A model in which your mother.
To determine if the line of best fit is appropriate for the data, examine the residual plot for randomness. If the residuals are randomly scattered around the horizontal axis without any discernible pattern, it suggests that the linear model is suitable. Conversely, if the residuals display a pattern (such as a curve), it indicates that a linear model may not be the best fit for the data.
Calculus
by figuring out the equation
If a linear model accurately reflects the measured data, then the linear model makes it easy to predict what outcomes will occur given any input within the range for which the model is valid. I chose the word valid, because many physical occurences may only be linear within a certain range. Consider applying force to stretch a spring. Within a certain distance, the spring will move a linear distance proportional to the force applied. Outside that range, the relationship is no longer linear, so we restrict our model to the range where it does work.
A model in which your mother.
One of the main reasons for doing so is to check that the assumptions of the errors being independent and identically distributed is true. If that is not the case then the simple linear regression is not an appropriate model.
It is a linear model.
There is not enough information to say much. To start with, the correlation may not be significant. Furthermore, a linear relationship may not be an appropriate model. If you assume that a linear model is appropriate and if you assume that there is evidence to indicate that the correlation is significant (by this time you might as well assume anything you want!) then you could say that the dependent variable increases by 1.67 for every unit change in the independent variable - within the range of the independent variable.
There is not enough information to say much. To start with, the correlation may not be significant. Furthermore, a linear relationship may not be an appropriate model. If you assume that a linear model is appropriate and if you assume that there is evidence to indicate that the correlation is significant (by this time you might as well assume anything you want!) then you could say that the dependent variable decreases by 0.13 units for every unit change in the independent variable - within the range of the independent variable.
It's a measure of how well a simple linear model accounts for observed variation.
when does it make sense to choose a linear function to model a set of data
Calculus
A model in which your mother.
Yes.
Depends on your definition of "linear" For someone taking basic math - algebra, trigonometry, etc - yes. Linear means "on the same line." For a statistician/econometrician? No. "Linear" has nothing to do with lines. A "linear" model means that the terms of the model are additive. The "general linear model" has a probability density as a solution set, not a line...
When you use linear regression to model the data, there will typically be some amount of error between the predicted value as calculated from your model, and each data point. These differences are called "residuals". If those residuals appear to be essentially random noise (i.e. they resemble a normal (a.k.a. "Gaussian") distribution), then that offers support that your linear model is a good one for the data. However, if your errors are not normally distributed, then they are likely correlated in some way which indicates that your model is not adequately taking into consideration some factor in your data. It could mean that your data is non-linear and that linear regression is not the appropriate modeling technique.