4 replies to this topic

[Qaz]

can anybody be so kind as too help me with some of these questions?

1. explain how to decide which line best approximates the scattered plot. describe the position of data points with respect to your linear model.
-i seriously don't have much of an idea. this isn't my regular hw.

2. can all plots be modeled by a line? is a linear model suitable or fits all scattered plots?
-no, right? because some plots can have no relation. right?
3. how do you know which linear model give the best predictions?

if someone can reply soon that would be great.

thanks

[Hater_Boxden]

Some of these questions sound like they come with pictures.....if so you should post them if you want help.
For number 2 yea if i plot something that has absolutely no relation to each other then the dots will be all over the place. And no linear models cannot determine anything, there are tons of other relations such as inverse, exponential, quadratic, and other stat plots.

[Melchoire]

can anybody be so kind as too help me with some of these questions?

1. explain how to decide which line best approximates the scattered plot. describe the position of data points with respect to your linear model.
-i seriously don't have much of an idea. this isn't my regular hw.

2. can all plots be modeled by a line? is a linear model suitable or fits all scattered plots?
-no, right? because some plots can have no relation. right?
3. how do you know which linear model give the best predictions?

if someone can reply soon that would be great.

thanks

1. Draw the line that comes closest to all the dots on the graph. It doesn't necessarily have to touch any dots.
2. No if the they're too scattered the linear model won't be useful and wouldn't give an approximation.
3. Whichever model best fits the plot. This could be linear, quadratic, cubic, quintic....

Edited by FlashGM, 08 March 2008 - 10:15 PM.

[Qaz]

thanks for the help everyone

[Silk]

For #1, to get a more accurate line, take the average of each set of data and plot them on the scatter graph. Use that point as a fulcrum for drawing the line.

For #3, you can also get the correlation coefficient (a number between -1 and 1). -1 is a perfect negative correlation, +1 a perfect one and 0 means there isn't any relation.

Dunno what grade you're in, but you may as well use the more complicated answer and make yourself seem smarter.