IvoDinov: New page: == SOCR CensusAtSchool International Workshop - Analyze bivariate data using linear regression methods== ===Fit regression lines to pairs of numeric variables a...

2008-07-24T23:14:42Z

New page: == SOCR CensusAtSchool International Workshop - Analyze bivariate data using linear regression methods== ===Fit regression lines to pairs of numeric variables a...

New page

== [[SOCR_Events_July2008 | SOCR CensusAtSchool International Workshop]] - Analyze bivariate data using linear regression methods==

===Fit regression lines to pairs of numeric variables and calculate the means and standard deviations of the two variables and the correlation coefficient, using technology===

* Copy the first 200 measurements of the [[SOCR_Data_Dinov_020108_HeightsWeights | Human Height and Weight Dataset]] into the [http://socr.ucla.edu/htmls/SOCR_Analyses.html SOCR Analysis (Simple Linear Regression)].

<center>[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig7.jpg|500px]]</center>

* Paste these in SOCR and Map the Heights and Weights columns to the X and Y variables.

<center>[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig2.jpg|500px]]
[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig3.jpg|500px]]</center>

* Click CALCULATE and see the output numerical results (in the Results tab) and the graphical outputs (in the Graphs tab).

<center>[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig4.jpg|500px]]
[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig5.jpg|500px]]</center>

===Compute predictions of y-values for given x-values using a regression equation, and recognize the limitations of such predictions===

The regression line we computed and illustrated on the [http://wiki.stat.ucla.edu/socr/uploads/d/d1/SOCR_Events_May2008_C9_S1_Dinov_020808_Fig4.jpg image above] shows the following linear relation between Heights (inches) and Weights (pounds):
<center> <math>Height = 56.457 + 0.09033384003691448\times Weight</math></center>

What would be the regression of the Weight on Height (the reverse regression)? You can compute this by hand to be <center> <math>Weight = -105.959 + 3.431662594473845\times Height.</math></center>

Or you can simply re-maps (swap) the dependent and independent variables in [http://socr.ucla.edu/htmls/SOCR_Analyses.html SOCR Analysis (Simple Linear Regression)] and recalculate the linear relation.

<center>[[Image:SOCR_Events_May2008_C9_S1_Dinov_020808_Fig6.jpg|500px]]</center>

==References==
* [[EBook | Interactive Statistics Education EBook]]
* SOCR Home page: http://www.socr.ucla.edu

{{translate|pageName=http://wiki.stat.ucla.edu/socr/index.php?title=SOCR_Events_July2008_C9_S1}}

SOCR Events July2008 C9 S1 - Revision history

IvoDinov: New page: == SOCR CensusAtSchool International Workshop - Analyze bivariate data using linear regression methods== ===Fit regression lines to pairs of numeric variables a...