Linear Regression

      Day                        date       Item

• 10           T              9/27       Project plan due (grp)
• 15           Th           10/20     Proj  rep draft  due (ind)
• 20           T              11/8       Project report due (ind)
• 25           T              11/29     Pres draft (grp)
• 27           T              12/6       Project Presentations
• 28           Th           12/8       Project Presentations

• Instructions
• sample surveys
• Examples
• Project-Plan
• Sample from spring 12: (has one major flaw, excel spreadsheet does not contain all the raw data from the survey)sample_proj_2.docx, sample_proj_2.xls,sample_proj_2_survey

Linear regression involves 2 variables that are linked. The variables can not be categorical (such as a person’s favorite color), but instead must be numeric and the numbers should be meaningful. There must be at least 10  possible values for each variable so size of a household might not be distributed widely enough (most values would be between 1 and 7). A good example would be arms_width_vs_height. The example has units in cm. Each variable has a sufficient variety of values. The same example done with inches might not have enough variety of values.  To do psychology studies, you may combine the outcomes from a variety of questions to get the requisite variety of values, e.g.,  the OCD survey in the “sample surveys”.