BUSINESS RESEARCH METHODS LECTURE 7 SAMPLING SAMPLING TECHNIQUES

BUSINESS RESEARCH METHODS LECTURE 7 SAMPLING & SAMPLING TECHNIQUES 1

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING • Population – the universe of units from which one selects the sample. The units can be people, nations, cities, regions, organisations, etc. – The population has much wider meaning than what is ordinarily used to identify an entire population • Sample – the section of the population one selects for the study or investigation. – It is a compartment or subset within the population based on predetermined criterion 2

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Sampling frame – The listing of all units in the population from which the sample will be chosen, e. g. Individuals, institutions, household • Representative sample – a sample that truly mirrors the population so that it is seen as a microcosm of the population • Sampling bias – a misrepresentation or distortion in the representativeness of the sample that results when members of the population have little or no chance of being chosen for inclusion the sample 3

ILLUSTRATION 4

FEATURES OF A POPULATION – Homogeneous – all cases are similar, e. g. bottles of beer on a production line; – Stratified – contain strata or layers, e. g. people with different levels of income: low, medium, high; – Proportionally stratified – contains strata of known proportions e. g. percentages of different nationalities of students in a university; – Categorised by type – contains distinctive groups, e. g. of apartment buildings – towers, slabs, villas, tenement blocks; – Classified by location – different groups according to where they are e. g. animals in different habitats – desert, equatorial forest, savannah, tundra, etc.

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Probability sample – a sample that has been selected using random selection such that each unit in the population has a known chance of being chosen • Non-probability sample – a sample that has not been selected using a random selection technique. This means that the selection of some units in the population are more probable than others 6

PROBABILITY AND NONPROBABILITY SAMPLING FORMS TYPES OF PROBABILITY SAMPLE TYPES OF NONPROBABILITY SAMPLE • Simple Random • • Systematic Sample • Convenience Sampling • Stratified Random Sampling • Quota Sampling • Cluster Sampling • Judgemental • Snowball Sampling 7

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Sampling error – the difference between a sample and the population from which it is selected, in spite of selecting a probability sample, e. g. unrepresentativeness of the taken sample • Non-sampling error – Identifies the differences between the population and the sample that arise either from deficiencies in the sampling method. This can result from insufficient sampling frame or nonresponse rate due to such problems as poor interviewing, bad data processing or poor wording of questions. 8

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Simple Random Sample – each unit of the population has an equal probability of inclusion in the sample • Systematic Sample – Process of selecting units directly from the sample frame without using any sense of randomness, such as 6, 26, 46, 66, 86, etc. • Stratified Sample – it is a proportional representation of the different categories within the selected unit. 9

ILLUSTRATION OF STRATIFIED SAMPLING TECHNIQUE Source: Explorable. com 10

PROBABILITY SAMPLING • Cluster sampling – Involves the selection groupings within the units but not the primary sampling unit, e. g. 10 out 100 firms for interviews involving 5000 employees in which 500 each of from the 10 companies would be interviewed. 11

EXAMPLE OF CLUSTER SAMPLING source: explorable. com 12

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Convenience sampling – A sample that is readily available to the investigator in terms of accessibility, flexibility, etc. • Snowball sampling – a situation of making an initial contact with few small group of people who are important to the research topic. The researcher then uses the small group to get in touch with others – Facilitated by current sample elements – The sample may well not be representative – It is not advisable to use in quantitative study, rather it can perfectly be used in a qualitative research strategy 13

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Quota sampling – Necessary features or characteristics are applied to divide the sample to enhance its representativeness – Normally used in market research or political opinion poll – The rationale of this sampling technique is to produce a sample that reflects a population in terms of the relative proportion in different categories such as age groups, gender, tribal groupings, etc. • Judgemental – Intentionally choose sample to conform to some criteria or conditions of the study 14

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Non-response – a source of non-sampling error that is particularly likely to happen when individuals are being sampled. This happens whenever some members of the sample refuse to cooperate, cannot be reached or cannot provide the needed data for some reason. • Census – The process of enumerating the total population 15

KEY TERMINOLOGIES & CONCEPTS IN SAMPLING CONT’D • Response rate – the percentage of a sample that does actually participate. Calculating the response rate is not pretty straightforward. • One should bear the following in mind: – Not every response will be included, e. g. Uncompleted questionnaire, suspicion of lack of seriousness, • It can be worked out thus: Total Number of responses x 100 Total number in sample – ineligible 16

EXAMPLE OF RESPONSE RATE CALCULATION • Kofi administered an email questionnaire to students who had left the supply chain and information systems department over the past 10 years. He obtained a list of the 520 past students from the registrar over this period and selected 50% sample. He obtained a response from 140. In addition, his list of the former students was inaccurate and ten of those he contacted were ineligible to respond, having left the department over five years earlier. • What would Kofi’s response rate be? 17

SOLUTION Kofi’s total response rate can be calculated thus: Response rate = Total Number of responses x 100 Total number in sample – ineligible 140 = 260 – 10 140 x 100 = 56% 250 18