Changes

'''Robert M. Groves''' is a liberal professor from the University of Michigan and a survey expert who has authored several books on the topic.  President Obama has chosen him to direct [[Census 2010]] as Secretary of Commerce.

'''Robert M. Groves''' is Research Professor at the University of Michigan and a survey expert who has authored many technical articles and books on the topic.  President Obama has nominated him to head the U.S. Census Bureau, with responsibility for [[Census 2010]].

Based on court precedents and the [[Census Act]], the Secretary of Commerce has power to conduct the census in whatever way he deems will be most accurate, including the use of adjustment methods.  The Census Act states, "Except for the determination of population for purposes of apportionment of Representatives in Congress among the several States, the Secretary shall, if he considers it feasible, authorize the use of the statistical method known as 'sampling' in carrying out the provisions of this title.""

Based on court precedents and the [[Census Act]], the Secretary of Commerce has power to conduct the census in whatever way he deems will be most accurate, including the use of adjustment methods.  The Census Act states, "Except for the determination of population for purposes of apportionment of Representatives in Congress among the several States, the Secretary shall, if he considers it feasible, authorize the use of the statistical method known as 'sampling' in carrying out the provisions of this title.""

==Attempted Survey Readjustment in 1990==

==Attempted Survey Readjustment in 1990==

"When he was the bureau's associate director, Groves recommended that the 1990 census be statistically adjusted to make up for an undercount of roughly 5 million people, many of them minorities in dense urban areas who tend to vote for Democrats. " <ref>http://www.breitbart.com/article.php?id=D97AG7201&show_article=1&catnum=0</ref>  Thankfully, this "political tampering" was overruled by Republican Commerce Secretary Robert Mosbacher.   

In 1990-92 Groves associate director of statistical design for the Census Bureau.   

"When Groves was associate director at the Census Bureau he recommended statistical sampling to make up for the projected “undercount” of 5 million people in 2000. These essentially made-up people – intended to represent the undercounted population – are assigned a gender, race, income, and generalized region based on the Census Bureau’s formula." <ref>http://cnsnews.com/public/content/article.aspx?RsrcID=48234</ref>

He recommended that the 1990 census be statistically adjusted to make up for an undercount of roughly 5 million people, who otherwise would be lieft out. <ref>http://www.breitbart.com/article.php?id=D97AG7201&show_article=1&catnum=0</ref>  He was overruled by Republican Commerce Secretary Robert Mosbacher. <ref>http://cnsnews.com/public/content/article.aspx?RsrcID=48234</ref>

When sampling is used to correct an undercount, the Census Bureau must create a formula to determine the characteristics of the "made-up" individuals.  "How you structure the formula affects the results. It's possible a party in power might determine the desired result first, then determine the sampling formula needed to achieve it." <ref>http://www.blnz.com/news/2009/04/13/Counting_Illegal_Aliens_6141.html</ref>

When sampling is used to correct an undercount, the Census Bureau must create a formula to determine the characteristics of the "made-up" individuals.  <ref>http://www.blnz.com/news/2009/04/13/Counting_Illegal_Aliens_6141.html</ref>

Since Groves had last served in the census bureau, in 1990 ''Doc v. United States House of Representatives'', 525 U.S. 316 (1999) ruled that "statistical sampling cannot be used to apportion congressional seats, but can be used for other purposes, such as legislative redistricting and, of course, doling out dollars." <ref>http://www.time.com/time/nation/article/0,8599,1889793,00.html</ref>

In ''Doc v. United States House of Representatives'', 525 U.S. 316 (1999) the Supreme Court ruled that "statistical sampling cannot be used to apportion congressional seats, but can be used for other purposes.<ref>http://www.time.com/time/nation/article/0,8599,1889793,00.html</ref>

==Books and Articles==

==Books and Articles==

A complete list can be found here: http://www.psc.isr.umich.edu/people/cv/groves_robert_cv.pdf

A complete list can be found here: http://www.psc.isr.umich.edu/people/cv/groves_robert_cv.pdf

Groves' books recent books and articles focus heavily on the psychology and methodology of telephone surveys.  Many of these discuss methods for dealing with survey non response, and Groves suggests various adjustments to make up for nonresponse.

Groves' books recent books and articles focus heavily on the psychology and methodology of telephone surveys.  Many of these discuss methods for dealing with survey non-response, and Groves suggests various adjustments to make up for nonresponse.

<blockquote>

<blockquote>

"The expressions above concern sample values and estimates, unadjusted in

"The expressions above concern sample values and estimates, unadjusted in any way to compensate for nonresponse. It is common, however, to use

any way to compensate for nonresponse. It is common, however, to use

weighting class adjustments (Bethlehem 2002), raking (Deville, Särndal, and

weighting class adjustments (Bethlehem 2002), raking (Deville, Särndal, and

Sautory 1993), calibration methods (Deville and Särndal 1992; Lundström

Sautory 1993), calibration methods (Deville and Särndal 1992; Lundström

and Särndal 1999), or propensity models (Ekholm and Laaksonen 1991) to

and Särndal 1999), or propensity models (Ekholm and Laaksonen 1991) to reduce the biasing effects of response propensities correlated with the survey variables."<ref>http://poq.oxfordjournals.org/cgi/reprint/70/5/646.pdf</ref>

reduce the biasing effects of response propensities correlated with the survey

variables."<ref>http://poq.oxfordjournals.org/cgi/reprint/70/5/646.pdf</ref>

</blockquote>

</blockquote>

Nevertheless, he says, "Since probability samples have the advantage of eliminating bias at the selection step, it is useful to consider assembling auxiliary variables to compensate for nonresponse bias through the use of strong postsurvey adjustment models."<ref>http://poq.oxfordjournals.org/cgi/reprint/70/5/646.pdf</ref>

Nevertheless, he says, "Since probability samples have the advantage of eliminating bias at the selection step, it is useful to consider assembling auxiliary variables to compensate for nonresponse bias through the use of strong postsurvey adjustment models."<ref>http://poq.oxfordjournals.org/cgi/reprint/70/5/646.pdf</ref>

===Books===

===Books===