Here's the massaged data from the class survey. - Jim Mahoney, Wed night, Sep 17 2003. The files and directories are unix/ text with line-ends in Unix/MacOSX format. survey-raw.txt The various data emailed to me by students. survey.csv All the data in one common format. windows/ text with line-ends in windows format. survey-raw.txt same as above survey.csv excel-survey.xls save file from Microsoft Excel::mac v10.0.0 excel-survey.html HTML save from Excel. The differences between the unix and windows versions aren't all that significant and may be handled by your software without you noticing - I just tried to cover all the bases. To do the conversion, I used both Excel and Mathematica (particularly the "Import[]" and "Transpose[]" commands) and a lot of editing by hand to convert the raw.txt to the .csv file. And since things weren't all that consistent, I had to make some arbitrary choices, including * everthing in lowercase * "yes" or "no" answers rather than "Y", "N" or other variations * gender given as "m" or "f" * handedness given as "right", "left", or "ambi" * multiple plan areas with slashes, e.g. sociology/religion * year at marlboro is given as a number, 1-4 means freshman-senior * I added a "pollster" field so that we could tell who collected which data. * missing entries is given by a single dash, e.g. "-" rather than "?" or "N/A". The questions we asked and what I'm taking as the most consistent version of an answer looks like this: * gender m, f * age [yr] number * height [in] number * father's height number * mother's height number * 5th ssn digit digit * 9th ssn digit digit * hair color text (eg brown) * eye color text (eg blue) * handedness right, left, ambi * smoker yes, no * plan area this/that/some other * year at marlboro 1, 2, 3, or 4 * number of siblings number * god yes, no * vegetarian yes, no * favorite ice cream flavor text * attended first Town Meeting yes, no * been in a fight yes, no * drink yes, no * use pot yes, no * pollster statistics class member So that's what we have so far.