Advantages of Sampling
- more economical
- more time efficient
- can be more accurate because there is greater control over the measurements and procedures
What is sampling bias
When the members of a sample over or under represent attributes of the population that are related to areas being studied.
Random Sampling
- equal chance of selection
- reduces risk of systemic bias
Systematic sampling
Similar to random sampling but every nth subject from a population list is chosen instead
Stratified Sampling
-population split into groups of similar individuals from which a simple random sample is drawn
If the population in Brisbane consists of 300 physios, 60% female and 40% male, then a sample of 60 physios stratified by gender would involve selecting 36 females and 24 males for your sample.
Disproportionate Sampling
For example, if a sample consisted of 90% females we might want to still select an equal number of females and males in the sample.
This is called disproportionate sampling. It is still probability sampling but the probabilities of a subject being selected are now not equal.
Cluster Sampling
The successive random sampling of a series of units in a population
-may compound the effects of bias
- Select 5 Australian states
- Choose a random sample of 10 hospitals in each state 3. Randomly select 2 physiotherapists from each hospital
Convenience Sampling
Simply chosen on the basis of avaliability
Quota Sampling
- Researcher guides the sampling process until the participant quota is met
- eg Volunteers could be called for until quota of males and females met
Purposive Sampling
participants ‘hand picked’ based on certain criteria
Snowballing Sampling
When the desired characteristics of the required sample are rare.
-relies on original participants identifying or referring other people with same characteristics
RCT
-one group receives treatment other doesn’t compare results
Crossover design
In a crossover design each subject acts as his/her own control but the order of treatments is randomised.
Factorial design
- Several factors compared at same time
- Each subject receives a combination of all factors such that all combination are received by some subjects
Single subject designs
- unit of study is a single patient
- OM before & after
- Weak generalisation and external validity
Retrospective
-studies look at past data, such as medical records, for potentially influential factors and outcomes
Prospective
studies follow participants into the future, observing exposure and then tracking outcomes
Longitudinal study
A population of subjects is identified by a common link.
Attrition
Loss of subjects to a study and is likely in studies with multiple follow up points
Case control studies
- Retrospective in time’
- similar individuals are matched (one with condition and one without), and then the presence or absence of an event of interest.
Cross sectional studies
‘Snapshot’ of a certain fact or issue at a given point in time
-often used to determine the prevalence of a disease/disability
-
Questionnaire
Adv
- efficiency and convenience
- can reduce bias
- respondents may be more candid in responses
Dis
-potential for misunderstandings questions
Histograms
- Useful for visualising large numbers of observations
- good picture of location, spread and shape
- somewhat subjective (based on number of bins)
- Difficult to compare more than two groups
Bar charts
Categorical variables (nominal or ordinal) we use bar charts to show counts or proportions
