China
Africa
Explore chapters and articles related to this topic
Communication
Published in Walter DeGrange, Lucia Darrow, Field Guide to Compelling Analytics, 2022
When selecting participants for a focus group, it is important to consider who will be most likely to contribute useful information. A few different ways to select participants are: Random sampling: In this method, everyone in the population has an equal chance of being selected.Stratified sampling: In this method, the population is divided into groups (or strata) and people are selected from each group at random.Cluster sampling: In this method, clusters of people are randomly selected, and then individual participants are chosen from within the clusters.Quota sampling: In this method, a certain number of participants from specific groups are chosen.Convenience sampling: In this method, participants are selected based on convenience (for example, people who are available at a specific time and place).
Thematic Map Accuracy Assessment Considerations
Published in Russell G. Congalton, Kass Green, Assessing the Accuracy of Remotely Sensed Data, 2019
Russell G. Congalton, Kass Green
In addition, autocorrelation may affect the size and number of samples used in a cluster sampling approach, because each sample unit may be contributing not new, independent information but rather, redundant information. Therefore, it would not be effective to collect information for many sample units in a large cluster, since the contribution of each new sample unit in the cluster could very quickly be reduced to very little because of this lack of independence. However, cluster sampling is a very cost-effective method, especially in the field, when the cost of traveling from one sample location to another can be very high. Even when the accuracy assessment samples are taken in the office from aerial imagery, cluster sampling can create savings in set-up time for each image. Therefore, it is important to consider spatial autocorrelation and balance the impact of having spatially autocorrelated samples against the efficiencies of cluster sampling. This can be done by limiting the number of samples taken in the cluster to between two and four, making sure that each sample unit in the cluster is taken in a different thematic class, and spreading the samples as far apart as possible. Without understanding these considerations, so that the statistical validity can be effectively balanced with practical application, the accuracy assessment process will not be as efficient as it should be.
Machine Learning
Published in Pedro Larrañaga, David Atienza, Javier Diaz-Rozo, Alberto Ogbechie, Carlos Puerto-Santana, Concha Bielza, Industrial Applications of Machine Learning, 2019
Pedro Larrañaga, David Atienza, Javier Diaz-Rozo, Alberto Ogbechie, Carlos Puerto-Santana, Concha Bielza
There are different random selection methods. However, if standard procedures are followed, mathematical expressions can be used to quantify the accuracy of the estimations. Cluster sampling is based on the idea that the whole population can be clustered into smaller subpopulations called clusters. Clusters are homogeneous and are treated as the sampling unit. Suppose that the factory has 1000 machines playing the role of clusters, cluster sampling can select 20 of these machines and inspect all the pieces manufactured by this smaller number of machines. Stratified sampling is used when the target population can be easily partitioned into subpopulations or strata. Strata are then chosen to divide the population into non-overlapping and homogeneous regions, where elements belonging to a given stratum are expected to be similar. Stratified sampling assumes that the different strata are very heterogeneous. Simple random samples are taken from each stratum. For example, if our factory has three types of machines, each producing different pieces, stratified sampling will select some pieces at random from each of these subpopulations.
A cross-cultural comparison of the self-efficacy of middle-school mathematics teachers across Turkey and the United States
Published in International Journal of Mathematical Education in Science and Technology, 2022
Sevim Sevgi, Giray Berberoglu, Paul Cobb, Thomas M. Smith
The samples of the study were selected concerning cluster sampling methodology where the schools were randomly selected and all the mathematics teachers within the selected school constituted the cohort of the study in both countries. The sample used in the USA comes from the Middle-school Mathematics and the Institutional Setting of Teaching (MIST, 2012) project. Schools in Turkey were selected in the Ankara district only. The self-efficacy scale was administered on an individual base in both countries. In the USA, 22 schools were selected in two states. In Turkey, 115 schools were selected, and they constitute 32.85% of all the schools in the middle-school level in the Ankara district. Thus, 22 schools with 245 American middle-school mathematics teachers and 115 schools with 379 Turkish middle-school mathematics teachers were used as the samples of the study. In the USA sample, there were 171 female and 69 male teachers, whereas, in the Turkish sample, there were 279 female and 100 male teachers. Age distributions of teachers are given at Table 1. 10.3% of the teachers in the Turkish sample and 3.3% of the teachers in the American sample did not provide information about their ages.