Geography
Sampling Techniques
Sampling techniques in geography refer to the methods used to select a representative subset of a larger population for study or analysis. Common techniques include random sampling, stratified sampling, and systematic sampling. These methods help ensure that the sample accurately reflects the characteristics of the larger population, allowing geographers to make inferences and draw conclusions based on the sample data.
Written by Perlego with AI-assistance
Related key terms
1 of 5
12 Key excerpts on "Sampling Techniques"
- Karen Kemp(Author)
- 2007(Publication Date)
- SAGE Publications, Inc(Publisher)
373 S AMPLING In GIS, the data that are used are often derived from some sort of sample. The term sampling refers to the process of establishing such samples. A sample is simply a subset of some population of entities that we wish to know about; this population is sometimes called the target population. Identifying the target population is not always easy, as it is defined by the varying objectives of each specific study. Sometimes, the entire population will be sufficiently small, which enables its inclusion in total in the study to give what is called a census. In this case, data are gathered on every member of the population. Usually, however, the population is too large for the researcher to attempt to survey all of its members, and a small, care-fully chosen sample can be used to represent it. Clearly, if it is to be of any value, the sample must rep-resent the whole of the population of interest. A sample that properly represents the population is said to be unbiased and is selected using a sampling design that has as its basis a list of all the subjects from which the sample is to be chosen. In work with human beings, populations are usually large but finite, for example, an electoral register, telephone directory, or other membership list. In such cases, it is meaning-ful to talk of the sampling fraction, which is the pro-portion of the population analyzed. In environmental work, the populations are often infinite, for example, all possible slope angles or the complete set of all pos-sible values for precipitation across the same area. In the first case, this might be sampled by field survey, and in the second, through a limited number of rain gauge sites. Simple Random Sampling We can recognize a series of general strategies for sampling from a population, all of which have a spa-tial equivalent. In all cases, it is assumed that we can isolate these samples from a defined population. Sampling methods are classified as either probability or nonprobability.
eBook - PDFData Collection in Fragile States
Innovations from Africa and Beyond
- Johannes Hoogeveen, Utz Pape(Authors)
- 2020(Publication Date)
- World Bank(Publisher)
103 1 Introduction Technological advances in geospatial data have the potential to change how survey data are collected. Long hampered by high costs, limited capacity, and difficulties in supervision, sample selection is often done using second-best or nonprobability approaches. As geospatial technol-ogy has improved and become more widespread, costs have come down and the number of available tools have increased, making Geographic Information Systems (GIS)-based sampling approaches accessible to more users. This chapter presents experiences with GIS-based sampling from three different settings: (i) where no sampling frame is present because the census is outdated; (ii) sampling pastoralist communities; and (iii) rapid listing of enumeration areas to reduce exposure of field 7 Methods of Geo-Spatial Sampling Stephanie Eckman and Kristen Himelein © International Bank for Reconstruction and Development/The World Bank 2020 J. Hoogeveen and U. Pape (eds.), Data Collection in Fragile States , https://doi.org/10.1007/978-3-030-25120-8_7 S. Eckman RTI International, Washington, DC, USA K. Himelein ( * ) World Bank, Washington, DC, USA e-mail: [email protected] 104 S. Eckman and K. Himelein teams. The case studies focus on extreme situations, particularly those in conflict-prone areas, as innovation often takes place when few other options are available. The applications discussed here, however, are applicable to many less extreme situations. 2 Data Challenge and Innovation #1: Creating a Sampling Frame in the Absence of a Census For many studies, no sampling frame of the target population is avail-able. The most common approach to addressing this problem for large-scale household surveys in the developing world is to use a strat-ified two-stage design. In the first stage, census enumeration areas are selected as the Primary Sampling Unit (PSU), using probability propor-tional to estimated size.
- Kumar, K Nirmal Ravi(Authors)
- 2021(Publication Date)
- Daya Publishing House(Publisher)
Thus, the sampling technique of data collection is very much flexible and adaptable to the changes in conditions of the enquiry. Response rate from the respondents will be high. It is easier on the part of the researcher to supervise fewer (sample) respondents than the whole population. The results of census method can be checked with Sampling Techniques. That is, sample data is also used to check the accuracy of the census data. If the researcher goes for a census survey, there is no guarantee on the reliability or confidence about the quality of information generated. In such case, sampling technique assume greater advantage. That is, the This ebook is exclusively for this university only. Cannot be resold/distributed. inference about the population parameters is possible only when the sample data is collected from the selected sample. The sampling technique is very much scientific in its approach. Particularly, the techniques of random sampling are based on the Theory of Probability, which is a mathematical concept. This technique is based on certain important laws and principles viz ., Law of Statistical Regularity (Law of Statistical Regularity says that a moderately large number of the items chosen at random from the large group are almost sure on the average to possess the features of the large group) , Law of Inertia of Large Numbers ( the other things being equal – the larger the size of the sample; the more accurate the results are likely to be) , Principle of Persistence, Principle of Optimization, Principle of validity etc. This facilitates the researcher to compute the sampling error and degree of reliability of the results by using Sampling Techniques. The analytical results of sampling technique are more dependable than those derived from the census survey.
No longer available |Learn more- Frederick J Gravetter; Lori-Ann B. Forzano; Tim Rakow, Frederick Gravetter, Frederick Gravetter, Lori-Ann Forzano, Tim Rakow(Authors)
- 2021(Publication Date)
- Cengage Learning EMEA(Publisher)
Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it. Exercises 126 The two basic categories of Sampling Techniques are probability and non-probability sampling. In probability sampling, the odds of selecting a particular individual are known and can be calculated. Types of probability sampling are simple random sampling, systematic sampling, stratified sampling, proportionate stratified sampling and cluster sampling. In non-probability sampling, the probability of selecting a particular individual is not known because the researcher does not know the population size or the members of the population. Types of non-probability sampling are convenience and quota sampling. Each sampling method has advantages and limitations and differs in terms of the representa- tiveness of the sample obtained. K E Y WO R D S population sample representativeness representative sample biased sample selection bias, or sampling bias sampling probability sampling random process non-probability sampling E X E R C I S E S The exercises are identified with specific learning objectives and are intended to assess your mastery of the objectives. 1. In addition to the key words, you should also be able to define each of the following terms: target population accessible population law of large numbers sampling methods, or Sampling Techniques, or sampling procedures simple random sampling systematic sampling stratified random sampling proportionate random sampling proportionate stratified random sampling cluster sampling convenience sampling quota sampling 2. (LO1) A researcher conducting a political poll for a general election would like to know the attitudes of university students concerning the candidates.
eBook - PDFResearch Methods in Geography
A Critical Introduction
- Basil Gomez, John Paul Jones, John Paul Jones, III, Basil Gomez, John Paul Jones, Basil Gomez(Authors)
- 2010(Publication Date)
- Wiley-Blackwell(Publisher)
Systematic samples are chosen according to a rule (every 5 km, for example). This kind of sampling involves systematically sampling an area or individuals. For example, assume you were using soil point sampling in a large field to determine average soil pH. If you generate simple random points to calculate your average you might not account for pH in all portions of the field. Conversely, a systematic collection of points (one point every 10 meters, for example) would help to ensure that all parts of the field are sampled. A stratified sample is implemented when researchers know that the population contains different sub-populations and he/she samples within each of these. Stratified sampling helps to ensure that all of the variation present in a group is measured in the sample. This is done first by determining any specific groups or sub- groups that may exhibit different characteristics than the rest of the population. For example, when conducting a political poll, it might be necessary to ensure that all socio- demographic groups (income, race, gender, etc.) are sampled to reflect the real opinion of the people. In a stratified sample the samples can be done either randomly or systemati- cally within each sub-population. Sampling with probability proportional to size is done to ensure that the probability of being selected is not dependent on the size of a group or sub-group being sampled. For example, assume you are collecting surveys from several cities and towns within a state. This scheme helps maintain the probability that if you are one of 500,000 people within a Sampling Our World 83 large city that you have the same probability of being selected if you were one of 500 people in small town. When there are natural groupings within a population, cluster sampling can be employed. Clusters are first defined within a population, and a sample of these clusters is collected.
eBook - PDFMeasurements from Maps
Principles and Methods of Cartometry
- D H Maling(Author)
- 2016(Publication Date)
- Butterworth-Heinemann(Publisher)
However, he is at pains to emphasise that this should not be taken as licence for anyone to believe that the insistence of statisticians on random sampling can, in general, be ignored. Spatial Sampling and Cartometry 135 Stratified Sampling The second major subdivision of plane sampling methods are the derived strategies which involve stratification of the population. The basic intention of stratification is to combine the advantages of both systematic and random sampling, namely their accuracy and their statistical validity respectively. This type of sampling is most often used when accuracy statements are to be derived for each thematic class represented on a map, as well as an accuracy statement for an entire map. It has been noted that there are two basic criteria for defining the strata. One is simple geometrical dissection of the sampling frame into contiguous compartments, usually of equal size; the other is subdivision based upon prior knowledge of the study region, derived from field-work, photographic interpretation or simply from study of existing maps. Choice of the number of strata required is also a matter for judgement. The population variability and the purpose of the investigation are major considerations which influence this. We have already seen that there are two principal kinds of stratified sampling is common use. STRATIFIED RANDOM SAMPLING In this case, illustrated by Figs. 8.7 and 8.8, the sampling units are drawn at random within each compartment or stratum. Since stratification extends throughout the sampling frame, every unit comprising the population has the same chance of being drawn if every stratum is of the same size and contains the same number of units. Consequently this form of sampling strategy allows the exercise of statistical tests which are the same as those used for unrestricted random sampling.
eBook - PDF- M.R. Carter, E.G. Gregorich, M.R. Carter, E.G. Gregorich(Authors)
- 2007(Publication Date)
- CRC Press(Publisher)
The random order of treatment placement is achieved using random number tables or computer-generated randomizations. 1.4 SAMPLE LAYOUT AND SPACING Although many types of sampling designs exist (reviewed in Gilbert 1987; Mulla and McBratney 2000; de Gruijter 2002) only two main types (random and systematic) are commonly used in the soil and earth sciences. Inventory studies can be completed using any of the designs discussed in the following two sections. Pattern and geostatistical studies typically use transect or grid designs, as is discussed in more detail in Section 1.5. 1.4.1 S IMPLE R ANDOM AND S TRATIFIED R ANDOM S AMPLING In simple random sampling all samples of the specified size are equally likely to be the one chosen for sampling. In stratified random sampling, points are assigned to predefined groups or strata and a simple random sample chosen from each stratum. The probability of being selected can be weighted proportionally to the stratum size or the fraction of points sampled can vary from class to class in disproportionate sampling. Disproportionate sampling would be used if the degree of variability is believed to vary greatly between classes, in which case a higher number of samples should be drawn from the highly variable classes to ensure the same degree of accuracy in the statistical estimates. Stratified sampling (correctly applied) is likely to give a better result than simple random sampling, but four main requirements should be met before it is chosen (Williams 1984): 1 Population must be stratified in advance of the sampling. 2 Classes must be exhaustive and mutually exclusive (i.e., all elements of the population must fall into exactly one class). 3 Classes must differ in the attribute or property under study; otherwise there is no gain in precision over simple random sampling. 4 Selection of items to represent each class (i.e., the sample drawn from each class) must be random.
eBook - PDFSpatial Statistics
GeoSpatial Information Modeling and Thematic Mapping
- Mohammed A. Kalkhan(Author)
- 2011(Publication Date)
- CRC Press(Publisher)
so. efficiently . .Sampling.tech-niques.commonly.used.to.assess.the.accuracy.of.a.classification.procedure. includes. simple. random. sampling,. systematic. sampling,. stratified. random. sampling,. and. cluster. sampling. (Congalton. 1991;. Kalkhan. 1994;. Kalkhan. et.al . .1997) . .Multiphase.(i .e., .double).sampling.is.also.used.to.assess.thematic. mapping. accuracy. (Kalkhan. et. al . . 1998) . . The. following. provide. simplistic. descriptions. of. the. most. common. sampling. designs. used. by. researchers,. resource.management,.and.so.forth.into.the.integration.of.geospatial.infor-mation.for.natural,.landscape-scale,.and.environmental.studies . Simple Random Sampling Simple. random. sampling. (SRS). is. the. fundamental. selection. method. (Husch.et.al . .1982) . .Husch.et.al . .(1982,.p . .162).point.out.that.all.other.sam-pling. procedures. are. modifications. of. simple. random. sampling,. which. are.designed.to.achieve.greater.economy.or.precision . .A.simple.random. 42 Spatial Statistics sample.is.one.in.which.the. n .sample.points.are.chosen.independently.and. uniformly.within.the.region.“1.or.2.or.3 .” .Simple.random.sampling.has.the. advantage.over.other.designs.in.that.it.is.easy.to.apply.and.provides.satis-factory.results.in.evaluating.the.accuracy.assessment.of.remotely.sensed. data. .SRS.is.the.most.basic.design.and.should.work.well.in.describing.the. spatial. continuity. of. almost. any. variable. and. allows. one. to. make. infer-ences. about. the. population. of. interest . . There. some. disadvantage. when. using.SRS . .First,.the.sample.size.within.each.thematic.class.is.proportional. to.its.area.(Congalton.1991) . .Second,.there.is.increased.cost.associated.with. traveling.between.sample.points . .Also.some.areas.within.the.landscape. (forest.landscape.such.as.rare.habitat,.aspen.stand).may.not.be.sampled . . For. example,. if. you. are. sampling. a. population.
eBook - PDFMeasuring Customer Satisfaction and Loyalty
Survey Design, Use, and Statistical Analysis Methods
- Bob E. Hayes(Author)
- 2008(Publication Date)
- ASQ Quality Press(Publisher)
In this case, this sample is referred to as a biased sample, one that is not representative of the population. Statistical sampling increases the chances that the sample is representative of the population. Table 5.1 defines and details important points of the three sampling methods just explained. Census sampling is relatively expensive and judgmental sampling carries the potential for unintended bias; these methods are less often used than statistical sampling. Statistical sampling is the most reliable method of gathering data and providing useful information about the population. The next section will focus on the different statistical 86 Chapter Five Table 5.1 Sampling methods. Method Definition Important Points Census Select all cases from the • Sample is representative of the population • population because it is the • population • Could be used when feedback from all customers is important • Could be used if all data are easily accessible (data are already on computer) • Cost may be high Judgment Select subset of cases • Degree to which results from based on discretion of • sample can generalize to popula- person creating the sample • tion is questionable • Easy approach Statistical Select subset of cases • Can calculate probability that our based on chance • sample is not representative of the population • Can generalize results to the population sampling methods: simple random sampling, stratified sampling, and cluster sampling. Simple random sampling. This is the simplest approach to determining which cases are to be included in the sample. In simple random sampling, every case in the population has an equal chance of being included. In our newspaper scenario, we may be interested in the overall satisfaction level of all readers (subscribers and those buying the paper at a newsstand), but we want to survey only 100 customers. To generate a random sample from our population of customers, we start with a list of all readers (N 9000).
eBook - ePubField Sampling
Principles and Practices in Environmental Analysis
- Alfred R. Conklin, Jr.(Authors)
- 2017(Publication Date)
- CRC Press(Publisher)
The sample thus obtained is subjected to laboratory analysis. Both methods are valuable and sometimes essential in field sampling. Both have limitations, and the best sampling results are obtained when all ways of investigating the medium of concern are used. In all cases knowing the location of sampling activities is essential. Determining the location is best done using a global positioning system (GPS). Because of its importance, GPS will be discussed before examining either noninvasive or invasive sampling.In many cases noninvasive sampling is done before invasive or physical samples are taken, mainly because these methods show surface and subsurface features affecting the invasive sampling pattern. In some cases these methods are necessary for the safety of sampling personnel and buried utilities. They will thus be used to develop the invasive sampling plan. Because they will be used first they will be discussed before discussing actual invasive sampling.Data obtained by all sampling methods can be combined in a geographical information system (GIS). Combining data allows them to be both stored and displayed on maps of the area of interest (e.g., thematic maps). As many or as few data as needed can be displayed at one time. Additionally, a GIS system can relate data obtained from sampling to data obtained from other sources, such as aerial maps and soil surveys.5.1 General Sampling Considerations
Analytical and instrumental methods of analysis are very precise and accurate, but the results of analyses of environmental samples are not. This means that the variability comes either from the sampling procedures or during sample handling and storage prior to analysis. Sampling is the most important source of variability, and both the sampling and handling process must be carried out with great attention to detail.Sampling is the act of isolating a portion of a larger entity, analyzing it, and using the analytical results to describe the characteristics of the whole entity. In statistical terms the sample is an individual, which is a member of a family of similar individuals. Analysis shows that even environmental samples taken from sites close together may not be individuals or members of a family. When samples are taken at progressively increasing distances from a specific location they change in one of several ways. Characteristics can change gradually until a very different individual is identified, or they may change in a repetitive manner over long distances.
eBook - PDFResearch Methods For Business
A Skill Building Approach
- Uma Sekaran, Roger Bougie(Authors)
- 2020(Publication Date)
- Wiley(Publisher)
Easy to use if sampling frame is available. Systematic biases are possible. 3. Stratified random sampling (Str.R.S.) Proportionate Str.R.S. Disproportionate Str.R.S. Population is first divided into meaningful segments; thereafter subjects are drawn in proportion to their original numbers in the population. Based on criteria other than their original population numbers. Most efficient among all probability designs. All groups are adequately sampled and comparisons among groups are possible. Stratification must be meaningful. More time consuming than simple random sampling or systematic sampling. Sampling frame for each stratum is essential. 4. Cluster sampling Groups that have heteroge-neous members are first identified; then some are chosen at random; all the members in each of the randomly chosen groups are studied. In geographic clusters, costs of data collection are low. The least reliable and efficient among all probability sampling designs since subsets of clusters are more homogeneous than heterogeneous. 5. Area sampling Cluster sampling within a particular area or locality. Cost-effective. Useful for decisions relating to a particular location. Takes time to collect data from an area. 6. Double sampling The same sample or a subset of the sample is studied twice. Offers more detailed informa-tion on the topic of study. Original biases, if any, will be carried over. Individuals may not be happy responding a second time. Non-probability sampling 7. Convenience sampling The most easily accessible members are chosen as subjects. Quick, convenient, less expensive. Not generalizable at all. 8. Judgment sampling Subjects selected on the basis of their expertise in the subject investigated. Sometimes, the only mean-ingful way to investigate. Generalizability is questionable; not generalizable to entire population. 9. Quota sampling Subjects are conveniently chosen from targeted groups according to some predetermined number or quota.
eBook - PDFResearch Methods For Business
A Skill Building Approach
- Roger Bougie, Uma Sekaran(Authors)
- 2021(Publication Date)
- Wiley(Publisher)
Easy to use if sampling frame is available. Systematic biases are possible. 3. Stratified random sampling (Str.R.S.) Proportionate Str.R.S. Disproportionate Str.R.S. Population is first divided into meaningful segments; thereafter subjects are drawn in proportion to their original numbers in the population. Based on criteria other than their original population numbers. Most efficient among all probability designs. All groups are adequately sampled and comparisons among groups are possible. Stratification must be meaningful. More time consuming than simple random sampling or systematic sampling. Sampling frame for each stratum is essential. 4. Cluster sampling Groups that have heteroge- neous members are first identified; then some are chosen at random; all the members in each of the randomly chosen groups are studied. In geographic clusters, costs of data collection are low. The least reliable and efficient among all probability sampling designs since subsets of clusters are more homogeneous than heterogeneous. 5. Area sampling Cluster sampling within a particular area or locality. Cost-effective. Useful for decisions relating to a particular location. Takes time to collect data from an area. 6. Double sampling The same sample or a subset of the sample is studied twice. Offers more detailed informa- tion on the topic of study. Original biases, if any, will be carried over. Individuals may not be happy responding a second time. Non-probability sampling 7. Convenience sampling The most easily accessible members are chosen as subjects. Quick, convenient, less expensive. Not generalizable at all. 8. Judgment sampling Subjects selected on the basis of their expertise in the subject investigated. Sometimes, the only mean- ingful way to investigate. Generalizability is questionable; not generalizable to entire population. 9. Quota sampling Subjects are conveniently chosen from targeted groups according to some predetermined number or quota.
Index pages curate the most relevant extracts from our library of academic textbooks. They’ve been created using an in-house natural language model (NLM), each adding context and meaning to key research topics.
