Goucher College Releases Poll of Md. Likely Voters

A poll of likely Maryland voters finds a lot of support for the Biden-Harris ticket.
Goucher College surveyed 1,002 Marylanders and 776 were identified as likely voters.
61% of those surveyed said they would vote for Joe Biden for President, and 30% favored President Donald Trump.
Goucher College also asked likely voters their preferred method of casting a ballot.
51% said they would vote in-person either during early voting or on Election Day.
48% said they would vote by mail or by using a ballot drop box.
Broken down by party:
72% of Republican likely voters and 39 percent of Democrats plan to vote in person.
59% of Democratic voters may vote by mail, compared to only 28-percent of Republican likely voters who were polled.

Goucher College released its poll methodology:

Survey Methodology
To ensure all Maryland residents are represented, the poll is conducted using random digit dialing (RDD) of a
county-level stratified random sample using landline and cellular telephone numbers. The sample of telephone
numbers and interviewing were provided by Dynata (https://www.dynata.com/). The survey was conducted
Wednesday, September 30, to Sunday, October 4. During this time, interviews were conducted from 1 to 9
p.m.
Interviewers attempted to reach respondents with working phone numbers a maximum of five times. Only
Maryland adults—residents aged 18 years or older—were eligible to participate. Interviews were not conducted
with adults who were reached at business or work numbers. Ninety-two percent of the interviews were
conducted on a cell phone and 8 percent were conducted on a landline.
Interviews for this survey were completed with 1,002 Maryland adults. For a sample size of 1,002, there is a 95
percent probability the survey results have a plus or minus 3.1 percentage point sampling error from the actual
population distribution for any given survey question. For the sample of 776 Maryland adults identified as
likely voters, there is a 95 percent probability the survey results have a plus or minus 3.5 percentage point
sampling error from the actual population distribution for any given survey question. Margins of error are
higher for subsamples. In addition to sampling error, all surveys are subject to sources of non-sampling error,
including question wording effects, question order effects, and non-response bias. Margin of error is not
adjusted for design effects. Data is weighted by gender, age, race, region, and educational attainment of the state
to represent adult population targets established by the most recent American Community Survey (ACS).