Presentation is loading. Please wait.

Presentation is loading. Please wait.

Sampling An Introduction to Scientific Research Methods in Geography GEOG 4020.

Similar presentations


Presentation on theme: "Sampling An Introduction to Scientific Research Methods in Geography GEOG 4020."— Presentation transcript:

1 Sampling An Introduction to Scientific Research Methods in Geography GEOG 4020

2 Sampling Without the ability to access all entities available to research, scientists are necessarily limited to small numbers of cases, places, times, variables, and measures. Without the ability to access all entities available to research, scientists are necessarily limited to small numbers of cases, places, times, variables, and measures. Sampling – the way of selecting a subset from the population Sampling – the way of selecting a subset from the population Virtually anything can be understood as a sample from a larger set of possible choices. Virtually anything can be understood as a sample from a larger set of possible choices. Introduces complexity and uncertainty to the research process Introduces complexity and uncertainty to the research process

3 Sampling Frame Sampling frame – the subset of the population from which cases are drawn to become part of the sample Sampling frame – the subset of the population from which cases are drawn to become part of the sample Example: when sampling from a list of registered voters, a researcher solicits participation through phone calls. Example: when sampling from a list of registered voters, a researcher solicits participation through phone calls. The sampling frame is those on the list whose phone number is correct, are available to receive the phone call, and are willing to participate. The sampling frame is those on the list whose phone number is correct, are available to receive the phone call, and are willing to participate.

4 Sampling Design Sampling Design – the procedure used to identify cases from the sampling frame to go into the sample Sampling Design – the procedure used to identify cases from the sampling frame to go into the sample Nonprobability sampling Nonprobability sampling Convenience sampling Convenience sampling Snowball sampling Snowball sampling Probability sampling Probability sampling Simple random sampling Simple random sampling Systematic random sampling Systematic random sampling Stratified random sampling Stratified random sampling Cluster sampling Cluster sampling Multistage area sampling Multistage area sampling

5 Implications Representativeness – the degree to which the smaller set (sampling frame or sample) resembles the larger set (population or sampling frame, respectively) Representativeness – the degree to which the smaller set (sampling frame or sample) resembles the larger set (population or sampling frame, respectively) Generalizability – what larger set can conclusions be validly drawn about from the evidence of the smaller set? Generalizability – what larger set can conclusions be validly drawn about from the evidence of the smaller set? Shortcomings of sampling can be more or less important in different research and are typically not fatal flaws. Shortcomings of sampling can be more or less important in different research and are typically not fatal flaws.

6 Nonparticipation and Volunteer Biases Nonparticipation bias – the sample becomes less representative of the sampling frame when cases given the opportunity to participate choose not to do so. The bias becomes greater when there is a difference between participants and nonparticipants. Nonparticipation bias – the sample becomes less representative of the sampling frame when cases given the opportunity to participate choose not to do so. The bias becomes greater when there is a difference between participants and nonparticipants. Record number and characteristics of nonparticipants Record number and characteristics of nonparticipants Maximize participation Maximize participation Reminders Reminders Incentives Incentives Shorten, clarify a study or make a study more interesting or relevant Shorten, clarify a study or make a study more interesting or relevant Volunteer bias – cases select themselves Volunteer bias – cases select themselves

7 Spatial Sampling From Continuous Fields Spatial sampling of points from fields

8 Continued Independent vs. nonindependent (dependent) spatial sampling – sampling on the basis of trends in the data Independent vs. nonindependent (dependent) spatial sampling – sampling on the basis of trends in the data Spatial interpolation – The method to make inferences from samples back to the continuous field Spatial interpolation – The method to make inferences from samples back to the continuous field

9 Sample Size The compromise: larger samples are more likely to be representative of the sample frame, however they are more likely to cost more money, time, and effort to obtain. The compromise: larger samples are more likely to be representative of the sample frame, however they are more likely to cost more money, time, and effort to obtain. Some studies require one case to make a conclusion (is it possible for an artificial reef to sustain populations of fish species?) Some studies require one case to make a conclusion (is it possible for an artificial reef to sustain populations of fish species?) Sometimes thousands are required (political polling) Sometimes thousands are required (political polling) Studies testing causal relationships (regression) require 20-30 or more. Studies testing causal relationships (regression) require 20-30 or more. More variables require more samples for the same level of confidence More variables require more samples for the same level of confidence


Download ppt "Sampling An Introduction to Scientific Research Methods in Geography GEOG 4020."

Similar presentations


Ads by Google