NATIONAL POLL REPORT
THE ROBERT MORRIS UNIVERSITY
POLLING INSTITUTE
POWERED BY:
September 2014
Robert Morris University Polling Institute
Statement of Confidentiality and Ownership
All of the analyses, findings and recommendations contained within this report are the
exclusive property of Robert Morris University.
As required by the Code of Ethics of the National Council on Public Polls and the United
States Privacy Act of 1974, The Robert Morris University Polling Institute maintains the
anonymity of respondents to surveys the Institute conducts. No information will be released
that might, in any way, reveal the identity of the respondent.
Moreover, no information regarding these findings will be released without the written
consent of an authorized representative of Robert Morris University.
Robert Morris University Polling Institute
Page 2
TABLE OF CONTENTS
1
2
SECTION
Introduction .............................................................................................................................. Page 4
SECTION
Methodology ............................................................................................................................. Page 5
3
4
SECTION
Highlights ................................................................................................................................... Page 7
SECTION
Summary of Findings ............................................................................................................. Page 9
Election 2014.......................................................... 9
Demographics ...................................................... 15
5
SECTION
Appendix .................................................................................................................................. Page 18
Survey Instrument
Composite Aggregate Data
Robert Morris University Polling Institute
Page 3
1
INTRODUCTION
The Polling Institute at Robert Morris University is pleased to present the results of a brushfire
poll of 500 likely Pennsylvania voters.
The poll was designed to assess voter intent in the November 2014 election along with views
on election related issues.
The research study included survey responses from 500 respondents from the state of
Pennsylvania who were 18 years of age or older and vote all the time or most of the time.
The national poll included the following areas for investigation:
 How respondents plan to vote in the race for Governor of Pennsylvania;
 Reasons for their respective candidate decisions or reasons the remain undecided;
 Likelihood President Obama’s involvement would move a voter to or away from a
candidate;
 Measuring the degree President Obama’s record would impact election decision
making;
 Plans for ticket splitting;
 Measure the percentage of respondents who have been polled on the PA election in
2014;
 Measure the percent of those admitting having lied to a pollster during this election
cycle;
 Views on the impact of the Affordable Healthcare Act or Obamacare;
 And views on the Sandusky investigation/scandal and the impact on the
gubernatorial election; and
 Demographics.
Section II of this report discusses the Methodology used in the study, while Section III
includes Highlights derived from an analysis of the quantitative research. Section IV is a
Summary of Findings from the online survey.
Section V is an Appendix to the report containing the composite aggregate data and the survey
instrument employed.
Robert Morris University Polling Institute
Page 4
2
METHODOLOGY
Using a quantitative research design, the Institute completed 500 online surveys among
Pennsylvania residents.
Survey design input was provided by Robert Morris University officials.
Survey design is a careful, deliberative process to ensure fair, objective and balanced surveys.
Staff members, with years of survey design experience, edit out any bias. Further, all scales
used by the Institute (either numeric, such as one through ten, or wording such as strongly
agree, somewhat agree, somewhat disagree, or strongly agree) are balanced evenly.
Additionally, placement of questions is carefully accomplished so that order has minimal
impact.
This survey was conducted September 26 - 30, 2014.
Respondents qualified for the survey if they were a resident of Pennsylvania and 18 years of
age or older.
All facets of the study were completed by the Institute’s senior staff and researchers. These
aspects include: survey design, pre-test, computer programming, fielding, coding, editing,
verification, validation and logic checks, computer analysis, analysis, and report writing.
Statistically, a sample of 500 has an associated margin for error of +/-4.5% at a 95%
confidence level.
Results throughout this report are presented for composite results – all 500 cases.
Readers of this report should note that any survey is analogous to a snapshot in time and
results are only reflective of the time period in which the survey was undertaken. Should
concerted public relations or information campaigns be undertaken during or shortly after the
fielding of the survey, the results contained herein may be expected to change and should be,
therefore, carefully interpreted and extrapolated.
Robert Morris University Polling Institute
Page 5
Furthermore, it is important to note that all surveys contain some component of “sampling
error”. Error that is attributable to systematic bias has been significantly reduced by utilizing
strict random probability procedures. This sample was strictly random in that selection of
each potential respondent was an independent event based on known probabilities.
Each qualified online panel member within the State of Pennsylvania had an equal chance for
participating in the study. Statistical random error, however, can never be eliminated but may
be significantly reduced by increasing sample size.
Robert Morris University Polling Institute
Page 6
3
HIGHLIGHTS
ON THE GUBERNATORIAL CONTEST IN PENNSYLVANIA…
 One month out, if the election was held today, Democratic candidate Tom
Wolf would overcome the incumbent Governor Tom Corbett by 50.0% to
29.2% with 20.8% undecided among likely voters.
 When those leaning toward the two candidates are included, the new results
show Tom Wolf leading Governor Corbett 56.8% to 33.8% with fewer (9.4%)
still undecided
 Large majorities of voters (84.3%) are extremely, 43.0%, or
somewhat/moderately, 41.3%, committed to their candidate of choice for
Governor.
 By more than a two-to-one margin, President Obama’s involvement on behalf
of a gubernatorial candidate would make voters less likely to support that
candidate than more likely – 30.0% to 14.2%, respectively. For just over half,
50.6%, the President’s support would not make a difference in their vote
plans.
 Resident voters are evenly split on whether President Obama’s record is
relevant to their decisions on candidates during this election cycle in
Pennsylvania. While 47.0% suggest the President’s record is very (22.2%) or
somewhat (24.8%) relevant, another 48.8% suggested his record is somewhat
irrelevant (12.2%) or not at all relevant (36.8%).
 Most Pennsylvania voters plan to split their respective tickets while voting.
While 12.6% will vote straight Republican and 16.4% plan to vote straight
Democratic, the remainder (56.8%) will split their votes. Some, 14.2%, are
unsure if they will split their votes among the two parties.
 One-fifth of likely voters surveyed indicated, prior to this poll, that they have
been polled one or more times. Of this group of polled voters, 17.8% admitted
they have lied to a pollster.
Robert Morris University Polling Institute
Page 7
 Over one-third, 36.4%, of all those surveyed indicated they or family members
have faced healthcare insurance challenges or changes. Of this group, 78.0%
describe these changes/challenges as somewhat or very negative.
 The Governor’s handling of the Sandusky investigation while he was Attorney
General is likely to make 12.4% of voters more likely to support him while
26.8% suggest his actions will make them less likely to support him for reelection.
 Just over one-half, 50.4%, agreed that Tom Corbett, when Attorney General,
took too long to prosecute the Sandusky case. Another 41.4% agreed that the
firing of Joe Paterno was the wrong decision while 38.6% indicated the NCAA
should never have sanctioned the Nittany Lions Football Team. About onethird, 37.8%, suggested campaign contributions to the Governor’s re-election
campaign from Sandusky charity board members should have been returned.
Robert Morris University Polling Institute
Page 8
4
SUMMARY OF FINDINGS
Readers are reminded that the narrative throughout this report refers to composite aggregate
data – the 500 respondents to a survey of Pennsylvania residents.
ELECTION 2014
If the election for Governor was held today, the Democratic candidate, Tom Wolf, would
receive 50.0% of the vote to 29.2% for Governor Tom Corbett with 20.8% undecided.
Results are presented below.
Vote for Governor Today?
50
29.2
20.8
TOM WOLF
TOM CORBETT
UNDECIDED
Support
Robert Morris University Polling Institute
Page 9
Those undecided, 20.8%, were asked who they would say they are leaning toward. When
“leaning” respondents are added to their respective candidate totals, Wolf leads Corbett
56.8% to 33.8% with 9.4% remaining undecided as depicted below.
Candidate Preference with Leaners
56.8
33.8
9.4
TOM WOLF
TOM CORBETT
REMAIN UNDECIDED
Support
Candidate commitment appears strong. Over four-fifths of all “decided or leaning”
respondents, 84.3% suggested they are extremely (43.0%) somewhat committed (41.3%) to
their candidate of choice. Composite results (all decided likely voters) as shown side by side
with Wolf and Corbett supporters in the following graph.
Extremely & Somewhat Committed
90
80
84.3
84.9
83.4
COMPOSITE
TOM WOLF SUPPORTERS
TOM CORBETT SUPPORTERS
70
60
50
40
30
20
10
0
Extremely & Somewhat Committed
Robert Morris University Polling Institute
Page 10
Respondents were asked: “If President Obama supported and campaigned for any candidate
for Governor in Pennsylvania, would that make you more or less likely to continue to
support that candidate or wouldn’t it make a difference?”
Results are presented here.
Would Support for Tom Wolf Be Impacted?
President
Obama
30.0
It would make me less likely to support that
candidate
It would make me more likely to support that
candidate
It would not make a difference in my support
Unsure
14.2
50.6
52
Resident voters are evenly split on whether President Obama’s record is relevant to their
decisions on candidates during this election cycle in Pennsylvania. While 47.0% suggest the
President’s record is very (22.2%) or somewhat (24.8%) relevant, another 48.8% suggested
his record is somewhat irrelevant (12.2%) or not at all relevant (36.8%). Results are
displayed in the following graph.
Relevance of President's Record on Vote?
36.8
24.8
22.2
12.2
4
VERY RELEVANT
SOMEWHAT
RELEVANT
SOMEWHAT
IRRELEVANT
Robert Morris University Polling Institute
NOT AT ALL
RELEVANT
UNSURE
Page 11
It appears most voters will be splitting their tickets in November 2014 between Republican
and Democratic candidates. Just 29.0% indicated they would be voting straight party line –
either 12.6% for all Republicans or 16.4% for all Democrats. Results are shown here.
Ticket Splitting?
23.63
20.8
16.4
14.2
12.6
ALL REPUBLICAN
12.4
MOSTLY
REPUBLICAN
EVEN SPLIT
BETWEEN PARTIES
MOSTLY
DEMOCRATIC
ALL DEMOCRATIC
UNSURE
Nearly one-fifth of all respondents, 18.0%, indicated they have answered a political poll
about the Pennsylvania 2014 election excluding the current poll. Some, 11.8%, indicated
one other while 4.8% suggested they done two to five other polls and 1.4% suggested they
have completed six or more polls. \
Respondents polled (18.0%) were asked: “Honestly now, have you ever lied to a pollster or
in a political survey in this election cycle or in past elections?” Nearly one-fifth, 17.8%,
indicated they have, indeed, lied to a pollster as show here.
Ever Lied to a Pollster?
4.4
17.8
77.8
Yes
No
Robert Morris University Polling Institute
Unsure
Page 12
Two-fifths of all Pennsylvania poll respondents, 36.4%, indicated they have faced healthcare
insurance changes or challenges over the past year or two such as a loss of coverage, higher
deductibles or higher employer deductibles that they would attribute to the new Affordable
Care Act or Obamacare.
Of this group, 36.5%, just 17.0% describe these changes as very or somewhat positive while
78.0% describe the changes as somewhat or very negative. The following graph depicts
results as collected.
Describing the Affordable Care Act
44.5
33.5
9.9
7.1
VERY POSITIVE
SOMEWHAT POSITIVE
4.9
SOMEWHAT
NEGATIVE
Robert Morris University Polling Institute
VERY NEGATIVE
UNSURE
Page 13
Respondents were asked how Governor Corbett’s handling of the Sandusky investigation –
which led to the imprisonment of Jerry Sandusky, the firing of Joe Paterno and sanctions
against Penn State’s football program – may have impacted their respective vote plans.
Results are presented here.
Impact of Sandusky Scandal on Voter Intent
54.6
26.8
12.4
6.2
MORE LIKELY TO SUPPORT LESS LIKELY TO SUPPORT
DIDN'T MAKE A
UNSURE / DIDN'T FOLLOW
THE GOVERNOR'S RETHE GOVERNOR'S RE- DIFFERENCE IN HOW I PLAN THE SANDUSKY SCANDAL
ELECTION
ELECTION
TO VOTE
Robert Morris University Polling Institute
Page 14
All respondents were provided with four statements related to the Sandusky scandal. Each
was asked to indicate which they agreed with, if any. Multiple responses were allowed.
The results are presented here.
Statements
It took Tom Corbett, then the PA Attorney General, too long to
prosecute the Sandusky case
The firing of Joe Paterno was a wrong decision
The NCAA should never have sanctioned the Nittany Lions Football
Team
Campaign contributions to Tom Corbett’s gubernatorial campaign
from Sandusky charity board members should have been returned
Percent
50.4
41.4
38.6
37.8
DEMOGRAPHICS
Age
18 to 44
45 to 64
65 +
PA
44.6
41.2
14.2
Income
Under $9,999
$10,000 to less than $40,000
$40,000 to less than $70,000
$70,000 to less than $100,000
$100,000 to less than $130,000
$130,000 to less than $160,000
$160,000 or more
Unsure
PA
2.2
19.2
27.3
21.0
11.9
6.1
7.9
4.4
Robert Morris University Polling Institute
Page 15
Party Affiliation
Republican
Democrat
Independent
Some other party
Unsure
PA
37.0
47.0
13.2
0.8
2.0
Marital Status
Single, never married
Married or domestic partner
Widowed
Divorced
Separated
Other
PA
24.8
61.5
3.2
7.7
1.8
1.0
Education
8th grade or less
Some high school
High school graduate or GED
Some technical school
Technical school graduate
Some college
College graduate
Post graduate or professional degree
Refused
PA
--1.8
15.9
1.6
5.6
17.3
34.6
23.1
---
Hispanic, Latin American, Puerto Rican,
Cuban or Mexican
Yes
No
PA
Robert Morris University Polling Institute
6.2
93.4
Page 16
Ethnicity (Among Non-Hispanics)
White
Black, African-American
Asian, Pacific Islander
Aleutian, Eskimo or American Indian
Other
Native Hawaiian
Two or more races
Refused
Don’t know/unsure
PA
80.6
12.2
3.2
0.6
3.4
----------
Children under 18 living at home
None
One
Two
Three
Four
Five or more
Don’t know
Refused
PA
34.1
18.9
26.9
12.2
5.6
2.2
-----
Gender
Male
Female
PA
50.8
49.2
Robert Morris University Polling Institute
Page 17
5
APPENDIX
INTERPRETATION OF AGGREGATE RESULTS
The computer processed data for this survey are presented in the following frequency
distributions. It is important to note that the wordings of the variable labels and value labels
in the computer-processed data are largely abbreviated descriptions of the Questionnaire items
and available response categories.
The frequency distributions include the category or response for the question items.
Responses deemed not appropriate for classification have been grouped together under the
“Other” code.
The “NA” category label refers to “No Answer” or “Not Applicable.” This code is also used
to classify ambiguous responses. In addition, the “DK/RF” category includes those
respondents who did not know their answer to a question or declined to answer it. In many
of the tables, a group of responses may be tagged as “Missing” – occasionally, certain
individual’s responses may not be required to specific questions and thus are excluded.
Although when this category of response is used, the computations of percentages are
presented in two (2) ways in the frequency distributions: 1) with their inclusion (as a proportion
of the total sample), and 2) their exclusion (as a proportion of a sample sub-group).
Each frequency distribution includes the absolute observed occurrence of each response (i.e.
the total number of cases in each category). Immediately adjacent to the right of the column
of absolute frequencies is the column of relative frequencies. These are the percentages of
cases falling in each category response, including those cases designated as missing data. To
the right of the relative frequency column is the adjusted frequency distribution column that
contains the relative frequencies based on the legitimate (i.e. non-missing) cases. That is, the
total base for the adjusted frequency distribution excludes the missing data. For many
Questionnaire items, the relative frequencies and the adjusted frequencies will be nearly the
same. However, some items that elicit a sizable number of missing data will produce quite
substantial percentage differences between the two columns of frequencies. The careful
analyst will cautiously consider both distributions.
The last column of data within the frequency distribution is the cumulative frequency
distribution (Cum Freq.). This column is simply an adjusted frequency distribution of the
sum of all previous categories of response and the current category of response. Its primary
usefulness is to gauge some ordered or ranked meaning.
Robert Morris University Polling Institute
Page 18