Sampling, Recruiting, And Retaining Diverse Samples
Sampling, Recruiting, andRetaining Diverse SamplesMethodology Application SeriesDr. Lorey WheelerResearch Assistant ProfessorNovember 20, 2015
Session Overview Research Questions and Target Populations Problems with Common Methods Sampling Strategies– Common Methods Random sampling Stratified sampling Convenience sampling Recruitment and Retention Strategies Issues Related to Study ValidityCopyright Lorey Wheeler (2015)2
Why is this topic important? Sampling and recruiting participants are basic steps inalmost every research enterprise and are fundamentalto determining the quality of the resulting research Need to be sure that we have studied the grouptargeted by our research Well-established research sampling and recruitmentmethods developed and used successfully with middleclass European American groups may not work as wellwith ethnic minority or economically disadvantagedpopulationsCopyright Lorey Wheeler (2015)3
RESEARCH QUESTIONS AND IDENTIFICATIONOF THE TARGET POPULATION4
Research Questions How best to answer research question with design– With disadvantaged populations, ethnical considerationsfor cultural relevance of research questions Group Comparative Designs– Examining cross-group comparisons– E.g., on incidence rates or relationships between variables Within Group (Ethnic Homogenous) Designs– Examining variability or heterogeneity within a group tounderstand specific mechanisms operating– E.g., culturally-relevant risk and protective factors thatmight inform culturally relevant preventive interventionsCopyright Lorey Wheeler (2015)5
Target Population Before you implement study to answer researchquestions, make sure that you can generalize yourresults– Define target population Population of interest– Determine accessible population Those who can be located and contacted from the targetpopulation– Determine validity of chosen population The degree of similarity between target and accessiblepopulations– Determine validity of sample The degree of similarity between the sample and the targetpopulationCopyright Lorey Wheeler (2015)6
Defining Ethnic Minority and EconomicDisadvantage Ignoring or dealing simplistically with thedefinition of target population contributes to themisrepresentation of these populations Reliance on participant self-reports to determinegroup membership Researchers should carefully determine whichindicator best fits research question beforedefining group membershipCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 20097
Defining Ethnic Minority and EconomicDisadvantage Studies focusing on specific research questions or on certain groupsmight be improved if more attention were paid to the issue ofmembership– For some groups, such as American Indians and Alaska Natives, thereare legal definitions that vary across group that determine formaltribal membership– Interracial families and how to determine membership under theseconditions Definitions of group membership determined on a study by studybasis after considering how various options might affect studyresults and ability to interpret and generalize resultsCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 20098
PROBLEMS WITH COMMON SAMPLINGAND RECRUITING METHODS9
Assumptions Underlying Common MethodsMay Not Apply1. Research is conducted for the common goodand potential participants should be willingto contribute to the common good– Common good usually refers to that whichbenefits most of society, generally “mainstream”and middle class populations– Ethnic minority and economically disadvantagedpopulations may not consider themselves as partof “mainstream” societyCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200910
Assumptions Underlying Common MethodsMay Not Apply2. Universities and other research institutions aregenerally well known, universally trusted, anddeserving of support–––Members of middle class more likely to have attendedcollege or to be aware of research institutionsIndividuals who attended college may be morecomfortable or familiar with research, and less fearful ofharm because of their experiences with or knowledge ofresearchMembers of disadvantaged populations may be less likelyto understand purposes of research or to be connectedto or familiar with institutions that conduct researchCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200911
Assumptions Underlying Common MethodsMay Not Apply3. Most White, middle-class persons think ofparticipation in social science research asquite benign; assume lack of harm fromresearch– Does not necessarily apply to other groups– Have experienced various forms of discriminationin society, and, at times, in research E.g., Tuskegee Study of syphilis with Black malesCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200912
Misrepresentation in Research Studies too often fail in the basic goal of samplingand recruitment: to represent the population– E.g., research on African American families, children,or adolescents, typically over represents those livingbelow the poverty level– African Americans, and most other ethnic minoritygroups, are more likely to be living in poverty thanEuropean Americans, the majority of AfricanAmericans and most other ethnic minority groups arenot living in poverty Studies are also needed that represent thediversity of groups to make accurate inferencesCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200913
Misrepresentation in Research Ethnic Comparative Designs– Designs that compare ethnically diverse populations to, most commonly, amiddle-class European American group (often considered to be the normativecomparison group)– Sample of the ethnic minority group often economically disadvantaged– Results of such comparisons often contribute to a “deficit model” of the ethnicminority groups Comparative research designs that confound social class andethnicity– Produced a literature with a biased perspective of ethnic minority individuals– Provides little insight into successful members of the groups Need for studies that disentangle ethnicity and socioeconomicstatus– E.g., take into account context of poverty when comparing groups– Examining within-group heterogeneityCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200914
Misrepresentation in Research Pan-ethnic research designs– Researchers ignore distinctions among ethnicminority subgroups– Treat groups of people as being very similarbecause they come from the same Continent (e.g., Africans, Asians) General geographic area (e.g., Latinos) Share a label developed by U.S. Census Bureau (e.g.,Hispanics, Asians and Pacific Islanders)Copyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200915
Misrepresentation in Research Pan-ethnic research designs problematicwhen – Research tests for differences (e.g., comparingapples to oranges)– Treats mixed groups as homogenous, ignoringdifferences in history, culture, immigrationexperiences, variations in race or skin color,variations in languageCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200916
SAMPLING STRATEGIES17
Sampling Operations used to select a small group ofpeople from a target population of interest In quantitative research, goal to selectmembers of targeted population who arerepresentative of, or very similar to thatpopulation Sampling methods determine if findings canbe generalizedCopyright Lorey Wheeler (2015)18
Sampling There are no unique sampling strategies forstudying ethnic minority or economicallydisadvantaged populations Pay close attention to implications of usingcertain strategies or interpreting results in suchstudies that include these groups Typical methods common to social science can beequally valid, but often must be combined withtailored recruitment processes to be successfulCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200919
Random (Probability) Sampling Samples are selected in accord withprobability theory Used when researchers want precise,statistical descriptions of large populations A sample of individuals from a populationmust contain the same variations that exist inthe populationCopyright Lorey Wheeler (2015)20
Random Sampling Each member of the target population has anequal probability of being selected into thesample Findings based on a sample represent theaggregation of elements that compose thesampling frame– A sampling frame is a list of elements in thepopulationCopyright Lorey Wheeler (2015)21
General Strengths of Random Sampling Proper use of random sampling generates asample more likely to be representative of thetargeted population than any other method– Assumes reasonably high and similar rates ofsuccessful recruitment for all segments ofpopulation Ideal for determining how many people in agiven group have experienced a given eventCopyright Lorey Wheeler (2015)22
Random Sampling:In Context of Ethnic Minority PopulationsComparative Research Designs– Random samples of general population are goodfor making cross-group comparisons on incidencerates and relationships between variables With successful recruitment rates , large sample size And when subpopulations of specific economic, racial,or ethnic groups large enoughCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200923
Random Sampling:In Context of Ethnic Minority PopulationsWithin-Group Designs– Strong foundation for studying within-groupdiversity on incidence rates or the utility oftheoretical models for that group When random sampling is applied exclusively to asingle economic, racial, or ethnic group Create sampling frame that includescommunities/schools that represent diversity in culturaland economic conditions related to target populationCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200924
Problems with Random Sampling May not capture groups that constitute a verysmall proportion of the general population ormay represent these groups in such smallnumbers that meaningful comparisons are notpossible– E.g., Sudanese immigrants Rarely represent all segments of the populationequally– Because of difficulties in recruiting disadvantagedpopulations, urban dwellers, those discriminatedagainst, those distrustful of researchCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200925
Problems with Random Sampling Well-executed random sampling designs mightnot fully represent marginalized ethnicminority or economically disadvantagedpopulations Research using random sampling may produceresults with somewhat biased perspective ofthese groupsCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200926
Summary of Random Sampling To overcome some problems of randomsampling – Researchers may want to tailor recruitmentmethods for each targeted subgroup– Monitor recruitment rates for subgroups– Use caution in interpretation and generalization ofresults to the degree that response rates are lowor uneven among one or more subgroupsCopyright Lorey Wheeler (2015)27
Stratified Random Sampling Rather than selecting a sample from apopulation, the researcher draws fromhomogenous subsets of the population.– Stratum/Strata Grouping of units composing a population intohomogenous groups before sampling Selected subgroups are proportionally represented infinal sampleCopyright Lorey Wheeler (2015)28
Use of Stratified Random Sampling Often used when important groups have asmall n in the population Used to be sure that each subgroup of samplehas the same proportion as the population Thus, may need to “oversample” from certaingroupsCopyright Lorey Wheeler (2015)29
General Advantages of Stratified RandomSampling When successful, strongest sampling design for studiesthat want to make comparisons across groups on:– Incident rates– Means– Relations between variables Example: The National Longitudinal Study ofAdolescent to Adult Health (Add Health) Nationally representative sample of U.S. schools for adolescents withrespect to region of country, ubanicity, school size, school type, andethnicity Selected using unequal probability sampling, and implicit stratificationCopyright Lorey Wheeler (2015)Harris, 200930
Stratified Random Sampling:In Context of Ethnic Minority Populations Use pre-identified characteristics on whichthere is substantial diversity within the groupas the stratifying variable– Family income– Family structure– Acculturation level– Language useCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200931
Stratified Random Sampling:In Context of Ethnic Minority Populations Modify the stratified random samplingmethod and oversample strata– Over represent groups that make up only smallportion of general population– Use when group comparisons are planned andone or more subgroups represent such smallportions of general population that comparisonswould have little statistical power to identifydifferences that exist E.g., Latinos in U.S.– Oversample country of originCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200932
Convenience Sampling Nonrandom sampling design Elements are selected for convenience sampling because they’reavailable or easy to find– Examples: intro psychology students, magazine surveys, online chatroom, school children in Lincoln This sampling method is also known as a haphazard, accidental, oravailability sample Obtained more easily, inexpensively, and quickly than othermethods Less desirable, but MOST common– Systematic differences between participants who volunteer forresearch and those who do notCopyright Lorey Wheeler (2015)33
Problems with Nonrandom Sampling Designs Rarely, if ever, are representative of the population In general, important subgroups of the population maybe represented poorly or not at all Research participants compared to nonparticipants aremore likely to be––––––College l, nonconforming, and sociableHigher need for social approvalCopyright Lorey Wheeler (2015)34
RECRUITMENT AND RETENTIONSTRATEGIES35
Recruitment and Retention Recruitment– Process of trying to get individuals selected duringsampling to participate in study– Together with sampling, determines initial quality ofthe sample studied (sample representativeness ofpopulation) Retention– For longitudinal studies, critical for maintaining qualityof sample– If particular subgroup of sample drops out, difficult tomake inferences about that subgroupCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200936
Questions What will motivate a member of anunderrepresented ethnic minority oreconomically disadvantaged group to participatein a study? How can a representative sample be obtained? Once recruited, what can a researcher do toconvince participants from these populations tocontinue in the study?Copyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200937
Practical Issues Hindering Recruitment Individuals and families often more mobile thangeneral population––––May rent on weekly or month-to-month basisDisruptions in income may result in evictionsSeasonal farm workersMay differ from less-mobile participants on key personal orfamily characteristics Less likely to have access to working telephones or tobe listed in directories– May use pre-paid cell phones with number changing often– May rely on neighbors or relatives for access to phoneCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200938
Practical Issues Hindering Recruitment Low-income ethnic minority populations more oftenlive in high-density inner-city communities– Higher income members of ethnic minority groups areoften widely scattered throughout urban and suburbanareas More cost-effective to recruit ethnic minoritypopulations in inner-city neighborhoods– Produces biased samples that may be less representativeof the target population than is desirable– E.g., interest in the experiences of Latino parents acrossdemographic characteristics, inner-city sample wouldunderrepresent the portion of the population that ismiddle to upper-classCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200939
Practical Issues Hindering Recruitment Researchers interested in immigrantpopulations need to be prepared to work inindividual’s native language– This is a considerable undertaking withimplications for measurement and translationprocesses– Need for bilingual recruitment staff and flexibilityCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200940
Practical Issues Hindering Recruitment Broader societal issues related to discriminatoryexperiences of the group Adversarial relationships between some members ofthese groups and authority figures Groups may have numerous reasons not to trustresearchers– Families may react defensively when approached by arecruiter because of assumptions that the recruiter is anauthority figure from the police, immigration, or socialservice agency, who too often represent a threat for manylow-income familiesCopyright Lorey Wheeler (2015)Knight, Roosa, & Umaña-Taylor, 200941
Overcoming Challenges to Recruitment Research on recruitment practices tends to bedescriptive and informally comparative at best Strategies––––––Understanding and acknowledging participants’ valuesMinimize barriers to participationCreate culturally relevant materialsSeek out the gatekeepersMaintain a presence at local community eventsProvide participant leadership and volunteer trainingopportunitiesCopyright Lorey Wheeler (2015)National Resource Center for Healthy Marriage and Families (2015)42
Understanding and Acknowledging Participants’ Values Treat all potential participants respectfully When working with middle-class groups, manyresearchers also have a middle- or upper-classsocioeconomic status, and, thus, understand the rulesof social interactions of their own social class group When working with ethnic minority or economicallydisadvantaged populations, become familiar with thetargeted group prior to trying to engage groupmembers in research– Rules of social interaction– Attitudes toward authority figures, and– Values and belief systemsCopyright Lorey Wheeler (2015)National Resource Center for Healthy Marriage and Families (2015)Knight, Roosa, & Umaña-Taylor, 200943
Minimize Barriers to Participation Minimize logistical barriers– Providing transportation and child-care– Offer schedule and location choices Increase opportunities for two-waycommunication– Personal contact– Present family with benefits and goals of study; letthem ask questions– Incorporate family in decision making processCopyright Lorey Wheeler (2015)Clarke, Sheridan, Sommerhalder, Wheeler, & Bhatia (2015, August)44
Minimize Barriers to Participation Increase cultural awareness among projectstaff– If possible, match participants and researchpersonnel on demographic characteristics Provide direct (face-to-face) contact andpersonalized interactions across all phases ofthe study– Conducting interviews in participant’s home orother locations of preferenceCopyright Lorey Wheeler (2015)Clarke, Sheridan, Sommerhalder, Wheeler, & Bhatia (2015, August)45
Create Culturally Relevant Materials Materials visually appealing, have useful content, anddeliver consistent message Use photos and cultural and ethnic symbols thatresonate with the target population Use the same or similar visuals and graphicsthroughout materials to make them easily recognizable Use appropriate language and consider the audience’sliteracy levels and learning styles– Write text at a fourth-grade reading level– Create bilingual materialsCopyright Lorey Wheeler (2015)National Resource Center for Healthy Marriage and Families (2015)46
Create Culturally Relevant Materials Be positive in your message and emphasize benefits tothe family, not just the i
middle-class European American group (often considered to be the normative comparison group) –Sample of the ethnic minority group often economically disadvantaged –Results of such comparisons often contribute to a “deficit model” of the ethnic minority groups Comparative research designs
REL Northwest Recruiting, Training, and Retaining Bilingual and Diverse Teachers 3 Partnership and Recruitment Successful partnerships establish a fully articulated system for identifying, recruiting, training, and retaining bilingual and diverse teacher candidates. These systems are based on a shared vision
Oracle Taleo Enterprise Recruiting User Guide Chapter 1 Recruiting Overview 1 1 Recruiting Overview Recruiting Overview Recruiting Process Oracle Taleo Recruiting is a solution that helps enterprises source, assess, and hire the best talent. The Oracle Taleo Enterprise suite includ
Sampling, Sampling Methods, Sampling Plans, Sampling Error, Sampling Distributions: Theory and Design of Sample Survey, Census Vs Sample Enumerations, Objectives and Principles of Sampling, Types of Sampling, Sampling and Non-Sampling Errors. UNIT-IV (Total Topics- 7 and Hrs- 10)
Oracle Taleo Enterprise Fluid Recruiting User Guide Chapter 1 Overview 1 1 Overview Overview Fluid Recruiting Overview Fluid Recruiting is an HTML recruiting solution that helps enterprises assess and hire the best talent. Fluid Recruiting offers a simple and modern user interface usable on desktop, laptop, tablet, and smartphone. Fluid
SoURCE: SoCIAL RECRUITING SURvEy, jobvITE 2011. ESSENTIAL RECRUITING STATS 22 yES 63.6% NO 36.4% 2/3 of Companies have Successfully hired Through Social Networks successfullY hireD from social networKs in 2010 58% SoURCE: SoCIAL RECRUITING SURvEy, jobvITE 2011. ESSENTIAL RECRUITING STATS 23
earth retaining structures. This dissertation aims to study the behavior of retaining walls executed under historical facades, maintained through a retaining structure, based on a case study related to the construction of a hotel at Avenida Duque de Loulé. For the numerical study of the behavior of the peripheral and facade retaining .
Retaining Wall: Walls on both sides of the track that hold back the surrounding earth where there are changes in grade Buffer around retaining walls Portal roof Retaining wall Tunnel Figure 2: Diagram depicting the various components of a typical portal and retaining walls Portals & Retaining Walls TRANSIT DESIGN GUIDE
2.4. Data quality objectives 7 3. Environmental sampling considerations 9 3.1. Types of samples 9 4. Objectives of sampling programs 10 4.1. The process of assessing site contamination 10 4.2. Characterisation and validation 11 4.3. Sampling objectives 11 5. Sampling design 12 5.1. Probabilistic and judgmental sampling design 12 5.2. Sampling .
