key concepts and definitions - 1.1 & 1.2

units/cases

the people or things that we collect the value of variables from (subjects, patients, etc.)

variable

specific piece of information we collect from the cases; will often be a category or numerical value

units/cases

the people or things that we collect the value of variables from (subjects, patients, etc.)

variable

specific piece of information we collect from the cases; will often be a category or numerical value

categorical/qualitative variable

if the variable is not a number or does not have numerical meaning

numerical/quantitative variable

if the variable is a number value and has numerical meaning

discrete quantitative

means there are only certain values that can be attained, no in-between values (ex: shoe sizes)

continuous quantitative

can attain in-between values

data

all the variable values we collect

population

all of the people or things we could collect data about

sample

subset of the population we collect data from

statistical inference

to gain information about the population from a sample

bias

anything that causes a sample to differ from the population (in a meaningful way)

convenience sample

a sample taken only from cases that are easily available

parameter

a number that is a property of the population

statistic

a number that is a property of a sample

level of measurement: nominal

only be categorized (no order or numerical value)

level of measurement: ordinal

can be categorized and ranked, can be numerical or categorical grades

level of measurement: interval

categorized, ranked, evenly spaced (no natural zero)

level of measurement: ratio

categorized, ranked, evenly spaced, natural zero

experiment

a study where the researchers assign the value of a variable

observational study

a study where the researchers don’t assign the value of any variables

explanatory variable

variable you use to predict the value of another variable

response variable

the variable predicted; usually the one cared about

confounding/lurking variable

a variable not accounted for by a study that could affect the outcome