PMCC Flashcards
What is an independent variable and dependant varaible ?
- independant: traditionally drawn on x axis, not determined or affectd by the other
out the two which are independant variables and which are dependant?
weight of crop yeild or amount of rainfall while crop is growing
- amount of rainfall: independnant
- weight of crop yield: dependant
what are random variables?
not predicable
what are controlled variables?
non random, used to determine the effect on the other varible
what is correlation?
how close is a data to a straight line
perfect negative/positive = exaclty a straight line
* a type of association (linear association)
what does it mean for data to be ellipitcal?
can draw a lips /oval around it
misleading graphs?
what is the diffrence between correlation and association?
- correlation is linear association
- association can be non linear
what is the PMCC
- Product moment correlation coefficent (PMCC)
- gives a value betwen -1 and 1
what do the PMCC values represent?
what is a normal distribution?
bell shape symentric
* majority in the centre
what is bivariate normal distribution?
- two varaibles are coming from normal distributions (random)
- forming the eplitical shape
What is the formula for the product moment correlation coefficient?
steps for hypothsis testing for PMCC
one tail test
- define p eg: Let p be the (parent) population correlation coefficent betwen exam scors and number of hours spent revising
- H0: p=0
- H1: p>/< 0 (one tailed test)
- calculate the value of r
- look at data shet table to find the critical value for n eg: For n=8 the 5% critical valye is 0.6215
- see if calculated value falls within region. If r > critical value, then r is significant, so we reject the null hypothesis
- write concluding statement
*eg: There is evidence to suggest that ther is a positive correlation between exam scores and number of hours spent revising *
steps for hypothsis testing for PMCC
two tail test
- define p eg: Let p be the (parent) population correlation coefficent betwen exam scors and number of hours spent revising
- H0: p=0
- H1: p ≠ 0 (two tailed test)
- calculate the value of r
- look at data shet table to find the critical value for n eg: For n=8 the 5% critical valye is 0.6215
- see if calculated value falls within region. If r > critical value, then r is significant, so we reject the null hypothesis
- write concluding statement
*eg: There is evidence to suggest that ther is a any correlation between exam scores and number of hours spent revising *
what is the siginificance level?
definition
the probability of rejecting the null hypothesis when in fact it is true
what is the p value?
NOT THE R VALUE
* used to compared against significant level purley (no need for data sheet)
* for two tailed test half the given significance level
what are cohens guidelines for effect size?
- r=0.1 : small effect size
- r=0.3 : medium effect size
- r=0.5 : large effect size
What modelling assumptions must be satisfied before it is valid to use PMCC in a hypothesis test?
- both variables must be random (not control variables since they are not random)
- data must be drawn from a normal bivariate population
PMCC is only valid for random on random data and the speeds are controlled (so not random).
Since data is not random on random, it cannot be used
Scatter graph not roughly in an elliptical shape (instead two islands)
This suggests that the underlying population did not have a bivariate normal distribution.
What is significance level (example) ?
Eg: (significance level 1%) If there is no association between x and y only about 1 sample in 100 would lead to the conclusion that there is an association between x and y