Census and Sampling • WHAT IS A CENSUS? • WHAT IS SAMPLING? AC 1.2 PRESENT THE SURVEY METHODOLOGY AND SAMPLING FRAME USED ASSIGNMENT TASK 1.2 Learning Outcomes 2 By the end of the session, learners would be able to • • • • • describe: Sampling and sampling terms Types of sampling methods Rationale for sampling Sampling methods Sampling errors and biasness POPULATION & SAMPLING 3 Census or population The sampling method refers to the entire market segment we want to get information from. Sampling is a representative of the market segment or population selected must be fair & accurate for the information to be statiscally reliable If the sampling method is incomplete and inaccurate, it is said to be biased. What exactly is a “sample”? 4 A subset of the population, selected by either “probability” or “nonprobability” methods. If you have a “probability sample” you simply know the Probablity of any member of the population being included (not necessarily that it is “random.”) Sampling 5 Who do you want to generalize to? The Theoretical Population What population can you get access to? The Study Population How can you get The Sampling Frame access to them? Who is in your study? The Sample Sample Design 6 Aim and Stages of Sampling Benefits of Sampling Aims at avoiding bias and Saves time achieving maximum precision from a given outlay of time and money Stages in sample design: Decide the objective of the research Define the study population Choose a sample method Decide sample size Faster results Saves money Enables more surveys to be carried out Can concentrate on a small number of units A carefully chosen sample can yield statistically valid information about the population as a whole Sampling Methods 7 Nature of the Population Sample population, time and money Unbiased (each object in the population to be equally chosen as part of the sample) Representative of the population (e.g. more female than male …) Sampling frame 8 Random sampling requires a sampling frame. The means of identifying sampling units. “A list of members of a population…” (Morris, 2000, p41) from which a sample can be chosen. Eg: Electoral Roll, telephone directory, etc. It should be comprehensive, complete, accurate and up to date. Care should be taken for any ‘Bias’: eg: ownership of telephone may indicate certain social class only. Sample Bias 9 A Sample is Biased if it Selection Bias (exclude differs from the population in a systematic way To be able to generalize from a sample population without bias, select random sample or under-represent part of the population) Measurement or response bias (measurement faulty ) Non response bias Choice of Sampling Method 10 Is a sample frame available? If Available budget so, random sampling is possible Is the population homogeneous or heterogeneous? What degree of precision is required? To lower the margin of error the sample size must be increased What is the budget for the exercise? Accuracy required How quickly the information is needed Accessibility of the population Types of Sampling 11 Probability Methods Random sampling Does not involve human judgement Requires a sampling frame Simple random sampling Systematic sampling Stratified random sampling Cluster sampling Non-probability Methods Non random sampling Involves human judgement No sampling frame required Quota sampling Convenience sampling Judgement sampling RANDOM SAMPLING METHODS 12 SIMPLE RANDOM SAMPLING: This means selecting a sample with every item or respondent in the sampling frame having an equal chance of being selected. SYSTEMATIC SAMPLING: selecting items from the list at regular intervals e.g. every 5th customer that buys hamburger or every 10th car that buys petrol. Cluster Sampling 13 Population is divided into clusters then chosen at random (e.g. department of a business) Within a cluster all the objects are included in the sample If clusters are different from each other regarding to the element we measure it can introduce bias or non representative. More convenient than simple random sampling Stratified Sampling 14 Similar to Cluster Complex to administer To make the sample more Students Females Males Strata Random subsamples of n/N representative the population is divided into a number of strata (groups or levels). If the list of 600 students consists of 360 males and 240 females, the sample of 50 students is more representative if it reflects these proportions as follows: Number of males 360/600x50 =30 Number of females 240/600x50 = 20 Stratified Sampling 15 Advantages Disadvantages • Every unit in a stratum has the same chance of being selected. • The sampling frame of the entire population has to be prepared separately for each stratum. • Using the same sampling fraction for all strata ensures proportionate representation in the sample of the characteristic being stratified. • Adequate representation of minority subgroups of interest can be ensured by stratification and by varying the sampling fraction between strata as required. • Varying the sampling fraction between strata, to ensure selection of sufficient numbers in minority subgroups for study, affects the proportional representativeness of the subgroups in the sample as a whole. Non-probability Methods 16 Convenience Sampling This involves gathering information from anyone available for the interview, no matter their background, this is not a very reliable method because the respondents may not really be the ideal people to provide the information we require. Quota Sampling Getting information only from respondents exhibiting certain characteristics e.g. sex, age, socio-economic group or other demographic details up to a maximum quota or number Judgement Sampling 17 This means the researcher using his or her own judgement to select respondents based on his belief that they fit quite into the profile of the people he wishes to get information from HOW TO HAVE A GOOD SAMPLE 18 Have an accurate and comprehensive sample frame or records such as electoral or census data, sales data etc. The sample population must be potential customers Choose a suitable sample size Identify the person who buys the product (always differentiate between customers & consumers) Errors in research 19 Sampling error Non response error Data collection error Data analysis error Sampling Error 20 The difference between the estimate of value obtained from a sample and the actual value Arises because a sample cannot exactly represent the population as a whole Bias as a consequence of the way in which a sample is structured or the way it was selected Sampling error can be reduced either by increasing the size of the sample or by improving the amount of knowledge on the structure of the market prior to drawing up the sample 21 Data Collection Errors Errors in Analysis Leading questions Omission of important factors Clerical errors Misrepresentation Inarticulate respondents Ambiguous questions Unintended interviewer bias Errors in completing precoded answers in questionnaire design Error in statistical analysis Drawing the wrong conclusion Confusion between cause and correlation Misrepresentation of the data False definitions
© Copyright 2026 Paperzz