pre-midterm Flashcards
clinical research
evaluates the best way to prevent, diagnose and treat adverse health issues that adversely affect individuals and families
population health research
focuses on the health outcomes and the determinants of health in groups of humans (populations)
biological research
looks at changes at the human cellular level that can be related to health outcomes
brainstorm
generating long lists of spontaneous ideas about possible research questions
concept mapping
visual listing of ideas and grouping them to reveal relationships & connections
exposure
personal characteristic, behaviour, environmental encounter, or intervention that might change the likelihood of developing a health condition
outcome
an observed event such as the presence of disease in a participant in an observational study or the measured endpoint in an experimental study
population
a group of individuals, communities, or organizations with identifiable similar characteristics
PICOT
patient/population
intervention
compared to
outcome
timeframe
a good research question is
real
testable
generalizable
purposeful
SMART
specific
measurable
attainable
realistic or relevant
timely
conceptual model
researcher sketches using boxes and arrows show relationships
probability sampling
probability of selecting each sampling unit is known
SSSCM (types)
simple random sampling
random 12/36
systematic sampling
random start then follow a frame
(every 3rd)
stratified sampling
stratas are groups divided based on geography, sex, culture etc. so you take random sampling from these distinct groups
cluster sampling
natural clusters like schools or neighbourhoods, you observe the entire cluster
multistage sampling
primary sampling units are selected, then secondary, then tertiary….
ex. municipalities, cities, neighbourhoods, individuals
non-probability based sample
convenience, and purposive sampling
convenience sample
selection from a source population due to ease of access
purposive sampling
chosen because of the insights they can provide
(key informants) selected to participate because they have expertise relevant to the study
sampling bias
(berksons, healthy worker, exclusion)
selection bias
healthier or educated people are more likely to volunteer for research
berksons bias
recruit from hospitals and therefore they’re more likely to have comorbid conditions
healthy worker bias
recruit from occupational settings: they’re more likely to be healthier than the general population
exclusion bias
occurs when different eligibility criteria are applied to cases and controls ex. the controls with health conditions related to an exposure are excluded but cases with those comorbidities are not???
type 1 å error
false positive
- yields statistically significant when its not
type 2 Beta error
false negative
- yields no significance when there is a significant difference
skip logic
codes automatically hide questions irrelevant to that participant
ex. if they click no they are not employed, further questions about employment will be SKIPPED
back translation
crucial in international: translation from one language to another and then back to the original language
pilot testing
evaluates feasibility of a research project
internal validity
do the observed results accurately reflect the true association
external validity (generalizability)
who the results can be applied to
requires internal validity
selection bias
systematic error in the way participants have been chosen
information bias
due to measurement error (where in the table they are applied - as exposed, diseased)
3 threats to validity
chance, bias, confounding
volunteer selection bias
volunteers are typically more health conscious people and from a different SES group
non-response selection bias
those suffering from a disease with a particular belief
membership selection bias
healthy worker effect
loss to follow up selection bias
sickest usually leave study early
what can be done to fix bias once it has occurred
little to nothing
confounding
distortion of the actual association due to a mixing of effects between the exposure and an incidental variable
- threatens internal validity of study
why does the confounding occur
the exposed and unexposed group are not exchangeable, they differ by other factors than just exposure status
confounders are usually more of a problem in which study
observational
how to identify a confounder
literature, consult experts, statistical tests
restriction
limit study inclusion criteria with respect to confounding factors
ex. study only men or only women
matching
produce case-control groups that have similar characteristics
randomization
?
random sampling error
?
descriptive research
monitor the publics health
evaluate the success of program
generate hypotheses about cause of disease
- cross sectional, correlational (can also be analytic)
analytic research
evaluate hypothesis about cause of disease
evaluate success of intervention program
- experimental, case control, cohort study
source population
population you are interested in knowing more about
study population
population you enrolled in your study to represent the source population
case report
health issues in one patient
case series
examine one health issue in a group of people
experimental study
investigator actively manipulates which groups receive agents (clinical and community trial)
observational study
investigator observes as nature takes its course
(cross sectional, cohort study, case-control study)
cross sectional study
group of people examined at one point in time
point prevalence
proportion of population with a characteristic at one point in time
limitation of cross sectional studies
cannot assess causality because it has no time dimension.
can be said to be associated or related but can not “cause” a disease
repeated cross sectional study
does NOT track the same individuals forward in time
correlational/ecological studies
the unit of analysis is the group, not the individual
ecological fallacy
the incorrect assumption that individuals follow the trends observed in population level data
cohort study
group of individuals followed forward in time
prospective
you start the study and follow into future
retrospective
using data from past to do the study
t/f? both retrospective and prospective, the exposure is determined before the outcome happens
true
perks of retrospective
cheap, fast, good for diseases with long latent period
however they’re more vulnerable to bias
pros of cohort study
valuable when EXPOSURE is rare
examine multiple effects of a single exposure
easy to determine temporal relationship between exposure and outcome
allows measurement of incidence
cons of cohort studies
validity affected by losses to follow up
confounding
inefficient for evaluation of rare diseases
can be expensive and time consuming
if retrospective they require good records
2 necessary requirements of controls
- must come from same source population as the cases
- must be selected independently of exposure
case control study
?
case control pros
more efficient than cohort study
suited to diseases with long latent period
optimal for rare disease
can examine multiple sources
cons of case control study
- exposure assessed after development of disease or outcome
- recall bias
- prone to selection bias in control choice
- can usually only study one disease or outcome
- inefficient for rare exposures
- cannot calculate absolute measure of association
epidemiology
the process of making a study group and a comparison group comparable with respect to extraneous factors
randomization
each study participant has the same probability of receiving treatment
- balances confounders
- creates 2 groups that are the same except one group has the treatment and one does not
minimizing bias (blinding or masking)
method of ensuring that participants or study investigators have no knowledge of whether a study participant has been assigned to the treatment or comparison group
single blind
study participant does not know whether they are receiving treatment or no treatment
double blind
neither the study participant or investigator administering treatment knows who receives it or not
triple blind
neither the study participant, investigator administering, or investigator monitoring effects knows who is receiving treatment
t/f blinding is always possible
false
t/f placebo is type of blinding
true
minimize bias?
compliance (follow protocol exactly as required)
minimize bias?
compliance (follow protocol exactly as required)
crossover trial
randomize so one group receives intervention then control… other receives control then intervention
- each person acts as their own control
