Correlations pt.2

if you know 1 of the 2 variable values, how confident can you be predicting the value of the second variable

high correlation → confident; moderate → more uncertainty; weak→ low to none

whats a good way to estimate correlation strength

ask how comfortable you are making an assumption between the 2 variables

for correlations can we say the relationsjip has an effect or not?

no only if its significant → cannont establish causal relationship as it doesnt mean causation

how would we expressive our results

using was/wasn’t a significant relationsip betwee….

what does the r mean in:

there was not a significant relationship between studying time and exam enjoyment, r(89) = .68, p= .59

symbol for pearson correlation coefficient, tells this is a correlational analysis

what does the 89 mean in:

there was not a significant relationship between studying time and exam enjoyment, r(89) = .68, p= .59

degrees of freedom, # of (x,y) data pairs in study minus 1

what does the .68 mean in:

there was not a significant relationship between studying time and exam enjoyment, r(89) = .68, p= .59

value of correlation coefficient, test statistic

what does the ,59 mean in:

there was not a significant relationship between studying time and exam enjoyment, r(89) = .68, p= .59

likehood of gettng a correlation this strong if null were true

do we talk about effect sizes in correlations

no the strenght is a good way to talk instead

what is r² in a correlation

acts as the “effect size” measure; % of change in variable 1 tha can be accounted for by change in v2

what about Confidence intervals

the same → no 0 in range =reject the null and are confident in the estimate of correlation, 0 in range = fail to reject and not confident