Ch 5 Flashcards
What are measures of dispersion?
- Average absolute deviation
- Average squared deviation (variance s^2)
- Standard deviation (square root of s^2)
What are box and whisker plots?
- box plots give you the median (the line), the box represents the interquartile range
- the whiskers (bars at the end) take 1.5 times the interquartile range and stop there, (whiskers) indicate variability outside the upper and lower quartiles,
Statistical Relationship
a relationship between two variables means we can express changes in one variable as a function of changes in the other
-expressed by 2-dimensional scatterplot, illustrating bivariate distribution
Relationship form
the form of the relationship describes the shape of the relationship between the variables. With two quasi-continuous variables, there are 3 possibilities
- linear (positively related)
- non-linear (no relation)
- flat line (
Relationship strength (between two quasi-continuous variables)
How strong two variables go together
(estimate of strength is dependant on the form of the relationship-using the wrong estimate of strength will misrepresent the form)
Conditional mean function
is the set of conditional means. It is the regression of Y on X, the form of the relationship between X and Y is equivalent to the shape of the conditional mean function
Measures of strength
-Covariance
-Correlation
they capture the degree to which two variables vary together; more specifically, how closer the observed scores are to the regression mean
Pearson Product Moment Correlation Coefficient (PPMC)
- bounded between -1 and +1
- values greater than zero indicate a positive linear relationship (1=perfect positive linear relationship)
- Values less than zero indicate a negative linear relationship (-1 =perfect negative relationship)
- value of 0=no relationship
General type preposition
asserts something presumably true of each and every member of a designable class
Aggregate type proposition
asserts something presumably true of the class considered as an aggregate
Aggregate type proposition
asserts something presumably true of the class considered as an aggregate (making a claim of an average in respect to a class)
Why transform?
- to bring empirical distribution in line with a theoretical one
- change the scale (e.g., cm to mm, farehnheit to celsius)
- ‘centre’ the data
- make something more interpretable
impact of linear transformation on measures of location
Xbar= aXbar+b
same for every other value