STATS 3 - establishing associatons i Flashcards
DEPENDENT VARIABLE - Y
the variable we seek to explain /predict
INDEPNDENT VARIABLE -X
the variable/s that Explain or predictthe dependent variable
3 components of association
NATURE OR DIRECTION
STRENGTH OF ASSOCIATION
STATISTICAL SIGNIFICANCE
Nature of direction
of the relationship between 2 variables e.g resp with high education more likely to turn out in elections
statistical significance
how likely it is that the association you observe in sample will generalise to population
LINEAR REGRESSION
designed for quantitative dependent variable
-scatterplot
least squares estimation
establishes the line that under some assumption best fits the data trend - often also called the sum of squared errors
errors= difference between data point and regression line
LINEAR ASSOSCIATIONS
-what does the sign of the slopes tell us ?
NATURE / DIRECTION OF LINEAR ASSOSCIATIONS
-B>0=positive relationship - as x inc y inc
B<0 = negative relationship - as x inc y dec
B=0 independence
strength of linear association- regression slop and intercept
PEARSONS R
- what does the level of R show ??
R > 0.40 - STRONG RELATIONSHIP
R>0.20 = MODERATE
R<0.20 = WEAK
r=1.00 -perfect correlation
higher r value = Stronger association
smaller P value ?
reject null hypothesis
what is T test
for regression co efficient - collects all evidence against Null hypothesis - derive p value SMALLER
what are confidence intervals ?
same purpose as p value - but provides more info
-represents a range of values in which parameter will fall in population with a given probability
sign =??
coefficient size of pearson r=??
p value =??
SIGN = DIRECTION
coefficient sized of pearson r = STRENGTH of association
P value SMALLER= STATISTICAL SIGNIFICANCE
