Study design Flashcards
What is a cohort study?
Longitudinal forward prospective study
Cross sectional at points- follow up
State 3 strengths and 3 weaknesses of cohort study?
Strengths
1) Demonstrate causal effects
2) Multiple diseases studied
3) Multiple exposure studied
Disadvantages
1) Expensive and time consuming
2) Not suitable rare diseases
3) Need deal confounding factors
What is cross sectional study?
Investigate what happening at certain point in time
Outcome and exposures measured simultaneously
State 3 strengths and weaknesses of cross sectional?
Strength
1) Relatively fast and inexpensive
2) Rapid feedback current events
3) Multiple outcome and exposure studied
Disadvantage
1) Not suitable rare disease
2) Limited to study causation
3) Can effected selection bias and confounding factors
If investigators assign exposure to patients what trial?
Rct
Depend random allocation
What are two types study design?
Experimental
Observational
Explain case control study?
Retrospective
Past exposures
Sample taken based disease
State 4 advantages and 4 disadvantages of case control ?
Strength
1) Quick and inexpensive
2) Suitable rare diseases
3) Multiple exposures studied
4) Suitable transient exposures (outbreaks)
Weaknesses
1) Only single disease studied
2) Exposure data needed before start study
3) Can affected selection bias and confounding factors
Explain RCT?
Have entry criteria-randomly split groups
Double blind
What is gold standard?
RCT
State 4 weaknesses of RCT?
1) Expensive
2) Not practical show long term
3) Affected by non compliance
4) Might not ethical
Give 3 examples of confounding factors?
Income
Sex
Age
Define association
Define causation
Association- statistical link between exposure and disease- may not reflect cause and effect relationship
Causation- statistical link where disease directly caused by exposure
What are 2 types of sampling?
Probability (random) methods
Non-probability (non random) methods
State 2 disadvantages for non probability (non random) methods?
Sampling error can’t measured
High potential bia
Convenient
State advantage for probability (random) methods?
Sampling error can be measured
Explain simple random sampling?
Disadvantages?
Each member of pop has equal prob being selected
Eg lottery
Disadvantages
Not always good rep- minority groups
Not as precise stratified sampling
Explain systematic random sampling?
Disadvantages?
Members of population selected at equal intervals
K=N/n
Disadvantages
Need complete list pop
Large standard error
Periodicity pattern- characteristics occur regular intervals in pop
Easy to do
Explain stratified random sampling?
Pop split into groups- samples taken each group
Eg 3 ppl from each of 20 group
State advantages and disadvantages of stratified random sampling?
Advantages
Increase representation
Can prod population estimates
Smaller standard error
Disadvantages
Require prior info
Expensive
Explain cluster sampling?
Population participated clusters- sample clusters selected
Clusters made to be similar
List advantages and disadvantages of cluster sampling?
Advantages
Don’t need complete list population
Cheap
Disadvantages
High standard error- decrease statistical precision