Week 6 Lectures Flashcards
summarize characteristics of a dataset
ex) demographic factors (age, sex, race)
descriptive statistics
allow you to test a hypothesis, determine associations, or assess whether data is generalizable to the broader population
inferential statistics
2 or more groups being measured
nominal- descriptive, no order ex) sex
ordinal- “ordered”, can give number ex) strongly agree, neutral, strongly disagree
categorical variable
can be quantified as a number
continuous- any number is possible btw 2 integers ex) age, weight
interval- degree of difference btw 2 values ex) temperature
discrete- whole integers ex) # of children
numerical variables
statistical test calculate ___ ___ - a # describing how much the relationship btw variables in your test differs from the H0
test statistic
indicates likelihood (probability) of obtaining a result at least as extreme as that observed in a study by chance alone
significant at 0.05 or less
doesn’t really give us enough info
p value
used to make inferences about population parameters
parametric
data that does not fit a normal or known distribution
nonparametric
used to compare the MEANS of 2 groups
tells you how significant the difference btw the group means are
T test
used to compare differences btw means of 3 or more groups
ANOVA
used when you have the same measure that participants were rated on at more than 2 time points
ex) performing training program study, want to measure participants resting HR one month before they start, at midpoint, and one month after the program ends to see if there is significant difference in mean resting HR across the 3 time points
repeated measures ANOVA
has 2 independent variables (ex, eye color and BMR category)
main effect: each factors effect considered separately
interaction effect: all factors considered at same time
Have 3 hypotheses
have to calculate an F value for each hypothesis
two way ANOVA
used to test whether 2 or more categorical variables are related to each other (binary, nominal, or ordinal)
non-parametric hypothesis test of independence, inferential statistical test
best way to organize data is in a contingency table (2x2)
more accurate for and used for LARGE sample
Chi-square (x^2) test
if the X^2 value is greater than the critical value (found in table/software), then the difference between the observed and expected distributions is:
statistically significant
if the X^2 value is less than the critical value (found in table/software), then the difference between the observed and expected distributions is:
not statistically significant
similar to X^2 in that it tests for nonrandom association or relationships btw 2 categorical variables
used for SMALL samples
ex) if total n < 20 or if n is btw 20 and 40 and one of the true expected cell frequencies is < or = 5
Fisher’s Exact Test
measure of the linear correlation btw 2 variables
denoted by “r”
btw -1 and 1- measure the strength and direction of the relationship btw 2 variables
Pearson correlation coefficient
descriptive statistic, describing the strength and direction of the linear relationship btw 2 quantitative variables
also an inferential statistic so it can be used to test statistical hypotheses– whether there is a significant relationship btw 2 variables
Pearson correlation coefficient
used to describe relationships between variables by fitting a line to observed data
allows you to estimate how a dependent variable changes as the independent variable(s) change
linear- has only one independent variable
regression models
a statistical technique that can be used to analyze the relationship btw a single dependent variable and several independent variable
can be linear or non linear
use when you want to know how strong relationship is btw 2 or more independent variables and one dependent variable (ex) how rainfall and temperature affect crop yield)
multiple regression
kind of t-test, groups come from same population (ex before and after treatment)
paired t-test
kind of t-test, groups come from 2 different populations
two sample t-test
kind of t-test, group is compared against a standard value
one sample t-test
kind of t-test, assesses whether one population mean is greater or less than the other
one tailed t-test