TO: Board of Elections and Ethics
FROM: Alice P. Miller
Executive Director
SUBJECT: Petition Verification Results for Initiative Measure #59: "Legalization of Marijuana for Medical Treatment Initiative of 1998"; Recommendation to Reject the Petition as Numerically Insufficient.
After completing the Board's standard procedure for verifying registered voter signatures contained in Initiative Measure #59, Legalization of Marijuana for Medical Treatment Initiative of 1998, the Board's staff, in conjunction with the Office of Planning, Data Management Division, has determined that this initiative does not meet the statutory requirements for certification to the ballot, in accordance with D.C. Code Sec. 1-282(a).
It is therefore my recommendation that the Board reject the Initiative Measure No. 59 petition as insufficient, in accordance with D.C. Code 51-1320(o)(1), because it does not contain a sufficient number of valid signatures of duly qualified electors of the District of Columbia.
An overview of the petition verification process and a summary of its results are entailed in this report. Additional detail on the verification procedures, the mathematical formulas used in calculating statistical confidence levels, and related documentation are contained in several attachments to this report.
The petition verification process is conducted to determine whether an initiative petition contains the required number of registered voter signatures needed to qualify for placement on the ballot. The required minimum number is 5% of the registered voters citywide and 5% of those registered in at least 5 of the 8 election wards, based on the published registration totals in effect 30 days before the petition was filed. The signature requirement for Initiative Measure No. 59 was based on registration totals published as of May 31, 1998. The numerical requirement for Initiative Measure No. 59 is shown in Table A below:
Page 2
Initiative Measure No. 59
August 3, 1998
|
|
|
1 | 41,264 | 2,063 |
2 | 41,344 | 2,067 |
3 | 45,732 | 2,287 |
4 | 47,106 | 2,355 |
5 | 45,786 | 2,289 |
6 | 43,945 | 2,197 |
7 | 43,110 | 2,156 |
8 | 31,664 | 1,583 |
Citywide | 339,951 | 16,997 |
The petition verification process that the Board follows
-- detailed in the Standard Procedures for Verification of Initiative
Petition, attached-has four basic steps:
. First, the Board's staff uses the voter registration computer
system to verify that the circulator of each petition page was
registered to vote in the District from the address listed on
the Board's records, during the time of circulation, as required
by law. Pages not circulated by duly registered voters are rejected
at the outset.
. Second, the name of each petition signer is checked against
the voter registration computer file to determine if the petitioner
was a duly registered voter in the District at the time the petition
was signed, as required by law. Only those petition signers whose
names and addresses are found to match the Board's registered
voter file are entered into the petition-checking program as "verified
registrants".
. Third, the totals of verified registrants are compiled, citywide
and by ward, to determine whether the petition contains enough
apparent registered voter signatures to proceed to the signature
verification stage. The total number of verified registrants is
provided to the of fice of Planning. Data Management Division.
Subsequently, the Of fice of Planning numerically provides the
Board with the names to be used in the random sample process.
Page 3
Initiative Measure No. 59
August 5, 1998
* Fourth, random samples of 100 signatures from each ward are drawn, for comparison to the original voter signatures on file in the Board's records. To be accepted as valid, each sampled signature must reasonably match the Board's records, and the signer must be registered at the address on the petition as of the date signed. The validity rate of the random samples is then used to determine the sufficiency of the petition as a whole, at confidence level of 95%.
The Initiative Measure No. 59 petition, containing 1,816 pages, was filed on July 6, 1998. After completing the procedure of verifying the voter registration status of petition circulators and signers - steps 1, 2, and 3 in the process outlined above - the total number of registered voters listed in the petition was found to be 17,092. This total exceeded, by 95, the required 5% minimum of 16,997 registrants citywide. In Addition, the minimum was attained in Wards 1, 2, 3, 6, and 8 - as shown in Table B on the following page. While Wards 4, 5, and 7 - as also indicated in Table B - fell below the required minimum, sufficient names of registered voters had been identified in the petition, at both citywide and ward levels, to proceed to the signature random sampling.
The random sample procedure is conducted to determine, with the required confidence level of 95%, whether or not a sufficient number of valid signatures of registered voters (as distinct from registrant names and addresses) are contained in the petition. To facilitate the sampling procedure, the Board's staff produced computer listings of the registrants identified in each ward, from which the Office of Planning Data Management Division randomly selected eight ward samples of 100 petition signatures each. To do this, the registrants contained in each ward listing were serially numbered by the Board, so that a set of 100 random numbers, generated by the Data Management Division for each ward, would identify the ward's 100 selected registrants. The listings also contained the petition page number on which the registrant's signature to be validated would be found. The staff then pulled each selected registrant's voter registration application and compared the randomly selected petition signature with the registrant's original signature on file. The total number of validated signatures in each ward sample of 100 was then reported to the Data Management Division.
The effect of the random sampling process was thus to establish a validity rate for the registered voter signatures in each ward. The Data Management Division applied this validity rate mathematically to the number of ward registrants identified in the petition and the ward's numerical signature requirement, to determine with 95% confidence whether or not the ward should be accepted as having a sufficient number of valid signatures. Sufficiency for the District as a whole was then assessed using a stratified random sampling formula, which incorporated the results of the various ward samples and applied them to the citywide criteria.
Page 4
Initiative Measure No. 59
August 5, 1998
These mathematical calculations were made by the Data Management Division -- as reported in the attached memorandum of Herbert Bixhorn, Chief, Office of Planning, Data Management Division and Mathematical Statistician -- and are summarized in the table below. As shown in Table B. Wards 1, 2, 3, 6 and Ward 8 were accepted as having a sufficient number of valid signatures based on the random sample. Wards 4, 5 and 7 were rejected as insufficient. However, at the citywide level, the District as a whole was rejected with 95% confidence as containing an insufficient number of valid signatures.
Ward |
|
|
|
|
1 | 2,063 | 2,501 | 97 of 100 | Accept |
2 | 2,067 | 2,254 | 99 of 100 | Accept |
3 | 2,287 | 2,604 | 100 of 100 | Accept |
4 | 2,355 | 1,572 | 99 of 100 | Reject |
5 | 2,289 | 1,971 | 98 of 100 | Reject |
6 | 2,197 | 3,091 | 98 of 100 | Accept |
7 | 2,156 | 1,199 | 100 of 100 | Reject |
8 | 1,583 | 1,900 | 98 of 100 | Accept |
Cityward | 16,997 | 17,092 | Reject |
While the petition thus met some of the ward level requirements, it clearly failed the random sample test at the citywide contained on the level -- primarily because the total number of registered voters petition was barely above the minimum requirement (16,997 vs.17,092). This meant that the random samples needed to be perfectly validated in order for sufficiency to be closer to being achieved. As the statistician noted, "because the number of verified registrants for the District was only slightly above the required number, the proportion of valid signatures needed for acceptance would have been over 99 percent." Although the random samples showed validity rates ranging above 95%, sufficiency of the petition still could not be achieved at the citywide level.
Based on the foregoing, the results of the petition verification process for Initiative Measure No. 59, "Legalization of Marijuana for Medical Treatment Initiative of 1998" are the determinations that the measure does not meet the statutory requirements for certification to the ballot and that it should be rejected by the Board because it does not contain a sufficient number of signatures of registered voters of the District.
Page 5
Initiative Measure No. 59
August 5, 1998
Attachments:
Summary Report for period ending May 31, 1998
Standard Procedures for Verification of Initiative Petitions
Memorandum to Herbert Bixhorn, Chief, Data Management Division
Offfice Planning; Requesting Random Sample Designation Re. Petition
#59,
Memorandum to Herbert Bixhorn, Chief, Data Management Division,
Office of Planning, Results of Random Sample Verification Re.
Petition #59
Memorandum from Herbert J. Bixhorn, Chief, Data Management Division
Statistical Summary and Findings on Initiative Measure #59 Petition
cc: Kenneth J. McGhie,
General Counsel
|
|
|
|
|
|
|
|
|
2,424 | 30,751 | 567 | 117 | 7,335 | 70 | 41,264 |
|
5,101 | 28,007 | 384 | 90 | 7,683 | 79 | 41,344 |
|
8,179 | 29,392 | 172 | 19 | 7,911 | 59 | 45,732 |
|
1,837 | 39,240 | 512 | 198 | 5,289 | 30 | 47,106 |
|
1,629 | 38,613 | 593 | 188 | 4,740 | 23 | 45,786 |
|
2,959 | 34,734 | 573 | 293 | 5,343 | 43 | 43,945 |
|
1,325 | 36,790 | 557 | 151 | 4,271 | 16 | 43,110 |
|
1,128 | 26,275 | 611 | 120 | 3,523 | 7 | 31,664 |
TOTALS | 24,582 | 263,802 | 3,969 | 1,176 | 46,095 | 327 | 339,951 |
When an Initiative Measure (or Measures) is (are) filed, the Registration Division staff will initially perforrn the following steps in the formal RECEIPT phase:
Step 1:
(a) Count the number of pages submitted by the proponent; and
(b) Issue a receipt for number of pages actually submitted.
Step 2:
Using an Automatic Numbering machine, serially number ALL petition pages.
Step 3:
After eliminating those lines which are completely blank, the Registrar of Voters determines if petition contains the required number of signatures to permit continuation of the acceptance and verification process. (See note below for handling of multiple initiative measures)
Step 4:
Registration Division staff will make photocopies of each petition page, front and back, before moving to the verification phase.
NOTE: In the event multiple Initiative Measures are filed and processed at the same time, the Registrar of Voters will ensure that great care is taken to keep the petition pages for these measures separated and will color-code, by highlighter, the Measure Number on the face of each petition page.
After the Registrar of Voters has determined that the petition, at face value, contains the required number of signatures to permit continuation of the acceptance and verification process, and the Registration Division staff has made the required copies of all petition pages, preparation of the petition for registration verification will proceed.
Each petition page will be reviewed, and using a highlighter (different colors when dealing with multiple Initiative Measures), the Registration Division staff will strike through any petition line that falls into any of the following categories:
(a) The line has been
obliterated by the proponent
(b) The line is blank
(c) The line has neither a
printed name nor a signature
(d) The line has no address
(e) The line has no date of
signature
When any of the above conditions exist, strike through the line with the highlighter
(f) ID Circulator's Affidavit: Yellow out any instance where the Circulator's signature. address. or dates of circulation is missing. Set any pages ID this category aside
After the staff has completed the Verification Preparation phase and identified all those petition lines which are to be excluded from the Registration Verification process, all petition pages will be passed to a central control point, normally the Registration Processing Manager, for handling through the next phase.
Page 2
Standard Procedures for Verification of Initiative Petitions
Section 3- Registration Verification
The Registration Processing Manager will ensure the equitable assignment of petition pages for verification; maintaining an accurate record of the petition pages assigned to each staff member.
Upon initial receipt of assigned petition pages, each staff member will first be performing registration verification on the circulator of the petition as follows:
Step 1: (a) See if the circulator of the
petition is a registered voter
at the address listed.
(b) If "YES", write the voter
registration number and date of
registration (in RED INK) next
to the circulator's name on the petition.
(c) If "NO", write "NR"
next to
the signature of the petition circulator.
(d) If the circulator has a transfer date
on the computer that is later than the first date of the circulation
period stated in the Circulator's Affidavit, put these pages aside,
in their own pile. Give this pile to the Registration Processing
Manager.
(e) Repeat Step 1 for each assigned page until it has been determined
how many pages are eligible for further review. In other words,
how many pages were circulated by a properly registered circulator.
Set aside those pages with "NR" because they are automatically
excluded from any further review.
Step 2:
(a) On each of the pages retained for further processing, look at each "DATE SIGNED" on the petition page and compare it with the "DATE OF REGISTRATION" of the circulator.
(b) If the DATE SIGNED" by the voter is prior to the "DATE OF REGISTRATION" of the circulator, put an "XC" on that line. This makes the line invalid
Page 3
Standard Procedures for Verification of
Initiative Petitions
Section 3 - Registration Verification - Circulator (Cont.) Step
3 - For Registration Processing Manager Only:
In cases where there is a transfer date on the after the initial
circulation date, Registration Processing Manager will determine
whether the circulator was properly registered at the address
listed an the dates circulated and signed.
(a) If "YES", Registration Processing Manager writes
the voter registration number and date of registration (in RED
INK) next to the circulator's name on the petition. The Manager
then reviews the signed lines of the petition, comparing the dates
of the signatures against the Transfer Date for the circulator
and placing an 'XC" on those lines which are invalid.
(b) If "NO" Registration Processing Manager writes "NA"
next to the signature of the petition circulator and puts these
pages in the same pile as the "NR" pages.
After each group of assigned pages has bees completely through
the above preliminary process, they should be returned to the
central control point, normally the Registration Processing Manager,
for later distribution to the staff who will complete the verification
process on the signers of the petition.
Section 4- Registration Verification - Signers
NOTE: As this next phase of the verification process continues,
remember that with the advent of NVRA, the status of a voter record
could be any one of the following: ACTIVE-INACTIVE -- DELETED.
After Registration Verification of the petition circulators has
been completed, and all petition pages have been resumed to the
Registration Processing Manager, those pages with the "NR"
and "NA" indicating an ineligible circulator will be
removed and the remainder distributed to the staff for the next
phase-registration verification of the signers of the petition.
The Registration Processing Manager will again ensure the equitable
distribution of petition pages to the staff involved, and will
again maintain a record of the petition pages assigned to each
staff member. To begin, the processing staff should display the
"Main Menu" on the Voter Registration System Terminal.
From that queen:
Choose "13" to begin entry into the petition checking
module;
- Choose "2" to enter the "Signatures to be Defined"
segment.
In response to the system request, complete the entries for the
following lines:
ELECTION CODE:
CONTEST NUMBER: .**
CANDIDATE NUMBER: 1
SECTION NUMBER: ****
DATE FILED: **/**/**
NON-REG. VOTERS: 0
Page 4
Standard Procedures for Verification of Initiative Petitions Section
4- Registration Verification - Signers (Con't)
The Data Systems Manager will already have provided each member
of the processing staff with the pre-determined entries for the
"ELECTION CODE" (*) and "CONTEST NUMBER" (***).
Because this petition checking procedure is for Initiative Measures,
"CANDIDATE NUMBER" will always be "1".
The "SECTION NUMBER" (****) will be the 4-digit page
number assigned by the Board when the petition was received (See
Section 1, Step 2)
"DATE FILED" (**/**/**) will be the date the Initiative
Measure was filed, and "NON-REG. VOTERS" will always
be ZERO.
When the above entries are complete, press "A" to accept
and proceed to the next step, press "C" to cancel, or
press "M" to modify (correct or change a field).
After the initial entry of data in the above fields, only the
"SECTION NUMBER" (Page number) needs to be entered.
Having completed the initial entry into the petition checking
module and recorded the various codes, proceed as follows:
1. Select "1" to obtain an Alphabetic look-up screen;
2. Select "3" to obtain a screen which will include
the "Last Transfer Date" along with other key elements
of each voter's record;
3. Press "RETURN" to start the Alpha Search process.
Once into the Alpha Search and verifying the signers of the petition,
great care should be taken in matching signers with registered
voters being displayed on the terminal screen.
Although the address listed on the petition MUST match the address
on the Board's records at the time of signing, registrants will
sometimes sign their names slightly different, such as -no middle
initial. Reasonable judgement should be used to determine whether
a petitioner is a registered voter. When in doubt, call the supervisor.
Of equal importance is the need to remember to CHANGE THE "SECTION
NUMBER" (PAGE NUMBER) IN THE PETITION CHECKING MODULE EACH
TIME YOU FINISH ONE PETITION SHEET AND GO ON TO THE NEXT.
Page 5
Standard Procedures for Verification of Initiative Petitions
Section 4- Registration Verification - Signers (son 't)
The alpha search proceeds as follows:
1. Type in the name as it appears on the petition and hit "Return".
2. If a match of the voter's name and address on the petition
is found-and there is no transfer date or there is a transfer
date which is prior to the date the petition was signed-enter
the screen line number to call the voter's record onto the screen.
3. Confirm the name and address and make sure that the voter's
registration status is ACTIVE - not inactive or deleted.
4. Enter "A" to accept the entry. Put a check mark next
to the accepted name on the petition page before you go to the
next name on the sheet.
5. If no match is confirmed for name and address, even after alternative
spellings have been checked, OR, status is inactive or deleted,
put an "X" by that line number. Enter "C"
to cancel and continue.
6. For cases where the voter's name is illegible, place an "I"
by that line.
Try to read both the printed name and the signature.
Special requirements for voter records which have a "transfer
Date" which is later than the date the petition was signed.
This includes active, inactive, or deleted records.
In those instances where a match is achieved for both name and
address,or name only but not address, and the "Last Transfer
Date" is after the date the petition was signed, place an
"M" by that line number.
Upon completion of each page, the staff member will initial the
bottom night hand corner of the petition page.
Page6
Standard Procedures for Verification of Initiative Petitions
Section 4- Registration Verification - Signers (Con 't)
To start a new page, select "P". The computer will then
return you to the "Enter Signers of Petition for Validating.
screen where everything will appear automatically EXCEPT- the
Section Number, which is where you must enter the new Page Number.
IT IS ESSENTIAL THAT THE NEW PA GE NUMBER BE ENTERED BEFORE YOU
TRY TO VERIFY ANY NAMES ON THE NEW PA GE.
Repeat the verification process as detailed above until all assigned
pages have been verified, then return the pages to the Registration
Processing Manager.
Re-Check Procedures
Upon completion of all petition pages, all "X" and "I"
entries will be rechecked-but not by the same staff member who
made them during the initial pass through.
Use the same verification steps as before, but this time use GREEN
ink to mark the petition page. Where a match is found in one of
the entries originally rejected, place an "OK" next
to the previous mark and accept the voter's record.
| Section 5- Production of Tabulation Printouts
Following the re-check procedure, the Data
Systems Branch will tabulate the data on verified registrants
accepted into the petition checking module and produce the following
reports:
a. Alphabetical listings of all registered petition signers, by
Ward, whose names and
addresses have been verified. The reports produced will also provide
an exact count of each Ward's verified registrants, indicating
the page number of the petition where each can be found, and the
current registration status of each verified registrant.
b. A listing of all duplicate names found on the petition which
will include the petition page number where the duplicates may
be found.
In the case of multiple petitions being processed concurrently,
separate reporting and listing preparation will be maintained
for each petition.
The reported totals of verified registrants citywide and by Ward
will be compared to the minimum signature requirements established
for the measure. If the registered voter totals reflect numerical
sufficiency of the petition, the procedure will move to the next
step- random sample signature verification.
Page 7
Standard Procedures for Verification of Initiative Petitions Section
5- Production of Tabulation (ContJ
If the registered voter totals appear potentially insufficient-either
at the citywide or ward levels - an assessment of the number of
"M" entries on the petition will be made. If there are
enough potential "M" entries to affect the sufficiency
of the petition, these will be researched to determine if the
voter was registered at the address on the petition on the date
the petition was signed. Procedures for this are as follows:
Checking the "M" Entries ("M''= Last Transfer Date
is after the date signed)
The relatively few entries on the petition that were marked with
an ''M'' during the verification procedure are cases where a potential
match may exist but further research is required to make a determination.
The "M's" are registration records where a "transfer
date" indicating a change in the voter's record-has occurred
after the date on which the petition was signed.
For example, a voter may have filed a change of address, been
reinstated to the active file, or have been cancelled from the
roll after the petition was signed. A search of prior record data
will be made for these entries if the outcome can affect the sufficiency
of the petition.
The purpose of the search is to determine the registration status
and address that was in effect for the voter on the date the petition
was signed. This process will be completed using a third color
of ink, something other than RED or GREEN.
1. Check the "M" entry on the system to see if either
the current or any previous address matches the address on the
petition.
a. If "no", place an "X" next to the "M"
entry on the page;
b. If "yes", check the date of registration and the
date of active status to determine if the registrant was unregistered,
inactive, or deleted as of the date signed. If any of these are
the case, place an "X";
c. If the registrant was in active status on the date signed,
determine if the matching address was in effect on the date signed;
if it was not, place an "X";
d. If the registrant was in active status at the matching address
on the date signed, place an "OK" next to the "M"
entry on the page.
2. If the number of accepted "M" entries is significant
to the further review of the petition, these will be incorporated
into the universe of verified registrants to be examined in the
random sample procedure. Addition of accepted "M" entries
to the listing of verified registrants may be accomplished either
by production of a new listing or attachment of the additional
entries to the existing list.
Memorandum
TO:Herbert Bixhorn, Chief,
Office of Planning, Data Management Division
FROM:Alice P. Miller Executive Director
SUBJECT: Provision of Random Sample Universe; Request for Random Sample Designations Be: Initiative Measure #59: "Legalization of Marijuana for Medical Treatment Initiative of 1998"
Pursuant to 3 DCMR 1008.2, the table below supplies the number of apparent registered voter signatures, both District-wide and in each Ward, that have been established as the random sample universe for the Initiative # 59 petition. These totals are shown in the column entitled "Total Signatures to be Utilized for Random Sampling". The table also provides the signature requirement for determining acceptance or rejection of this measure.
The names of the verified registrants in each Ward have been listed alphabetically and are serially numbered. In accordance with 3 DCMR l 008.3, please designate random samples of l 00 signatures from each ward listing for validation against the Board's records.
Ward Signatures of Registered Voters
RequiredTotal Signatures to be Utilized for Random
Sampling1 2,063 2,501 2 2,067 2,254 3 2,287 2,604 4 2,355 1,572 5 2,289 1,971 6 2,197 3,091 7 2,156 1,199 8 1,583 1,900 Citywide l6,997 17,092
441 4th St., NW, suite 250 * Washington, D.C. 20001-2745 * Telephone (202) 727-2525 * Fax: (202) 347-2648
DISTRICT OF COLUMBIA
BOARD OF ELECTIONS AND ETHICS
WASHINGTON, D.C.
July 31,1998Memorandum
TO: Herbert Bixhorn, Chief,
Office of Planning, Data Management DivisionFROM: Alice P. Alice P. Miller, Executive Director
SUBJECT: Results of Random Sample Verification Conducted on Petition Signatures Re: Initiative Measure #59: "Legalization of Marijuana for Medical Treatment Initiative of 1998"
Pursuant to 3 DCMR 1008.4, we have completed signature verification of the randomly selected names appearing on the petition filed in support of Initiative Measure # 59. The results of these findings are shown in the table below, in the column headed "Number of Valid Signatures".
In accordance with 3 DCMR 1008.5, please provide your analysis as to whether the initiative petition has met the requirements necessary for the measure to be placed on the ballot. If you have any questions or if additional samples are needed to make a determination please call me at 7272525.
WARD Signatures of
Registered Voters
RequiredNo. of Registrants
in the Random
Sampling UniverseNumber of
Signatures in
Random SamplingNumber of Valid
Signatures1 2,063 2,501 100 97 2 2,067 2,254 100 99 3 2,287 2,604 100 100 4 2,355 1,572 100 99 5 2,289 1,971 100 98 6 2,197 3,091 100 98 7 2,156 1,199 100 100 8 1,583 1,900 100 98 Citywide 16,997 17,092
441 4th St., NW, Suite 250 Washington, D C 20001-2745 * Telephone: (202) 727-2525 * FAX: (202) 347-2648Government of the District of Columbia Office the Director July 31, 1998 MEMORANDUM
TO: Kathryn Fairley
Registrar of VotersFROM: Herbert J. Bixhorn, Chief
Data Management DivisionSUBJECT: Statistical Summary and Findings on Initiative Measure #59 Petition
Office of Planning
415 12th Street, N.W.
Washington, D.C. 20024Following is a statistical summary resulting from the random sample verification of signatures on the Initiative Measure #59 petition. (Refer to attached technical document "Criteria for Acceptance or Rejection of Petitions with 95% Confidence".) Our finding is that the initiative petitionisrejected with 95% confidence, as it did not have a sufficient number of valid signatures of registered voters for the District as a whole.
Table l shows the required number of valid signatures and the number of registrant signatures by ward and the District as a whole that were subject to random sampling. As shown in Table 2, five wards were accepted with 95% confidence. Wards 4, 5 and 7 were rejected at the outset because the number of verified registrants in each was less than the required number. Nevertheless, these three wards were sampled so that they could contribute with the other wards to the District-wide figures.
Table 3, which presents the District-wide results, shows that the District as a whole was rejected with 95% confidence.TABLE 1
WARD SIGNATURES OF
REGISTERED VOTERS
REQUIREDNUMBER OF REGISTRANT
SIGNATURES SUBJECT TO
RANDOM SAMPLING1 2,063 2,501 2 2,067 2,254 3 2,287 2,604 4 2,355 1,572 5 2,289 1,971 6 2,197 3,091 7 2,156 1,199 8 1,583 1,900 DISTRICT WIDE 16,997 17,092 TABLE, 2
WARD ACCEPTANCE/REJECTION
PARAMETERS
a1------------------- b1NUMBER OF VALID
SIGNATURES IN
SAMPLES OF 100DECISION WITH 95% CONFIDENCE 1 75------------------- 89 97 Accept 2 86------------------- 97 99 Accept 3 81------------------- 94 100 Accept 4 *---------------------* 99 Reject 5 *---------------------* 98 Reject 6 62-------------------- 80 98 Accept 7 *---------------------* 100 Reject 8 76------------------- 90 98 Accept
TABLE 3
District as a Whole Z(R) = -2.08 Z(R-1) = -2.07 Decision with 95% Confidence: Reject
. _