Biostats
- can be descriptive or explanatory
- can be retrospective or prospective
- can have a comparison or not
- can be hypothesis testing or generating
- NO INTERVENTION is designed or implemented
observational studies
examples of observational studies:
case reports
case series
cohort study
cross-sectional study
case-control study
2 types of categorical (qualitative) data:
nominal (has a name; ex. ABO blood types, race)
ordinal (ordered; ex. pain scales, Likert scale)
2 types of numerical (quantitative) data:
discrete data (whole numbers; ex. office visits per year)
continuous data (continuum, decimals, precision of instrument)
ex. height, weight, BP, age
data collection that occurs after study objectives and methods are defined; collected
going forward
prospective
data collection in which pre-existing data is reviewed; data collected for other
purposes is examined
retrospective
brief, objective, descriptive report that details the outcome or characteristic from
group of clinical subjects; typically describes treatment outcomes, disease
symptoms, and disease endpoints
retrospective; no comparison group; limited statistical significance or generalizability
case series
study that takes a group without the outcome(s) of interest and follows them over
time; attempts to determine which factors produce the outcome(s) of interest
cohort study
what type of study determines relative risk of disease (RR)?
cohort study
risk of disease among all exposed/risk of disease among unexposed
1/100 = (risk/odds)
, 1:99 = (risk/odds)
risk
odds
the focus of cohort studies=
exposures
what are case-control studies?
retrospective study comparing cases (those with outcome of interest) to controls
(similar individuals to the cases without the outcome of interest)
only evaluates one outcome (concerned with disease)
what kind of study identifies odds ratios (OR)?
case-control studies
case-control studies are best for (common/rare) diseases
cohort studies are best for (common/rare) diseases
rare
common
can case-control studies determine risk of disease?
no
"snapshot" study, most common study design, typically survey-based
cross-sectional study
review types of studies
to make generalization possible, the sample must be a representative sample;
ideally, some form of ___________ sample is used
random
the group you want to generalize to
population
ones you can actually access= study population
a list from where you can draw your sample
sampling frame
ex. electoral register, school class lists
sampling in which each individual selected from the population is chosen randomly
and entirely by chance, such that each individual has the same probability of being
- can be descriptive or explanatory
- can be retrospective or prospective
- can have a comparison or not
- can be hypothesis testing or generating
- NO INTERVENTION is designed or implemented
observational studies
examples of observational studies:
case reports
case series
cohort study
cross-sectional study
case-control study
2 types of categorical (qualitative) data:
nominal (has a name; ex. ABO blood types, race)
ordinal (ordered; ex. pain scales, Likert scale)
2 types of numerical (quantitative) data:
discrete data (whole numbers; ex. office visits per year)
continuous data (continuum, decimals, precision of instrument)
ex. height, weight, BP, age
data collection that occurs after study objectives and methods are defined; collected
going forward
prospective
data collection in which pre-existing data is reviewed; data collected for other
purposes is examined
retrospective
brief, objective, descriptive report that details the outcome or characteristic from
group of clinical subjects; typically describes treatment outcomes, disease
symptoms, and disease endpoints
retrospective; no comparison group; limited statistical significance or generalizability
case series
study that takes a group without the outcome(s) of interest and follows them over
time; attempts to determine which factors produce the outcome(s) of interest
cohort study
what type of study determines relative risk of disease (RR)?
cohort study
risk of disease among all exposed/risk of disease among unexposed
1/100 = (risk/odds)
, 1:99 = (risk/odds)
risk
odds
the focus of cohort studies=
exposures
what are case-control studies?
retrospective study comparing cases (those with outcome of interest) to controls
(similar individuals to the cases without the outcome of interest)
only evaluates one outcome (concerned with disease)
what kind of study identifies odds ratios (OR)?
case-control studies
case-control studies are best for (common/rare) diseases
cohort studies are best for (common/rare) diseases
rare
common
can case-control studies determine risk of disease?
no
"snapshot" study, most common study design, typically survey-based
cross-sectional study
review types of studies
to make generalization possible, the sample must be a representative sample;
ideally, some form of ___________ sample is used
random
the group you want to generalize to
population
ones you can actually access= study population
a list from where you can draw your sample
sampling frame
ex. electoral register, school class lists
sampling in which each individual selected from the population is chosen randomly
and entirely by chance, such that each individual has the same probability of being
