I got these searching "Linear Regression Online Calculators."  There are probably many more, if you find one you like, email me the link, please!

 

1) Alcula - Pros:  Easy to read, provides graph, sample size (n), x-bar, y-bar, a, b and regression equation in proper y = a + bx form, provides list of data entered, can calculate predicted values (y-hat) for a given x, 14 non-zero digits of accuracy.  Cons: Does calculate Correlation coefficient (r) but have to go to 2nd page, doesn't provide r-squared (can get from r), seems kinda slow to process.

Link:  http://www.alcula.com/calculators/statistics/linear-regression/

 

2) MathPortal.Org - Pros:  Easy to read, provides graph, lists data input, proper y = a + bx format, provides explanation of work.  Cons: Have to use 2nd page for Correlation coefficient (r), doesn't provide r-squared (can get from r), doesn't list a and b separately, only gives values to 1000ths (won't work well with real small numbers), no predicted values.

Link: http://www.mathportal.org/calculators/statistics-calculator/correlation-and-regression-calculator.php

 

3) GraphPad - Pros: Easy to use, enter data in lists like calculator, , provides Coefficient of Determination (r-squared),   provides info for later sections (95% C.I. for slope, x and y intercepts, P-values)   Cons:  Uses y = bx + a format (a and b not given, just in equation), doesn't provide data list (useful to check accuracy of data input), worked great once, can get to work again.

Link: http://www.graphpad.com/quickcalcs/linear1/

 

4) Correlation and Regression applet - Pros: from publisher of book, quick visual exploration of regression - does not use numbers, can easily explore data such as outliers, changing one observation or adjusting clustering.  Good for testing conjectures about data.  Cons:  No input of numbers, not for use in getting a regression equation for specific data.

Link: http://bcs.whfreeman.com/tps4e/#628644__925297__