The Coalition for the Protection of Marriage
The Coalition for the Protection of Marriage (“Coalition”) is sponsoring this webpage for the purpose of facilitating the sharing of data, analysis, and discussion on an important statistical report (“Report”). The Report was filed in the Ninth Circuit Court of Appeals on October 13, 2014, in Nevada’s same-sex marriage case, Sevcik v. Sandoval, Case No. 12-17668, and has rightly attracted serious attention.
THE COALITION AND THE NEVADA SAME-SEX MARRIAGE CASE
As a party in the Nevada same-sex marriage case since it started, the Coalition is defending Nevada’s marriage laws. On October 7, 2014, a three-judge panel of the Ninth Circuit Court of Appeals declared those laws unconstitutional. That decision can be found at this link. Less than a week later, the Coalition asked the full Ninth Circuit to rehear the case before a randomly selected panel of eleven judges (“Petition”). The Petition can be found at this link. In doing so, the Coalition demonstrated the strong appearance that the Ninth Circuit did not follow its neutral, impartial assignment process in assigning this particular three-judge panel to the Nevada case and its companion Idaho case. The Report gave strong statistical confirmation to that disturbing appearance. The Report and related documents can be found at this link.
DATA SHARING POLICY
We feel that open discussion of calendaring practices within the federal courts is beneficial to citizens, attorneys, judges, and the entire judicial process. To that end, we want to provide the calendaring data collected from the public website of the Ninth Circuit of Appeals and used as the basis of the Report.
We recognize that replication and validation are important parts of data analysis, and we hope that providing these data will lead to open, civil dialogue. Undoubtedly, different analysts will approach in different ways the calendaring issue addressed by the Petition and the attached Report. This is beneficial and will undoubtedly lead to additional questions, discussions, and research. The end result may be increased transparency and confidence in the judicial system.
In order to avoid the pitfalls of “data snooping”—the process of analyzing data until a desired outcome is achieved—we will provide the full data to any individual who provides to us his or her real name and genuine contact information, along with any professional affiliations, as well as the code of his or her analysis. We also ask each of those persons to provide to us his or her results for posting on this website.
The first step is to send to firstname.lastname@example.org a message containing your real name, genuine contact information (including e-mail address and telephone number), professional affiliations, and your code. Here are guidelines for your code:
1. Write your code so that the analysis can be performed by replacing the sample dataset with the full dataset. The sample dataset can be found at this link; the data dictionary, at this link; information about the Relevant Cases, at this link.
2. The output should be an estimated probability or related measure.
3. Annotate each step with comments so that the logic of the analysis is clear.
4. R code is preferred. SAS or Matlab code are acceptable. Other statistical languages will be considered as needed. Only scripted languages (not Excel) will be accepted.
ADDITIONAL INFORMATION ABOUT THE REPORT
An important document entitled “Addressing Misunderstandings About The Report” can be found at this link.
Another document entitled “A Summary of Calendaring Practices in the Ninth Circuit” can be found at this link.