A Political Survey
Part 1 -
Comments on Methodology

Today I received in the mail a survey from the US Representative for the Michigan 7th District, Mr. Nick Smith.  Perhaps a little explanation is in order.  Corpus Callosum is listed as an Ann Arbor blog, but Nick Smith does not represent Ann Arbor.  I actually live in Lenawee County.  But I was born in the U Mich Hospital, lived in Ann Arbor until age 5, then again from ages 18 to 37, and I still work there.  My philosophy and polictitical tendencies and hairstyle are more congruent with Ann Arbor than Lenawee County.   Hence the Ann Arbor affiliation.

Anyway, I got this survey from Representative Smith.  Please read the survey before reading the rest of this article; otherwise, you won't know what I am writing about.  I expect that all of my comments will take two or three articles to get through, so bear with me please. 

The survey was mailed to my house.  There also is one on the website.  My first comment is that this survey, in the format as mailed, is different than the one on the website.  The mailed version inlcudes a cover letter.  In the cover letter, Mr. Smith expounds upon his beliefs regarding some of the topics that are in the questionnare.  This, of course, could bias the responses.  The cover letter is not seen on the website version.  The mailed version includes a picture of Mr. Smith shaking hands with Colin Powell.  Again, this is not seen on the webstie.  Another difference between the two versions is that item 8 on the mailed version asks respondents to rank five choices in order of priority.  The web version askes only for the most important priority.  Also, there is nothing to keep anyone who wants to from filling out the online survey.  So my results could end up aggregated with results from citizens of Mozambique, Uzbekistan, and Botswana.  I could fill out the form multiple times -- unless there is an IP address checking function, which I doubt  -- and stack the results any way I want.  Yet another methodological problem is that the online version permits the respondent to view the results that have been collected so far.  If the respondent loks at the resullts before taking the survey, it could influence the responses.  My final objection to the methodology is that the wording of the possible responses is, in some cases, going to bias the results.  For example:

8. In the Middle East, what should be the most important priority?
Establishing a democratic Iraq.
Reducing U.S. responsibility and control in Iraq by securing greater United Nations and international cooperation.
Withdrawal from Iraq as soon as possible.
Continue aggressive U.S. effort to hunt down terrorists.
Securing peace between Israel and Palestine to reduce Islamic unrest.

Choices 1,3, and 4 each consists of a simple description of an action.  Choice 2 describes an objective, then specifies the means to that objective.  Choice 5 describes an objective and gives a reason for pursuit of that objective.  This inconguent wording can affect the choices.  I once read about a little experiment that demonstrated this point.  The experiment involved asking people for a favor using one method of phrasing in some cases; another in other cases.  The experimenter had people go up to someone who was using a copy machine and ask "May I use the Xerox?"  On other occasions, the question was "May I use the Xerox to make copies?"  These two statements are functionally equivalent, since the only thing you can do with a copy machine is to make copies.  Yet, the people who asked "May I use the Xerox to make copies?" were more likely to get a favorable response.  It is hard to know why this was the case, but it seems likely that the addition of the explanation makes the request seem more reasonable. 

These factors would seem to indicate that the survey is not scientifically valid.  I am sure that Mr. Smith knows this.  Whether the lack of validity is important depends upon what he intends to do with the results.  I intend to ask him, eventually.  First I am going to blog about my impression of the survey, make some comments in summary, then put the articles together as a letter to Mr. Smith. 

He said he wanted my input.

---

A Restaurant -
Comments on Food

There are no pictures of me with President Bush, and I haven't been to Antartica.  However, this afternoon, I had lunch with the Senior Editor of Die Niemandsblog.  We went to:

Lotus Thai
2803 Oak Valley (Oak Valley Centre) map
668-2828

The link is from the Ann Arbor Observer.  The review on the Observer website is not a review of Thai Garden.  It is a review of Beijing Restuarant, which is the establishment that occupied the same building until recently.  The Ann Arbor News published a review on January 22, 2004, by Ana Wagner.  Her review is valuable for people who have a sense of taste.  The review that you are reading now is intended for the rest of us.  You see, I am homozygous for the gene that makes people like any food that someone else prepares.  I do not care if "Everything is strikingly presented, from the artfully arranged wedges of fried tofu to the diminutive dish of tamarind sauce decoratively sprinkled with crushed peanuts."  What I care about is: 1) Are the people nice? 2) Is the place clean? 3) Is it quiet? and 4) Am I still hungry when I leave?  Yes, the people are nice.  Yes, it is clean.  Yes, it is quiet.  And, no, they gave me enough food.  This is in contast to some Ann Arbor restuarants (eg. Amadeus, Old Siam) that serve very good food, but not enough of it.  Other amenities: plenty of parking.  Close to Media Play, Office Max, and Target.  The interior is very nicely decorated, too; much more tasteful than one would expect in a strip mall, and comparable to most of the places on Main Street.  Contrary to the findings of Ms. Wagner, I felt that the mount of spiciness in the food was just right.  Remeber, though, that this was food prepared by someone else.