Biostatistics: Sampling Methods

The Research Process

• The research process involves several stages, including planning, study design, data collection, data processing, data presentation, data analysis, interpretation, and publication.

Planning a Research

• Planning a research involves considering the participants/subjects of the study.
• Population: a complete set of items or subjects which can be studied.
• Target Population: a collection of items that have something in common for which we wish to draw conclusions at a particular time.
• Study Population: the specific population from which data are collected.
• Sample: A subset of the study population.

Sampling

• Sampling is vital to allow for generalizability of the research.
• Generalizability is a two-stage process: which allows the researcher to generalize conclusions from the sample to the study population and then from the study population to the target population.
• Sampling methods can be classified into two types: probability sampling and non-probability sampling.

Non-Probability Sampling

• Non-probability sampling is a sampling technique in which the researcher selects the sample based on subjective judgment rather than random selection.
• Not all the members of the population have a chance to participate in the study.
• Types of non-probability sampling include convenience sampling, quota sampling, judgmental sampling, and snowball sampling.

Convenience Sampling

• Convenience sampling involves selecting samples from the population directly because they are most accessible and available for the researcher.
• Advantages: very low cost, extensively used/understood, no need for list of population elements.
• Disadvantages: variability and bias cannot be measured or controlled, volunteer bias.

Quota Sampling

• Quota sampling involves generating homogenous groups, based on specific qualities and traits, that represent the population.
• Percentage composition of each group in the population (similar to stratified sampling) is calculated.
• Convenient procedure is used to select the final subjects.
• Advantages: moderate cost, very extensively used/understood, no need for list of population elements, introduces some elements of stratification, representative with regard to known characteristics.
• Disadvantages: variability and bias cannot be measured or controlled, volunteer bias.

Judgmental Sampling

• Judgmental sampling involves selecting subjects purposively on the basis of having particular features.
• Most case studies are done in this manner.
• Sample size may not be large but an in-depth study of the cases is the main focus.
• Advantages: moderate cost, commonly used/understood, sample will meet a specific objective, useful for qualitative research, useful for rare characteristics.
• Disadvantages: bias.

Snowball Sampling

• Snowball sampling involves the researcher asking other known contacts to find people who would like to participate in the study; i.e., referrals.
• Enables researcher to reach groups that are otherwise hard to reach.
• Used in rare events, e.g., when studying rare behaviors in the population.

Error Types

• Non-sampling error: statistical error that arises from the result of data collection, which causes the data to differ from the true values.
• Types of non-sampling error include systematic error, random error, non-response error, measurement error, interviewer error, adjustment error, and processing error.

Non-Sampling Error: Systematic Error

• Systematic error arises when the sample is not drawn properly.
• Affects the sample of the study and often creates useless data.
• Usually, the fault of the researcher.
• Consistent and repeatable.

Non-Sampling Error: Random Error

• Random error arises from the degree to which the sample is not perfectly representative of the population.
• Errors that cannot be accounted for and just happen.
• Not the fault of the researcher.
• Each error offsets the other, hence, they are often of little to no concern.

Reducing Errors

• Errors can be eliminated by increasing the sample size or the number of samples.
• Using valid research designs.
• Applying statistical corrections.