Intro to Biostats Terms

population

class of people they are trying to make generalizations about, everyone

sample

portion of the population,

inference

process of studying the sample and then using inductive reasoning to to make generalizations about the population

parameters

describe population characteristics with numerical facts, estimated

sample statistic/estimate

number that can be computed from the sample

varience

spread of observations around a center point

descriptive statistics

deals with methods for summarizing and or presenting data

inferential statistics

deals with methods for making generalizations about a population based on info contained in the sample

categorical data

data that takes on a distinct value

nominal or ordinal

nominal data

type of categorical data

there is no apparent ordering to the categories

yes/no

ordinal data

type of categorical data

there is a natural ordering to the categories

severity is low, med, high

quantitative data

data that takes on numeric values

continuous data

type of continuous data

set of all possible values consists of all real numbers over some interval or continuum

decimals and blood pressure

discrete data

type of continuous data

the set of all possible values that consists of distinct numeric values

count of number of tumors

a row represents…

a different study unit - patient, community, etc.

a column represents…

a different varibaleva

variable

each different kinds of data that describe some feature of your study

absolute frequency

the number of observations in a category, raw data

relative frequency

percentage of observations in a category, relative to the whole sample

contingency table

summarizes categorical variables

mean

sum over total

highly sensitive to outliers

population mean

(mu) observations over observation in population

sample mean

(x bar) smaller dataset info

median

value in which is larger than half and smaller than half

robust to outliers

for skew

skew is measured by the tail, right skewed = tail is on the right

right skewed

mean is larger

left skewed

median is larger

measures of dispersion

number designed to reflect degree of spread or variability within a dataset

range, variance, standard deviation, interquartile range

range

difference between the largest and smallest value in a dataset

not robust to outliers