Table of Contents
Graphing and Linear Relationships
The blue curve represent normal growth for girls (95&5 profiles)
The growth of 90% of all girls will be fall in the yellow area
Sara’s growth curve is below normal with a smaller slope
Sera’s growth rate is 4 cm/yr. compared to normal (6 cm/yr.)
Growth hormones were recommended
Sara is catching up
The slope of the line is greater after treatment than before
Jason is also short
However, Jason has a genetic history for short size
and a normal slope indicated that the hormone treated would not help
A scatter plot both the x value and the y value must be input
In 1977 there were 447,000 boat registration and 13 manatee deaths
In 1978 there were 460,000 boat registration and 21 manatee deaths
Each point is a Case
Clearly the manatee deaths increase with registrations
The association could be negative
The scatter plot could be linear
Or exponential
or no association
Watch for outliers
Here is a plot of cavities as a function of fluoride
Green for small towns Red for towns above 10,000
A linear plot the variable change is constant, in this case one week
A linear curve can be represented by the equation y = a + bx
In this case the inercept is 0
The slope is Dy/Dx
The equation y = 0 + 500,000x can be used to plot line from two points
The residual are the distances from the points to the line
The residue can be positive
The points below the line have a negative residuals
A residual plot can indicate if the is a problem
This curve can be fit with a linear fit
A plot of the residuals indicate that they are not random, and the linear fit is not appropriate
Extending a linear plot from 6 weeks to 8 year can be a problem
|
Author: Compaq User
Email: Grow@adm.njit.edu
Home Page: http://www-ec.njit.edu/~grow/statisti.htm
Download presentation source
|