What Potomac Can Do For You What Potomac Can Do For You
Click here to return to the Potomac home page.
 
Who We Are and What We DoWinning StrategiesSurveys, Focus Groups, & AnalysesMedia, PR, & Public EducationSecure LoginCall Today


The Maryland Poll (N=1,000 likely voters)

April 11-13, 2005

2005 Post-Legislative Session - 2006 Election Benchmark

 Methodology


Objectives

Potomac Survey Research was commissioned by The (Baltimore) Sun to conduct a telephone survey on issues and politics among likely voters in Maryland.

The survey instrument was designed to measure:

  • Economic outlook
  • Job performance of Governor Ehrlich
  • Impressions of the Steffen controversy
  • Awareness and rating of the 2005 legislative session
  • Slots
  • Stem cell funding
  • Sale of public lands
  • Elected official favorability ratings
  • 2006 Democratic primary match-ups for Governor and U.S. Senator
  • 2006 general election match-ups for Governor and U.S. Senator
  • Orioles v. Nationals

Overview and Sampling Error
Potomac completed a total of 1,000 telephone interviews statewide, April 11–13, 2005, among a random sampling of likely-voting Maryland registered voters. 

According to customary statistical standards, the statewide sample produces a margin of error of no more than ± 3.2% at a 95% confidence level.  This means that, 95 percent of the time, the “true” figure would fall within this range if every likely voter in the state had been interviewed. 

For smaller subgroups of the overall sample, the potential sampling error is larger.  For example, the sub-sample of 491 likely Democratic primary voters has a potential margin of sampling error of ± 4.5%. The following chart outlines the maximum sampling error associated with sub-samples of varying sizes:

Sample Size

Sampling Error*

800

+/- 3.5

600

+/- 4.1

400

+/- 5.0

300

+/- 5.8

200

+/- 7.1

150

+/- 8.2

100

+/- 9.8

*At the 95% confidence limit.

Margin of sampling error is only one of many potential sources of error in this or any other public opinion survey.  The practical difficulties of conducting any survey of public opinion introduce other sources of error into the study.  Variations in question wording or the order of the questions, for instance, can lead to somewhat different results.

Sampling
The survey sample was drawn from an electronic database of Maryland registered voters and matched with telephone numbers by a commercial vendor.  The use of a voter file is preferred because it ensures the exclusion from the survey sample of residents who are not registered to vote.  Voter history codes on the file allow a pre-election sample to be targeted to voters who have a proven track-record of voting in past elections.

Strict geographic and partisan sampling quotas were adhered to throughout the interviewing process to reflect the proper distribution of registered voters statewide.  These quotas are outlined later in this section of the report.

Telephone Interviewing
Our subcontractor’s staff of professionally-trained and supervised callers completed the telephone interviews on the evenings of April 11–13.  Calls were made between 5:00 and 9:30 p.m., Eastern time.  The survey length averaged approximately 7.5 minutes.

Data Weighting
After field interviewing was complete, the data were further weighted to bring them into strict balance with the underlying demographics of the state’s electorate.  Those weights are an attachment to this section of the report.

Suggested Language for Statement of Methodology
This Maryland Poll, sponsored by The (Baltimore) Sun, was conducted by telephone among a random cross-section of 1,000 likely Maryland voters, April 11–13, 2005.  A statewide list of registered voters was used and matched with telephone numbers to ensure that non-registrants were excluded from the sample.  Throughout the interviewing process, strict geographic and political party sampling quotas were maintained, and the sample was weighted so that it accurately reflected the likely demographic make-up of the statewide electorate.  The maximum margin of error for the statewide sample is plus or minus 3.2% at a 95% confidence level.  For smaller subgroups the potential sampling error is larger.  The survey was conducted by Potomac Survey Research of Bethesda, Maryland.

2005 Maryland Poll Methodology Information
301-656-7900

 

(04-21-05) Download the poll Summary Charts as an Adobe Acrobat .PDF file

(04-21-05) Read "O'Malley ahead of Ehrlich" as an Adobe Acrobat .PDF file
(04-21-05) Read "Orioles versus Nationals in World Series" as an Adobe Acrobat .PDF file
(04-21-05) Read "Slots Support Down and Stem Cell Funding Up" as an Adobe Acrobat .PDF file
(04-21-05) Read "Steele Senate Race Support" as an Adobe Acrobat .PDF file

 

For complete newspaper coverage, visit The (Baltimore) Sun website.

 

 

More News
 

[ home  |  about us   |  strategies  |  research  |  communications  |  client area  |  contact us ]