Slide 1Slide 2The Value of Sampling in Information ResearchSampling & The Research ProcessSamplingThe Sampling ProcessIdentify the Sampling FramePopulations and CensusSamplesChoosing the Sampling MethodMethods for Determining Sample SizeStatistical Method for Computing Sample SizeMethods to Estimate the Population Standard DeviationDetermining Appropriate Sample SizesDetermining Sample Size: MeansDetermining Sample Size: ProportionsMeasures of DispersionMeasures of Dispersion – cont’dNormal DistributionConfidence IntervalSampling Distribution & The CLTSampling & Nonsampling ErrorDetermining Statistical PrecisionTypes of Sampling MethodsProbabilistic SamplingProbabilistic Sampling – cont’dProbability Sampling MethodsSlide 28Non-probabilistic SamplingNon-probabilistic Sampling – cont’dNonprobability Sampling MethodsSlide 32Slide 33A Summary of Comparative DifferencesSlide 35Factors to Consider when Selecting a DesignSlide 37The Steps for Developing a Sampling PlanSummary of Learning Objectives1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. CCHHAAPPTTEERR1234 0001 897251 000001111Sampling: Theory, Sampling: Theory, Designs, and Issues in Designs, and Issues in Marketing ResearchMarketing Research11-211-21995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. The Value of Sampling in Information Research“Sampling” is an essential part of what it means to be “human.” Therefore, when we “sample” we’re doing what comes “natural.”The sampling process associated with marketing research is quite complex, based on the scientific method rather than intuition.Sampling is efficient, enabling a research team to project outcomes from a small group out to a larger target population, saving time and money.The information gathered from the small group (or sample) allows the research team to make judgments about the larger target population, helping management address the information problem or market opportunity.11-311-31995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Sampling & The Research ProcessA census is a research endeavor that seeks to include data about every member of the defined target population.A sample is a research endeavor the seeks to make judgments about a larger group of respondents by communicating with a smaller number of people drawn from the total target population.11-411-41995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. SamplingCost and time considerations usually make surveying an entire population impractical, so companies take a sample of elements to represent the target population.It is undeniable to most statistical experts that both small and large samples can be highly accurate—provided the sampling plan is sound.An important part of the plan is to use a correct sampling frame.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. The Sampling ProcessDefine the Target PopulationIdentify the Sampling FrameChoose the Sampling MethodDetermine the Sample SizeGather the Data1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Identify the Sampling FrameSampling Frame: Actual list of each element or member of the target population.Sampling BiasesNonresponse Bias: (a.k.a., nonresponse error) Occurs when a high percentage of respondents choose not to participate in the study.NoncooperationNoncontactPassive RefusalInaccurate Sampling Methods: An inaccurate sampling frame happens when, for example, samples for surveying a firm’s customers are based on a list provided by the firm. The list may be a list of accounts rather than a list of customers. This would result in a customer with three accounts having a triple probability of being drawn into a sample.To reduce the chance of sampling bias and to allow for statistical analysis, a random sample should be taken.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Populations and CensusPopulation: (a.k.a., universe) Is the entire group of people, markets, companies, or products that is being investigated by the researcher.Finite Population: Has a limited or fixed number of individuals or objects .Infinite Population: Has an unlimited or non-fixed number of individuals or objects.Parameter: A measurement used to describe some characteristic of a population.Census: When a population is sampled in its entirety.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. SamplesSubset of representative units from the population. The sample is used to represent the population for statistical study, and the findings from the sample are used as the basis for estimating or predicting the characteristics of the population.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Choosing the Sampling MethodProbability Sample: A subset of the population in which the probability of selection is known and nonzero for every sampling unit in the population. Nonprobability Sampling: Any subset of a population in which the probability of obtaining the sample cannot be computed.1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Methods for Determining Sample SizeBlind guessingIndustry rules of thumbAffordability methodStatistical method1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Statistical Method for Computing Sample SizeDesired Precision (+E). Asks, “How precise does the measurement need to be?”Value Associated with Desired Confidence Level (Z). Asks, “How confident do you want to be that the specified confidence interval takes in the population mean?”Estimator of the Standard Deviation of the Population (s). Asks, “How heterogeneous are the members that are being investigated?”1995 7888 4320 000 000001 00023Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Methods to Estimate the PopulationStandard