Chapter 3 Flashcards
What makes something an independent variable markscheme
X is independent because it’s NOT SUBJECT TO RANDOM VARIATION
Why is something an outlier
Does not appear to fit the trend , where all the points appear to be on this trend except this point, thus outlier
Difference between corraltion and association
Correlation is LINEAR ASSOCIATION, association describes the trend between two variables but it could be exponential etc
Thurs stronger correlation means stronger close to a LINEAR trend
Difference between random and non random variable which are both independent
Random is something like rainfall which you can’t control non random is like amount of fertiliser
The point of using a scatter graph? When we have PMCC
Allows you to confirm the trend
- removes outliers
- allows you to see if the data fits a BIVARIATE NORMAL, which allows you to use data , here if data is roughly elliptical it came from a BIVARIATE normal which allows us to do PMCC hyptjbridd
Also tells us trned which allows us to see what to do, linear then PMCC else quadratic might have to do rank
Remember scatte graph what’s the point
We use to see if data is roughly elitpvi which suggests it came from a BIVARIATE normal and hence we can use data in PMCC
Remember r look out for wacky scatter diagrams
1) two islands
- for example when you have two variables that both increase with a THIRD VARIABLE, then they kinda appear correlated , but by each island no correlation so can’t use
- IQ and height , bith infrease with age pretty much, DOESNT MEAN IQ LINKS WITH HEIGHT
2) if the scatter is influenced by a few outliers
- remove outliers if still ELIPTICAL calm, else no
2)
What does r give you
The scale of correlation
Closer to +-1 means strong negative positive correlation
Closer to 0 means NO CORRELATION at all
1 and -1 is PERFECT CORRELATION
What does data coming from BIVARIATE normal mean
Two sets of data
Which are both NORMALLY DISTRIBUTED
When plotted as scatter , thus show a roughly eliptical shape
