Intro to Biostats Flashcards
population
study subjects of all individuals
what is the study population
final group of individuals selected for a study
sample
subset or portion of the full population
“representatives”
when is it useful to choose a sample
when studying complete population is not feasible
random process commonly utilized to draw sample
what are the measurements in a study?
dependent and independent variables
which are outcome variables, dependent or independent?
dependent
how are comparisons made between study groups?
statistical analyses
how are inferences made about the sample?
deprived from measurements and their comparisons
also utilized for entire populations by making inferences for generalizability
null hypothesis
research perspective which states there is no true difference between groups being compared
most conservative and commonly utilized
researchers accept or reject this perspective based on results
null hypothesis
what statistical perspectives are taken by researchers with null hypothesis
superiority
non inferiority
equivalency
alternative hypothesis
research perspective that states there will be a true difference between groups being compared
2 key attributes of data measurement
magnitude (dimensionality)
consistency of scale (fixed interval)
also include rational zero
a scale is consistent if
there are equal and measurable spacing between the units
3 key levels and attributes of measures
nominal
ordinal
interval
nominal attributes
NO magnitude
NO consistency of scale
NO rational zero
labeled variables without qualitative characteristics (i.e. career, gender)
ordinal attributes
YES magnitude
NO consistency of scale
NO rational zero
(i.e. pain scales, “strongly agree, strongly disagree, agree, neutral)
interval/ratio attribute
YES magnitude
YES consistency of scale
no or yes rational zero
(i.e. number of living siblings, personal age, physiological measurement)
changing levels of data: can you go up or down in specificity?
you can only go DOWN in specificity, never up
how can you convert blood pressure down the scale?
interval: actual numbers (i.e. 120/80)
ordinal: hypertensive, hypotensive, normal
nominal; high blood pressure, low blood pressure
which attributes are descrete? interval?
nominal and ordinal descrete (only these choices)
interval: continous
descriptive statistics
non comparative, simple description of various elements of study’s data
measures of central tendency/dispersion
mode/median/mean
outliers
minimum/maximum/range
IQR
variance
from the mean
difference in each individual measurement value and the group’s mean
standard deviation
square root of the variance value (restores units of mean)