# Thread: Deciphering linear, nonlinear and random relationships from scatter plots

1. ## Deciphering linear, nonlinear and random relationships from scatter plots

Hello All!!

This may sound not so bright and I hope none are too badly offended, but I can't for the life of me figure out if my hunches about the linear relationship of two variables is correct when looking at their scatter plot.

I'm working on some homework and I have to generate scatter plots for two variable data set and determine if the scatter plots suggest linear, nonlinear or random relationships.

Trouble is I am not clear on what exactly is random and how it differs from just plain old nonlinear.

A little help---Please.

2. An example of a non-linear relationship would be a quadratic function. The idea of the scatter plot is to see if it suggests profitable avenues of inquiry. It is more qualitative than quantitative.

3. Yea, and that helps me how?

Rather or not the scatter plot is qualitative or quantitative is of litte significance. I need to know and have a way to be fairly certain of what the scatter plot suggest about the data, either it is linear, nonlinear or random. I think that is simple. I am not trying to develope an equation, if I was I would just use a regression tool to determin the best variables to include in a simple or multi-linear regression equation to explain the varibility in the dependent variable. However all I need is an clear example of what nonlinear and random data sets graphed in a scatter plot would look like.

Thank you.

4. Well ...

Does it look like a n or a u or a s-curve ? or a sine wave

does it look like an "organized" line of some sort that isn't just a plain ol' diagonal?

(please don't shoot me like the last poster if my response isn't exactly what you need - I can't afford to lose this job - the benefits are amazing!)

5. Originally Posted by soon-to-grad
Yea, and that helps me how?

Rather or not the scatter plot is qualitative or quantitative is of litte significance. I need to know and have a way to be fairly certain of what the scatter plot suggest about the data, either it is linear, nonlinear or random. I think that is simple. I am not trying to develope an equation, if I was I would just use a regression tool to determin the best variables to include in a simple or multi-linear regression equation to explain the varibility in the dependent variable. However all I need is an clear example of what nonlinear and random data sets graphed in a scatter plot would look like.

Thank you.
If it looks like a parabola (an n or a u) then you don't have to bother trying a linear model.

If the points are all over a rectangular area then it is a safe guess that the data is random. Have you ever seen a function that is everywhere continuous, but is nowhere differentiable? It is somertimes called a Wiener process.

The scatter plot is a pattern recognition thing. Either you see it or you don't. If you don't see it, then I guess it doesn't help you even one little bit.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts