Unit 4c SAMPLING Design • Sampling is the process or technique of selecting a suitable sample for the purpose of determining parameters or characteristics of the whole population • To carry out a study, one might bear in mind what size the sample should be, and whether the size is statistically justified and lastly, what method of sampling is to be used 1/17/2024 1 1/17/2024 2 The concept of sampling in qualitative research • In qualitative research the issue of sampling has little significance as the main aim of most qualitative inquiries is either to explore or describe the diversity in a situation, phenomenon or issue. • Qualitative research does not make an attempt to either quantify or determine the extent of this diversity 1/17/2024 3 Contd. • The class, families living the city or electorates from which you select a few students, families, electors to question in order to find answers to your research questions are called the population or study population, and are usually denoted by the letter (N). • The small group of students, families or electors from whom you collect the required information to estimate the average age of the class, average income or the election outcome is called the sample. • The number of students, families or electors from whom you obtain the required information is called the sample size and 1/17/2024 4 is usually denoted by the letter (n). • The way you select students, families or electors is called the sampling design or strategy. • Each student, family or elector that becomes the basis for selecting your sample is called the sampling unit or sampling element • A list identifying each student, family or elector in the study population is called the sampling frame. 1/17/2024 5 • Your findings based on the information obtained from your respondents (sample) are called sample statistics. • From sample statistics we make so estimate of the answers to our research questions in the study population. • The estimates arrived at from sample statistics are called population parameters or the population mean. 1/17/2024 6 Principles of sampling • Principle one-In a majority of cases of sampling there will be a difference between the sample statistics and the True population mean, which is attributable to the selection of random elements in the sample. • Principle two-the greater the sample size, the more accurate will be the estimate of the true population mean 1/17/2024 7 • Principle three -The greater the deference in the variable under study in a population for a given sample size, the greater will be the difference between the sample statistics and the true population mean 1/17/2024 8 Factors affecting the inferences drawn from a sample • The size of the sample -As a rule, the larger the sample size, the more accurate will be the findings • The extent of variation in the sampling population-As a rule, the higher the variation with respect to the characteristics under study in the study population, the greater will be the uncertainty for a given sample size. 1/17/2024 9 1/17/2024 10 • How ‘representative’ is one’s sample may be a common question. Researchers always try to draw a representative sample to draw any conclusion about the ‘real world’ • This is a part of the researcher’s responsibility • There are two basic sampling techniques: probability and non-probability sampling. 1/17/2024 11 • A probability sample is defined as a sample in which every element of the population has an equal chance of being selected. • if sample units are selected on the basis of personal judgment, the sample method is a non probability sample • A sampling frame is the list of elements from which the sample may be drawn • sampling frame might be a list of all members of an institute or workers in a company 1/17/2024 12 • The sampling unit is a single element or group of elements subject to selection in the sample • flights can be selected as sampling units. The term primary sampling units (PSUs) designates units selected in the first stage of sampling. If successive stages of sampling are conducted, sampling units are called secondary sampling units, or tertiary sampling units (if three stages are necessary). 1/17/2024 13 REPRESENTATIVE SAMPLING PLANS • Simple Random Sample A random sample is defined as follows: • Selections are made from a specified and defined population (i.e., the frame is known). • Each unit is selected with known and non-zero probability, so that every unit in the population has an equal (known) chance of selection. • The method of selection is specified, objective and replicable 1/17/2024 14 Stratified Random Sampling • the population is observed to be heterogeneous in nature • in order to apply simple random techniques to such a heterogeneous population, we have to group them as homogeneouslyas possible, where each group is termed a ‘stratum’ (in plural ‘strata’). • Then samples are drawn equally or proportionately from each stratum and, therefore, the procedure is called stratified random sampling 1/17/2024 15 • The procedure for selecting a stratified Sample • Step 1 Identify all elements or sampling units in the sampling population. • Step 2 Decide upon the different strata (k) into which you want to stratify the population. • Step 3 Place each element into the appropriate stratum. • Step 4 Number every element in each stratum separately. • Step 5 Decide the total sample size (n) • Step 6 Decide whether you want to select proportionate or disproportionate stratified sampling and follow the steps below 1/17/2024 16 1/17/2024 17 Contd. • Systematic (Quasi-random) Sampling • In systematic random sampling, we range the population from which selections are to be made in a list or series, choose a random staring point and then count through the list selecting every n-th unit • Systematic sampling has been classified under the `mixed' sampling category because it has the characteristics of both random and nonrandom sampling designs 1/17/2024 18 Cluster (Multistage) Sampling • In cluster sampling we have to have a number of clusters which are characterized by heterogeneity in between and homogeneity within. • Cluster sampling is used for a variety of purposes particularly for large sample surveys or a nation-wide survey • If we consider two stages to conduct the survey, then it is called two-stage cluster sampling. If someone considers more than two stages to collect the data, then it is called multistage sampling. 1/17/2024 19 1/17/2024 20 Sequential (Multiphase) Sampling • This is a sampling scheme where the researcher is allowed to draw sample on more than one occasions. • It may be economically more convenient to collect information by a sample and then use this information as a basis for selecting a sub-sample for further study. This procedure is called double sampling, multiphase sampling or sequential sampling. This is a technique frequently used to draw samples in industries for ensuring the quality of their products 1/17/2024 21 Non-probability Sampling Methods • In non-probability sampling, the probability of selecting population elements is unknown • Convenience Sampling • Non-probability samples that are unrestricted are called convenience samples • They are the least reliable design but, normally, the cheapest and easiest to conduct. Interviewers have the sole freedom to choose whomever they find, thus the name convenience 1/17/2024 22 Purposive Sampling • Judgement sampling and • Quota sampling. • Snowball (Network or Chain) Sampling 1/17/2024 23 1/17/2024 24 SAMPLE SIZE DETERMINATION • Sample size is associated with time and cost • A more relevant issue is how to judge whether the sample size is adequate in relation to the goals of the study. Strictly speaking, exact tests to check whether sample size is adequate for the analysis • required can be carried out by using statistical methods such as significance tests, but, in many studies, readers who do not have the required statistical skills can use a more common-sense approach to the problem 1/17/2024 25 • a sample which has to be able to produce results that are statistically significant, statistically robust or statistically justified, but, more importantly, representative of the whole population • determination of sample sizes for mean and proportion can be calculated under the normality conditions where standard error plays an important role 1/17/2024 26 Contd. • The formula for determining sample size in the case of testing hypothesis of population means can be • expressed as:n=zα/2(sd)2 d2 • where n = sample size, • Z = Standardised normal value, usually taken as 1.96 for a 95 per cent confidence interval, • α = Level of significance, • SD = Standard deviation (assumed to be known from prior survey or can be guessed or other published • studies can inform on this), • d = Precision range (the required confidence interval) 1/17/2024 27 • Assume that a researcher wants the estimate to be within ±£25 of the true population value and he/she wishes to be 95 per cent confident it will contain the true population mean. Also assume that early studies have demonstrated the standard deviation to be around £100. What would be the required sample size? • given Z = 1.96, SD = 100, d = 25 • n=64 1/17/2024 28 KEY STATISTICAL CONCEPTS • Statistical Estimates • It is relatively straightforward to gather information about small, subsets of populations (for example,employees of a particular small- or medium-sized company • Population or Universe • The population consists of any well-defined set of elements 1/17/2024 29 • The characteristics of a population such as its population mean (μ), and population standard deviation (σ) are population parameters. • The characteristics of the sample such as sample mean and sample variance are called sample statistics • Why Sample? 1/17/2024 30 • • • • • • Cost Relevance and Flexibility Speed Practicality and Feasibility Higher Data Quality Public Acceptability 1/17/2024 31 Variance and Bias • The results of sample surveys are not completely precise, but suffer from random sampling variability. • This means that the same sampling method applied repeatedly to the same population will not produce identical results each time. • Bias is different from random variation • A biased method will systematically misrepresent the population, no matter how large the sample 1/17/2024 32 The main sources of bias are • • • • Imperfect Coverage Sampling Bias Non-response Bias Response and Other Data Collection and Processing Biases 1/17/2024 33 Methods of drawing a random sample • The fishbowl draw-if your total population is small, an easy procedure is to number each element using separate slips of paper for each element, put all the slips into a box, and then pick them out one by one without looking, until the number of slips selected equals the sample size you decided upon. This method is used in some lotteries. 1/17/2024 34 • Computer program-there is a number of programs that can help you to select a random sample 1/17/2024 35 1/17/2024 36