SAMPLING & DATA COLLECTION Flashcards
Define POPULATION
Well-defined group with specific characteristics
All the individuals the researcher is interested in studying.
Define SAMPLE
Subset of overall population
Set of elements that make up population
Define CONVENIENCE SAMPLE.
All members of the population with the relevant characteristics who can be readily found (and consent).
Define SNOWBALL SAMPLING
A participant refers the researcher to more potential participants, who may then refer researcher to further potential participants (snowballing).
Define PURPOSIVE SAMPLING
An intentional (purposeful) approach is made by the researcher to select participants with specific characteristics or participants within a specific area.
Define QUOTA SAMPLING
A sample gathered to represent population as closely as possible. E.g. 40% of population is male so try to make sure 40% of sample is male.
Define SIMPLE RANDOM SAMPLING
Participants allocated ‘randomly’ to the study or part of a study: ‘pulled out of a hat’ chosen by computer.
Define STRATIFIED RANDOM SAMPLING
Members of the population allocated to groups according to characteristics important to the study and then subjects randomly chosen from these groups.
Why are ELIGIBILITY CRITERIA so important?
Characteristics specific to allow generalisability of findings.
What are the MAIN purposes of SAMPLING?
Increase efficiency of a study
Maintain representativeness of sample
Name the two MAJOR headings under which sampling falls:
Probability
Non-probability
What type of samples are PROBABILITY suited to?
Simple random sample
Cluster random sample
Systematic sample
What type of samples are NON-PROBABILITY suited to?
Quota
Purposive sample
Convenience sample
What are the advantages of random sampling?
No researcher bias.
Maximise representativeness
What is the aim of stratified random sampling?
increase representativeness
What are the disadvantages of non-probability sampling?
less rigorous
limits generalisability
not representative
Name 4 qualitative data collection methods
in-depth interviews: may be structured, semi-structured or unstructured
focus groups: involve multiple participants discussing an issue
Secondary data/document review: diaries, written accounts of past events, photographs
Observations: may be on site, or under lab conditions, e.g. where participants are asked to role-play a situation to show what they might do.
How is trustworthiness/rigor assured in qualitative research?
credibility (truthfulness)
auditability (consistency)
transferability (fittingness/applicability)
confirmability (no bias or distortion)
Member checking; audit trails; triangulation
Why is rigor so important?
Need to know methods can be trusted and have confidence in results and using them to apply to your clinical practice.
List 4 quantitative data collection methods
physiologic/laboratory-based: clinical/experimental trials
observations: observing and recording well-defined events (such as counting the number of pts. waiting in ED at specified time of day)
Questions & self-report scales - questionnaires: Administer surveys with closed-ended questions
Interviews: face-to-face and telephone interviews
Define reliability and validity in relation to measurement error.
Reliability means a measure that can be relied upon consistently to give the same result if the aspect being measured has not changed.
Validity reflects how accurately the measure yields information about the true or real variable being measured. A measure is valid if it measures correctly and accurately what it is intended to measure.
Descriptive statistics allow researchers to:
Describe, organise and summarise raw data
Inferential statistics allow researchers to:
Estimate how reliably they can make predictions and generalise their findings based on the data.
The purpose of descriptive statistics is to:
Organise and summarise the data.
Name 4 levels of measurement in quantitative data analysis and briefly define each of these:
Nominal: discrete categories
Ordinal: relative ranking
Interval: specific numerical distance between scores - treated as equal; continuous
Ratio: as above but has absolute zero
Name and briefly describe the 3 most common measures of central tendency:
Mean: average score
Median: middle score
Mode: most common score
Briefly define cross-sectional studies:
Collect all data at one point in time
Longitudinal studies collect data at different points in time
Retrospective studies collect data on past events
Prospective studies collect data as they occur
What is an independent variable?
Manipulated variable (cause) Used to predict outcome of interest i.e. dependent variable
What is a dependent variable?
Measured variable (effect) Consequence/presumed effect that changes with change in independent variable
Name two types of validity and briefly define each.
Internal validity: does the independent variable accurately measure what it says it will measure. Asks whether independent variable really made the difference - refers to the causal relationship
External validity: deals with the problem of generalisability of findings to other populations and other environmental conditions
