Populations and sampling
Populations and sampling 456
Popular in Introduction to Research Methods
Popular in AXED
This 7 page Class Notes was uploaded by Selena Blanco on Monday October 17, 2016. The Class Notes belongs to 456 at New Mexico State University taught by Brenda Seevers in Fall 2016. Since its upload, it has received 2 views. For similar materials see Introduction to Research Methods in AXED at New Mexico State University.
Reviews for Populations and sampling
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 10/17/16
Populations and Sampling Objectives Objectives (cont.) n Define terms related to population and n List the uses, characteristics, and sampling limitations of each kind of sampling n Distinguish between target and procedure accessible populations n Given a chart, determine the sample n Distinguish between a census and a size for a given population sample n Explain the steps in conducting a n Distinguish between probability and sample non probability sampling Specifying the Population Definition of Terms and the Sample n Based on process of “inductive n Population (target population) – the larger reasoning” group to which a researcher wishes to generalize; all members of a well defined – Making observations and then drawing class of people, events, or objects. conclusions about them - Samples must be representative of the n Accessible population – the population of population if one is able to generalize subjects that is accessible to the researcher with confidence from the sample to the population for the study, and to whom findings can be generalized to 1 Definitions (cont.) Direct Route Inference n Census – a survey that includes the entire Population population of interest ↓ n Sample – a portion of the population Describe Data ↓ n Statistical Inference – procedures used to make generalizations from sample data to Parameter the population from which the sample was drawn Indirect Route Inference (Sampling) To generalize back to the target population with any degree of confidence, it is critical that the Population → Sample sample be representative of the ↑ ↓ population. The strength is not in Generalize Describe Data the size of the sample but in the representativeness. ↑ ↓ Parameter ← Statistic Definitions (cont.) Example 1 n Sampling Error – the difference between the n A national magazine hn 1) Who is the target sample estimate and the true population one million subscriberspopulation? know which aspects of the score magazine are liked and 2) Who is the accessible which are not. The staffopulation? decides that a personal n Accepting sample – those individuals from interview is the bestn 3) Who is the sample? the sample that participated method to obtain and economic reasons aical total of only 500 people in n Usable sample – scores or data from the five cities will be surveyed. accepting sample that is complete or usable 2 Types of Sampling Example 2 Procedures In a study of the 1) Who is the target n Probability volunteer management population? competencies of 4-H agents in the United 2) Who is the accessible population? n Non Probability States, the researcher 3) Who is the sample? selected 400 agents to 4) Who is the accepting send a mail sample? questionnaire. 279 questionnaires were 5) Who is the usable returned. 275 had sample? complete data. Non-Probability Sampling Non-Probability Sampling n NOT GENERALIZABLE n Three types n Does not rely on random selection n Cannot determine representativeness – Accidental, Chunk, Convenience to population – Quota n May be the only option available – Purposive Accidental Quota n Involves using subjects from diverse strata n Also known as chunk or convenience n Involves using available subjects of the population, based on known characteristics n Regarded as the weakest of all n Goal is to match sample and population sampling techniques proportions on characteristics n Often uses intact groups (classrooms, n Problems conference attendees, etc.) – Must know characteristics – Not random 3 Purposive Limitations n Also known as a “judgment” sample n Results cannot be inferred or n Researcher attempts to select units generalized to a larger population that are judged to be representative of n Sampling error cannot be estimated the typical population n An estimate of representativeness n Selected for a specific purpose due to cannot be made specific characteristics Probability Sampling Probability Sampling n GENERALIZABLE n 5 approaches – able to generalize results to target – Simple Random population – Systematic – Stratified n Based on laws of probability nProportional nNon-Proportional – Cluster – Multi-Stage Using a Random Simple Random Table of Numbers n Every subject in the population has n Numbers are set up in rows and and equal and independent chance of columns being selected n You must know population and n Procedures: sample sizes – Draw a name or number from a hat n Select an arbitrary starting point – Use a random table of numbers 4 Example Systematic Sampling n Use a random start n Choose every nth case - calculate interval by dividing sample size into population size; 2000/333 = n Population size is 247 6 (interval of 6) n Sample size is 50 n Use an accurate frame with all names numbered from 1-247 n Cases should be random order n Look at consecutive 3 numbers (first 3 or last 3) n Working down, choose first 50 numbers under 247 Stratified Sample Stratified (cont.) n Involves using subgroups that differ in n Ensures the most representative the characteristics being studied sample from the population n May take multiple samples based on n Proportional Stratified – proportion of the number of subgroups the sample is in direct proportion to n Must have a valid reason for stratifying the population – Differs from quota in that subjects are n Results are inferred to strata rather selected randomly from population than the larger population Cluster Sampling Multi-stage Sampling n Uses all elements in randomly selected n Sampling is done more than once to clusters arrive at a final sampling list n Used when it is difficult to list all n Example: use cluster followed by members of a population simple random n Examples: n Generalization of results can only be – All households in a range of city blocks made back to the “units’ sampled – School buidlings 5 Benefits of Probability Steps in Probability Sampling Sampling 1. Define the population (target and accessible) n Can specify the probability of a unit 2. Obtain an accurate frame being selected before sample is drawn 3. Determine sample size (see handout) n Can estimate the sampling error 4. Select sample 5. Conduct Study (collect data) n Can make unbiased inferences to 6. Determine accepting and usable sample populations 7. Draw conclusions based on sample 8. Infer conclusions back to population Sample Size Sampling Review n Larger are more likely to be a better A. Probability __ Accidental representation than smaller ones __ Stratified n Representativeness is more important __ Purposive __ Generalizable than size B. Non Probability n Size alone will not guarantee accuracy __ Quota __ Systematic n If conducting correlations, it is recommended to have a minimum __ Non-generalizable __ Simple Random sample size of 30 __Cluster __ Convenience Sampling Review Sampling Review _ equal & independent chance Accidental T/F Randomness guarantees representativeness Quota _ uses available subjects T/F Purposeful samples rely on random selection Purposive _ every nth case T/F Sampling is based on the process of inductive _ judgement sample _ uses a random table of reasoning Simple Random numbers T/F Direct route inference uses parameters to _ study existing groups rather Systematic than individuals describe the population Stratified _ weakest of all designs T/F Inferential statistical procedures are used to Cluster _ ex. Uses cluster sampling, make tentative generalizations from sample data followed by simple random Multistage about the population from which the sample was drawn 6 7