what is another term for r?
correlation coefficient
what is another term for r^2?
coefficient of determination
how would you prove reliable data?
interpolated data
r value
r^2 value
amount of data
r^2 comment
(r^2 percentage) of the variation in y can be associated with the variation in x
r comment
There is a strong/moderate/weak positive/negative correlation between the variables. There is strong/some/little evidence to show that the greater the level of x the greater the level of y
which axis is the dependent variable
y
which axis is the independent variable
x
how do you find percentage from a two way table?
number/total x 100 = ans
what is the residual?
distance from the trend (predicted) line
how do you find the residual?
subtracting predicted y value from actual y value
when can’t you eliminate a point? (outlier)
if there isn’t enough data
it is checked and is real data
it is the end point of the data
how do you find a percentage from a two way table?
amount/total x 100 = %
comment for a percentage graph from two way table
x have a slight/major preference for the blank, whereas y prefer blank (include percentages)
what are the types of time series data (explain)
seasonal - pattern of data repeats over a regular time period
cyclic - the same pattern repeats but not over a regular time period
random - there is no real pattern
how would you deseasonalise data
find average for each cycle
divide actual data by cycle averages
what is the equation for deseasonalised data
actual data/seasonal index = deseasonalised data
or
a/i = d
deseasonalised data comment
on (day) sales were higher/lower than expected according to deseasonalised data
how would you forecast actual data
find trend line (of deseasonalised data)
sub number into x
use ans as d in a=di
*this number is never reliable as it is extrapolated data
how would you smooth data (3 point average)
(data+data+data)/3
a point is lost either end