11 Populations and Samples
Learning Objectives Define Population And Sample Distinguish Between Target And Accessible Population Discuss Probability And Nonprobability Sampling Procedures Compare Four Methods Of Probability Sampling
Learning Objectives Compare Three Methods Of Nonprobability Sampling Determine Which Sampling Technique To Use In Various Types Of Research Studies Compare Longitudinal And Cross-Sectional Studies Enumerate Factors To Be Considered In Deciding The Size Of The Sample 3
Learning Objectives Discuss Sampling Error And Sampling Bias Critique The Sampling Procedure Described In Research Reports And Articles 4
Learning Objective One Define Population And Sample
Population Complete set of persons or objects Common characteristic Of interest to the researcher
Sample Subset of a population Sample represents the population.
Sample Selection Representation of the population Method for getting the sample Sample size for the study
Sample Terms Element Sampling frame Single member of a population Listing of all elements Study sample, if from this frame
Learning Objective Two Distinguish Between Target And Accessible Population 10
Population Terms Target population Accessible population
Target Population Definition Entire group of people or objects People or objects meet designated set of criteria. Generalization of the findings
Accessible Population Definition Group of people or objects Researcher has access to them.
Population Importance Conclusions based on data Data from accessible population Decisions made from study results
Learning Objective Three Discuss Probability And Nonprobability Sampling Procedures 15
Types of Sampling Methods Probability Nonprobability
Probability Sampling Uses random sampling procedures Selects sample from elements or members of population Types Simple Stratified Cluster Systematic
Nonprobability Sampling Uses nonrandom sampling procedures Selects sample from elements or members of population Types Convenience Quota Purposive
Learning Objective Four Compare Four Methods Of Probability Sampling 19
Probability Sampling Simple random Stratified Cluster Systematic
Random Selection Key word in sample selection Every subject has an equal chance.
Probability Sampling Allows researcher to estimate the chance Helps with inferential statistics with greater confidence Gives the ability to generalize the findings
Simple Random Sampling Type of probability sampling Importance of this sampling Equal chance of selection Independent chance of selection
Advantages of Simple Random Sampling Little knowledge of population is needed. Most unbiased of probability method Easy to analyze data and compute errors
Disadvantages of Simple Random Sampling Complete listing of population is necessary. It is time consuming to use.
Steps for Simple Random Sampling Identify the accessible population or list of elements Choose the method for getting the sample Note how easy it is through this example Names of elements on slips of paper Papers are placed into a hat. Individual draws a slip of paper. Individual continues until sample number is met.
Stratified Random Sampling Type of probability sampling Population is divided into subgroups or strata. Simple random sample taken from each strata
Approaches for Stratified Random Sampling Proportional stratified sampling Determine sampling fraction for each stratum Ensure that this stratum is equal Proportion in total population Disproportional stratified sampling Determine stratum is represented Used when strata are very unequal Note the key word disproportional
Advantages of Stratified Random Sampling (cont’d) Increases probability of being representative Ensures adequate number of cases for strata
Disadvantages of Stratified Random Sampling Requires accurate knowledge of population May be costly to prepare stratified lists Statistics are more complicated.
Cluster Random Stratified Sampling Large groups or clusters, not people, are selected from population. Simple, stratified or systematic random sampling may be used during each phase of sampling.
Advantages of Cluster Random Sampling Saves time and money Arrangements made with small number sampling units Characteristics of clusters or population can be estimated.
Disadvantages of Cluster Random Sampling Causes a larger sampling error Requires each member assignment of population to cluster Uses a more complicated statistic analysis
Systematic Random Sampling Type of probability sampling Every kth element is selected. Process Obtain a listing of population Determine the sample size Determine the sampling interval (k = N/n) Select random starting point Select every kth element
Advantages of Systematic Random Sampling Easy to draw sample Economical Time-saving technique
Disadvantages of Systematic Random Sampling Samples may be biased. After first sample is chosen, no longer equal chance
Learning Objective Five Compare Three Methods Of Nonprobability Sampling 37
Nonprobability Sampling Sample elements are chosen nonrandomly. Produces biased sample Each element of the population may not be included in the sample. Restricts generalizations made about study findings
Nonprobability Sampling Convenience Quota Purposive
Convenience Sampling Chooses the most readily available subject or object Does not guarantee that the subject or object is typical of the population
Snowball Sampling Type of convenience sampling method Study subjects recruit other potential subjects. May be called network sampling May find people reluctant to volunteer
Quota Sampling Type of nonprobability sampling Researcher selects sample to reflect characteristics. Examples of stratum
Quota Sampling Age Gender Educational background Number of elements in each stratum Number is in proportion to size of total population.
Purposive Sampling Type of nonprobability sampling Researcher uses personal judgment in subject selection. Each subject chosen is considered representative of population. Many qualitative studies use this technique.
Nonprobability Sampling Procedures Advantages Time Money Disadvantages Nonrandom Not able to generalize findings
Learning Objective Six Determine Which Sampling Technique To Use In Various Types Of Research Studies 46
Research Studies Use voluntary subjects Follow the ethics of research Subjects must voluntarily agree. Subjects may refuse to participate.
Research Data Based on voluntary responses from subjects Biased sample occurs if subjects do not participate.
Volunteers As Subjects Sample selection varies. Subjects volunteer for a study. Researcher approaches subjects.
Random Sampling or Random Assignment Each subject has equal probability of being included. Random assignment Procedure to ensure that each subject has equal chance
Threefold Randomization Process Used for experimental studies Helps represent the ideal study procedure Steps to ensure the process Subjects randomly selected from population Subjects randomly assigned to groups Experimental treatments randomly assigned to groups
Learning Objective Seven Compare Longitudinal And Cross-Sectional Studies 52
Classification of Research Studies Longitudinal Cross-sectional
Longitudinal Research Study Subjects are followed over time. A cohort study is one example. Subjects are studied based on Similar age group Similar background
Longitudinal Research Study (cont’d) Data are gathered. Same subjects Several times Tells influence of time
Cross-Sectional Study Subjects checked at one point in time Data collected from groups of people Data may represent differences in Ages Time periods Developmental states Important considerations
Longitudinal Versus Cross-Sectional Studies Longitudinal studies Accurate means of studying changes over time Studies take a long time to perform. Cross-sectional studies Less expensive Take less time Easier to conduct
Learning Objective Eight Enumerate Factors To Be Considered In Deciding The Size Of The Sample 58
Sample Size No simple rules Qualitative studies use much smaller samples than quantitative studies. Factors to consider for sample sizes in quantitative studies Homogeneity of population Degree of precision desired by the researcher Type of sampling procedure that is used
Sample Size (cont’d) Central limit theorem Sampling distribution of the mean
Larger Sample Sizes Many uncontrolled variables are present. Small differences are expected in members. Population must be divided into subgroups. Dropout rate among subjects is expected to be high. Statistical tests are used that require a minimum sample size.
Power Analysis Helps to determine sample size May prevent type II error Helps to detect statistical significance
Power Analysis (cont’d) Low power; type II error high External funding sources require it. Helps determine the optimum sample size
Nursing Research Studies Usually limited to small convenience samples Generalizations to total population difficult Small sample sizes warrant replication studies. Similar results from replication help with generalization.
Learning Objective Nine Discuss Sampling Error And Sampling Bias 65
Sampling Error Random fluctuations in data Not under the control of the researcher Chance variations occur when sample is chosen.
Sampling Bias Bias when samples are not carefully selected All nonprobability sampling methods have it. May occur in probability sampling methods Subjects decide not to participate when chosen. Final sample is now not representative of population.
Learning Objective Ten Critique The Sampling Procedure Described In Research Reports And Articles 68
Critiquing the Population and Samples Is the target population identified? Is the accessible population identified? Was a probability or nonprobability sampling method used? Is the specific sampling method named? Is the sampling method described? Is the sampling method appropriate for the study?
Critiquing the Population and Samples (cont’d) Are the demographic characteristics of the sample presented? Is the sample size adequate? Was power analysis used to determine the sample size? Is the sample representative of the population?
Critiquing the Population and Samples (cont’d) Are potential sampling biases identified? Is subject dropout discussed?