Sampling: Design and Procedures
The objective of most marketing research is to obtain information about the characteristics of population.
A population is the aggregate of all the elements that share some common set of characteristics and that comprise the universe for purpose of the marketing research problem.
A census involves a complete enumeration of the elements of a population. A sample is a subgroup of the population selected for participation in the study. Sample characteristics, called statistics, are used to make inferences about the population parameters.
Sample versus Census
Budget Small Large
Time available Short Long Population size Large Variance in charact. Small Large Cost of sampling errors Low High Cost of non-samp. er. High Low Nat. of measurem. Destructive,Non Attention to individual cases Yes No If the cost of sampling errors is high (e.g. if the sample amitted a major manufactures like Swarzędz, the results could be misleading), a census, which eliminates such errors, is desirable.
High cost of non- sampling errors would be our sampling.
In practice, non-sampling errors are the major contributor to total error, whereas sampling errors have been relatively small.
Define the population
Sampling design begins by specifying the target population. The target population is the collection of elements or objects that posses the information sought by the researcher and about which inferences are to be made.
Defining the target population involves translating the problem definition into a precise statement of who should and should not be included in the sample. The target population should be defined in terms of elements, sampling units extent, and time.
A sampling unit is an element that is available for selection at some stage of the sampling process.
Determine the Sampling Frame
A sampling frame is a representation of the elements of the target population. Sampling frame = complete specification of sampling units. It consist of a list or set of directions for identifying the target population.
/całóść w wersji do ściągnięcia/
Examples: a telephone book, a official registers (REGON, PESEL). A list which enables one to identify the sampling units unambiguously (bezpośredni). Such a list or directory sometimes needs to be set for the purpose of a particular survey or at best it may exist (e.g. city addresses in many countries).
Other necessary (1-2) or desirable (3-6) properties which the sampling frame should satisfy:
1.The frame must provide observational access to all elements in the target population.
2.It must allow the researcher to determine how the units in the frame are associated with the elements in the population.
List of statistical regions and census districts
This territorial register consist of:
1.a set of maps for rural and urban areas
2.a list of numbers of statistical regions and census districts, the number of each unit, precise addresses of all blocks and houses, and some basic data about census units
3.a list of housting properties and flats in each census districts with the addresses all flats, names of the owners (or lodgers), number of persons living in each flat and for the farmers the area of the farm
4.territorial drafts for each statistical region with is division into census districts and the specification of each resident property
Business register (REGON) - region gospodarki narodowej
The business register REGON has been designed to cover the following three groups of units:
- legal entities
- organisational units without status of legal personality
- natural persons running economic activities
Information about each unit comprises: the name, address and telephone number, legal status, type of ownership , form of financing, basic activity, testiary and subsidiary activities, number of employees and branch of the national economy. This information is collected and codified by respective regional statistical office and the unit is given a nine - digit - identifier.
Select a Sampling Technique
Non probability sampling relies on the personal judgement of the researcher rather than on chance to select sample elements.
Non probability samples may yield good estimates of the population characteristics but they do not allow for objective evaluation of the precision of the sample results.
Because there is no way of determining the probability of selecting any particular element for inclusion in the sample, the estimates obtained are not statistically projectable to the population.
Probability sampling requires not only a precise definition of a target population. Probability sampling requires not only a precise definition of a target population but also a general specification of the sampling frame. Every potential sample need not to have the same probability of selection, but the following conditions must be satisfied:
- for each population element it is possible to determine a non-zero probability of selection the element to the sample
- for each set of population elements it is possible to determine a non-zero probability that the whole set will be drawn to the sample.
Nonprobability Sampling Techniques
Convenience sampling attempts to obtain a sample of convenient elements. The selection of sampling units is left primarily to the interviewer. Respondents are selected because they happen to be in the right place at the right time.
Examples:
- use of students, members of social organisations, “people on the street”
- tear-out questionnaires included in a magazines
- TV and radio surveys
Convenience samples are not representative of any definable population.
Judgmental sampling is a form of convenience sampling in which the population elements are selected on the judgement of the researcher. He/she chooses the elements to be included in the sample because he or she believes that they are representative of the population of interest or are otherwise appropriate.
Ex:select.of household based o sec.dat
Quota sampling may be viewed as two-stage restricted judgmental sampling. The main ideas: Quotas (subsets of elements) ensure that the composition of the sample is the same as the composition of the population with respect to the characteristics of interest.
First stage - developing control categories of population elements fe. age, place of living, income groups.
Second stage - elements are selected to the sample in such a way that the proposition of the sample elements possessing the control characteristics is the same as the proportion of population elements with these characteristics.
Once the quotas have been assigned, there is considerable freedom in selecting the elements to be included in the sample. The only requirement is that the elements selected fit the control characteristics.
Quota sampling attempts to obtain representative samples at a relatively low costs.
In snowball sampling on initial group of respondents is selected, usually at random. After being interviewed, these respondents are asked to identify others respondents, these process may be carried out in wares leading to a snowballing effects. A major objective of snowball sampling is to estimate characteristics that are rate in the population (eg. widowed males under 35, members of a scatter minority population). Even though probability sampling is used to select initial respondents, the final sample is a nonprobability sample.
Probability Sampling Techniques
In sample random sampling each element in the population has a known and equal probability of selection. Every element is selected independently of every other element.
The researcher compiles a sampling frame in which each element is assigned a unique identification number. Then random numbers are generated to determine which elements to include in the sample.
Advantages:
- easily understood
- the sample results may be projected to the target population, as most of statistical inference methods assume that the data have been collected by simple random sampling.
Disadvantages:
- it is often difficult and costly to construct a sampling frame
- it can result in samples that are spread over large geographic areas
- it often results in lower precision than other probability techniques
- it may generate non-representative samples
In systematic sampling, the sample is chosen by selecting a random starting point and then picking every i-th element in succession from the sampling frame.
Sampling interval (i) = population size (N) / sample size (n)
If element number k is selected from the sampling interval, the following elements will be included in the sample: k+i, k+2i, k+3i, …
Ordering population elements is a crucial issue. If they are arranged in a manner unrelated to the characteristic of interest systematic sampling will yield results similar to simple random sampling. If the ordering of the elements is related to the characteristic of interest, systematic sampling increases the representativeness of the sample.
Advantages:
- less costly and easier than simple random sampling because random selection is done only once
- if additional information related to the characteristic of interest is available for the population, systematic sampling will generate more representative and more reliable samples
Disadvantages:
To construct a sampling frame may be costly
Systematic sampling is frequently employed in consumer mail, telephone and mail intercept interviews.
Stratified sampling is a two-step process in which the population is positioned into subpopulation, or strata. The strata should be mutually exclusive (wzajemnie wyłączające się) and collectively exhaustive (wyczerpuje całą zbiorowość).
The variables used to partition the population into strata are referred to as stratification variables. They should be closely related to the characteristic of interest.
Researcher errors.
a) surrogate information error, when the information needed for the making research is different than the information sought by the researcher.
b) measurement error, we measure sth different that is required
c) population definition error, variation between the actual population relevant to the list of actual population and actual population (when some) population is not covered.
d)sampling frame error, different between the list of actual population and actual population (when some) population is not covered.
e)data analysis error, when wrong data from questionnaires are transformed into research findings.
Total error - variation between what you have obtained in the sample and the real value of population.
1.Random sampling error - occurs because the particular sample is imperfect representation of the population of interest.
2.Nonsampling error - has various sources: errors in problem definition, in approach, scales, questionnaire design, interviewing methods, data preparation and analysis.
a)nonresponse error- if respondents do not respond
b)response error - when respondent gives inaccurate answers or them answers are misrecorded or misanalyses.