Question 140773
The equation y=mx+b tells you a significant amount of information: 1) the slope, m, tells you how fast y is changing per unit increase in x; 2) b tells you, absent of any "x" (that is to say, 0 "x"), how much "y" would be present--- namely, b is the y-intercept. With regression analysis, we seek to "explain" some quantitative phenomena. The equation you supplied is a linear equation. Continuing, the equation would seek to "explain" via a linear relationship. Simply, linear regressions attempt to assign a line to data that are often not in an exact line graphically. By doing so, and having established statistical significance, we can analyze the strength of a relationship and determine whether or not, or to what degree, x affects y. 

Let me give you an example. Let's suppose I wanted to see if watching morning television has an affect on what time of day someone places his vote for President of the U.S. What I would do is collect data regarding how much morning television (in exact times) each person watched over a certain period preceding an election. I would then find out afterward at what time each person voted. I would graph these data and see how well a line could describe the relationship. Perhaps, the more television watched (x being the amount of television watched), the more likely someone is to vote later at night (y being time of day). This is just a simple example, but I hope it allows you to better see why anyone would want to do this. By the way, regression analysis is done in many sciences and disciplines. 

If you need anymore help, or clarification, please E-mail me by going to my "me" page.