comment on the decision not to use face-to-face interviews or phone

interviews.
(4)
(ii) The team is considering the use of either questionnaires sent by post or an
internet survey. What would be your advice and why?
(7)
5 6
7 8

Rss Ordinary Cert Module1 Print

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Rss Ordinary Cert Module1 Print

Uploaded by

Copyright:

Available Formats

EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY

(formerly the Examinations of the Institute of Statisticians)

ORDINARY CERTIFICATE IN STATISTICS, 2001

Time Allowed: Three Hours

There is no restriction on the number of questions that a candidate may attempt,

The number of marks allotted to each question or part-question is shown in brackets.

Graph paper and Official tables are provided.

Candidates may use silent, cordless, non-programmable electronic calculators.

Where a calculator is used the method of calculation should be stated in full.

There are 8 questions altogether in the paper.

simple random sampling

stratified random sampling (stratified by ward)

quota sampling (quotas for each ward).

Name and address including postcode

Home telephone number

Number and type of accounts (e.g. savings, current, loan) held

Paper I Questions 1, 2, 3 concern the responses to this survey.

Time Allowed: Three Hours

There are 8 questions altogether in the paper. 2

(ii) discuss the problems involved in collecting data on prices;

(iii) mention any difficulties in drawing conclusions from such an index

ORDINARY CERTIFICATE IN STATISTICS, 2003

There are 8 questions altogether in the paper. 2

6. (i) Briefly explain the difference between longitudinal and cross-sectional

2. A survey about leisure activities is planned in a community with about 3000

3. A pharmacist wishes to keep details of regular customers on a database to help ensure

There are 8 questions altogether in the paper. 2

(a) systematic sampling;

(b) cluster sampling.

Outline the advantages and disadvantages of each of these methods in this

(ii) Would stratified sampling be a suitable method in this application? Justify

1. Your answers to this question should relate to the proposed survey.

There are 8 questions altogether in the paper. 2

There are 8 questions altogether in the paper. 2

3. The information obtained from farmers, as detailed in question 2, is to be stored in

(b) Explain why it is particularly important to pilot the questionnaire in

Candidates may use calculators in accordance with the regulations published in

There are 8 questions altogether in the paper.

Ward Number of patients Cost of sampling one patient SD of length

6. (i) Present arguments in favour of conducting interviews over the telephone

1. A professional organisation wishes to discover its members' opinions about its

3. A professional organisation has three grades of member: student, ordinary, and

There are 8 questions altogether in the paper.

(i) Suggest two different two-stage sampling methods of selecting a sample of

Candidates may attempt all the questions.

Age group, sex, marital status and occupation of respondents

There are 8 questions altogether in the paper.

(6 marks for each of part (i) and part (ii))

(ii) A chain of self-service restaurants has outlets throughout a country. A survey

A. Leave paper questionnaires near to the pay desks with an invitation to

Candidates may attempt all the questions.

There are 8 questions altogether in the paper.

5. A researcher is interested in finding out whether the views of teachers concerning

7. A survey is being planned in a community with about 5000 households. A sample of

The total for the whole paper is 100.

A pass may be obtained by scoring at least 50 marks.

There are 7 questions altogether in the paper.

(v) How can sampling accuracy be improved?

(a) Stratified random sampling.

(b) Cluster sampling.

• Your recommendation regarding the mechanism by which the data should be

• No more than 5 questions to determine appropriate personal details

ORDINARY CERTIFICATE IN STATISTICS, 2012