THE ASSOCIATIONS AMONG PERSONALITY FACTORS THE THEORY OF PLANNED

the associations among personality factors, the theory of planned behavior and voting blair nicole lynch ================== b.s.,

THE ASSOCIATIONS AMONG PERSONALITY FACTORS, THE THEORY OF PLANNED
BEHAVIOR AND VOTING
Blair Nicole Lynch
==================
B.S., Southern Oregon University, 2008
======================================
THESIS
Submitted in partial satisfaction of
the requirements for the degree of
MASTER OF ARTS
in
PSYCHOLOGY
at
CALIFORNIA STATE UNIVERSITY, SACRAMENTO
=======================================
SPRING
2011
THE ASSOCIATIONS AMONG PERSONALITY FACTORS, THE THEORY OF PLANNED
BEHAVIOR AND VOTING
A Thesis
by
Blair Nicole Lynch
Approved by:
__________________________________, Committee Chair
Lisa Bohon, Ph. D.
__________________________________, Second Reader
Rebecca Cameron, Ph. D.
_____________________________________, Third Reader
Emily Wickelgren, Ph. D.
____________________________
Date
Student: Blair Nicole Lynch
I certify that this student has met the requirements for format
contained in the University format manual, and that this thesis is
suitable for shelving in the Library and credit is to be awarded for
the thesis.
__________________________, Graduate Coordinator ___________________
Jianjian Qin, Ph. D. Date
Department of Psychology
Abstract
========
of
THE ASSOCIATIONS AMONG PERSONALITY FACTORS, THE THEORY OF PLANNED
BEHAVIOR AND VOTING
by
Blair Nicole Lynch
The Theory of Planned Behavior (TPB) (Ajzen, 1991) and personality
were used to predict intention to vote in the 2010 California general
election. The TPB considers Attitudes, Subjective Norms and Perceived
Behavioral Control beliefs in relation to intention and subsequent
action. Personality was measured using the Big Five Questionnaire
(John and Srivastana, 1991). Extraversion, Conscientiousness and
Agreeableness were assessed in relation to intention. The participants
(N =64) were able to participate using an online research website
(surveymonkey.com) that required them to complete the measurements
before Election Day and return after Election Day to indicate their
voting action. The findings show that attitude was a significant in
predicting intention (p < .01) over subjective norms and perceived
behavioral control, and that intention is a significant predictor of
voting action. Overall, the findings suggest that intention (p < .01)
and attitude (p < .05) are the only significant predictors of voting
action and affiliation is the only significant predictor of intention
(p < .01).
Keywords: theory of planned behavior, personality, vote, intention
_______________________, Committee Chair
Lisa Bohon, Ph. D.
_______________________
Date
ACKNOWLEDGEMENTS
================
This research was made possible by the assistance and dedication of
Dr. Lisa Bohon, and her willingness to take on a student who was
interested in studying the psychological aspect of politics. I would
also like to thank Dr. Rebecca Cameron, and Dr. Emily Wickelgren, for
assisting in the completion of my thesis and guiding me along the
research process.
TABLE OF CONTENTS
Page
Acknowledgements v
List of Tables vii
List of Figures viii
Chapter
1. INTRODUCTION 1
Theory of Planned Behavior 1
Personality Factors 3
Hypotheses 6
2. METHODS 8
Overview 8
Participants 8
Materials 9
Procedure 12
3. RESULTS 14
4. DISCUSSION 20
References 25
LIST OF TABLES
==============
Page
1. Table 1 Reliability of TPB Questionnaire Subscales………………………….. 11
2. Table 2 Hierarchical Regression Predicting Intention to
Vote.……………..... 15
3. Table 3 Intention and TPB Predicting Vote Action…………..……………….. 16
4. Table 4 Interaction of Personality and TPB Components……………………... 18
LIST OF FIGURES
===============
Page
1. Figure 1 Illustrated Theory of Planned Behavior…………………………....... 2
Chapter 1
=========
INTRODUCTION
Democracy is the backbone of how American society functions. When
individuals cast their vote in an election, they are becoming part of
the democratic process of America. American society is composed of
many different cultures and values which influence the political
climate. However, individual factors such as personality, attitudes,
and intention can also influence voting. In the present research, I
investigated the nature of individual characteristics that are
associated with voting.
Theory of Planned Behavior
The Theory of Planned Behavior (TPB) uses a series of components that
comprise intention to predict behavior. The TPB was originally
developed by Ajzen in 1985. The theory was designed to predict and
explain human behavior in specific contexts. It is based on the idea
that behavior is a function of three beliefs about the specific
situation. These three beliefs are attitudes (A), subjective norms
(SN) and perceived behavioral control (PBC) (Ajzen, 1991).
Attitude is the overall beliefs about something—the evaluative opinion
the individual holds in regard to the behavior in question. It is
measured through the combination of the direct attitude beliefs and
the salient behavioral beliefs (Ajzen, 1991). Subjective norms are the
beliefs that one holds about the normative expectations of others and
consists of perceived pressure, approval or disapproval from the
influential people in one’s life (Chatzisarantis & Hagger, 2008), the
pressure to engage in the given behavior, and the individuals
motivation to comply (Ajzen, 1991). Perceived behavioral control is
the belief about the perceived ease or difficulty in engaging in the
behavior, as well as any previous experiences with carrying out the
specific behavior and any anticipated hindrances. The influence of PBC
depends on the strength of the control beliefs and the power to combat
any resistance to the intended action (Ajzen, 1991).
The more favorable the attitude and subjective norms are for the
specific context, and the greater the perceived control, the stronger
the intention to behave will be for the specific individual. The
theory is situation specific and the three contributing factors are
not constant for any individual (Ajzen, 1991). This means that every
new situation requires relevant information to form the best predictor
of behavior. The TPB is illustrated in Figure 1.
Figure 1. Illustrated Theory of
Planned Behavior.
Research by Harder and Krosnick (2008) examined psychological studies
that focused on why people vote and found that the psychological
disposition of an individual shapes his or her motivation to vote, and
is influenced by the individual’s motivation and ability [or
difficulty] in relation to the specific vote. This finding supports
the idea of perceived control having an influence on voting behavior
because PBC encompasses motivation and perceived difficulty when being
examined as an influence on intention. The research shows that aspects
of PBC are measured in other studies, and parts of TPB are present,
even when the model itself is not being utilized.
The Theory of Planned Behavior has been used to successfully predict
real world behaviors; for example, it has been used to predict
political behavior. One such study used the theory to predict state
legislator’s intentions to vote on tobacco control. It was predicted
that the intention-voting relation was dependent upon the level of
perceived behavioral control the participant experienced. The results
indicated that a relation does exist between PBC and intention-voting.
Overall, those who had higher intentions were 9.5 times more likely to
vote in comparison to those with low intentions (Flynn, et al., 1997).
Personality Factors
Personality has been studied in relation to the psychology of politics
on many different occasions and is found to be a reliable predictor of
political participation and political party affiliation (Caprara,
Barbaranelli and Zimbardo, 1999; Caprara & Zimbardo, 2004; Digman,
1990; Schoen & Schumann, 2007). Early research about personality and
politics focused on psychoanalysis, however, today the political
psychologies are focusing on the Big Five factors of Personality
(Caprara & Zimbardo, 2004). The five factors are: Extraversion,
Agreeableness, Conscientiousness, Neuroticism and Openness (Digman,
1990).
The five factors were later detailed and discussed in order to
understand what each of the factors represents within the individual’s
personality (John & Srivastavna, 1999). Extraversion is noted as the
opposite of introversion and includes gregariousness, assertiveness,
excitement seeking, enthusiasm, and being outgoing. Agreeableness is
the opposite of antagonism and includes trust, forgiveness, altruism,
compliance, modesty, and – being straight forward but not demanding.
Conscientiousness is having direction; it includes competency; and
being efficient, organized, and thorough, self-disciplined, and
deliberate. Neuroticism is defined as the opposite of emotional
stability and encompasses anxiety, hostility, depression,
impulsiveness, and vulnerability. Lastly there is openness, which is
seen as the opposite of closedness to experience. Openness includes
being curious, imaginative, artistic, and excitable, and having
unconventional values (John & Srivastavna, 1999).
Gordon Allport (1945) found that personality is closely related to
participation and gave rise to the research of personality factors
influencing political participation. Personality has mostly been
studied in relation to political party affiliation and voting behavior
in terms of motivation and values (Caprara, Barbaranelli & Zimbardo,
1999, Barnea & Schwartz, 1998); however, the current research is
designed to assess any relation between personality and intention and
use it as a predictor of behavioral intentions and action.
A link between personality and voting behavior was found by Schoen and
Schumann (2007) when they discovered that each factor of personality
is linked with different political agendas. The research found that
openness was indicative of liberal views because of their openness to
change. Conscientiousness was indicative of obedience to social rules,
and therefore would represent either party given their adherence to
the rules they follow. A high level of neuroticism indicated a support
for welfare policies, while low levels supported social and economic
liberalism. Agreeableness was positively correlated with a support for
collective society. Lastly, extraversion did not show associations
with political attitudes or vote choice and the relations were argued
to be hard to predict since no aspect is logically related to
political involvement.
As mentioned earlier, Harder and Krosnick (2008) found that an
individual’s psychological state influences the development of his or
her motivation, ability and likelihood of voting. Furthermore, the
repeated use of personality in predicting political behavior assures
its reliability and association to politics, and allows it to be
utilized as a moderating variable when assessing intention and action
based on the Theory of Planned Behavior. Therefore, I believe that
personality can be assessed as an influence on the intention,
motivation and desire to participate in an activity, and more
specifically, voting.
TPB has been used in conjunction with personality to enhance the
prediction of real life behaviors. For example, Chatzisrantis and
Hagger (2008) studied intention to continue with a physical activity.
The goal was to understand the role of personality as a moderator of
intention in the continuation of engaging in physical activity. The
results showed that conscientiousness moderated the effect of
intention to engage in the physical activity, however, other
personality factors were not effective predictors (Chatzisrantis &
Hagger, 2008). Despite the underwhelming relation between some
personality factors and intention in this specific study, it can still
be argued that personality has an important association with the
intention to engage in physical activity, and might also have an
impact on intention to vote.
The present study was designed to use personality and the Theory of
Planned Behavior to predict intention to vote, and then studied the
association between intention and actual voting behavior. Although the
TPB is not widely used to study voting behavior, it is the type of
situation for which it was developed. Moreover, personality was added
to investigate any increases in predictive power.
I used the political party to understand if being associated with a
political party is related to intention to vote and subsequent action.
All individuals who were registered or identify themselves as a member
of a political party were considered affiliated, and those who did not
note any such membership were not considered affiliated.
Hypotheses
Overall, I hypothesized that intention would be a good predictor of
voting behavior. In addition, I predicted that personality would have
a strong interaction with the aspects of intention, through its
relation to specific aspects of attitude, subjective norms, and
perceived behavioral controls.
H1: I predict that the TPB will be able to predict intention to vote
and the act of voting (H1a). Furthermore, that the influence of PBC
will have the strongest influence on the act of voting (H1b) based on
the model developed by Ajzen (1991).
H2: I predict that extraversion will have a significant interaction
with intention to vote, through the component of attitude. This is
consistent with research by Schoen and Schumann (2007) who found that
extraversion is related to a person’s attitudes regarding politics.
H3: I predict that agreeableness will have a significant interaction
with intention to vote. This is consistent with research by Schoen and
Schumann (2007) who found that individuals with high scores for
agreeableness are found to show support for the collective society.
This relates directly to the subjective norms and its measurement of
the influence of other people’s belief on the participant as well as
their motivation to comply (Ajzen, 1991).
H4: I predict that conscientiousness will have a significant
interaction with intention to vote, through the component of
subjective norms and the motivation to comply. This is consistent with
the findings of Schoen and Schumann (2007) who found that
conscientious individual’s are more obedient to social rules.
H5: In each election cycle, voters are motivated to participate by the
candidates, specific propositions, and perceived outcomes. In this
research, the specifics pertaining to this election are unknown,
however, it is predicted that being affiliated with a political party
will have a strong relation with intention to vote and vote action.
Chapter 2
METHODS
Overview
The purpose of this research was to understand how the Theory of
Planned Behavior, personality factors, and political party could be
used to predict and assess intention to vote. Intention to vote was
then used to predict actual voting behavior (action). Predictor
variables were attitude about voting, subjective norms, perceived
behavioral control, intention to vote, and personality traits
(neuroticism, extroversion, openness, agreeableness and
conscientiousness), and political party affiliation measured using
self-report (indicating membership as a Democrat, Republican and
others). The criterion variable was voting behavior.
Participants
A convenience sample was selected from the California State
University, Sacramento subject pool as well as undergraduate students
in communication and ethics courses at Sacramento City College after
obtaining permission to recruit participants from their instructors.
The CSUS students were able to find the study online through the
psychology department’s research website which detailed the
instructions for finding the online inventory, and the SCC students
were recruited during the first few minutes of their classes and given
a flyer that contained information identical to that found on the CSUS
website. All of the participants were treated according to the ethical
and confidentiality guidelines outlined by the American Psychological
Association. This included being fully informed prior to beginning
their participation, being allowed to withdraw at any time, being
assured that their answers and identity were kept separate and
confidential, and they were fully debriefed at the end of their
participation.
A total of 106 participants completed the first survey and 84 returned
to complete the follow up. Out of the 84 participants, 21 did not have
a matching identifier so the responses were omitted, bringing the
total number of participants to 63. The participants were
predominantly female (70%) and had a mean age of 22 years. Those who
participated were more often affiliated with a political party (86%)
than not (14%). Of the 64 participants, 30 (47.6%) did not vote and 33
(54.2%) did vote and less that 13% said there was a slightly strong to
strong influence of participation in the study on their decision to
vote. The responses collected for the first phase were collected
between October 25th, 2010 and November 1st, 2010 and the second phase
responses were collected between November 2nd, 2010 and November 13th,
2010 to allow the respondents enough time to return as well as enough
time for the data to be thoroughly analyzed before being submitted for
review.
Materials
Big Five Questionnaire—Personality Factors
Personality was measured using the Big Five Questionnaire, as
developed by John (1991). It measures personality by assessing five
main factors: Extraversion, Neuroticism, Openness, Conscientiousness
and Agreeableness. The questionnaire consists of 44 items that are
representative of the different factors. Responses are made on a
five-point scale in which the participant can indicate how much the
items reflect their self-perception (strongly agree – strongly
disagree). The BFQ is found to be a sound measure of the five factors
of personality, and uses a Lie (L) scale to account for the Social
Desirability response (Digman, 1994). Construct validity is upheld
with high correlations to other personality measurements such as the
NEO-PI (mean r = .68) (John, Naumann & Soto, 2008) in other samples
from different cultures (Caprara, et. al., 1999) and the Cronbach’s
alpha scores for each of the five subscales ranged from .82 to .86
when tested for reliability (John, et. al, 1991). Furthermore, a
chronbach’s alpha of the five domain scales with the current data
showed that the questionnaire has a high internal validity and
reliability with alpha scores ranging from .77 to .80.
Theory of Planned Behavior Questionnaire—Intention
Attitude, subjective norms, perceived behavioral control and intention
were measured using instructions from Ajzen
(http://people.umass.edu/aizen, 2006) which have been modified to
measure intention to vote in the 2010 California general election. The
items were developed according to the Theory of Planned Behavior. The
measurement consists of 42 items which have a 7-point response scale
with anchors that vary depending on how the question is worded, making
it specific to one of the three components of intentions: attitude,
subjective norms and perceived behavioral control. See Appendix A for
a complete version of the developed questionnaire. The anchors range
from good-bad, to extremely possible-extremely impossible, to
true-false, with a center anchor of neither, and included other
anchors depending on the wording of the question and the component
being measured. The range of response options identify the direct
attitude, subjective norm beliefs and perceived behavioral control as
well as the underlying behavioral beliefs, normative and power beliefs
of those the components, respectively. The responses range from
strongly agree-strongly disagree to extremely likely-extremely
unlikely depending on how the question is worded. A Cronbach’s alpha
analysis was conducted to be sure the newly developed TPB
questionnaire was reliable. Upon the initial analysis, certain items
measuring direct attitude and perceived behavioral control beliefs
were found to not significantly contribute to the individual subscales
and were thus omitted from the analysis. Before the items were
omitted, the alpha levels were .64 for direct attitude and .55 for
perceived behavioral control beliefs. After this omission, the alpha
levels ranged from .70 to .98 for the subscales of attitude,
subjective norms, and intention (see Table 1 for a complete list of
alpha levels).
Table 1
Reliability of TPB Questionnaire Subscales
_____Salient Beliefs_______ Direct Measure
Attitude .74 .88*
Subjective Norms .70 .72
Perceived Behavioral Control .87** .70*
Intention .98
Note. The reliability was determined by analyzing the Cronbach’s alpha
of each subscale. *Indicates alpha after items were omitted. N = 63.
**Perceived behavioral control salient beliefs are the power of those
control beliefs (Ajzen, 1991). All significance levels were at the p <
.05 level.
Political and demographic information were gathered through
self-report from the participants after completing both
questionnaires. I asked them for their age, sex, and political party
affiliation. Ethnicity was not included because when reading the
previous TPB research, it was not utilized in predicting intention. To
report their political party they were simply asked, “Which political
party are you a representative of (which party did you designate when
you registered to vote)? Example: Republican, Democrat.” Actual voting
behavior was determined through the use of a follow up set of
questions which not only asked if the participant voted, but asked
about any possible impact that answering the questions in the first
part of the study had on their desire or intention to vote.
Procedure
Participants from the CSUS subject pool were made aware of the
research website in their undergraduate psychology courses which
requires them to participate in a psychology research study. When they
went to the website, the current research was available as “A Study of
Politics (Online Survey)” and when they selected this research they
were given an overview about the two part participation, and given the
dates by which to complete the first survey, as well as the link to
complete the first surveys. The participants from Sacramento City
College (SCC) were given the same instructions; however, they were on
a flyer rather than posted on a website run by the school. SCC
students are not allowed to access the CSUS research website.
The first round of participation consisted of going through four pages
on the website link, the first being the informed consent, followed by
the BFQ, then the TPB Questionnaire and lastly the demographic
information. Upon completing all of these items, participants were
automatically redirected to a new site in which they were asked to
enter their email address for correspondence about the follow up
items. Their emails were collected on a different site to ensure the
confidentiality of their responses and keep them protected under the
APA ethical guidelines. The first part of the participation closed at
7:00am on November 2, 2010 when the polls opened to be sure no one
could respond after voting on Election Day.
After the first part of the data collection was closed, the emails
were collected and a follow up email was sent out asking participants
to answer a few follow-up questions. The email gave them instructions
similar to the ones they had in the first phase, and provided the link
to the follow up items (i. e. “Did you vote in the 2010 California
general election?”). Upon completing the questions they were debriefed
online, and asked to acknowledge their understanding of the study.
They were also given information about how to contact the researcher
with any questions. Upon acknowledging the debriefing, they were
redirected to another website to collect their emails once again. The
email was then used to contact them for their names in order to assign
credit to the research website at CSUS or to their instructors at SCC.
Only participants who completed both parts one and two were included
in the data analysis.
Chapter 3
=========
RESULTS
Those who did not respond to every question were kept in the sample
and the missing score was calculated with the mean score for the
specific item. This was done to keep the sample size at a number that
would still provide enough power to carry out the regression analysis.
The missing data was not greater than 3.2% of the respondents for any
item in the TPB questionnaire and 4.7% for any item in the BFQ and any
individual who did not answer all of the questions was omitted from
the sample. The total scores for the direct and salient subscales of
attitude, subjective norm and perceived behavioral control were used
for the main analyses
A hierarchical multiple regression analysis was used to assess the
ability of the Theory of Planned Behavior (Attitude, Subjective Norms,
Perceived Behavioral Control) to predict intention to vote (H1), after
controlling for age, sex, party affiliation and influence of the
participation in this research. Age, sex, affiliation and influence of
participation were entered at Step 1, explaining 31.8% of the variance
in intention to vote. After the entry of Attitude, Subjective Norms,
and Perceived Behavioral Control at Step 2, the total variance
explained by the model as a whole was 39.3%, F(7,54) = 5.00, p < .001.
Attitude, Subjective Norms and Perceived Behavioral Control accounted
for an additional 7.6% of the variance in intention, after controlling
for age, sex, affiliation and influence which were not statistically
significant, R square change = .076, F change (3,54) = 2.24, p >.05.
When looking at the model as whole, the only variable that was
statistically significant was affiliation (β = .47, p < .001), which
was controlled for in Step 1 (see Table 2 for a complete list of beta
levels and confidence intervals for each step).
Table 2
Hierarchical Regression Predicting Intention to Vote
Model 1 Model 2
Variable β 95%CI β 95%CI
Age .26 (.04, .59) .19 (-.06, .51)
Sex .05 (-2.36, 3.77) .09 (-1.86, 4.31)
Affiliation .51* (5.05, 12.94) .47* (4.57, 12.27)
Participation Influence -.03 (-1.31, 1.04) .00 (-1.17, 1.18)
Attitude .18 (-.02, .19)
Subjective Norms .22 (-.01, .34)
Perceived Behavioral Control -.06 (-.24, .14)
Note. N = 63. CI = confidence interval. Total scores for attitude,
subjective norms and perceived behavioral control.
*p < .001.
A logistic regression analysis was conducted to assess the ability of
intention to predict vote action (H1a) as well as PBC being the
strongest predictor of vote action(H1b) —over Attitude and Subjective
Norms. The model contained four independent variables (Intention,
Perceived Behavioral Control, Attitude and Subjective Norms) and vote
action as the categorical, dependent variable. The full model,
containing all predictors was statistically significant, Χ2(4, N = 63)
= 39.26, p<.001, indicating that the model was able to distinguish
between participants who reported and who had not reported voting. The
model as whole explained between 46.4 % (Cox and Snell R square) and
61.9% (Nagelkerke R square) of the variance in vote action, and
correctly classified 79.4% of the cases. As shown in Table 3,
intention and attitude made a statistically significant contribution
to the model. The strongest predictor of voting action was intention
recording an odds ratio of 1.57, indicating that those who have higher
intention to vote are more likely to have voted than those who did
not. The odds ratio for attitude was .92, a value less than one which
would indicate those with a more favorable attitude would be less
likely to vote, however, given the proximity to a value of one there
is room for speculation on this result (see Table 3 for a complete
list of odds ratios).
Table 3
Intention and TPB Predicting Vote Action
Vote Action
Variable B S.E. OR
Attitude -.09 .03 .92*
Subjective Norms .12 .06 1.11
Perceived Behavioral Control -.05 .05 .96
Intention .45 .13 1.57**
Χ2 39.26**
Note. N = 63.Predictor variables include overall scores for attitude,
subjective norms, perceived behavioral control and intention. OR =
Odds ratio.
*p < .05. **p < .001.
A second hierarchical multiple regression was conducted to assess the
interaction of certain personality factors with aspects of intention
(H2, H3, H4). Age, sex, affiliation and influence of participation
were entered at Step 1, explaining 31.8% of the variance in the
variance of intention to vote. After the entry of the centered values
of Attitude, Agreeableness, Conscientiousness, Extraversion and
Subjective Norms were entered at Step 2, the variance explained this
far in the model was 42.3%, F(8,53) = 4.86, p <.001. This step
accounted for an additional 10.6% of the variance in intention to
vote, R square change = .11, F change (4,53) = 1.91, p > .05. After
the interaction of Extraversion and Attitudes (H2), Subjective Norms
and Agreeableness (H3) and Subjective Norms and Conscientiousness (H4)
were entered at Step 3 the total variance explained by the model as a
whole was 44.2%, F(11,50) = 3.60, p <.01. The interaction of
personality factors and intention sub factors explained an additional
1.8% of the variance in intention to vote, R square change = .018, F
change (3,50) = .54, p >.05. The overall model was statistically
significant; however, the findings of the overall model in assessing
the ability of the interaction between specific personality factors
and aspects of intention were not statistically significant. In the
final model, the only variable that was found to be statistically
significant was affiliation, which was controlled for in Step 1 (β =
.47, p < .001) and not part of the interaction analysis (see Table 4
for complete list of beta levels and confidence intervals for the
variables in each model).
The correlation between party affiliation (being affiliated with a
party vs. not being affiliated) and intention was investigated using
Pearson r correlation (H5). There was a strong correlation between
affiliation and intention, r = .50, N = 63, p < .001, indicating that
those who are affiliated with a party would have higher intentions to
vote than those who were not affiliated with a political party.
Table 4
Interaction of Personality and TPB Components
Model 1 Model 2 Model 3
Variable β 95%CI β 95%CI β 95%CI
Age .26 [.04, .59] .15 [-.11, .46] .16 [-.10, .49]
Sex .05 [-2.36, 3.77] .05 [-2.33,3.80] .06 [-2.40, 4.00]
Affiliation .51* [5.08, 12.44] .48* [4.73, 12.44] .47 [4.38, 12.30]
Participation Influence -.03 [-1.31, 1.0]) -.06 [-1.52, .87] -.03
[-1.42, 1.08]
Attitude*Extraversion .12 [-.08, .25]
Subjective Norms*Conscientiousness -.08 [-.37, .18]
Subjective Norms*Agreeableness -.01 [-.29, .27]
Note. N = 63.Total scores for attitude, subjective norms and perceived
behavioral control. CI = confidence interval.
*p < .001.
Furthermore, a multiple hierarchical regression was conducted to
assess the ability of specific party affiliation (Democratic,
Republican, and others) to predict intention after controlling for age
and sex. The overall model was not significant, accounting for only
7.2% of the variance in intention to vote, F(4,57) = 1.12, p >.05. A
logistic regression was also conducted to assess the impact of
specific party affiliation (Democratic, Republican, and others) on
voting action. The model contained the specific party as a categorical
independent variable and voting action as the categorical dependent
variable. The model was not significant, Χ2(5, N = 63) = 3.36, p >.05.
Additional logistic regressions were conducted to assess the impact of
multiple other factors on vote action to see if other variables would
better predict voting action. The first model was set up to examine
any possible predictability the three personality factors associated
with the components of intention might have had on voting action. The
model contained three independent variables (Extraversion,
Agreeableness and Conscientiousness) and voting action as the
categorical, dependent variable. The model was not significant, Χ2(3,
N = 63) = 3.74, p >.05.
The second model examined the predictive ability of political party
affiliation on voting action, given its significant correlation with
both intention and voting action. The model contained political party
affiliation as the categorical independent variable, and voting action
as the dependent variable. The model was significant in predicting
voting action, Χ2(1, N = 63) = 3.83. p =.05. The significance level
was not very strong, and the model was only able to predict between
6.1% (Cox and Snell R square) and 8.2% (Nagelkerke R square) of the
variance in vote action and only classified 60.3% of the cases
correctly. Affiliation with a political party did not have a
significant odds ratio with a value of 4.71, p > .05.
Chapter 4
=========
DISCUSSION
The Theory of Planned Behavior was a disappointing predictor of
intention to vote and voting action and the influence of personality
did not prove to have any influence on the components of intention.
The application of the TPB in predicting voting should not be ruled
out in the future, given the limitations to this research.
The TPB was hypothesized to be able to predict intention to vote (H1)
and the act of voting (H1a) however, this hypothesis was not supported
by the data. The model developed to test this hypothesis assessed the
impact of attitudes, subjective norms and perceived behavioral control
on intention. The model as a whole was significant, which allows for
the speculation that with a greater sample size, the theory would be
able to predict intention to vote. In addition, intention to vote was
a significant predictor of voting action, which supports the TPB
overall, and the predictive ability of intention for the specific
situation of voting in the 2010 California General Election.
I also predicted that Perceived Behavioral Control would have had the
strongest influence on the act of voting (H1b) based on its unique
relation to actions outside of the model. This hypothesis was not
supported, and it was found that attitude, not perceived behavioral
control was a significant contributor to predicting vote action.
Previous research has argued that the nature of participation digs
into the inner workings of the person’s subjectivity, values and
motivation (Allport, 1954). It is conceivable that attitude would
relate to these factors and override the influence of the individual’s
control in being able to cast their vote given the fluctuation of
influence of the different components of intention depending upon the
behavior in questions (Ajzen, 1991).
Overall, the findings do not correspond with previous research which
has found the TPB to be a successful predictor of legislative voting
(Flynn, et al., 1997), continuing with an exercise regimen
(Chatzisrantis & Hagger, 2008), and even in predicting participation
in recreational activities like hunting (Hubres, Ajzen and Diagle,
2001). The lack of support for the predictive ability of the TPB
suggests that the findings are a result of the limited power of the
analyses used given the small sample size. Perhaps, if the sample size
were increased, the TPB-model would have found the TPB successful in
predicting intention to vote.
To further improve the use of the TPB in predicting intention to vote,
it is believed that the TPB questionnaire should be validated
extensively. In this specific study, four items from the PBC direct
measure were omitted. The items related to unanticipated events,
family obligations, work demands on time, and school demands on time
all inhibiting the participant’s ability to cast their vote in the
election. Also, two items from the salient attitude measure were
omitted—the items asked the participant if voting was important enough
to miss work, and important enough to miss out on social plans. All of
the items omitted were left out because they decreased the strength of
the measure and the remaining items had a strong enough reliability to
be used.
The hypotheses regarding personality having an interaction with
intention to vote were not supported. Extraversion did not have a
significant interaction with attitude in the predicting of intention
to vote, agreeableness did not have a significant interaction with
subjective norms in the predicting of intention to vote, nor did
conscientiousness. The analysis looked at the associations between the
personality factors as they interacted with components of intention.
Based on previous research findings these factors were related to
specific components of intention (Schoen and Schumann, 2007; Ajzen,
1991).
I believe that the lack of statistical support for the model can be
related to limitations in power of the analyses based on the lacking
sample size. Since the data was collected in such a short time span
the sample size was smaller than it would have been if the online data
collection site was available to the research pool at an earlier date.
It is also possible that students from CSUS psychology classes had not
yet begun to focus on their research participation and by the time
they got involved in projects from the research website, this research
option was no longer available since it ended a month before the
semester in which students needed to complete their participation. In
the future, the data collection should begin as soon as possible
before the election. If an undergraduate subject pool is used the
limitations of having to collect the data within a timeframe set by
the academic institution should be considered when determining the
time frame of collecting responses.
Although the research is limited in this area, by understanding that
personality is a guiding force in political party affiliation based on
the research noted above, it is suggested that this research be taken
seriously and utilized in the future. By understanding how personality
can influence intention to vote, and through which specific aspects of
intention (Attitude, Subjective Norms, and Perceived Behavioral
Control) researchers might be able to understand how politicians can
get individuals motivated to vote better than before.
The research of political psychology has focused on personality in
regards to specific political party affiliation has found strong
support for the belief that personality is a strong predictor of which
political party an individual will belong to (Schoen and Schumann,
2007; Caprara, Barbaranelli and Zimbardo, 1999; Caprara & Zimbardo,
2004; Digman, 1990) however, the use of personality with regard to
predicting intention to vote is novel. The model laid out in this
research should not be discredited, however, because the relation
between the two is highlighted in political psychology research even
though it is not studied directly.
The hypothesis that political party affiliation would have a strong
relation to intention with vote was supported by the data. The
relation between affiliation and vote action approached significance,
and with a greater sample size would likely be supported. Specific
party affiliation was also assessed for an influence on intention and
vote action and in both cases, no specific party affiliation was found
to be a significant predictor. Previous research has found that
personality traits influenced political party preference (Caprara, et
al., 1999) but has not focused on specific party membership predicting
intention to vote or voting altogether. Since this research focused on
personality and intention, affiliation was used controlled for to
eliminate any possible effect it could have had on intention to better
understand the components of intention and personality. However, it is
possible that voter turnout is based solely on the items on the ballot
and in the future intention could be assessed in relation to specific
ballot items.
It is important to note, that given the specific nature of the Theory
of Planned Behavior Questionnaire, these predictions are based only on
voting behavior in the 2010 California general election and that if
voting in any other election were assessed, the three components of
intention might have different influence on intention and behavior. As
a result of the specificity of the questionnaire, it is feasible that
in some uses of the TPB attitudes have the most significant impact on
intentions, or that attitudes and perceived behavioral control account
for intentions, and yet for others all three independently contribute
to the resulting intention (Ajzen, 1991). The models developed here
lay the framework for future research to be conducted in forthcoming
elections to see if with a greater sample size, and a new version of
the TPB questionnaire, personality interacts with intention to vote
and voting action.
REFERENCES
Allport, Gordon (1945). The Psychology of Participation. The
Psychological Review. 53, 117-131.
Ajzen, I. (1991). The theory of planned behavior. Organizational
Behavior and Human Decision Processes, 50, 179-211.
Allport, Gordon (1945). The Psychology of Participation. The
Psychological Review. 53, 117-131.
Barnea, Marina F., & Schwartz, Shalom H. (1998). Values and Voting.
Political Psychology. 1, 17-40.
Caprara, Gian V. , & Zimbardo, Phillip G. (2004). Personalizing
Politics: A Congruency Model of Political Preference. American
Psychologist. 59, 581-594.
Caprara, G.V., Barbaranelli, C., & Zimbardo, P.G. (1999). Personality
profiles and political parties. Political Psychology, 20, 175-197.
Chatzisarantis, N. L. D., & Hagger, M. S. (2008). Influences of
personality traits and continuation intentions on physical activity
participation within the theory of planned behaviour. Psychology and
Health, 23, 347-367.
Digman, J. M. (1990). Personality structure: Emergence of the
five-factor model. Annual Review of Psychology, 41, 417-440.
Flynn, B. S., Goldstein, A. O., Solomon, L. J., Bauman, K. E.,
Gottlieb, N. H., Cohen, J. E., Munger, M. C., & S., D. G. (1998).
Predictors of state legislators' intentions to vote for cigarette tax
increases. Preventive Medicine, 27, 157-165.
Harder, Joshua, & Krosnick, Jon A. (2008). Why Do People Vote? A
Psychological Analysis of the Causes of Voter Turnout. Journal of
Social Issues. 64, 525-549.
Hrubes, D., Ajzen, I., & Daigle, J. (2001). Predicting hunting
intentions and behavior: An application of the theory of planned
behavior. Leisure Sciences, 23, 165-178.
John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to
the integrative Big Five trait taxonomy: History, measurement, and
conceptual issues. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.),
Handbook of personality: Theory and research (pp. 114-158). New York,
NY: Guilford Press.
John, O.P., & Srivastana, S. (Ed.). (1999). Handbookof personality:
theory and research. New York, New York: The Guildord Press, pp.
102-138.
McCrae, R. R. (1996). Social consequences of experiential openness.
Psychological Bulletin, 120, 323-337.
Schoen, Harald, & Schumann, Seigfreid (2007). Personality Traits,
Partisan Attitudes, and Voting Behavior. Evidence from Germany.
Political Psychology. 4, 471-498.
viii