Lecture 7: General Methodologies Of Study Design Flashcards
Two types of quantitative studies
Interventional and observational
What occurs in an interventional study?
Researchers forces people into groups— typically by a type of randomization
Leads to increasing evidence and hopefully causation
What’s the basis of observational studies?
“Natural” study— people are NOT forced into groups.
Useful for when group allocation is unethical.
What is a weakness of observational studies?
Most are not able to prove causation
How do you decide which type of study to use?
It all depends on the RESEARCH QUESTION that you are trying to answer.
How are subjects chosen for certain studies?
Depends on: Research question Population of interest Inclusion/exclusion criteria Case/control OR exposed/unexposed groups
How does selection criteria impact a study?
Selection criteria is inclusion/exclusion criteria. It affects generalizability and external validity
EX: mental health exchangeability
Types of Null hypotheses
Superiority
Non-inferiority
Equivalency
Superiority question/null hypothesis
Question: is my drug SUPERIOR to placebo/another drug?
Null hypothesis: drug is NOT superior to placebo/another drug
Non-inferiority question/null hypothesis
Question: is my drug NOT WORSE than another drug?
Null hypothesis: drug IS worse than other drug
Equivalency question/null hypothesis
Question: is my drug equal to other drug?
Null Hypothesis: drug is NOT equal to other drug
What are the different types of sampling/randomization?
Simple random Systematic random Stratified simple random Stratified disproportionate random Multi-stage random Cluster multi-stage random Quasi-systematic
Simple random
Assigned numbers then randomly pick numbers
Systematic random
Assign random numbers and then randomly sort the numbers, then systematically choose desired sample
Ex: top 10, bottom 10, every 3rd, etc.
Stratified simple random
Framed by desired characteristic, then use simple random to select desired sample size
Ex: stratified by gender
Stratified disproportionate random
Disproportionately uses stratified simple random sampling when baseline pop is not as desired proportional percentages to the referent population
“Weighted” to return to baseline
Useful for over-sampling
Ex: adding/removing based on race to mimic USA population
Multi-stage random
Simple at many stages in succession
EX: random region —> random city —> random hospital —> random incident
Cluster multi-stage random
ALL elements are clustered together are selected for inclusion
EX: ALL hospitals in Overland Park, ALL households in community
Quasi-systematic
A NON-PROBABILITY sampling scheme
First come first served; mostly based on convenience— not really completely random
Weakness: opportunity for BIAS (selection bias)
What is a research question?
“I wonder if…” statement
Helps frame your study and direct researchers to selecting and developing an effective stud design to answer the question.
What is study design based on?
Hypothesis
Ability/desire of researcher to force groups (randomization)
Ethics of methodoligy
Efficacy/practicality
Costs
Validity of acquired info (internal validity)
Applicability of acquired information (external validity
Population
All individuals making up a common group from which a smaller sample can be obtained, if desired
Sample
“Study population”
A subset or portion of the full, complete population
internal validity
methods inside study
assessments
accurate and reproduciblt
equipoise
genuine confidence that an intervention may be worthwhile (risk vs. benefit) in order to use it in humans
what are the 4 principles of bioethics?
autonomy, beneficence, justice, nonmaleficence
autonomy
must decide for oneself to participate and have full and complete understanding of risk/benefits
beneficence
to benefit, or do good for the patient (NOT society)
justice
equal and fair treatment regardless of patient characteristics
nonmaleficence
do no harm; researchers must not….
withhold information, provide false information, or exhibit professional incompetence
consent
agreement to participate based on being fully informed and of legal consenting age
–adults age 18+
assent
agreement to participate, based on being fully informed and given by mentally-capable individuals NOT able to give consent
–children and adolescents
IRB
institutional review board–> ethics committee
GOAL: to protect human subjects
must be done prior to start of study
risk/benefit AND 4 ethical principles MUST be met.
who develops IRB laws
the department of health and human services (DHHS)
who enforces IRB laws
office of human research protections (OHRP)
what are the different levels of IRB review
full board
expedited
exempt
full board review
used for ALL interventional trials with more than minimal risk to patients
LOTS of review is necessary.
used when: pills given, needles used, CTs needed, etc.
expedited
minimal risk and no patient identifiers
used for: surveys –> ensures no traumatic questions or extreme measures are taken
exempt
no patient identifiers, low/no risk, uses existing data/specimens (de-identified)
example: crunching numbers from CDC
who decides what level of review to use?
data safety and monitoring board (DSMB)
can stop study early if too much harm/too much benefit is occurring during study.
