PHAR 628 Lecture 5: Sampling - Research Methods & Biostatistics

PHAR 628 Research Methods & Biostatistics Lecture 5: Sampling Lixian Zhong, Ph.D. Feb 5th, 2025 1 Quick Review Steps in Scientific Research Process Pose...

PHAR 628 Research Methods & Biostatistics Lecture 5: Sampling Lixian Zhong, Ph.D. Feb 5th, 2025 1 Quick Review Steps in Scientific Research Process Pose a research question and hypothesis 1 Develop and implement a research plan 2 Perform data collection and analysis 3 Prepare a research report 4 2 Quick Review Pose a Research Question and Hypothesis Population, Intervention, Comparator, Outcomes, Timeline and Setting (PICOTS) framework can be used to develop a good clinical question. Economical, Clinical and Humanistic Outcomes (ECHO) are usually evaluated in pharmaceutical practice and policy research. 3 Objectives ❖ Examine the relationship(s) between population and samples ❖ Introduce terms related to research sampling: - Sampling bias - Inclusion and Exclusion Criteria ❖ Describe major sampling methods 4  Population The larger group to which research results are generalized is called the population A population is a defined aggregate of persons that meet a specified set of criteria 5  Populations and Samples A researcher chooses a sub-group of the population through a process of sampling. The sample serves as the reference group for estimating characteristics of or drawing conclusions about the population. sampling generalization A good sample should be representative of the population 6  Target Population The target population refers to the group of individuals to which the results of the study will apply. Because it is not possible to gain access to every unit, some portion of the target population that has a chance to be selected must be identified. This is the accessible population. The study sample will be chosen from this accessible population 7  Sample Levels of the sampling process 8 Hierarchy for selection of a study sample 9 What is our sample? 10 What is our sample? 11 Inclusion and Exclusion Criteria In defining the target population, an investigator must first specify selection criteria called eligibility criteria that will govern who will and will not be subjects. Eligibility criteria include both inclusion criteria and exclusion criteria. 12  Inclusion and Exclusion Criteria Inclusion criteria describe the primary traits of the target and accessible populations that will qualify someone as a subject. Exclusion criteria indicate those factors that would preclude someone from being a subject. These factors will generally be considered potentially confounding to the results – That is, they are likely to interfere with interpretation of the findings 13 Inclusion and Exclusion Criteria Used in Clinical Research - Age - Gender - Type and stage of a disease - Treatment history and ongoing treatments - Other medical conditions (comorbidities) … 14 What are the inclusion and exclusion criteria? 15 What are the inclusion and exclusion criteria? 16 Sampling Bias: What is it? To make generalizations, the researcher must assume that the responses of sample members will be representative of how the population members would respond in similar circumstances. Populations are, by nature, heterogeneous, and the variations that exist in behavioral, psychological, or physical attributes should also be present in a sample. Theoretically, a good sample reflects the relevant characteristics and variations of the population in the same proportions as they exist in the population. 17  Sampling Bias: What is it? Bias refers to systematic error introduced into the research process. Sampling bias occurs when the individuals selected for a sample overrepresent or underrepresent certain population attributes that are related to the phenomenon under study Such biases can be conscious or unconscious – Conscious biases occur when a sample is selected purposefully 18 In order to estimate the height of TAMU students… Can you identify bias in this sampling? 19  Minimize Sampling Bias Although there is no way to guarantee that a sample will be representative of a population, sampling procedures can minimize the degree of bias in choosing a sample. It is not so much the size of a sample that is of concern. - Simply increasing sample size does not reduce bias - A small representative sample may be preferable to a larger but unrepresentative sample. 20 Types of Sampling Methods https://www.youtube.com/watch?v=be9e-Q-jC-0 21 Types of Sampling Methods Samples Non-Probability Probability Samples Samples Simple Random Stratified Purposive Convenience Cluster Systematic Quota 22  Probability Sampling Probability samples are created through a process of random selection. It means that every unit in the population has an equal chance, or probability, of being chosen. This also means that every unit that is chosen has an equal chance of having some of the characteristics or exposures that are present throughout the population. 23  Probability Sampling With probability sampling, the sample should be free of any known bias and is considered representative of the population from which it was drawn. Because this process involves the operation of chance, there is always the possibility that a sample's characteristics will be different from those of its parent population. The difference between sample averages (called statistics) and population averages (called parameters) is sampling error. 24  Simple Random Sampling Sampling without replacement: once a unit is selected it has no further chance of being selected A simple random sample is unbiased in that each selection is independent No one member of the population has any more chance of being chosen than any other member E.g. Assign a number to each of the subjects in the accessible population and use computer to generate random numbers to select a sample 25  Systematic Sampling Systematic sampling is a type of probability sampling method in which sample are selected from the accessible population according to a random starting point and a fixed periodic interval (sampling interval). Systematic sampling is generally considered equivalent to random sampling, as long as no recurring pattern or particular order exists in the listing E.g. to select a sample of 100 from a list of 1,000 students, every tenth person on the list is selected. The starting point on the list is determined at random. 26 Stratified Random Sampling Stratified random sampling is a method of probability sampling that involves identifying relevant population characteristics and dividing members of a population into more homogeneous, non-overlapping subsets, or strata, based on these characteristics. Then conduct simple random sampling within each stratum. E.g. Stratify COP students into P1, P2, P3 and P4 strata and conduct simple random sampling from each stratum 27 Cluster Sampling Cluster sampling is method of probability sampling in which researchers divides the population into separate groups. Then a simple random sample is selected from each cluster. It is used when mutually homogeneous, yet internally heterogeneous groupings are evident in a population. Clusters are often generated by geographical areas. E.g. Simple random samples are selected from each county (cluster) in Texas to study the health status of Texas residents 28 Types of Sampling Methods Samples Non-Probability Probability Samples Samples Simple Random Stratified Purposive Convenience Cluster Systematic Quota 29  Nonprobability Sampling Nonprobability sampling is created by choosing samples on some basis other than random selection Because all the elements of the population do not have an equal chance of being selected under these circumstances, we cannot readily assume that the sample represents the target population. The probability exists that some segment of the population may be disproportionately represented. 30  Convenience Sampling The most common form of nonprobability sample is a convenience sample, or accidental sample With this method, subjects are chosen on the basis of availability – Ex. Recruiting patients for clinical trials based on inclusion and exclusion criteria 31 Purposive Sampling Purposive sampling is a method of non-probability sampling that select a sample based on characteristics of a population and the objective of the study. Tend to produce a biased sample May be used in marketing research E.g. Select females between 20-50 in a mall to conduct research on opinions on skincare products 32 Quota Sampling Quota sampling is a non-probabilistic method for selecting a sample via strata. In quota sampling, a population is first segmented into mutually exclusive subsets, just as in stratified sampling. Then a non-random sampling method is used to select the subjects from each subset based on a specified proportion. E.g. Stratify COP students into P1, P2, P3 and P4 strata and conduct convenience sampling from each stratum 33