University of Southern California Dissertations and Theses

A closer look at intention to use condoms and risky sexual behaviors among homeless people: application of the theories of reasoned action and planned behavior

(USC Thesis Other)

A closer look at intention to use condoms and risky sexual behaviors among homeless people: application of the theories of reasoned action and planned behavior

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running head: RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE

A CLOSER LOOK AT INTENTION TO USE CONDOMS AND RISKY SEXUAL
BEHAVIORS AMONG HOMELESS PEOPLE: APPLICATION OF THE THEORIES OF
REASONED ACTION AND PLANNED BEHAVIOR

by
Ahyoung Song

______________________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(SOCIAL WORK)
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE

ii
TABLE OF CONTENTS

ABSTRACT ................................................................................................................................... vi

CHAPTER 1. INTRODUCTION ................................................................................................... 1
Background ..................................................................................................................................... 1
Study Aims...................................................................................................................................... 3
Aim 1 .................................................................................................................................. 3
Aim 2 .................................................................................................................................. 4

CHAPTER 2. THEORIES OF REASONED ACTION AND PLANNED BEHAVIOR .............. 6

CHAPTER 3. PSYCHOLOGICAL AND PARTNER-RELATED FACTORS ASSOCIATED
WITH CONDOM USE AMONG HOMELESS WOMEN WITH INTENTION TO USE A
CONDOM: AN EVENT-BASED APPROACH ............................................................................ 9
Background ..................................................................................................................................... 9
Present Study .................................................................................................................... 13
Methods......................................................................................................................................... 15
Study Sample .................................................................................................................... 15
Measures ........................................................................................................................... 16
Analytic Plan ..................................................................................................................... 19
Results ........................................................................................................................................... 20
Sample Characteristics ...................................................................................................... 20
Women with and without Intention .................................................................................. 23
Logistic Regression of Condom Use ................................................................................ 26
Discussion ..................................................................................................................................... 29
Strengths and Limitations ................................................................................................. 33
Conclusion ........................................................................................................................ 35

CHAPTER 4. A MULTILEVEL PATH ANALYSIS OF CONSISTENT CONDOM USE
AMONG HOMELESS PEOPLE: AN APPLICATION OF THE THEORIES OF REASONED
ACTION AND PLANNED BEHAVIOR .................................................................................... 36
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE iii
Background ................................................................................................................................... 36
Present Study .................................................................................................................... 41
Methods......................................................................................................................................... 42
Study Sample .................................................................................................................... 42
Homeless women. ............................................................................................................. 42
Homeless men. .................................................................................................................. 43
Sampling for multilevel modeling. ................................................................................... 44
Study Variables ................................................................................................................. 44
Analytic Plan ..................................................................................................................... 47
Results ........................................................................................................................................... 50
Descriptive Characteristics ............................................................................................... 50
Between and Within Variance of Within-Group Variables .............................................. 53
Model Prediction Results .................................................................................................. 56
Significance Tests of Indirect Effects ............................................................................... 60
Discussion ..................................................................................................................................... 61
Strengths and Limitations ................................................................................................. 65

CHAPTER 5. CONCLUSIONS AND IMPLICATIONS ............................................................ 67

REFERENCES ............................................................................................................................. 70

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE iv
LIST OF TABLES
Table 1 Descriptive Statistics of Weighted Sample: Psychological, Relationship, and Contextual
Factors during the most Recent Sexual Event (N = 445) .................................................. 21
Table 2 Group Difference Tests by Women's Intention to Use a Condom .................................. 24
Table 3 Logistic Regression of Condom Use among Women with Intention to Use a Condom . 27
Table 4 Descriptive Statistics of Sample by Gender .................................................................... 51
Table 5 Between and Within Variance of Within-Group Variables among Homeless Men ........ 53
Table 6 Between and Within Variance of Within-Group Variables among Homeless Men ........ 55
Table 7 Model Prediction Results for Homeless Women ............................................................. 58
Table 8 Model Prediction Results for Homeless Men .................................................................. 59
Table 9 Significance Test of Indirect Effects ............................................................................... 60

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE v
LIST OF FIGURES

Figure 1 Theory of Reasoned Action .............................................................................................. 7
Figure 2 Theory of Planned Behavior ............................................................................................. 7
Figure 3 A Path Model to Predict Condom Use among Homeless People ................................... 42
Figure 4 Model Prediction of Homeless Women with Odds Ratios ............................................. 56
Figure 5 Model Prediction of Homeless Men with Odds Ratios .................................................. 58

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE vi
ABSTRACT
This dissertation aimed to examine factors associated with condom use among homeless
people based on the theory of reasoned action (TRA) and theory of planned behavior (TPB). To
achieve this goal, two separate studies were conducted.
The first study aimed to identify factors associated with condom use among homeless
women who had intention to use a condom during their most recent sexual event using an event-
based approach. This study featured a randomly collected sample of 445 women recruited from
temporary shelters in Los Angeles. Women were divided by their intention to use a condom
during their most recent sexual event. To identify factors that differed by group, a series of
group-difference tests (t-test and chi-square test) were conducted. Results indicated that women
with intention were more likely to use condom and showed more positive attitudes toward
condom use. Women who did not have intention to use a condom were more likely to be in a
committed or long-term relationship compared to women who had intention to use a condom.
Women who had intention to use a condom reported stronger perceptions that their partner
intended to use a condom. Furthermore, they were more likely to have discussed condom use
with their partner than women who did not have intention. Women in the two groups (intention
vs. no intention) showed significant differences in partner type, psychological violence
victimization, pressured sex, need-based sex, and condom availability. To identify which
variables were associated with condom use among homeless women, a logistic regression was
conducted. The group of women without intention was excluded in logistic regression due to low
observations for several study variables. Logistic regression results indicated that for women
who had intention to use a condom, partner-related factors were more important than
psychological factors in terms of explicating their condom use. Women used condoms less often
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE vii
with partners in committed relationships. However, they were more likely to use condoms when
their partners were perceived to have intention to use a condom and when they had need-based
sex. Furthermore, if condoms were available and women communicated about condoms with
their partner, it increased the likelihood of actual condom use among women who had intention
to use a condom. To be more effective, HIV prevention interventions to promote condom use
should focus on intention due to its association with increased condom use according to group-
difference tests. There were some partner-related factors that promoted or decreased condom use
among women with condom use intention. Focusing on those factors in interventions would help
promote consistent condom use. Practical and research implications and study strengths and
limitations are discussed.
The second study focused on testing a model guided by TRA and TPB to explain
consistent condom use among homeless men and women. The study model was constructed
including condom attitudes, subjective norms, HIV vulnerability, intention to use a condom, and
consistent condom use with sexual partners during the previous 6 months. In addition, to
examine factors quantifying the association between intention and condom use, two moderators
were included: primary partner status and suggesting condom use. Using multilevel path
modeling with multilevel structural equation modeling (MSEM), two separate path analyses
were conducted with data for homeless men and women. Men’s data were collected randomly
via meal lines in the Los Angeles area. MSEM results indicated strong associations between
intention and condom use regardless of gender. For women, intention was predicted by
subjective norms, whereas condom attitudes prediction intention for men. Furthermore, no
proposed moderators were associated with condom use among women, but for men there were
significant association between moderators and condom use. The results of this study indicated
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE viii
that increasing intention to use a condom could be a key to promote consistent condom use
among homeless people, regardless of gender. However, to improve intention to use a condom,
strategies may need to be different by gender. For men, addressing condom attitudes would be
helpful, whereas for women, addressing subjective norms may promote consistent condom use.
Furthermore, for men, addressing partner type and efforts to suggest condom use would improve
the association between intention and condom use. Practical and research implications and study
strengths and limitations are discussed.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE

1
CHAPTER 1. INTRODUCTION
Background
Homelessness exposes individuals to substantial health-threatening risks (American
Psychological Association, 2013). HIV infection is among the main concerns because of its
prevalence; evidence has consistently indicated that HIV infection is dramatically more prevalent
among homeless people than general population (Beijer, Wolf, & Fazel, 2012; Song, 1999).
According to the National Coalition for the Homeless (2009), the HIV infection prevalence rate
among homeless people was 3.4% compared to 0.4% in the general population. Frequent risky
sexual behavior among homeless people contributes to HIV infection. In particular, unprotected
sex is one of the leading causes of HIV/AIDS (Centers for Disease Control and Prevention
[CDC], 2013a). Therefore, promoting consistent condom use has been a main goal of many HIV
prevention interventions.
To promote consistent condom use, understanding important factors associated with
consistent condom use is needed. Some theories provide direction on how to achieve that goal,
helping to clearly describe the relationship between factors and condom use. Furthermore, good
theories help frame effective interventions and design evaluation. Among theories explaining
factors related to consistent condom use, many previous studies have focused on the theory of
reasoned action (TRA) and the theory of planned behavior (TPB; Albarrací n, Johnson, Fishbein,
& Muellerleile, 2001). Many HIV prevention interventions have been designed based on these
theories, which have proven useful in this context (Albarrací n et al., 2005; Basen-Engquist, 1994;
Sanderson & Jemmott, 2006).
Both TRA and TPB assume that an individual’s intention and willingness to engage in a
behavior is a strong motivation for actual behavior. For example, individuals with strong
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 2
intentions to use condoms are more likely to use a condom during sexual events (Albarrací n et
al., 2001). HIV prevention interventions designed based on those theories usually expect
participants to increase their consistent condom use mainly by promoting their intention to use a
condom (Albarrací n et al., 2005; Basen-Engquist, 1994; Sanderson & Jemmott, 2006).
According to TRA, intention is predicted by attitudes and subjective norms regarding a behavior
(Ajzen, 1985). As an extended form of TRA, TPB includes individual perceptions of control that
predict intentions, which in turn lead to behavior. TRA and TPB have been widely applied to
explain a range of health behaviors, including substance use, exercise, smoking, and HIV and
sexually transmitted disease (STD) prevention behaviors (Albarrací n et al., 2001; Bandawe &
Foster, 1996; Blue, 1995; Bosompra, 2001; Morrison, Spencer, & Rogers Gillmore, 1998). TRA
and TPB have shown considerable relevance for understanding engagement in unprotected sex
by individuals (J. D. Fisher & Fisher, 2000; W. A. Fisher, Fisher, & Rye, 1995). In a meta-
analysis exploring TRA and TPB, results indicated that factors related to TRA and TPB were
plausible predictors of unprotected sex among diverse populations (Albarrací n et al., 2001).
Despite the well-recognized usefulness of TRA and TPB in terms of understanding
condom use, few studies have used these theories to explain condom use among homeless people.
Several studies partially relied on these theories to explain condom use among homeless people
by including a few variables from TRA and TPB (Kennedy, Wenzel, Brown, Tucker, & Golinelli,
2013; Liverpool, McGhee, Lollis, Beckford, & Levine, 2002; Somlai, Kelly, Wagstaff, &
Whitson, 1998; Sugerman, Hergenroeder, Chacko, & Parcel, 1991; Tucker et al., 2013).
However, most previous studies included only one or two variables from TRA and TPB. Thus, it
remains unclear if TRA and TPB have plausible applicability to explain unprotected sex among
individuals experiencing homelessness. Furthermore, previous studies guided by TRA and TPB
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 3
have various methodological limitations. Critically examining these theories may result in
strategies to improve condom use among homeless individuals.
Due to the limitations of previous studies, this dissertation focused on examining TRA
and TPB using data collected from homeless men and women in Los Angeles County, CA. This
study explored the association of core variables of these theories with unprotected sex among
homeless people who had heterosexual relationships. Previous studies on unprotected sex have
largely focused on a particular population at significant risk of HIV infection: men who have sex
with men. As a result, fewer studies have examined the HIV risks of heterosexual individuals.
According to a recent report, heterosexual intercourse is the main cause of 15% of newly
diagnosed HIV infections among men (CDC, 2013a). For women, unprotected heterosexual sex
is particularly risky; it is their leading cause of HIV infection (CDC, 2013b). Previous studies
have shown that most sexual interactions among homeless individuals are heterosexual in nature
and most homeless people identify as heterosexual (Kennedy, Wenzel, et al., 2013; Tucker et al.,
2013; Valente & Auerswald, 2013). Therefore, focusing on homeless men and women who had
heterosexual relationship with regard to HIV infection among homeless adults is imperative.
Study Aims
Aim 1
The first aim of this dissertation was to examine factors associated with condom use
among homeless women using an event-based approach. To accomplish this aim, factors were
examined in terms of how their association with condom use differed by intention to use a
condom among women. Women were divided into two groups: women who intended to use a
condom and women who did not intend to use a condom during their most recent sexual event.
Psychological, relationship-related, and contextual factors during the most recent sexual event
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 4
were included in study model. Because this study focused on participants’ most recent sexual
event, an event-based approach was used. This approach helps capture relationship dynamics and
contextual factors in terms of condom use, and has been increasingly applied in the study of
sexual behaviors. Condom use among women in particular needs to be examined in the context
of relationship dynamics because it is difficult for women to abide by their determination to use
condoms regardless of their partner’s reaction. First, group-difference tests of each study variable
were conducted. Then, a logistic regression with women who had intention to use a condom at
the event was analyzed to examine the associations of factors with condom use. More details
regarding this study can be found in Chapter 3.
Aim 2
This aim involved testing a conceptual path model derived from TRA and TPB with two
moderating predictors of gender and partner type to predict condom use among homeless adults
using a multilevel mediation approach. TRA and TPB have been relatively less studied in
conjunction with homeless populations. Thus, it remains unclear whether core concepts of TRA
and TPB can be used to successfully predict condom use among homeless people. Given this
background, a study model was developed to examine whether predictors from TRA and TPB
successfully explained condom use among homeless people. Condom attitudes, HIV
susceptibility, subjective norms regarding risky sexual behaviors, and intention to use a condom
were included in the model. Intention was included as a mediator between intention and condom
use. Additionally, this study included partner type (primary vs. nonprimary) and suggesting the
use of a condom to sexual partners as moderators that may differentiate the association of
intention with condom use. To test if the associations between variables varied by gender, the
study model was tested separately with homeless men and homeless women. This study adopted
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 5
multilevel path modeling, which has strength in analyzing nested data to overcome the
methodological limitations of previous studies. A detailed description of this study can be found
in Chapter 4.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 6
CHAPTER 2. THEORIES OF REASONED ACTION AND PLANNED BEHAVIOR
TRA (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975) is one of the main psychological
theories for predicting individual behaviors. One notable concept of TRA is its emphasis on the
importance of intention, a strong predictor of individual behaviors (Fishbein & Ajzen, 1975).
The theory proposes that intention serves as a proximal indicator that directly guides behaviors.
Individual intention to engage in a certain behavior is a function of several psychological traits,
according to TRA. First, individual intention can be determined by attitudes toward behaviors.
More positive attitudes toward behaviors are related to increased intention. The second
determinant of behavioral intention is subjective norms, which refer to individual perceptions of
social pressure to perform a certain behavior. This construct often has been operationalized as
how individuals perceive significant others in their social networks think or engage in certain
behaviors (Sheeran & Taylor, 1999). Fishbein and Ajzen (1975) described subjective norms as
social influences, which can be interpreted as social pressure or commitment to engage in a
behavior. Based on this description, subjective norms have been measured by asking respondents
about how many of their social network members engage in a certain behavior. Others have
recommended including descriptive subjective norms, or perceptions of the attitudes and
behaviors of significant others (Sheeran & Taylor, 1999). This can be measured by asking
respondents about their perceptions of the attitudes of peers or others toward a certain behavior.
In sum, according to TRA, individual behaviors can be determined by intention to perform the
behavior and intention can be determined by attitudes toward a behavior and subjective norms.
In terms of understanding condom use, TRA can be constructed as outlined in Figure 1.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 7
Figure 1
Theory of Reasoned Action

In 1985, Ajzen proposed another health psychological model, TPB, which is an extended
form of TRA. In TBP, an individual’s perceived ability to control a behavior is added as a
predictor of intention. Unlike two other psychological variables (attitudes toward a behavior and
subjective norms), a direct association between perceived behavioral control and condom use
was allowed in TPB. In both TRA and TPB, two psychological predictors, attitudes and
subjective norms, have only indirect associations with condom use through intention. When
applied to condom use, TPB can be constructed as illustrated in Figure 2.
Figure 2
Theory of Planned Behavior

In addition to efforts to integrate other psychological models into TRA and TPB, there
have been concerns that TRA and TPB have neglected important factors other than
psychological factors. As described by Fishbein and Middlestadt (1989), TRA and TPB were
developed based on the assumption that individual behaviors can be understood based on a set of
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 8
psychological processes. Researchers who study sexual decision making and HIV preventive
behaviors, however, argued that HIV preventive behaviors, particularly condom use, require
understanding interpersonal structure (Kippax & Crawford, 1993). Because consistent condom
use is fundamentally based on a decision between partners, by not incorporating relationship-
related factors, Kippax and Crawford, 1993) argued that theories only focused on psychological
traits, such as TRA and TPB, have inherent limitations in terms of explaining condom use.
In sum, TRA and TPB were constructed using psychological traits to explain individual
health behaviors. Both emphasize the role of intention and propose several psychological traits
that predict intention. According to these theories, attitudes toward a behavior and subjective
norms are psychological factors that predict intention. TPB added one additional factor,
perceived behavioral control, as an extension of TRA. Furthermore, there is an apparent
necessity to consider other psychological factors such as perceived HIV susceptibility and
relationship-related factors to strengthen TRA and TPB, particularly when applying these
theories to HIV preventive behaviors. TRA and TPB served as the main theoretical framework
for this dissertation, which aimed to identify factors explaining condom use among homeless
people.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 9
CHAPTER 3. PSYCHOLOGICAL AND PARTNER-RELATED FACTORS
ASSOCIATED WITH CONDOM USE AMONG HOMELESS WOMEN WITH
INTENTION TO USE A CONDOM: AN EVENT-BASED APPROACH
Background
One in 4 people living with HIV in the United States are women (CDC, 2013b).
Furthermore, HIV is one of the top 10 causes of death for African American and Hispanic
women (CDC, 2013b). Unprotected heterosexual intercourse with HIV-infected men is the main
cause of HIV infection among women (CDC, 2013b). However, it is often difficult for women to
detect their male partner’s HIV and STD status. Furthermore, unprotected vaginal sex is
considerably riskier in terms of contracting HIV for women than men (CDC, 2013b). Thus,
consistently using a condom remains a major strategy to prevent the spread of HIV among
women other than abstinence (CDC, 2013a). There have been efforts to find effective predictors
of condom use among women (Calsyn et al., 2013; Kennedy et al., 2010; Pulerwitz, Amaro, De
Jong, Gortmaker, & Rudd, 2002; Wechsberg et al., 2003; Wingood & DiClemente, 1998).
Being homeless is intricately associated with HIV. A previous study showed that the
prevalence of HIV among homeless people was at least 3 times higher than among housed
people (National Coalition for the Homeless, 2007). Thus, the intersection of being homeless and
being a woman may increase vulnerability to HIV infection. Homeless women engage in risky
sexual behaviors, such as having unprotected sex, engaging with multiple sexual partners, and
using substances (Nyamathi & Vasquez, 1989; Ryan et al., 2009; Sikkema et al., 1996; Wenzel
et al., 2004). Given the vulnerability to HIV infection of homeless women, efforts to identify
factors promoting consistent condom use are highly needed.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 10
Traditionally, individual psychological factors associated with condom use have been the
main focus of research. Sufficient evidence on the association between condom attitudes and
condom use has been accumulated (Bogale, Boer, & Seydel, 2010; Campbell et al., 2011; Heeren,
Jemmott, Madeya, & Tyler, 2007). Condom efficacy and perceived HIV susceptibility have also
been important psychological factors related to condom use (Lau, Li, Choi, & Gu, 2014;
Widman, Golin, Grodensky, & Suchindran, 2013).
Along with these psychological characteristics, TRA and TPB have also been applied to
individuals’ intention to use a condom. According to TRA and TPB, intention is the most
proximal predictor of health behaviors such as condom use (Ajzen & Fishbein, 1977; Ajzen &
Madden, 1986). Most studies of intended and actual condom use have relied on a strong
assumption that intention is an immediate proximal cause of individual behaviors. This belief led
researchers to focus largely on intention as a strategy to promote condom use among individuals
(Bryan, Aiken, & West, 1996; Couture, Soto, Akom, Joseph, & Zunzunegui, 2010; Fazekas,
Senn, & Ledgerwood, 2001; Greene, Hale, & Rubin, 1997; Jemmott & Jemmott, 1991). Previous
studies showed that intention is a plausible predictor to explain condom use with a variety of
populations (Albarrací n et al., 2001; Bandawe & Foster, 1996; Blue, 1995; Bosompra, 2001; J. D.
Fisher & Fisher, 2000; W. A. Fisher, Fisher, & Rye, 1995; Morrison et al., 1998). Thus, many
HIV prevention interventions were designed to increase intention to use a condom to promote
consistent condom use (Albarrací n et al., 2005; Basen-Engquist, 1994; Sanderson & Jemmott,
2006). However, intention to use a condom among homeless women has been largely neglected
in studies examining factors associated with condom use. Only a few studies have focused on
intention to use condoms as an important predictor of homeless women’s condom use (Rew,
Fouladi, & Yockey, 2002; Somlai et al., 1998; Stark et al., 1998).
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 11
Despite its importance in understanding condom use, deciding to use a condom during a
sexual event may not be a sole function of intention. Particularly for women, sexual decision
making, including decisions regarding the use of a condom, is a fairly complex process
associated with relationship dynamics (Widman, Golin, & Noar, 2013). Responding to this
argument, previous studies highlighted partner-related factors that influence condom use among
women.
Women’s condom use was associated with their partner’s reaction and intention to use
condoms (Bowleg, Lucas, & Tschann, 2004; Crosby et al., 2002; Gómez & Marí n, 1996;
VanderDrift, Agnew, Harvey, & Warren, 2013). Although previous studies did not directly
measure intention of sexual partners, they included similar concepts such as negative attitudes
and condom preferences of women’s sexual partners and achieved similar results. Negative
reactions by partners regarding condom use may decrease the likelihood that women will use a
condom (Cabral, Pulley, Artz, Brill, & Macaluso, 1998; Plichta, Weisman, Nathanson,
Ensminger, & Robinson, 1992; Wingood & DiClemente, 1998). However, have a male partner
who preferred to use condoms increased consistent condom use among women (Plichta et al.,
1992). Partner-related factors can strongly predict condom use among women. For instance,
dislike of condoms and feeling accused of having an STD among male partners decreased the
likelihood of condom use among women (Crosby et al., 2002).
Other than male partners’ intention to use a condom, some other relationship
characteristics with sex partners have been studied as predictors of consistent condom use among
women. Type of partner (e.g., casual or committed) also has been associated with condom use
among women. Plichta et al. (1992) showed that women reported more consistent condom use
with casual (short-term) sexual partners; similar results were found in a qualitative study with
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 12
low-income African American women (Noar et al., 2012). Other than partner type, commitment
to the relationship was another relationship-related factor associated with condom use (Bauman
& Berman, 2005; Civic, 1999; Wingood & DiClemente, 1998, 2000), as is the length of the
relationship (Catania, Coates, & Kegeles, 1994; Civic, 1999; Matson, Chung, Sander, Millstein,
& Ellen, 2012; Tucker et al., 2009).
Other relationship dynamics related to condom use among women have been studied
(Cabral et al., 1998; Kalichman, Williams, Cherry, Belcher, & Nachimson, 1998; Maman,
Campbell, Sweat, & Gielen, 2000; Noar, Carlyle, & Cole, 2006; Wingood & DiClemente, 1998).
Some studies have shown that communication with sexual partners regarding condoms may help
women use condoms with partners. Women who engage in more communication about HIV and
condoms with sexual partners are more likely to use condoms more frequently (Moore, Harrison,
Kay, Deren, & Doll, 1995). Research has shown a potential association between violent
victimization and condom use (Kalichman et al., 1998; Maman et al., 2000). Violence by sexual
partners may increase women’s risky sexual behavior, including not using condoms through
forced or coercive sexual intercourse, limiting their ability to negotiate condom use (Kalichman
et al., 1998). Women in violent relationships have reported fear of being beaten when they ask
their partner to use condoms (Kalichman et al., 1998).
For homeless women, partner-related factors could be important to condom use. With the
understanding that consistent condom use is the most effective HIV and STD prevention
behavior, studies have investigated condom use and predictors of condom use among homeless
women (Kennedy et al., 2010; Rew et al., 2002; Somlai et al., 1998; Stark et al., 1998;
Wechsberg et al., 2003; Wenzel et al., 2004). Results of studies examining partner-related factors
associated with condom use among homeless women have been fairly consistent with the results
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 13
of studies with other female populations. Partner type and communication were both associated
with condom use among homeless women (Kennedy et al., 2010). Homeless women are less
likely to use condoms with their primary (long-term) partner than casual partners. Engaging in
HIV communication is a protective factor that increases condom use among homeless women
(Kennedy et al., 2010). Substance use during sexual events has been shown to be an important
contextual predictor that decreases the likelihood of condom use by homeless women (Tucker et
al., 2012) and in other populations (Bailey, Camlin, & Ennett, 1998; Leigh et al., 2008;
Weinhardt & Carey, 2000).
Although previous studies have repeatedly shown that partner-related factors should be
considered when exploring condom use among homeless women, it is unclear if those partner-
related predictors operate similarly to explain condom use among women who have intention to
use condoms. Considering that HIV prevention interventions have focused on improving
intention to use a condom, identifying factors associated with condom use when women have
intention to use a condom would provide evidence to strengthen the effectiveness of HIV
interventions. Thus, this study focused on factors associated with condom use among homeless
women when they intended to use a condom.
Present Study
The present study aimed to identify factors associated with condom use among women
who had intended to use a condom during their most recent sexual event. This study included
individual psychological factors and partner-related factors. Partner-related factors were divided
into two categories: relationship quality and event-specific relationship factors. Relationship
quality factors were the general quality or characteristics of relationships with sex partners.
Event-specific relationship factors were relationship and partner-related factors related to the
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 14
most recent sexual event. Condom attitudes, condom efficacy, and HIV susceptibility were
included as psychological factors. Relationship quality factors were length of relationship and
commitment of participants to the relationship. Event-specific relationship factors were partner
type, perceived partner intention, need-based sex, condom availability, communication about
condom se, pressured sex, psychological violence, and substance use.
Because this study focused on the most recent sexual event, it adopted an event-based
approach that may improve the methodological power of condom use research that assesses
relationship dynamics in sexual decision making. Event-based analysis is preferred to other
methods because it ensures that predictors and condom use occurred at the same time. Compared
to examining the general associations between predictors and condom use, this method increases
causal inferences between predictors and condom use (Leigh, 2002). This approach may reveal
rich information regarding relationship dynamics and condom use, leading to a better
understanding of partner- and relationship-related determinants of condom use. Researchers have
increasingly recognized the usefulness of event-based approaches and have successfully applied
these methods to predict condom use among individuals (LaBrie, Earleywine, Schiffman,
Pedersen, & Marriot, 2005; Testa & Collins, 1997; Tucker et al., 2012; VanderDrift et al., 2013).
Event-based approaches, however, have been mostly used to explore the role of substance use in
condom use. Relatively fewer studies have explored partner-related barriers by using event-based
approaches. For example, Tucker et al. (2012) examined the association between substance use
and condom use during sexual events among homeless women, but did not explore whether the
intentions of women and those of their partners played a role in determining condom use. Event-
level analysis helps to examine particular dynamics of sexual relationships that may determine
condom use. To address these issues, this study used an event-based analysis to address the
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 15
associations of predictors with condom use among homeless women, particularly those who
intended to use a condom.
Methods
Study Sample
Participants in this study were 445 homeless women who were randomly sampled and
interviewed in temporary shelter settings in the central region of Los Angeles County during a
previous study of the social context of substance use and sexual risk (Wenzel et al., 2009). This
study recruited women using shelter-facility-based sampling approaches which have been used
to investigate substance use among homeless women in previous studies (Nyamathi, Leake,
Keenan, & Gelberg, 2000; Robertson, Zlotnick, & Westerfelt, 1997). To randomly select women,
stratified random sampling was used through 51 eligible facilities. The 51 selected facilities
served as sampling strata. A modified approach was used and corrected with sampling weights
(see Wenzel et al., 2009 for more information).
The survey was conducted between June 2007 and March 2008. Women were eligible to
participate if they were at least 18 years old, had vaginal or anal sex with a male partner during
the previous 6 months, were able to speak and understand English, and did not have significant
cognitive impairment. Of 472 women assessed eligible, 451 women were interviewed. Of these
451 women, five women were excluded because they only had oral sex with a partner and one
woman was excluded because she did not complete the questionnaires. Eligible women were
asked to complete a personal network survey (Green et al., 2012; Wenzel et al., 2009).
Computer-assisted, face-to-face interviews were completed with 445 women, each lasting
approximately 1 hour and 15 minutes and conducted by trained female interviewers. Women
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 16
who completed the interview were paid $20 each. The final sample size was 445 women, and the
completion rate was 94% (445 of 472).
Measures
Dependent variable.
Condom use. The dependent variable of this study was whether women consistently and
properly used a condom during vaginal or anal intercourse during their most recent sexual event
with a male partner. Women were asked to identify the most recent sexual event during which
they had vaginal or anal intercourse. To assess consistent condom use during the most recent
sexual event, three items were used. First, women were asked if they had vaginal sex more than
once during the event. If women had sex once, they were asked if they used a condom during
sexual intercourse. If women had sex more than once, they were asked if they used condoms
every time. Consistent condom use was coded as 1 and inconsistent condom use was coded as 0.
The final condom use outcome was dichotomized (0 = no condom use, 1 = consistent condom
use).
Intention to use a condom. Information on women’s intention to use a condom was used
to identify women who wanted to use a condom during their most recent sexual event. Women’s
intention to use a condom was measured by one item: “Did you want to use a condom?”
According to Ajzen (1991), this item may not be the best measurement of intention, which he
argued should be an indication of “how hard people are willing to try, of how much of an effort
they are planning to exert, in order to perform the behavior” (p. 181). Although a similar item, “I
want to use a condom every time I have sex,” was used to measure intention to use a condom in
a previous study (Roth, Ngugi, & Jansson, 2011), this issue is discussed later as a limitation of
this study.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 17
Independent variables.
Condom attitudes. Attitudes toward condom use were assessed by eight items:
“Condoms are too much trouble,” “You would be comfortable suggesting that you/you’re your
partner use a condom,” “Condoms are unreliable—sometimes they work, and sometimes they
don’t,” “You would be embarrassed to use a condom,” “Condoms are uncomfortable for both
partners,” “Condoms do not interfere with the enjoyment of sex,” “Condoms ruin sex,” and “It
would be OK with you if your partner suggested that he use a condom.” Those items were
adopted by Bogart et al. (2005). Each item was rated on a 4-point scale (1 = strongly disagree to
4 = strongly agree). Cronbach’s alpha for this scale was .73.
HIV susceptibility. Perceived HIV susceptibility was measured with one item: “It would
be easy for you to get the HIV infection or AIDS.” Responses were ranked on a 4-point scale (1
= strongly disagree to 4 = strongly agree). This item was developed by Gibson, Lovelle-Drache,
Young, & Chesney (1992).
Condom efficacy. Condom efficacy was assessed by three items, modified from the
refusal assertiveness subscale of the Sexual Assertiveness Scale (Morokoff et al., 1997) and
adopted in previous studies (Kennedy et al., 2010). The items were: “Give in and have sex if he
pressured you, even if you already said no?” “Have sex if he wanted you to, even if you didn’t
want to?” “Refuse to have sex even if he insisted?” Responses were rated on a 3-point scale (1 =
unlikely, 2 = somewhat likely, 3 = very likely). Cronbach’s alpha was .87.
Length of relationship. Women were asked to describe the length of relationship with
their partner during the most recent sexual event, measured in months.
Commitment to relationship. Commitment to the relationship with the partner with
whom women had sex most recently was measured by five items developed by Tucker, Elliot,
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 18
Wenzel, & Hambarsoomian (2007): “You are extremely committed in this relationship,”
“Generally speaking, you have invested a great deal in this relationship,” “You feel very
dependent on this relationship,” “Your life would be very disrupted if this relationship ended,”
and “You depend on this person for money, food, a place to stay, drugs, or something else.” Each
item was rated on 4-point scale (1 = strongly disagree to 4 = strongly agree). Cronbach’s alpha
was .85.
Partner type. Women were asked whether their most recent sexual partner was a current
sexual partner and whether they would identify the partner as a primary or casual partner.
Primary partner included a husband, boyfriend, or another man described as a “steady” partner.
Casual partner was defined as a partner who was not a long-term boyfriend or husband but rather
a casual “in the moment” or “just for fun” sexual partner.
Perceived partner intention. Perceived partner intention to use a condom was measured
by asking women one item: “Did he want to use a condom that time?” The answer was
dichotomized (0 = no, 1 = yes). Similar to the measurement of women’s intention to use a
condom, this may not be the best way to assess perceived partner intention. This is also stated as
a limitation in the discussion section.
Condom communication. Women were asked if they had discussed condom use with
their partner immediately before their most recent sexual event (0 = no, 1 = yes). This was
developed based on a qualitative study (Ryan et al. 2009) and has been adopted in previous
studies (Tucker et al., 2009).
Psychological violence. Psychological violent victimization by a sexual partner was
assessed by five items: “Has he made you feel unsafe in this relationship?” “Has he tried to keep
you from doing things to help yourself?” “Has treated you like you are lower than he is or
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 19
stupid?” “Has he made you tell him where you’ve been and what you’ve been doing?” “Has he
called you names or swore at you?” If women had experienced any psychological violence in
their relationship with the sex partner involved in their most recent sexual event, responses were
coded as 1. No psychological violence experience was coded as 0.
Pressured sex. Women were asked if they felt pressured to have sex with their most
recent partner with one item: “To what extent did you feel pressured to do anything sexually that
you did not want to do?” Responses were coded as 1 if women reported feeling pressured to have
sex or 0 if they reported no pressure.
Substance use. Women were asked whether they and their partner used alcohol and drugs
before or during their most recent sexual event. Respondent and partner use of alcohol and drugs
were included as separated variables. Alcohol use was coded as 1 and no alcohol use was coded
as 0. Similar coding was applied for drug use of respondents and partners.
Control variables.
Demographic characteristics. Demographic variables included age, race and ethnicity,
years of education, chronic homelessness, and total days of homelessness. Participants were
classified into one of three groups for analysis: African American, Hispanic, and others who are
neither African American nor Hispanic.
Analytic Plan
This study examined how different barriers to condom use among homeless women were
associated with condom use by women’s intention to use a condom during their most recent
sexual event. Using an event-based analysis approach, this study included psychological factors,
factors related to relationship quality, and factors related to women’s relationship with their
partner during their most recent sexual event.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 20
First, women were categorized into two groups: those with and without intention to use a
condom. Second, group-difference tests were conducted to determine whether there were
significant group differences by women’s intention for each study factors. A series of t-tests and
chi-square tests was conducted.
Third, logistic regressions were conducted with the subsample of women who intended to
use a condom during the most recent sexual event. Women without intention were excluded
because the results of group-difference tests indicated that many predictors did not have enough
observations to conduct logistic regression to predict condom use in this sample. All statistical
analyses were conducted in STATA 13.0. In addition, sampling weights were applied to all
analyses.
Results
Sample Characteristics
Individual characteristics. As shown in Table 1, African Americans were the largest
racial and ethnic group (38.65%), followed by White (26.52%), and Hispanic (23.82%). A
majority of the sample had a high school diploma or more education (66.85%). Only 15.38% of
women reported being married and approximately one quarter of women had been employed
during the previous 6 months (25.84%). Average monthly income of participants was $571.94.
Women had been homeless for an average of 34 months in their lifetime.
Condom use and intention to use a condom. About one third of women used a condom
during their most recent sexual event with a male partner (28.92%), whereas 34.16% of women
reported they wanted to use a condom during the event.
Partner-related characteristics. In terms of perceived partner intention to use a condom,
26.73% of partners had condom use intention during the most recent event. Most women had sex
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 21
with primary partners (71.24%) rather than casual partners during their most recent sexual event.
About half of the participants (49.22%) experienced any psychological violence from their most
recent sexual partner, 3.39% experienced pressured sex, and 4.29% felt threatened.
Contextual characteristics. Nearly half (44.27%) of the women reported that either they
or their male partners had a condom during the sexual event. In terms of substance use, 21.74%
of women reported that either they or their partners used alcohol before sex; the percentage of
drug use before sex was similar (21.50%). About one fifth of women (22.28%) had discussed
condom use before having sex.

Table 1
Descriptive Statistics of Weighted Sample: Psychological, Relationship, and Contextual Factors
during the most Recent Sexual Event (N = 445)
Variable % or M (SD)
Individual factors
Age 36.56 (11.10)
Race and ethnicity
White 26.52
Black 38.65
Hispanic 23.82
Asian or other 11.01
Education
Less than high school 33.15
High school or more 66.85
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 22

Married 15.38
Employed 25.84
Monthly income (US$) 571.94 (744.61)
Total months of homelessness 31.37 (60.21)
Condom use 28.92
Intention to use a condom 34.16
Relationship factors
Length of relationship (month) 74.7 (132.4)
Commitment on relationship 12.13 (4.92)
Event-specific relationship factors
Partner type
Primary or main partner 71.24
Nonprimary or casual partner 28.76
Perceived partner intention to use a condom 26.73
Condom availability 44.27
Condom communication 22.28
Psychological violence 49.22
Pressured sex 13.39
Substance use
Women’s alcohol use before sex 13.46
Women’s drug use before sex 16.67
Partners alcohol use before sex 16.42
Partners drug use before sex 15.31
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 23

Women with and without Intention
Table 2 features descriptive statistics by women’s intention to use a condom during their
most recent sexual event. To test statistical differences between groups for each variable, a series
of chi-square and t-tests was conducted. Women with intention to use a condom were more
likely to have achieved a high school or GED education level (p < .05). More women without
intention to use a condom were currently married at the time of the interview than women with
intention (p < .01). Most women with intention (80.53%) used a condom during their most recent
sexual event, whereas only 2.15% of women without intention used a condom (p < .001).
Women with intention had significantly higher scores regarding attitudes toward condoms than
women without intention (p < .001). Women without intention were more likely to have longer
relationships with their most recent sexual partner (M = 83.59 months, SD = 6.40) than women
with intention (M = 48.70 months, SD = 6.32). Women without intention also had greater
commitment to the relationship (M = 3.26, SD = 0.06) than women with intention (M = 2.74, SD
= 0.11).
Regarding contextual variables related to the most recent sexual event, the partners of
women with intention were more likely to intend to use a condom (73.68%) than the partners of
women without intention (2.05%). Women with intention were more likely to have had sex with
primary partners (46.58%) than women without intention (20.62%). Women without intention
were more likely to have experienced psychological violence from their most recent partners
than women with intention (p < .001). Women with intention were more likely to have
experienced pressure to have sex even though they did not want to have sex (p < .05) and were
more likely to have had need-based sex (p < .01). Women with intention were more likely to
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 24
have a condom during their most recent sexual event (83.51%) than women without intention
(23.38%). Furthermore, women with intention had more communication about condom use with
their partner before having sex (49.54%) than women without intention (8.14%). In terms of
substance use, only alcohol use differed between groups. Women with intention reported more
alcohol use before sex than women without intention (p < .05).

Table 2
Group Difference Tests by Women's Intention to Use a Condom (Weighted; N = 445)

Intention
(n = 152)
No intention
(n = 293)

Variable % or M (SD) % Sig.
Individual factors
Age 36.46 (11.27) 36.61 (11.03)
Race and ethnicity
White 25.69 25.95
Black 40.36 40.07
Hispanic 17.47 25.52
Asian or other 16.48 8.46 **
Education *
Less than high school 26.94 36.38
High school or more 73.06 63.62
Married 7.88 19.27 **
Employed 25.03 26.26
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 25
Monthly income (US$) 624.34 (997.13) 533.68 (571.03)
Outcome
Condom use 80.53 2.15 ***
Psychological characteristics
Condom attitudes 3.31 (0.44) 2.92 (0.37) ***
HIV susceptibility 1.98 (0.10) 2.08 (0.07)
Condom efficacy 6.24 (0.23) 6.00 (0.16)
Relationship factors
Length of relationship (months) 48.70 (6.32) 83.59 (6.40) ***
Commitment 2.74 (0.11) 3.26 (0.06) ***
Event-specific factors
Perceived partner’s intention to use a condom 73.68 2.05 ***
Primary or main partner 46.58 20.62 ***
Psychological violence 34.21 57.34 ***
Pressured sex 17.11 10.58 *
Need-based sex 9.87 2.39 **
Condom availability 83.51 23.38 ***
Substance use
Women’s alcohol use before sex 7.89 3.41 *
Drug use before sex 7.24 5.12
Partner alcohol use before sex 9.87 7.85
Partner drug use before sex 5.92 4.10
Condom communication 49.54 8.14 ***
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 26
Note. Sig., significance.
*p < .05, **p < .01, ***p < .001

Logistic Regression of Condom Use
The associations between variables and condom use during the most recent event were
analyzed using logistic regression.
The associations were analyzed with women who had intention to use a condom during
their most recent sexual event (Table 3). Women who did not have intention to use a condom
were excluded because many predictors did not have enough frequency, so the model failed to be
estimated. Analyzing associations between variables and condom use among women who had
intention to use a condom enabled an examination of which variables promote or hinder condom
use even if women intended to use a condom.
No demographic and individual psychological variables were associated with condom use
among women with intention. Women in committed relationships were less likely to use a
condom (OR = 0.23, 95% CI [0.07, 0.72]). Perceived partner intention was significantly
associated with condom use among women with intention (OR = 35.56, 95% CI [8.21, 17.83]).
In this subgroup, women who had sex because they needed something were more likely to use a
condom (OR = 14.01, 95% CI [1.47, 13.35]). Condom availability was also significantly
associated with condom use (OR = 73.91, 95% CI [2.93, 18.62]). Having a conversation about
condoms with partners before having sex also increased the likelihood of condom use among
women with intention (OR = 3.15, 95% CI [5.82, 17.08]).
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 27
Table 3
Logistic Regression of Condom Use among Women with Intention to Use a Condom (n = 152)
Model 1 Model 2 Model 3 Model 4
OR 95% CI OR 95% CI OR 95% CI OR 95% CI
Age 1.03 0.98, 1.08 1.03 0.98, 1.09 1.04 0.98, 1.10 1.06 0.91, 1.23
Black 1.53 0.59, 3.93 1.50 0.52, 4.34 1.12 0.37, 3.33 0.11 0.01, 1.29
Hispanic 1.59 0.53, 4.82 2.39 0.67, 8.51 2.34 0.63, 8.66 0.23 0.01, 3.83
High school or GED 1.59 0.62, 4.04 1.20 0.41, 3.50 1.02 0.34, 3.13 1.50 0.19, 11.73
Married 1.31 0.26, 6.73 1.37 0.27, 7.00 1.10 0.18, 6.68 0.13 0.00, 137.05
Employed 1.56 0.52, 4.77 1.41 0.46, 4.30 1.40 0.44, 4.45 0.71 0.04, 11.47
Condom attitudes 4.61 1.78, 11.91 4.60 1.57, 13.44 6.18 0.11, 33.52
HIV susceptibility 0.30 0.10, 0.91 0.30 0.09, 0.96 0.51 0.03, 8.84
Condom efficacy 1.09 0.94, 1.27 1.12 0.95, 1.34 1.73 0.77, 3.89
Length of relationship 1.00 0.99, 1.01 1.04 0.99, 1.10
Committed relationship 0.89 0.66, 1.20 0.23 0.07, 0.72
Primary partner 9.49 0.71, 0.72
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 28
Partner’s intention 35.56 8.21, 17.83
Need-based sex 14.01 1.47, 13.35
Condom availability 73.91 2.93, 18.62
Condom communication 3.15 5.82, 17.08
Pressured sex 4.16 0.90, 19.30
Psychological violence 0.37 0.04, 3.15
Any alcohol use 0.71 0.28, 1.78
Any drug use 1.56 0.30, 8.15
Note. Figures in bold are statistically significant at p < .01. Model 1 included individual factors; Model 2 included individual and
psychological factors; Model 3 included individual, psychological, and relationship factors; Model 4 included individual,
psychological, relationship, and contextual factor
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 29
Discussion
This study examined factors associated with condom use among homeless women using
an event-based approach. In particular, this study focused on factors related to condom use
among women who intended to use a condom during their most recent sexual event. According
to TRA and TPB, intention to use a condom is a proximal predictor of condom use. Studies that
employed TRA and TPB to explain condom use have typically explored factors predicting
intention or the relationship between intention and condom use. However, other factors may
hinder or promote condom use despite women’s intention.
In this study, only one third of women had intention to use a condom during their most
recent sexual event. According to the group-difference tests by intention (see Table 2), women
who intended to use a condom were much more likely to use a condom (80.53%) than who did
not (2.15%). This indicated a strong association between intention and condom use among
homeless women. Thus, to improve consistent condom use in this population, HIV prevention
interventions for homeless women should focus on increasing intention to use a condom.
Among psychological characteristics, condom attitude scores differed by group. Women
who had intended to use a condom had higher scores (M = 3.31, SD = 0.44) than women without
intention (M = 2.92, SD = 0.37). According to TRA and TPB, condom attitudes are an important
predictor of intention. Although no significant differences were found in terms of HIV
susceptibility and condom efficacy by group, results regarding attitudes about condoms
supported that condom-related beliefs are associated with intention to use a condom among
homeless women.
Fairly significant group differences emerged in terms of relationship and contextual
characteristics. Women without intention were more likely to have sex with partners in long-term
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 30
and committed relationships than women with intention. Given these results, it can be assumed
that women may want to use a condom with partners in a short-term or less committed
relationship. This is consistent with previous evidence (Bauman & Berman, 2005; Civic, 1999;
Noar et al., 2012; Plichta et al., 1992; Wingood & DiClemente, 1998, 2000). This result can be
interpreted as an indication that women may feel safe in terms of exposure to HIV infection with
sexual partners who they already know well. However, this result also highlights the need to
improve condom use among women in deeply engaged relationships with sexual partners,
because improving consistent condom use is a critical outcome of HIV preventive interventions.
Women’s intention to use a condom was closely associated with the intention of partners;
73.68% of women with intention perceived that their partners also intended to use a condom,
compared to only 2.05% of women without intention. This result indicates that women’s
intention to use a condom may be related to their partner’s intention. It is not clear whether one
partner’s intention is more superior than the other’s. However, considering that women with
intention were much more likely to use condoms, there might be strong association between
women’s intention and partners’ intention, particularly when they make a decision to use a
condom.. Therefore, particularly for women, interventions to promote condom use should
recognize that women’s intention to use a condom and actual condom use could be determined
by partners.
Women with intention were more likely to have sex with a primary partner (46.58%) than
women without intention (20.62%). Although women without intention experienced more
psychological violence from their most recent partner (57.34%) than women with intention
(34.21%), women with intention reported more pressure to have sex by their partner (17.11%). A
similar pattern was observed in terms of need-based sex; women with intention had more need-
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 31
based sex than women without intention. Additionally, they were more likely to have a condom
during their most recent sexual event and to have a conversation about condoms with their
partner than women without intention.
In terms of substance use, both women and their partners reported a low frequency of
substance use before the most recent sexual event. Only women’s alcohol use before sex differed
by groups (women with intention vs. without intention; 7.89% vs. 3.41%, respectively).
The results of logistic regression on condom use among women who intended to use a
condom showed that relationship and contextual characteristics were significantly associated
with condom use. Before entering these variables, women’s psychological characteristics were
significantly associated with condom use. However, once relationship-related and contextual
factors were entered, these significant associations between psychological factors and condom
use disappeared. This can be interpreted as an indication that relationship-related or contextual
factors during sexual events may have stronger power in determining condom use among women
with intention to use condoms. According to the results of the final model, relationship and
contextual factors play an important role in determining condom use among women who
intended to use condoms. Women in committed relationships were less likely to use condoms
despite their intention to use a condom. Similar results were found in previous studies (Bauman
& Berman, 2005; Civic, 1999; Wingood & DiClemente, 1998, 2000). Women often report
frequent unprotected sex with partners in stable relationships due to trust issues (Wingood &
DiClemente, 1998, 2000). In terms of HIV infection, women may feel safe with committed
partners. What distinguished this study from previous studies was its focus on women who had
intention to use a condom. Results indicated that even when women had intention to use a
condom, if they had sex with partners in a committed relationship, they may not use condoms
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 32
consistently. Because promoting consistent condom use regardless of partner type is the main
goal in HIV prevention intervention, efforts to improve condom use even with trusted and
committed partners and even when women have intention is critical.
As evidenced by the results of group-difference tests, perceived partner intention was
strongly associated with condom use in this subsample. When women wanted to use a condom,
the likelihood of using a condom increased if their partner also wanted to use a condom.
Previous studies consistently showed that sexual decisions cannot be made solely by one person
(Bauman & Berman, 2005; Civic, 1999; Noar et al., 2012; Plichta, 1992; Wingood &
DiClemente, 1998, 2000). Inherently, deciding to use a condom is a relationship-based process.
Particularly for women, due to power imbalance, their partner’s intention or reaction is important.
This indicates that understanding perceived partner intention is as important as understanding
women’s intention. On the other hand, women may withdraw their intention if their partner does
not want or intend to use a condom. Thus, HIV prevention intervention should educate women to
handle situations in which their partner does not want or intend to use a condom.
Furthermore, having condoms and talking about condoms during the most recent sexual
event increased the likelihood of condom use. Therefore, women should be encouraged to carry
condoms. In addition, unlike the results of the full sample, training women about how to discuss
condom use with sex partners is important. Women with intention were more likely to have
conversations about condoms with their partners before having sex than women without
intention. In addition, when women had sex because they needed something, they were more
likely to use condoms. Similar results were found in the full sample. It is likely that need-based
sex partners are more often casual partners or strangers than partners in a stable relationship.
Women may not know enough about the HIV and STD status of these individuals, which may
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 33
explain this result. This finding reinforces the need to increase condom use among homeless
women who engage in need-based sex regardless of their condom use intentions. In addition,
encouraging women to carry condoms would increase the likelihood of condom use. Providing
information regarding where women may obtain free condoms and improving accessibility to
free condoms would be good strategies to promote condom availability.
Strengths and Limitations
Important strengths of this study include the rigorous sampling design to collect
representative sample of homeless women in shelter settings. Additionally, this study focused on
factors associated with condom use among women who intended to use a condom using an
event-based approach. To accomplish this goal, various types of factors (e.g., psychological,
relationship, and contextual factors) were included in the analyses. However, this study also has
some notable limitations. Despite its rigorous sampling design, because this study focused on
women with condom use intentions during their most recent sexual event, generalizability issues
exist. Because only one third of women reported that they had intended to use a condom, the
sample size for logistic regression was 152. This may have caused fairly large odds ratios and
wide 95% confidence intervals in the associations of some factors with condom use (Prel et al.,
2009). This does not mean that the results of this study should be ignored. Prel et al. (2009)
advised distinguishing between statistical significance and clinical relevance when interpreting
odds ratios and confidence intervals. A similar argument can be found in Akobeng’s (2008)
study. Even results based on small samples can have important clinical relevance and should be
considered. Furthermore, they can provide direction for future studies.
Although the event-based approach generated plentiful information regarding relationship
and contextual factors and their role in condom use, a more advanced research design is needed
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 34
in future studies. Similar suggestions were made in previous studies using this event-based
approach (Tucker et al., 2012). Assessing multiple sexual events may be challenging with
homeless populations; however, several previous studies adopted this approach with vulnerable
populations such as substance users (Fortenberry, Orr, Katz, Brizendine, & Blythe, 1997;
Morrison et al., 2003). However, they did not analyze multiple sexual events, which may result
in more accurate associations between factors and condom use using more advanced statistical
methods such as multilevel modeling. As a result, multilevel modeling was adapted to test the
study model developed based on TRA and TPB, as described in the following chapter. Finally, it
should be noted that the measure used to assess intention to use a condom among women may
have limitations. Ajzen (1991) noted that “intentions are assumed to capture the motivational
factors that influence a behavior; they are indications of how hard people are willing to try, of
how much of an effort they are planning to exert, in order to perform the behavior” (p. 181).
Ajzen (2002) offered a few suggestions for constructing measurements of intention. For
example, “I intend to walk on a treadmill for at least 30 minutes each day in the forthcoming
month,” “I will try to walk on a treadmill for at least 30 minutes each day in the forthcoming
month,” and “I plan to walk on a treadmill for at least 30 minutes each day in the forthcoming
month.” Ajzen (2002) suggested measuring each item using a 7-point scale. It should be
acknowledged that the item used to measure the intention of women and their partners to use a
condom did not focus on how much effort they made to use a condom or how much they were
willing to use a condom. A similar item was applied in a previous study (Roth et al., 2011), but
future studies should adopt or develop more accurate measurements to capture homeless
women’s condom use intentions.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 35
Conclusion
Despite some limitations, this study provided evidence of factors that promote or hinder
condom use based on the condom use intentions of homeless women. The results of this study
can be used to develop effective HIV prevention interventions, particularly interventions
designed to promote intention as a strategy to improve condom use. Emphasizing factors that
promote condom use among women with intention is expected to strengthen the association
between intention and condom use.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 36
CHAPTER 4. A MULTILEVEL PATH ANALYSIS OF CONSISTENT CONDOM USE
AMONG HOMELESS PEOPLE: AN APPLICATION OF THE THEORIES OF
REASONED ACTION AND PLANNED BEHAVIOR
Background
Particular populations in the United States, such as homeless individuals, are at greater
risk of contracting HIV compared to other groups. Studies have established that homeless people
are more likely to engage in HIV risk behaviors, such as having unprotected sex and engaging in
sex with multiple partners (Valente & Auerswald, 2013; Wenzel et al., 2012) and substance use
(Rhoades et al., 2011).
Among those risky sexual behaviors, unprotected sex is a major route of HIV
transmission (CDC, 2013a). Frequent engagement in unprotected sex among homeless people
has been well documented in previous studies (Bailey et al., 1998; Kennedy, Brown, et al., 2013;
Kennedy, Wenzel, et al., 2013; Nyamathi & Vasquez, 1989; Sikkema et al., 1996; Tucker et al.,
2013; Valente & Auerswald, 2013). More than 50% of homeless men randomly recruited in the
Los Angeles area reported having unprotected sex in heterosexual relationships (Kennedy,
Wenzel et al., 2013; Tucker et al., 2013). Nearly half of the young homeless men (48.8%)
interviewed in another study reported having unprotected sex (Valente & Auerswald, 2013).
Frequent unprotected sex also has been observed among homeless women (Kennedy et al., 2010;
Nyamathi & Vasquez, 1989; Ryan et al., 2009; Sikkema et al., 1996; Wenzel et al., 2004).
Consistent condom use is the most effective and strategic method of preventing HIV
infection (CDC, 2013a). Enhancing knowledge about condom use among homeless people is an
important public health issue. Exploring this issue among heterosexual homeless people in
particular may be beneficial. Not only do most homeless men in large urban areas identify as
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 37
heterosexual (Robertson et al., 2004), but heterosexual intercourse is also the main cause of HIV
infection among women (CDC, 2013b). Therefore, understanding why heterosexual homeless
people frequently engage in unprotected sex may be crucial for developing effective HIV
prevention interventions to increase consistent condom use in this population.
Theory serves an important role in improving understanding of condom use and
developing effective interventions. Theory-based HIV interventions have shown their power in
terms of promoting condom use (Fishbein, 2000). Among many health theories related to
condom use, TRA and TPB have served as important theoretical frameworks to explain condom
use by individuals (Albarrací n et al., 2001; Armitage & Conner, 2001; Manstead, 2011; Sheeran
& Orbell, 1998). TRA and TPB hypothesize that individual health behaviors are the function of
intention. According to TRA, intention is a function of several antecedents: attitudes,
vulnerability, and norms. TPB expands the main assumption of TRA by embracing perceived
control as an antecedent of intention. TRA and TPB have shown considerable relevance for
understanding HIV risk behaviors, including condom use (J. D. Fisher & Fisher, 2000). Several
meta-analyses have evaluated the usefulness of TRA and TPB in predicting condom use
(Albarrací n et al., 2001; Armitage & Conner, 2001; Sheeran & Orbell, 1998), indicating that the
theories worked fairly well in terms of predicting condom use.
However, some studies have challenged TRA and TPB by revealing particular contexts in
which they may not work well. Sheeran and Orbell (1998) found that TRA and TPB alone may
not be enough to explain condom use by examining moderators that influence the effect of
intention on condom use. Furthermore, Manstead (2011) determined that a simple linear
relationship between intention and health behaviors may not be the inherent pattern of the
association. Identifying the limitations of TRA and TPB would serve to increase the power of the
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 38
theories. As Widman, Golin, & Noar (2013) noted, the sexual-decision-making process
regarding condom use is undoubtedly complex. It may be difficult to posit that this process is
solely determined by one individual because sexual behaviors are fundamentally based on
relationship dynamics. Thus, theories focusing solely on individual psychological factors often
do not work well in predicting condom use by individuals.
A series of studies has identified moderators and contexts associated with the link
between intention and condom use. Sheeran and Orbell (1998) identified six moderators to
quantify the relationship between intention and condom use. Among them, gender was the most
commonly studied moderator. Several researchers have suggested that gender differences in
intention and condom use may play a role in sexual decision making (Abraham et al., 1999;
Bowleg et al., 2004; Cabral et al., 1998; Crosby et al., 2002; Plichta et al., 1992; VanderDrift et
al., 2013; Wingood & DiClemente, 1998). Intention among men may possess more power in
determining condom use than that of women (Abraham et al., 1999; VanderDrift et al., 2013).
Women may withdraw their intention to use a condom if their male partner has a negative
reaction to condom use (Cabral et al., 1998; Plichta et al., 1992; Wingood & DiClemente, 1998).
Sheeran and Orbell (1998) identified partner type as a moderator of the link between
intention and condom use. More frequent use of condoms was found with casual partners than
primary partners (Gebhardt, Kuyper, & Greunsven, 2003; Plichta et al., 1992; Weinhardt et al.,
1998). With primary or committed partners, individuals may be less concerned about their
vulnerability to HIV infection and have greater trust in their partner (Weinhardt et al., 1998).
Predictors based on TRA and TPB, particularly intention, may operate better in the context of
primary partners (Sheeran & Orbell, 1998). Morrison, Rogers Gillmore, and Baker (1995) argued
that partner type may determine the power of TRA and TPB in predicting condom use. They
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 39
pointed out that TRA showed more power in predicting condom use with primary partners than
casual partners. Similar results were found by Galligan and Terry (1993).
In the sexual-decision-making process, studies have increasingly focused on condom
communication (Noar et al., 2006). In their meta-analysis, Noar et al. (2006) criticized previous
theories based on intrapersonal variables, such as TRA and TPB, for largely ignoring the
importance of interpersonal dynamics, such as condom communication. Because sexual decision
making is a complicated process in which intra- and interpersonal variables are integrated, efforts
to identify the role ofcondom communication may increase the power of these theories. Among
various types of condom communication, suggesting the use of a condom has been studied in
association with condom use (Noar et al., 2006; Peltzer, 2000; Sanderson, 1999). Abraham et al.
(1999) found that suggesting condom use facilitates the link between intention and condom use.
Although TRA and TPB have been commonly used in the area of individual sexual
decision making, less attention has been paid to these theories in the context of homelessness.
Considering the vulnerability of homeless individuals to HIV risks and frequent engagement in
unprotected sex, these theories may help identify effective predictors of condom use among
homeless people. Some studies showed that core concepts of TRA and TPB were associated with
condom use among homeless people. Intention to use condoms among homeless people was
associated with condom use (Rew et al., 2002; Somlai et al., 1998; Stark et al., 1998), as were
attitudes toward condoms (Kennedy et al., 2013; Somlai et al., 1998; Tucker et al., 2013).
Furthermore, previous studies have applied social network approaches to measure subjective
norms regarding condom use among homeless individuals (Ennett, Bailey, & Federman, 1999;
Tucker et al., 2012; Tyler & Melander, 2011; Wenzel et al., 2012). Evidence from those studies
has repeatedly shown that having more sexual risk-taking network members may lead to more
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 40
frequent engagement in unprotected sex and other risky sexual behaviors by homeless people.
However, previous studies that examined predictors of TRA and TPB with homeless people were
not fully based on the framework of these theories. Therefore, little is known about whether TRA
and TPB effectively predict condom use among this population.
Similar to studies with other groups, evidence has indicated that gender and partner type
may play a role in determining condom use among homeless people. Intention effectively
predicted condom use among homeless men but not women (Somlai et al., 1998). Furthermore,
for impoverished women, discussing condom use with a male partner may be difficult (Nyamathi,
Lewis, Leake, Flaskerud, & Bennett, 1995), indicating that men have more power in determining
condom use than women in sexual relationships. Partner type is another important factor related
to condom use among homeless people. Tucker et al. (2007) showed that homeless women
reported less consistent condom use with primary partners than with casual partners. Although
not conducted with homeless individuals, a study by Kapadia et al. (2011) with injection drug
users (nearly 40% of whom described themselves as homeless) found different predictors
associated with consistent condom use by partner type. Condom self-efficacy was only
associated with condom use with primary or committed partners.
Unfortunately, although those core predictors based on TRA and TPV and partner type
have been examined regarding their associations with condom use among homeless people in
some studies, they were studied separately in different studies. By critically examining the
effectiveness of TRA and TPB in predicting condom use among homeless people, more effective
HIV prevention interventions may be developed to promote consistent condom use among
homeless people. Therefore, examining a conceptual model guided by TRA and TPB to predict
condom use of homeless people is critical to accumulating this needed evidence.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 41
Present Study
This study examined whether condom use among homeless people is a function of predictors of
TRA and TPB. To accomplish this aim, a conceptual model based on TRA and TPB was
developed (Figure 3). Based on TRA and TPB, attitudes toward condoms, perceived HIV
vulnerability, and subjective norms were included as predictors of intention to use condoms.
Following the guidance of TRA and TPB, intention was included as a mediator between those
predictors and condom use, which was the outcome of this study. Furthermore, partner type and
suggesting condom use were included as potential moderators of the influence of intention on
condom use. To test gender differences in predicting condom use based on TRA and TPB among
homeless people, the conceptual model was analyzed by gender (homeless women and homeless
men).
Between-group variables included predictors, condom attitudes, HIV vulnerability, and
subjective norms. Within-group variables were a mediator (intention) and an outcome (condom
use). Therefore, the mediation model of this study followed a 2-1-1 design (Preacher, Zhang, &
Zyphur, 2011). Partner type and suggesting condom were also within-group variables included as
moderators between intention to use a condom and condom use. This final research model was
analyzed by gender to assess gender differences in applying TRA and TPB to homeless people.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 42
Figure 3
A Path Model to Predict Condom Use among Homeless People

Methods
Study Sample
This study examined the conceptual model developed based on TRA and TPB among
homeless people. To examine how predictors were associated with condom use by gender, two
distinctive datasets were used.
Homeless women. Homeless women in this study were randomly recruited via 51
temporary shelters in the central region of Los Angeles. The final sample size was 445 using
stratified random sampling methods. As described in Wenzel et al.’s (2009) study, a strict
proportionate-to-size stratified random sample would have been overly burdensome at larger
facilities, so a modified approach was used and sampling weights were created. The survey was
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 43
conducted between June 2007 and March 2008. To be eligible, women had to be at least 18 years
old, able to speak and understand English, have had vaginal or anal sex with at least one male
partner during the previous 6 months, and have no cognitive impairment. An initial sample of
472 women was deemed eligible, but 451 women completed the interview. Five women were
excluded because they only had oral sex with a partner and one woman was excluded because
she did not complete the questionnaires. Women completed a personal network survey; more
information can be obtained elsewhere (Green et al., 2012; Wenzel et al., 2009). Computer-
assisted, face-to-face interviews were completed with 445 women, each lasting approximately 1
hour and 15 minutes and conducted by trained female interviewers. Women who completed the
interview were paid $20 each. The final sample size was 445 women, and the completion rate
was 94% (445 of 472).
Homeless men. This study also featured a sample of 305 men randomly recruited from
13 meal programs in the Skid Row area of Los Angeles, a region with one of the highest
concentrations of homeless individuals in Los Angeles County. In order to obtain a
representative sample, a probability sampling strategy was implemented through 13 meal
programs. To create a list of meal programs in the area, the research team used existing
directories of services for homeless individuals in Los Angeles and interviewed with service
providers to identify other meal lines. Then the average number of men served daily at each
location were estimated. Based on this information, proportional quotas were assigned to each
site. Finally, a probability sample of homeless men from the 13 programs was generated. Men
were randomly selected by their position in line using statistically generated random number
tables. This study implemented sampling weights to maximize the representativeness of the
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 44
sample. Strictly speaking, the sampling strategy deviated from a proportionate-to-size stratified
random sample.
The survey was completed by 305 men for a completion rate of 91%; 33 men were
excluded for declining to participate, not completing the survey, or repeating the survey. Men
who completed the interview were paid $30 each; interviews took an average of 83 minutes to
complete.
Sampling for multilevel modeling. Both datasets included information on the social
networks of homeless people. Participants were asked to name 20 individuals aged 18 years or
older who they knew, who knew them, and with whom they had contact during the previous year.
Second, they were asked about support exchange and closeness among the network members
they named. Immediately after they generated names of 20 alters, they were also asked if they
had named any of their sex partners among the 20 alters. To characterize the composition of the
network, they were asked a series of questions about each alter, including background
characteristics, behaviors, and relationship with the respondent. Network alters were stratified
into sex partners and nonsex partners. Personal network information for sex partners provided
data for a multilevel analysis of sexual risk behaviors. This strategy to obtain multilevel data was
adopted from a previous study (Kennedy, Brown, et al., 2013).
Study Variables
Both datasets used for this data were originally designed to investigate risky sexual
behaviors in each population, i.e., homeless women and homeless men in the Los Angeles area.
Study variables derived from TRA and TPB were found in both studies. Some variables required
modification because although they measured same concepts, they may have been measured
differently. For example, condom attitudes were measured differently in both studies. Thus, a
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 45
standardization procedure to transform these variables to z-scores was conducted to enable
interpretation of the results.
Individual variables.
Condom attitudes. Attitudes regarding condoms among homeless people were assessed
using standardized scores. Women’s condom attitudes were assessed by eight items used in
previous studies (Bogart et al., 2005), such as (1) “condoms are too much trouble,” (2) “condoms
are unreliable—sometimes they work and sometimes they don’t,” and (3) “condoms ruin sex.”
The items were rated on a 4-point scale (1 = strongly disagree to 4 = strongly agree). Higher
scores indicated more negative attitudes regarding condom use. Men’s condom attitudes were
assessed by four items used in previous studies (Kennedy, Brown, et al., 2013; Mutchler et al.,
2008; Stall, Paul, Barrett, Crosby, & Bein, 1999; Tucker et al., 2013): (1) “using condoms makes
sex less enjoyable,” (2) “using condoms can be difficult,” (3) “condoms can make you lose your
hard on,” and (4) “condoms break too often.” These items were rated on a 4-point scale (1 =
strongly disagree to 4 = strongly agree). The sum of condom attitude scores were standardized
for men and women.
Perceived HIV vulnerability. Both men and women were asked if they thought it would
be easy for them to contract HIV or AIDS. This item was rated on a 4-point scale (1 = strongly
disagree to 4 = strongly agree). Perceived HIV vulnerability was included as a continuous
variable.
Subjective norms. Subjective norms regarding risky sexual behaviors were derived from
social network information. Men and women were asked to name 20 individuals aged 18 years or
older who they knew, who knew them, and with whom they had contact during the previous year.
Using 20 alters has been shown to capture structural and compositional variability present in
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 46
personal networks. Participants were asked to identify whether, during the previous 6 months,
their network members had multiple sex partners, had sex with someone they didn’t know, or
didn’t use a condom with a new partner to assess subjective norms regarding risky sexual
behaviors. Men were asked to rate their answers on a 2-point scale (0 = yes, 1 = no), whereas
women were asked to use a 3-point scale (1 = unlikely, 2 = somewhat likely, 3 = very likely). To
merge the data, women’s answers were dichotomized as 0 (unlikely) and 1 (somewhat or very
likely). The proportion of alters who engaged in risky sexual behaviors across all network
members was calculated to assess subjective norms concerning risky sexual behaviors among
homeless men and women.
Relationship and event variables.
Condom use. As a dependent variable, condom use among homeless people was derived
from two items that assessed the frequency of sexual events and condom use with each sexual
partners. The two items were as follows: (1) “During the past 6 months, about how many days
have you had vaginal or anal sex with a sexual partner?” (2) “How many of these times did you
use a condom?” Based on that information, a new variable for consistent condom use was
created. If a condom was consistently used during all sexual events with a partner (proportion of
condom use = 1), condom use was coded as 1. If a condom was not used at least once
(proportion of condom use < 1), condom use was coded as 0. As a result, condom use in this
study was binary, which required a multilevel approach.
Intention to use condoms. Both men and women were asked how often they wanted to
use a condom with each partner during sexual events. This item was rated on a 5-point scale (1 =
never, 2= less than half the time, 3 = about half the time, 4 = more than half the time, 5 = always).
This item was dichotomized as 1 (always) and 0 (not always). If respondents wanted to use a
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 47
condom during all sexual events with each sexual partner, intention to use a condom was coded
as 1. Otherwise, it was coded as 0.
Partner type. Men and women were asked whether their most recent partner was a
current sex partner and whether they would identify that partner as primary or casual. Primary
partners included a husband or wife, boyfriend or girlfriend, or other partners described as
“steady” or committed. Casual partners were defined as “in the moment” or “just for fun”
partners.
Suggesting condom use. Men and women were asked if they had suggested using a
condom to their sexual partners. If they ever asked their partners to use condoms, this variable
was coded as 1; otherwise it was coded as 0.
Control variables. Multilevel path analyses were conducted to control for individual
demographic characteristics and substance use.
Individual demographic characteristics. Demographic variables included age, race and
ethnicity, years of education, chronic homelessness, and total days of homelessness. Participants
were classified into one of three groups for analysis: African American, Hispanic, and others
who are neither African American nor Hispanic.
Substance use. Men and women were asked to identify whether they and their partner
used any substances before and during sexual events with sexual partners. Use of any substance
(alcohol or drugs) was coded as 1; otherwise it was coded as 0. Substance use by partners was
coded in a similar fashion.
Analytic Plan
This study examined the mediating effect of intention on the link between condom
attitudes, perceived HIV vulnerability, and subjective norms regarding risky sexual behaviors
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 48
and condom use among homeless people. Additionally, this study examined whether partner type
moderated the link between intention and condom use. To examine gender differences in the
association between variables, the same model was separately tested using two distinct datasets
for homeless women and homeless men.
To examine the conceptual model guided by TRA and TPB, multilevel path analyses
were used. The advantages of multilevel analyses over traditional analyses approaches for
clustered data have been well documented (Kenny, Korchmaros, & Bolger, 2003). First, unequal
group sizes and missing data issues can be handled in multilevel analyses. Second, in the nested
(clustered) data, multilevel analyses can test simultaneously the effects of between- and within-
group variables by accounting for any bias in standard errors resulting from the nonindependence
of observations (Krull & MacKinnon, 2001). Traditional analysis approaches, such as ordinal
regression, assume that each individual observation is independent of one another, but in nested
(clustered) data this independence assumption can be significantly violated. Thus, using a
traditional analysis approach to examine nested data may cause biased results.
Due to the advantages of multilevel analyses for nested data, this study analyzed data
using multilevel path analyses. Sex partners of homeless people in this study were nested
(clustered) in participants. That is, each homeless individual could have more than one sex
partner. This one-to-many study design (each participant to multiple sex partners) encouraged
the use of multilevel path models rather than other traditional analyses. Ignoring inherent
clustering and assessing general associations between variables may increase type I errors (Krull
& MacKinnon, 1999, 2001).
To develop a deeper and more systematic understanding of condom use, some studies
have relied on a multilevel analysis approach (Kennedy, Brown, et al., 2013; Kennedy et al.,
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 49
2010; McMahon, Pouget, & Tortu, 2007; Sherman & Latkin, 2001; Westercamp et al., 2010).
Relationship dynamics in terms of sexual decision making have inherent multilevel
characteristics. Applying a multilevel approach increases methodological power to explore
condom use with partners (Westercamp et al., 2010). Applying a multilevel approach is
beneficial to examining whether condom use is predicted either by individual or relationship and
contextual factors, or by dynamics between the two levels (Hofmann & Gavin, 1998). Several
studies with homeless populations have used this approach (Kennedy, Tucker, Green, Golinelli,
& Ewing, 2012; Kennedy, Brown, et al., 2013; Tucker et al., 2013). However, previous studies
have focused less on developing a theory-based conceptual model to predict condom use;
therefore, multilevel path analyses were rarely adapted. This study included a mediator (intention)
and a moderator (partner type), necessitating path analyses.
Applying a multilevel approach will more effectively examine whether condom use is
predicted by either individual or relationship and contextual predictors or dynamic interactions
between the two levels (Hofmann & Gavin, 1998). Given this background, this study applied
multilevel path analyses to examine the mediating effect of intention to use a condom. To test
mediating effects with nested data, traditional methods such as multiple linear regression or
logistic regression are not appropriate (Preacher et al., 2011). Although there is consensus on the
necessity of multilevel approach to assess mediating effects with nested data, how to assess these
effects with a multilevel approach remains controversial. Several papers have described
multilevel modeling to assess mediating effects (Bauer, Preacher, & Gil, 2006; Kenny et al.,
2003; Krull & MacKinnon, 1999; Zhang, Zyphur, & Preacher, 2009). For example, Kenny et al.
(2003) proposed a procedure to analyze multilevel mediation effects by conducting a series of
equations with a resampling method to obtain a value for this covariance. However, Preacher,
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 50
Zyphur, and Zhang (2010) criticized the ability of previous multilevel modeling to assess
mediating effects. Previous multilevel modeling only allows lower-level mediation, failing to
accommodate upper-level mediators or outcome variables. In addition and more importantly,
particularly for 2-1-1 mediation model, conflation issues between mediations and outcomes
could be produced. To overcome the limitations of previous multilevel modeling to assess
mediation, Preacher et al. (2010) suggested a new multilevel method called multilevel structural
equation modeling (MSEM). The advantages of using MSEM include the fact that sampling
error can be corrected by treating lower-level variables as latent; MSEM also includes
measurement error, similar to traditional structural equation modeling (Marsh et al., 2009). In
addition, in MSEM, the between- and within-group variables are separated, allowing for an
examination of direct and indirect effects at each level in addition to contextual effects across
levels. A previous study proved MSEM’s effectiveness in predicting mediation models compared
to traditional multilevel mediation modeling (Preacher et al., 2011).
Mplus software estimates the covariance among random effects by combining these
equations (Muthén & Muthén, 2006). Following recommendations by Zhang et al. (2009), the
path diagram of this study based on a multilevel approach is depicted in Figure 1. This study
adopted a multilevel mediation approach with moderators of a binary outcome using MPlus’s
TYPE=TWO LEVEL command to examine the mediation effect of condom use intentions and
the two moderation effects of gender and partner type.
Results
Descriptive Characteristics
Descriptive characteristics of the sample are depicted in Table 5. The average age of
homeless women was 36.17 years (SD = 10.86) and the average age of homeless men was 45.79
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 51
years (SD = 10.54). African Americans comprised the largest racial and ethnic group for both
genders (homeless men: 74.59%; homeless women: 38.65%). White participants were the
second-largest group of women (26.51%); for men, Hispanics comprised the second-largest
group (10.38%). Women reported more education than men; 66.74% of women graduated high
school or received more education, compared to 38.35% of men. About 15% of women were
currently married, compared to only 4.26% of men. Approximately one quarter of women and
4.26% of men were employed at the time of the survey. Average monthly income was $579.64
among women and $448.31 among men. Homeless men reported more time spent homeless
(66.59 months) than women (27.87 months). Women reported slightly more negative condom
attitudes (M = 3.04, SD = 0.66) than men (M = 2.11, SD = 0.83). Men perceived greater
susceptibility to HIV (M = 2.61, SD = 1.31) than women (M = 2.12, SD = 1.26). In terms of
subjective norms regarding risky sexual behaviors, which were measured as number of social
network members who had risky sex, women reported 5.85 of 20 alters engaged in risky sexual
behaviors, compared to 4.45 among men.
In terms of within-group variables, substance use of respondents and partners, condom
use during sexual events, intention to use a condom, and partner type were analyzed. Women
reported more frequent substance use (92.40%) than men (69.62%). In terms of partner substance
use, 88.60% of women reported that a partner used any substance during a sexual event. During
63.74% of all sexual partners, women reported intending to use a condom, a similar rate as
homeless men (65.11%). Women were more likely to have sex with a primary partner (52.34%)
than men (39.10%).

Table 4
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 52
Descriptive Statistics of Sample by Gender
Variable Women Men
% or M (SD) % or M (SD)
Between-group variables, n 445 305
Individual characteristics
Age 36.17 (10.86) 45.79 (10.54)
Race and ethnicity
White 26.51 9.32
African American 38.65 74.59
Hispanic 23.82 10.38
Asian or other 11.02 5.71
Education
Less than high school 33.26 61.64
High school or more 66.74 38.36
Married 14.61 4.26
Employed 25.39 15.74
Monthly income (US$) 579.64 (748.97) 448.31 (427.11)
Length of homelessness (months) 27.87 (58.50) 66.59 (72.57)
Condom attitudes 3.04 (0.66) 2.11 (0.83)
HIV susceptibility 2.12 (1.26) 2.61 (1.31)
Subjective norms regarding risky
sex
5.85 (4.87) 4.45 (5.56)
Within-group variables, n 1,026 605
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 53
Substance use 92.40 69.62
Partner substance use 88.60 72.93
Condom use 17.64 49.02
Intention to use a condom 63.74 65.11
Partner type
Primary or main 52.34 39.10
Suggesting condom use 52.34 35.94

Between and Within Variance of Within-Group Variables
Table 6 and Table 7 depict the between and within variance of within-group variables by
gender. Analyzing the between and within variance is necessary because a sufficient level of
within variance is needed to conduct multilevel analyses.
Homeless men reported 665 distinct sexual relationship with sexual partners and an
average of 2.18 sex partners. Condom use, the main outcome of this study, had more variance at
the between level (SD = 0.46) but still had enough variance at the within level (SD = 0.24). This
could indicate that men varied their condom use across all sexual partners. Similarly, men had
different levels of intention to use a condom across sexual partners (SD = 0.24) and suggesting
condom use across partners (SD = 0.21). Additionally, men had sex with both primary and casual
partners (SD = 0.29). Overall, the variances in these within-group variables provided evidence
that multilevel modeling should be adopted to explain homeless men’s sexual behaviors,
especially unprotected sex.

Table 5
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 54
Between and Within Variance of Within-Group Variables among Homeless Men
Variable SD Observation
Condom use
Between .46
Within .24
Intention to use a condom
Between .44
Within .24 N = 665
Primary partner
Between .44 n = 305
Within .29 t-bar = 2.18
Suggesting condom use
Between .46
Within .21

Women reported 1,026 sexual partners, with an average of 2.31 partners per participant.
In terms of variance, similar results were found among homeless women as men. Women also
may have varied their condom use across sexual partners during the previous 6 months (SD =
0.26). In addition, their condom use intention may have varied during certain sexual partners (SD
= 0.34). Women reported having sex with not only their primary partners but also casual or need-
based partners (SD = 0.36). Statistical difference tests were not conducted, but women showed
more within variance in condom use intention and partner type than men. Suggesting the use of
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 55
condoms also had within variance (SD = 0.30), indicating that women did not always suggest
using condoms to their partners.

Table 6
Between and Within Variance of Within-Group Variables among Homeless Men
Variable SD Observation
Condom use
Between .33
Within .26
Intention to use a condom
Between .41
Within .34 N = 1,026
Primary partner
Between .37 n = 445
Within .36 t-bar = 2.31
Suggesting condom use
Between .43
Within .30

Overall, regardless of gender, all within-group variables had sufficient variance across
sexual partners, justifying the use of multilevel modeling to explain condom use among
homeless people.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 56
Model Prediction Results
Multilevel path analyses using the MSEM approach were conducted with each dataset to
examine the associations among predictors, mediators, and the outcome (actual condom use).
First, MSEM for homeless women was conducted (Figure 4).
TRA and TPB emphasize the role of intention in predicting condom use, and intention to
use a condom was strongly associated with consistent condom use (OR = 127.502). Women who
reported intention to use a condom with a partner were more likely to use a condom with that
partner. Among psychological variables, only subjective norms regarding risky sexual behavior
were significantly associated with condom use. Women reporting more subjective norms in
terms of risky sexual behaviors, or having more nonsex partners in their social networks who
engaged in risky sex, were more likely to report intention to use a condom (OR = 1.011).
Perceived HIV vulnerability and condom attitudes were not statistically significantly associated
with intention. In addition, partner type and suggesting condom use, which were expected to play
moderating roles between intention and condom use, were not statistically significant moderators
in this study.
Figure 4
Model Prediction of Homeless Women with Odds Ratios
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 57

*p < .05, **p < .01, ***p < .001

Results for homeless men differed significantly compared to women (Figure 5). For
homeless men, intention to use a condom was strongly associated with condom use (OR =
21.128), similar to homeless women. Men who intended to use a condom were more likely to use
a condom during sexual partners. Among psychological predictors of intention, only condom
attitudes was significantly associated with intention to use a condom (OR = 1.718). Men who had
positive condom attitudes were more likely to have intention to use a condom. Unlike women,
two moderators of intended and actual condom use emerged. Men used condoms less often with
their primary partners (OR = 0.127) and the interaction between

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 58
Figure 5
Model Prediction of Homeless Men with Odds Ratios

*p < .05, **p < .01, ***p < .001

More detailed results on model prediction can be found in Table 8 and Table 9 for
homeless women and men, respectively.

Table 7
Model Prediction Results for Homeless Women
OR 95% CI p
Within
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 59
Condom use
(a) Intention to use a condom 127.502 22.7, 71.55 .001
(b) Primary partner 0.445 0.15, 1.36 .16
(c) Suggesting condom use 0.539 0.16, 1.78 .31
a × b 3.403 0.85, 13.65 .08
a × c 2.471 0.57, 10.72 .23
Between
Intention to use a condom
Condom attitudes 1.019 0.76, 3.12 .57
Subjective norms regarding risky
sex
1.011 1.12, 2.13 .002
HIV susceptibility 1.004 0.29, 2.28 .84

Table 8
Model Prediction Results for Homeless Men
OR 95% CI p
Within
Condom use
(a) Intention to use a condom 34.133 9.404, 100.705 < .001
(b) Primary partner 0.127 0.051, 0.271 < .001
(c) Suggesting condom use 7.343 2.771, 19.462 < .001
a × b 5.085 1.100, 23.511 .04
a × c 0.180 0.037, 0.874 .03
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 60
Between
Intention to use a condom
Condom attitudes 1.718 1.29, 3.98 < .001
Subjective norms regarding risky
sex
1.014 0.78, 1.21 .52
HIV susceptibility 0.933 0.57, 2.10 .70

Significance Tests of Indirect Effects
Indirect effects were examined to test their significance (see Table 10). For men, the path
from condom attitudes to condom use was statistically significantly mediated by intention to use
a condom (p < .01). Subjective norms and HIV susceptibility were not statistically significantly
associated with intention, contrary to expectations. On the other hand, for homeless women, the
link between condom attitude, intention to use a condom, and condom use was statistically
significant (p < .001).

Table 9
Significance Test of Indirect Effects
Indirect path p
Homeless men
Condom attitudes → intention to use a condom → condom use .002
Subjective norms → intention to use a condom → condom use .523
HIV susceptibility → intention to use a condom → condom use .694
Homeless women
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 61
Condom attitude → intention to use a condom → condom use .573
Subjective norms → intention to use a condom → condom use .004
HIV susceptibility → intention to use a condom → condom use .843

Discussion
This study examined a conceptual model guided by TRA and TPB to assess condom use
behaviors of homeless men and women. Several important predictors of condom use among
homeless people were included. Condom attitudes, perceived HIV vulnerability, subjective
norms regarding risky sexual behaviors, and condom use intention were included in the model.
Given the lack of studies using TRA and TPB to examine consistent condom use among
homeless people, this study was expected to provide empirical evidence of the applicability of
TRA and TPB to homeless people. Furthermore, although previous HIV prevention interventions
designed based on TRA and TPB have been shown to be effective in promoting positive sexual
behaviors (Montanaro & Bryan, 2013; Tyson, Covey, & Rosenthal, 2014; Webb & Sheeran,
2006), meta-analyses have noted that the effect sizes were fairly small (Tyson et al., 2014; Webb
& Sheeran, 2006). To address this issue of effect size, this study included two moderators,
partner type (primary partners vs. nonprimary partners) and suggesting condom use, to identify
whether the moderators quantified the link between intended and actual condom use.
One of strengths of this study is its use of multilevel modeling to examine the conceptual
model. Because the data were collected based on a one-to-many design, each homeless
individual could have had more than one sexual partner. The results of this study indicated that
homeless women had an average of 2.31 sexual partners and that homeless men had an average
of 2.18 sexual partners during the previous 6 months. Previous studies that used TRA and TPB to
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 62
explain condom use typically adopted a traditional analysis model due to the limitations of their
data. However, in terms of sexual behaviors, individuals may behave differently with different
sexual partners. Furthermore, the between and within variance of variables indicated that there
was enough variance at the within level to support multilevel modeling. Using a traditional
analysis approach for this nested data would likely increase estimation errors (Krull &
MacKinnon, 1999, 2001).
Regardless of gender, intention was a strong predictor of consistent condom use across
sexual partners. As emphasized by TRA and TPB, homeless people with intention to use
condoms were more likely to consistently use condoms with sexual partners during the previous
6 months. This indicates that encouraging intention to use a condom among homeless people can
be an effective way to promote consistent condom use. Promoting intention to use a condom has
been considered an important mediator of condom use in previous HIV interventions with other
populations (Bryan et al., 1996; Jemmott, Jemmott, & Fong, 1998; Thompson, Kyle, Swan,
Thomas, & Vrungos, 2002). Those intervention studies showed fairly consistent results that
intention resulted in changes in sexual behaviors. Given the results of this study, HIV
interventions focusing on improving condom use intention among homeless people can lead to
positive sexual behaviors (namely, consistent condom use).
In terms of moderators, two moderators affected the association between intention and
condom use among men. No moderators were detected among women. Men were less likely to
use condoms with primary partners. This supports previous evidence that individuals use
condoms less often with primary partners (Gebhardt et al., 2003; Plichta et al., 1992; Weinhardt
et al., 1998). Furthermore, this result indicated that the effect of intention to use a condom
increased when men had sex with a primary partner. That is, intention became more important in
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 63
determining condom use among homeless men when they were having sex with their primary
partner. Therefore, addressing intention would be more effective among homeless men who have
sex with primary partners. Suggesting the use of a condom was also significantly associated with
condom use among men. Men who suggested using condoms to their sexual partners were more
likely to use condoms. This also supported evidence from previous studies focusing on the role
of suggesting condom use in actual condom use (Noar et al., 2006; Peltzer, 2000; Sanderson,
1999). In addition to the direct association of suggesting condoms with actual condom use,
suggesting the use of a condom moderated the link between intention and condom use. Overall,
these results suggest that understanding moderators can be important when considering how to
promote condom use, particularly for homeless men. The effect of interactions between intention
and two proposed moderators, suggesting condom use and partner type, on condom use among
homeless men was statistically significant regardless of intention to use a condom. This indicates
that intention to use a condom use is important when predicting condom use among homeless
men; however, men could benefit from suggesting the use of condoms to their partners. The
likelihood of condom use was the highest when men had intention and suggested condom use to
their partners. Thus, for homeless men, integrating intention promotion and condom suggestion
skills in HIV preventive interventions may help them use condoms more consistently. This study
focused on partner type and suggesting condom use as moderators. Future studies may include
other moderators to identify the best scenario to promote condom use using intention.
However, these moderators were not associated with condom use for women in this study.
Nonetheless, concluding that partner type and suggesting the use of a condom may not be
effective moderators may be incorrect, because examining moderating effects by creating
interaction variables is often difficult (McClelland & Judd, 1993. Multigroup analyses represent
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 64
important alternative analytic strategies to test moderating effects. However, in multilevel path
analyses, to conduct multigroup analyses, the group should be the upper level. Because primary
partner was the lower-level variable in this study, multigroup analyses were not possible.
Therefore, to obtain conclusive evidence on moderators between intention and condom use for
homeless women, further studies should be conducted that permit multigroup analyses.
This study included three psychological variables to predict condom use intention based
on TRA and TPB. Significant associations of these variables with condom use differed by gender.
For men, condom attitudes were significantly associated with intention. Men who had positive
condom attitudes were more likely to intend to use a condom with sexual partners. Focusing on
increasing positive condom attitudes among homeless men can be an effective strategy to
promote intention to use a condom when they consider having sex with women. For women,
subjective norms regarding risky sexual behaviors were significantly associated with intention to
use a condom, although in the opposite direction as expected. Subjective norms regarding risky
sexual behaviors were expected to decrease the likelihood of intention to use a condom among
women. However, subjective norms increased the likelihood of intention to use a condom.
Similar results were found in a previous study. Having more network members who engaged in
risky sexual behavior decreased unprotected sex among homeless women (Kennedy et al.,
2010).It is difficult to provide a conclusive interpretation of this result but it may be possible
homeless women may be discouraged from engaging in unprotected sex by observing others
engaging in risky sexual behaviors and experiencing negative outcomes. To confirm this
hypothesis, further studies are strongly encouraged. However, it should be noted that the effects
of subjective norms on risky sexual behaviors were measured based on the perceptions of
participants. Thus, this information may not be accurate.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 65
Overall, this study supported the argument that intention to use a condom is a proximal
and important predictor of condom use among homeless people. Regardless of gender, intention
had a very strong association with condom use. This indicates that HIV prevention interventions
for homeless people should focus on increasing intention to use a condom. However, for men,
focusing on promoting positive condom attitudes may be particularly effective, as is promoting
consistent condom use even with primary partners.
Strengths and Limitations
This study contributed to the development of a theory-driven model to explain condom
use among homeless men and women. To overcome the methodological limitations of previous
studies, this study applied multilevel path modeling. The results of this study have important
implications for practice and the direction of future studies, as described earlier. Despite these
strengths, a few limitations should be noted. Although the two datasets used in this study were
collected using rigorously designed sampling procedures, they were collected in different
settings. Men’s data were collected at meal programs in Los Angeles, whereas women’s data
were collected in shelter setting in Los Angeles. This may cause differences in some background
variables (e.g., age, ethnicity and race). Those background variables were controlled when
analyzing data, but results should be interpreted in conjunction with this limitation. The study
model was developed based on TRA and TPB, but it is perhaps disingenuous to say that this
study rigorously tested the applicability of those theories to explaining condom use among
homeless people. The study did not included perceived control, as suggested in TPB (Ajzen,
1985), as a predictor of intention and condom use.
In addition, the cross-sectional nature of the data design should be mentioned.
Particularly in terms of condom use intention, a time-series design needs to be implemented.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 66
According to TRA and TPB, intention has been often measured as intention in the future or
during a certain time period. Future studies focusing on testing TRA and TPB to explain condom
use among homeless people should develop an advanced research design combining a time-
series design and multilevel modeling. This will result in more conclusive evidence to support
the development of effective HIV prevention interventions for this population. In addition, as
described in the previous chapter, the measurement of intention in this study may have been
limited. However, the association between intention and condom use in this study was
statistically significant with strong associations, providing support for the concepts of TRA and
TPB. Other studies have used similar measurements to indicate intention to use a condom among
other populations (Roth et al., 2011). As a recommendation for future studies, improved
measurement of intention to use a condom would improve the quality of evidence regarding the
applicability of TRA and TPB to homeless people.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 67
CHAPTER 5. CONCLUSIONS AND IMPLICATIONS
To prevent HIV infection among homeless people, social services professionals should
focus on improving consistent condom use. Studies have examined factors associated with
condom use among homeless people. However, few studies used TRA and TPB to explain
homeless people’s condom use. Some studies included some variables from TRA and TPV in
their study models, but it is still unclear if those theories have plausible applicability in terms of
explaining condom use among homeless people. Considering that TRA and TPB have been
shown to explain condom use among other populations and used to design HIV prevention
interventions, studying these theories with homeless people would provide evidence of condom
use behaviors and inform effective HIV interventions.
With this understanding, the current dissertation examined the association of main
variables from TRA and TPB with condom use among homeless people. To achieve this goal,
this dissertation involved two separate studies. The first study identified factors associated with
condom use among homeless women who intended to use a condom, using an event-level
approach. The second study examined a path model developed based on TRA and TPB to predict
homeless people’s condom use using multilevel path modeling.
The results from both studies showed that exploring intention to use a condom among
homeless people is important to understand their condom use. Logistic regression results from
Aim 1 showed that homeless women who had intention to use a condom during their most recent
sexual event showed more frequent condom use than women without intention. Multilevel path
analyses results also indicated that intention to use a condom was a strong indicator of consistent
condom use for both homeless men and women. In sum, individuals seeking to promote
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 68
consistent condom use among homeless people should consider promoting their intention to use
a condom.
Aim 2 examined the path to condom use with intention to use a condom as a mediator.
Based on TRA and TPB, a path model was developed and tested. Furthermore, two moderators
that may quantify the association between intention and condom use were included in the path
model. Intention to use a condom was a strong predictor of consistent condom use, regardless of
gender. However, predictors associated with intention differed by gender. For homeless men,
improving condom attitudes may promote their intention to use a condom. On the other hand, for
women, addressing subjective norms regarding risky sexual behavior may help to promote their
intention. In terms of moderators, statistically significant associations emerged only for men.
Encouraging homeless men to suggest using a condom to partners and considering partner type
when delivering HIV prevention interventions that focus on condom use intention may lead to
more consistent condom use among homeless men. No moderators were detected with homeless
women, but this does not mean that condom use among women is the sole function of intention.
There might be other explanations and further research is required.
The results of Aim 1 showed that there might be factors that promote or hinder condom
use, even when women reported intention to use a condom. The results showed that relationship-
related factors were significantly associated with condom use among women who had intention
to use a condom. Even if homeless women intend to use a condom, if those barriers are not
addressed, it might be difficult to expect better outcomes, including more frequent condom use
by women. Thus, addressing partner-related factors may encourage women to use condoms more
frequently.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 69
In sum, this study showed that TRA and TPB may be useful tools to explain condom use
among homeless people. In particular, this study addressed the importance of intention to use a
condom in terms of understanding condom use. Intention to use a condom is a core predictor that
is directly associated with condom use, according to TRA and TPB. Furthermore, this study
identified various factors that can quantify the association between intention and condom use.
The results indicated that intention can play a key role in promoting condom use among
homeless people and thus, HIV prevention interventions for homeless people need to focus on
improving their intention to use a condom.

RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 70
REFERENCES
Abraham, C., Sheeran, P., Norman, P., Conner, M., de Vries, N., & Otten, W. (1999). When
good intentions are not enough: Modeling postdecisional cognitive correlates of condom
use. Journal of Applied Social Psychology, 29, 2591–2612. doi:10.1111/j.1559-
1816.1999.tb00127.x
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J.
Beckman (Eds). Action control: From cognition to behavior (pp. 11–39). Berlin,
Germany: Springer-Verlag.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision
Processes, 50, 179–211. doi:10.1016/0749-5978(91)90020-T
Ajzen, I. (2002). Constructing a TPB questionnaire: Conceptual and methodological
considerations. Retrieved from Bielefeld University website:
http://www.unibielefeld.de/ikg/zick/ajzen%20construction%20a%20tpb%20questionnair
e.pdf
Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review
of empirical research. Psychological Bulletin, 84, 888–918. doi:10.1037/0033-
2909.84.5.888
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice-Hall.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behavior: Attitudes, intentions, and
perceived behavioral control. Journal of Experimental Social Psychology, 22, 453–474.
doi:10.1016/0022-1031(86)90045-4
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 71
Akobeng, A. K. (2008). Confidence intervals and p-values in clinical decision making. Acta
Paediatrica, 97, 1004–1007. doi:10.1111/j.1651-2227.2008.00836.x
Albarrací n, D., Johnson, B. T., Fishbein, M., & Muellerleile, P. A. (2001). Theories of reasoned
action and planned behavior as models of condom use: A meta-analysis. Psychological
Bulletin, 127, 142–161. doi:10.1037/0033-2909.127.1.142
Albarrací n, D., Gillette, J. C., Earl, A. N., Glasman, L. R., Durantini, M. R., & Ho, M.-H. (2005).
A test of major assumptions about behavior change: A comprehensive look at the effects
of passive and active HIV-prevention interventions since the beginning of the epidemic.
Psychological Bulletin, 131, 856–897. doi:10.1037/0033-2909.131.6.856
American Psychological Association. (2013). Health & homelessness. Retrieved from
http://www.apa.org/pi/ses/resources/publications/homelessness-health.aspx
Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behavior: A meta-
analytic review. British Journal of Social Psychology, 40, 471–499.
doi:10.1348/014466601164939
Bailey, S. L., Camlin, C. S., & Ennett, S. T. (1998). Substance use and risky sexual behavior
among homeless and runaway youth. Journal of Adolescent Health, 23, 378–388.
doi:10.1016/S1054-139X(98)00033-0
Bandawe, C. R., & Foster, D. (1996). AIDS-related beliefs, attitudes and intentions among
Malawian students in three secondary schools. AIDS Care, 8, 223–232.
doi:10.1080/09540129650125894
Bauer, D. J., Preacher, K. J., & Gil, K. M. (2006). Conceptualizing and testing random indirect
effects and moderated mediation in multilevel models: New procedures and
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 72
recommendations. Psychological Methods, 11, 142–163. doi:10.1037/1082-
989X.11.2.142
Bauman, L. J., & Berman, R. (2005). Adolescent relationships and condom use: Trust, love and
commitment. AIDS and Behavior, 9, 211–222. doi:10.1007/s10461-005-3902-2
Beijer, U., Wolf, A., & Fazel, S. (2012). Prevalence of tuberculosis, hepatitis C virus, and HIV in
homeless people: A systematic review and meta-analysis. Lancet Infectious Diseases, 12,
859–870. doi:10.1016/S1473-3099(12)70177-9
Blue, C. L. (1995). The predictive capacity of the theory of reasoned action and the theory of
planned behavior in exercise research: An integrated literature review. Research in
Nursing & Health, 18, 105–121. doi:10.1002/nur.4770180205
Bogale, G. W., Boer, H., & Seydel, E. R. (2010). Condom use among low-literate, rural females
in Ethiopia: The role of vulnerability to HIV infection, condom attitude, and self-efficacy.
AIDS Care, 22, 851–857. doi:10.1080/09540120903483026
Bogart, L. M., Kral, A. H., Scott, A., Anderson, R., Flynn, N., Gilbert, M. L., & Bluthenthal, R.
N. (2005). Condom attitudes and behaviors among injection drug users participating in
California syringe exchange programs. AIDS and Behavior, 9, 423–432.
doi:10.1007/s10461-005-9014-1
Bosompra, K. (2001). Determinants of condom use intentions of university students in Ghana:
An application of the theory of reasoned action. Social Science & Medicine, 52, 1057–
1069. doi:10.1016/S0277-9536(00)00213-6
Bowleg, L., Lucas, K. J., & Tschann, J. M. (2004). “The ball was always in his court”: An
exploratory analysis of relationship scripts, sexual scripts, and condom use among
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 73
African American women. Psychology of Women Quarterly, 28, 70–82.
doi:10.1111/j.1471-6402.2004.00124.x
Bryan, A. D., Aiken, L. S., & West, S. G. (1996). Increasing condom use: Evaluation of a
theory-based intervention to prevent sexually transmitted diseases in young women.
Health Psychology, 15, 371–382. doi:10.1037/0278-6133.15.5.371
Cabral, R. J., Pulley, L., Artz, L. M., Brill, I., & Macaluso, M. (1998). Women at risk of
HIV/STD: The importance of male partners as barriers to condom use. AIDS and
Behavior, 2, 75–85. doi:10.1023/A:1022363408723
Calsyn, D. A., Peavy, M., Wells, E. A., Campbell, A. N. C., Hatch-Maillette, M. A., Greenfield,
S. F., & Tross, S. (2013). Differences between men and women in condom use, attitudes,
and skills in substance abuse treatment seekers. American Journal on Addictions, 22,
150–157. doi:10.1111/j.1521-0391.2013.00312.x
Campbell, A. N. C., Tross, S., Hu, M.-C., Pavlicova, M., Kenney, J., & Nunes, E. V. (2011).
Female condom skill and attitude: Results from a NIDA Clinical Trials Network gender-
specific HIV risk reduction study. AIDS Education and Prevention, 23, 329–340.
doi:10.1521/aeap.2011.23.4.329
Catania, J. A., Coates, T. J., & Kegeles, S. (1994). A test of the AIDS risk reduction model:
Psychosocial correlates of condom use in the AMEN cohort survey. Health Psychology,
13, 548–555. doi:10.1037/0278-6133.13.6.548
Centers for Disease Control and Prevention. (2013a). Condoms and STDs: Fact sheet for public
health personnel. Retrieved from
http://www.cdc.gov/condomeffectiveness/docs/Condoms_and_STDS.pdf
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 74
Centers for Disease Control and Prevention. (2013b). HIV among women. Retrieved from
http://www.cdc.gov/hiv/pdf/risk_women.pdf
Civic, D. (1999). The association between characteristics of dating relationships and condom use
among heterosexual young adults. AIDS Education and Prevention, 11, 343–352.
Couture, M.-C., Soto, J. C., Akom, E., Joseph, G., & Zunzunegui, M.-V. (2010). Determinants of
intention to use condoms among clients of female sex workers in Haiti. AIDS Care, 22,
253–262. doi:10.1080/09540120903111478
Crosby, R. A., DiClemente, R. J., Wingood, G. M., Cobb, B. K., Harrington, K., Davies, S. L.,
… Oh, M. K. (2002). Condom use and correlates of African American adolescent
females’ infrequent communication with sex partners about preventing sexually
transmitted diseases and pregnancy. Health Education & Behavior, 29, 219–231.
doi:10.1177/109019810202900207
Elliott, M. N., Golinelli, D., Hambarsoomian, K., Perlman, J., & Wenzel, S. L. (2006). Sampling
with field burden constraints: An application to sheltered homeless and low-income
housed women. Field Methods, 18, 43–58. doi:10.1177/1525822X05284014
Ennett, S. T., Bailey, S. L., & Federman, E. B. (1999). Social network characteristics associated
with risky behaviors among runaway and homeless youth. Journal of Health and Social
Behavior, 40, 63–78. doi:10.2307/2676379
Fazekas, A., Senn, C. Y., & Ledgerwood, D. M. (2001). Predictors of intention to use condoms
among university women: An application and extension of the theory of planned behavior.
Canadian Journal of Behavioural Science, 33, 103–117. doi:10.1037/h0087133
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to
theory and research. MA: Addison-Wesley.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 75
Fishbein. M., & Middlestadt, S. (1989). Using the theory of reasoned action as a framework for
understanding and changing AIDS-related behavior. In V. M. Mays, G. W. Albee, & S.
Schneider (Eds). Primary prevention of AIDS: Psychological approaches. Newbury Park,
CA: Sage.
Fishbein, M. (2000). The role of theory in HIV prevention. AIDS Care, 12, 273–278.
doi:10.1080/09540120050042918
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to
theory and research. Reading, MA: Addison-Wesley.
Fisher, J. D., & Fisher, W. A. (2000). Theoretical approaches to individual-level change in HIV
risk behavior. In J. L. Peterson & R. J. DiClemente (Eds.), Handbook of HIV prevention
(pp. 3–55). New York, NY: Springer.
Fisher, W. A., Fisher, J. D., & Rye, B. J. (1995). Understanding and promoting AIDS-preventive
behavior: Insights from the theory of reasoned action. Health Psychology, 14, 255–264.
doi:10.1037/0278-6133.14.3.255
Fortenberry, D. J., Orr, D. P., Katz, B. P., Brizendine, E. J., & Blythe, M. J. (1997). Sex under
the influence: A diary self-report study of substance use and sexual behavior among
adolescent women. Sexually Transmitted Diseases, 24, 313–319. doi:10.1097/00007435-
199707000-00001
Galligan, R. F., & Terry, D. J. (1993). Romantic ideals, fear of negative implications, and the
practice of safe sex. Journal of Applied Social Psychology, 23, 1685–1711.
doi:10.1111/j.1559-1816.1993.tb01061.x
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 76
Gebhardt, W. A., Kuyper, L., & Greunsven, G. (2003). Need for intimacy in relationships and
motives for sex as determinants of adolescent condom use. Journal of Adolescent Health,
33, 154–164. doi:10.1016/S1054-139X(03)00137-X
Gibson, D. R., Lovelle-Drache, J., Young, M. T., & Chesney. M. (1992). HIV risk linked to
psychopathology in IV drug users. Paper presented at International Conference on AIDS,
Amsterdam, Netherlands.
Gómez, C. A., & Marí n, B. V. (1996). Gender, culture, and power: Barriers to HIV prevention
strategies for women. Journal of Sex Research, 33, 355–362.
doi:10.1080/00224499609551853
Green, H. D., Jr., Tucker, J. S., Wenzel, S. L., Golinelli, D., Kennedy, D. P., Ryan, G W., &
Zhou, A. J. (2012). Association of childhood abuse with homeless women’s social
networks. Child Abuse & Neglect, 36, 21–31. doi:10.1016/j.chiabu.2011.07.005
Greene, K., Hale, J. L., & Rubin, D. L. (1997). A test of the theory of reasoned action in the
context of condom use and AIDS. Communication Reports, 10, 21–33.
doi:10.1080/08934219709367656
Heeren, G. A., Jemmott, J. B., III, Mandeya, A., & Tyler, J. C. (2007). Theory-based predictors
of condom use among university students in the United States and South Africa. AIDS
Education and Prevention, 19, 1–12. doi:10.1521/aeap.2007.19.1.1
Hofmann, D. A., & Gavin, M. B. (1998). Centering decisions in hierarchical linear models:
Implications for research in organizations. Journal of Management, 24, 623–641.
doi:10.1177/014920639802400504
Janz, N. K., & Becker, M. H. (1984). The health belief model: A decade later. Health Education
& Behavior, 11, 1–47. doi:10.1177/109019818401100101
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 77
Jemmott, J. B., III, Jemmott, L. S., & Fong, G. T. (1998). Abstinence and safer sex HIV risk-
reduction interventions for African American adolescents: A randomized controlled trial.
Journal of the American Medical Association, 279, 1529–1536.
doi:10.1001/jama.279.19.1529
Jemmott, L. S., & Jemmott, J. B., III. (1991). Applying the theory of reasoned action to AIDS
risk behavior: Condom use among Black women. Nursing Research, 40, 228–234.
doi:10.1097/00006199-199107000-00008
Kalichman, S. C., Williams, E. A., Cherry, C., Belcher, L., & Nachimson, D. (1998). Sexual
coercion, domestic violence, and negotiating condom use among low-income African
American women. Journal of Women ’s Health, 7, 371–378. doi:10.1089/jwh.1998.7.371
Kapadia, F., Latka, M. H., Wu, Y., Strathdee, S. A., Mackesy-Amiti, M. E., Hudson, S. M., …
Garfein, R. S. (2011). Longitudinal determinants of consistent condom use by partner
type among young injection drug users: The role of personal and partner characteristics.
AIDS and Behavior, 15, 1309–1318. doi:10.1007/s10461-009-9569-3
Kennedy, D. P., Brown, R. A., Golinelli, D., Wenzel., S. L., Tucker, J. S., & Wertheimer, S. R.
(2013). Masculinity and HIV risk among homeless men in Los Angeles. Psychology of
Men & Masculinity, 14, 156–167. doi:10.1037/a0027570
Kennedy, D. P., Tucker, J. S., Green, H. D., Jr., Golinelli, D., & Ewing, B. (2012). Unprotected
sex of homeless youth: Results from a multilevel dyadic analysis of individual, social
network, and relationship factors. AIDS and Behavior, 16, 2015–2032.
doi:10.1007/s10461-012-0195-0
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 78
Kennedy, D. P., Wenzel, S. L., Brown, R., Tucker, J. S., & Golinelli, D. (2013). Unprotected sex
among heterosexually active homeless men: Results from a multi-level dyadic analysis.
AIDS and Behavior, 17, 1655–1667. doi:10.1007/s10461-012-0366-z
Kennedy, D. P., Wenzel., S. L., Tucker, J. S., Green, H. D., Jr., Golinelli, D., Ryan, G. W., …
Zhou, A. (2010). Unprotected sex of homeless women living in Los Angeles County: An
investigation of the multiple levels of risk. AIDS and Behavior, 14, 960–973.
doi:10.1007/s10461-009-9621-3
Kenny, D. A., Korchmaros, J. D., & Bolger, N. (2003). Lower level mediation in multilevel
models. Psychological Methods, 8, 115–128. doi:10.1037/1082-989X.8.2.115
Kippax, S., & Crawford, J. (1993). Flaws in the theory of reasoned action. In D. J. Terry, C.
Gallois, & M. McCamish (Eds.), The theory of reasoned action: Its application to AIDS-
preventive behavior (pp. 253–271). Tarrytown, NY: Pergamon Press.
Krull, J. L., & MacKinnon, D. P. (1999). Multilevel mediation modeling in group-based
intervention studies. Evaluation Review, 23, 418–444.
doi:10.1177/0193841X9902300404
Krull, J. L., & MacKinnon, D. P. (2001). Multilevel modeling of individual and group level
mediated effects. Multivariate Behavioral Research, 36, 249–277.
doi:10.1207/S15327906MBR3602_06
LaBrie, J., Earleywine, M., Schiffman, J., Pedersen, E., & Marriot, C. (2005). Effects of alcohol,
expectancies, and partner type on condom use in college males: Event-level analyses.
Journal of Sex Research, 42, 259–266. doi:10.1080/00224490509552280
Lau, J. T. F., Li, J., Choi, K. C., & Gu, J. (2014). A randomized controlled trial evaluating the
efficacy of a theory-based intervention promoting condom use among Chinese
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 79
monogamous female sexually transmitted infection patients. International Journal of
Behavioral Medicine. Advance online publication. doi:10.1007/s12529-014-9388-4
Leigh, B. C. (2002). Alcohol and condom use: A meta-analysis of event-level studies. Sexually
Transmitted Diseases, 29, 476–482. doi:10.1097/00007435-200208000-00008
Leigh, B. C., Vanslyke, J. G., Hoppe, M. J., Rainey, D. T., Morrison, D. M., & Rogers Gillmore,
M. (2008). Drinking and condom use: Results from an event-based daily diary. AIDS and
Behavior, 12, 104–112. doi:10.1007/s10461-007-9216-9
Liverpool, J., McGhee, M., Lollis, C., Beckford, M., & Levine, D. (2002). Knowledge, attitudes,
and behavior of homeless African-American adolescents: Implications for HIV/AIDS
prevention. Journal of the National Medical Association, 94, 257–263.
Maman, S., Campbell, J., Sweat, M. D., & Gielen, A. C. (2000). The intersections of HIV and
violence: Directions for future research and interventions. Social Science & Medicine, 50,
459–478. doi:10.1016/S0277-9536(99)00270-1
Manstead, A. S. R. (2011). The benefits of a critical stance: A reflection on past papers on the
theories of reasoned action and planned behaviour. British Journal of Social Psychology,
50, 366–373. doi:10.1111/j.2044-8309.2011.02043.x
Marsh, H. W., Lüdtke, O., Robitzsch, A., Trautwein, U., Asparouhov, T., Muthén, B. O., &
Nagengast, B. (2009). Doubly-latent models of school contextual effects: Integrating
multilevel and structural equation approaches to control measurement and sampling
errors. Multivariate Behavioral Research, 44, 764–802.
doi:10.1080/00273170903333665
Matson, P. A., Chung, S.-E., Sander, P., Millstein, S. G., & Ellen, J. M. (2012). The role of
feelings of intimacy on perceptions of risk for a sexually transmitted disease and condom
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 80
use in the sexual relationships of adolescent African-American females. Sexually
Transmitted Infections, 88, 617–621. doi:10.1136/sextrans-2012-050536
McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interactions and
moderator effects. Psychological Bulletin, 114, 376–390. doi:10.1037/0033-
2909.114.2.376
McMahon, J. M., Pouget, E. R., & Tortu, S. (2007). Individual and couple-level risk factors for
hepatitis C infection among heterosexual drug users: A multilevel dyadic analysis.
Journal of Infectious Diseases, 195, 1572–1581. doi:10.1086/516785
Montanaro, E. A., & Bryan, A. D. (2013). Comparing theory-based condom interventions:
Health belief model versus theory of planned behavior. Health Psychology. Advance
online publication. doi:10.1037/a0033969
Moore, J., Harrison, J. S., Kay, K. L., Deren, S., & Doll, L. S. (1995). Factors associated with
Hispanic women’s HIV-related communication and condom use with male partners.
AIDS Care, 7, 415–428. doi:10.1080/09540129550126371
Morokoff, P. J., Quina, K., Harlow, L. L., Whitmire, L., Grimley, D. M., Gibson, P. R., &
Burkholder, G. J. (1997). Sexual Assertiveness Scale (SAS) for women: Development
and validation. Journal of Personality and Social Psychology, 73, 790–804.
doi:10.1037/0022-3514.73.4.790
Morrison, D. M., Rogers Gillmore, M., & Baker, S. A. (1995). Determinants of condom use
among high-risk heterosexual adults: A test of the theory of reasoned action. Journal of
Applied Social Psychology, 25, 651–676. doi:10.1111/j.1559-1816.1995.tb01767.x
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 81
Morrison, D. M., Rogers Gillmore, M., Hoppe, M. J., Gaylord, J., Leigh, B.C., & Rainey, D.
(2003). Adolescent drinking and sex: Findings from a daily diary study. Perspectives on
Sexual and Reproductive Health, 35, 162–168. doi:10.1111/j.1931-2393.2003.tb00253.x
Morrison, D. M., Spencer, M. S., & Rogers Gillmore, M. (1998). Beliefs about substance use
among pregnant and parenting adolescents. Journal of Research on Adolescence, 8, 69–
95. doi:10.1207/s15327795jra0801_4
Mutchler, M. G., Bogart, L. M., Elliott, M. N., McKay, T., Suttorp, M. J., & Schuster, M. A.
(2008). Psychosocial correlates of unprotected sex without disclosure of HIV-positivity
among African-American, Latino, and White men who have sex with men and women.
Archives of Sexual Behavior, 37, 736–747. doi:10.1007/s10508-008-9363-8
Muthén, L. K., & Muthén, B. O. (2006). Mplus: The comprehensive modeling program for
applied researchers: User ’s guide. Los Angeles, CA: Muthén & Muthén.
National Coalition for the Homeless. (2007). HIV/AIDS and homelessness (NCH Fact Sheet No.
9). Retrieved from www.nationalhomeless.org/publications/facts/hiv.html
National Coalition for the Homeless. (2009). HIV/AIDS and homelessness. Retrieved from:
http://www.nationalhomeless.org/factsheets/hiv.html
Noar, S. M., Carlyle, K., & Cole, C. (2006). Why communication is crucial: Meta-analysis of the
relationship between safer sexual communication and condom use. Journal of Health
Communication, 11, 365–390. doi:10.1080/10810730600671862
Noar, S. M., Webb, E., Van Stee, S., Feist-Price, S., Crosby, R., Willoughby, J. F., & Troutman,
A. (2012). Sexual partnerships, risk behaviors, and condom use among low-income
heterosexual African American: A qualitative study. Archives of Sexual Behavior, 41,
959–970. doi:10.1007/s10508-011-9890-6
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 82
Nyamathi, A., Leake, B., Keenan, C., & Gelberg, L. (2000). Type of social support among
homeless women: Its impact on psychosocial resources, health and health behaviors, and
use of health services. Nursing Research, 49, 318–326. doi:10.1097/00006199-
200011000-00004
Nyamathi, A. M., Lewis, C., Leake, B., Flaskerud, J., & Bennett, C. (1995). Barriers to condom
use and needle cleaning among impoverished minority female injection drug users and
partners of injection drug users. Public Health Reports, 110, 166–172.
Nyamathi, A., & Vasquez, R. (1989). Impact of poverty, homelessness, and drugs on Hispanic
women at risk for HIV infection. Hispanic Journal of Behavioral Sciences, 11, 299–314.
doi:10.1177/07399863890114001
Peltzer, K. (2000). Factors affecting condom use among South African university students. East
African Medical Journal, 77, 46–52.
Plichta, S. B., Weisman, C. S., Nathanson, C. A., Ensminger, M. E., & Robinson, J. C. (1992).
Partner-specific condom use among adolescent women clients of a family planning clinic.
Journal of Adolescent Health, 13, 506–511. doi:10.1016/1054-139X(92)90015-4
Pulerwitz, J., Amaro, H., De Jong, W., Gortmaker, S. L., & Rudd, R. (2002). Relationship power,
condom use and HIV risk among women in the USA. AIDS Care, 14, 789–800.
doi:10.1080/0954012021000031868
Rew, L., Fouladi, R. T., & Yockey, R. D. (2002). Sexual health practices of homeless youth.
Journal of Nursing Scholarship, 34, 139–145. doi:10.1111/j.1547-5069.2002.00139.x
Rhoades, H., Wenzel, S. L., Golinelli, D., Tucker, J. S., Kennedy, D. P., Green, H. D., & Zhou,
A. (2011). The social context of homeless men’s substance use. Drug and Alcohol
Dependence, 118, 320–325. doi:10.1016/j.drugalcdep.2011.04.011
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 83
Robertson, M. J., Clark, R. A., Charlebois, E. D., Tulsky, J., Long, H. L., Bangsberg, D. R., &
Moss, A. R. (2004). HIV seroprevalence among homeless and marginally housed adults
in San Francisco. American Journal of Public Health, 94, 1207–1217.
doi:10.2105/AJPH.94.7.1207
Robertson, M. J., Zlotnick, C., & Westerfelt, A. (1997). Drug use disorders and treatment contact
among homeless adults in Alameda County, California. Journal of American Public
Health, 87, 221–228. doi:10.2105/AJPH.87.2.221
Roth, E. A., Ngugi, E. N., & Jansson, M. (2011). Pathways and barriers to condom use among
Ariaal agro-pastoralists of northern Kenya. Sexuality Research and Social Policy, 8, 93–
102. doi:10.1007/s13178-010-0033-y
Ryan, G. W., Stern, S. A., Hilton, L. Tucker, J. S., Kennedy, D. P., Golinelli, D., & Wenzel, S. L.
(2009). When, where, why and with whom homeless women engage in risky sexual
behaviors: A framework for understanding complex and varied decision-making
processes. Sex Roles, 61, 536–553. doi:10.1007/s11199-009-9610-z
Sanderson, C. A. (1999). Role of relationship context in influencing college students’
responsiveness to HIV prevention videos. Health Psychology, 18, 295–300.
doi:10.1037/0278-6133.18.3.295
Sanderson, C. A., & Jemmott, J. B., III. (1996). Moderation and mediation of HIV-prevention
interventions: Relationship status, intentions, and condom use among college students.
Journal of Applied Social Psychology, 26, 2076–2099. doi:10.1111/j.1559-
1816.1996.tb01788.x
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 84
Sheeran, P., & Orbell, S. (1998). Do intentions predict condom use? Metaanalysis and
examination of six moderator variables. British Journal of Social Psychology, 37, 231–
250. doi:10.1111/j.2044-8309.1998.tb01167.x
Sheeran, P., & Taylor, S. (1999). Predicting intentions to use condoms: A meta-analysis and
comparison of the theories of reasoned action and planned behavior. Journal of Applied
Psychology, 29, 1624–1675. doi:10.1111/j.1559-1816.1999.tb02045.x
Sherman, S. G., & Latkin, C. A. (2001). Intimate relationship characteristics associated with
condom use among drug users and their sex partners: A multilevel analysis. Drug and
Alcohol Dependence, 64, 97–104. doi:10.1016/S0376-8716(00)00236-2
Sikkema, K. J., Heckman, T. G., Kelly, J. A., Anderson, E. S., Winett, R. A., Solomon, L. J., …
Mercer, M. B. (1996). HIV risk behaviors among women living in low-income, inner-city
housing developments. American Journal of Public Health, 86, 1123–1128.
doi:10.2105/AJPH.86.8_Pt_1.1123
Somlai, A. M., Kelly, J. A., Wagstaff, D. A., & Whitson, D. P. (1998). Patterns, predictors, and
situational contexts of HIV risk behaviors among homeless men and women. Social Work,
43, 7–20. doi:10.1093/sw/43.1.7
Song, J. (1999). HIV/AIDS & homelessness: Recommendations for clinical practice and public
policy. Nashville, TN: National Health Care for the Homeless Council, Health Care for
the Homeless Clinicians’ Network. Retrieved from http://www.nhchc.org/wp-
content/uploads/2011/10/AIDS-Homelessness.pdf
Stall, R. D., Paul, J. P., Barrett, D. C., Crosby, G. M., & Bein, E. (1999). An outcome evaluation
to measure changes in sexual risk-taking among gay men undergoing substance use
disorder treatment. Journal of Studies on Alcohol, 60, 837–845.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 85
Stark, M. J., Tesselaar, H. M., O’Connell, A. A., Person, B., Galavotti, C., Cohen, A., & Walls,
C. (1998). Psychosocial factors associated with the stages of change for condom use
among women at risk for HIV and STDs: Implications for intervention development.
Journal of Consulting and Clinical Psychology, 66, 967–978. doi:10.1037/0022-
006X.66.6.967
Sugerman, S. T., Hergenroeder, A. C., Chacko, M. R., & Parcel, G. S. (1991). Acquired
immunodeficiency syndrome and adolescents: Knowledge, attitudes, and behaviors of
runaway and homeless youths. American Journal of Diseases of Children, 145, 431–436.
doi:10.1001/archpedi.1991.02160040089014
Testa, M., & Collins, R. L. (1997). Alcohol and risky sexual behavior: Event-based analyses
among a sample of high-risk women. Psychology of Addictive Behaviors, 11, 190–201.
doi:10.1037/0893-164X.11.3.190
Thompson, S. C., Kyle, D., Swan, J., Thomas, C., & Vrungos, S. (2002). Increasing condom use
by undermining perceived invulnerability to HIV. AIDS Education and Prevention, 14,
505–514. doi:10.1521/aeap.14.8.505.24115
Tucker, J. S., Elliott, M. N., Wenzel, S. L., & Hambarsoomian, K. (2007). Relationship
commitment and its implications for unprotected sex among impoverished women living
in shelters and low-income housing in Los Angeles County. Health Psychology, 26, 644–
649. doi:10.1037/0278-6133.26.5.644
Tucker, J. S., Kennedy, D., Ryan, G., Wenzel., S. L., Golinelli, D., & Zazzali, J. (2009).
Homeless women’s personal networks: Implications for understanding risk behavior.
Human Organization, 68, 129–140.
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 86
Tucker, J. S., Ryan, G. W., Golinelli, D., Ewing, B., Wenzel, S. L., Kennedy, D. P., … Zhou, A.
(2012). Substance use and other risk factors for unprotected sex: Results from an event-
based study of homeless youth. AIDS and Behavior, 16, 1699–1707. doi:10.1007/s10461-
011-0017-9
Tucker, J. S., Wenzel, S. L., Golinelli, D., Kennedy, D. P., Ewing, B., & Wertheimer, S. (2013).
Understanding heterosexual condom use among homeless men. AIDS and Behavior, 17,
1637–1644. doi:10.1007/s10461-012-0165-6
Tucker, J. S., Wenzel, S. L., Straus, J. B., Ryan, G. W., & Golinelli, D. (2005). Experiencing
interpersonal violence: Perspectives of sexually active, substance-using women living in
shelters and low-income housing. Violence Against Women, 11, 1319–1340.
doi:10.1177/1077801205280190
Tyler, K. A., & Melander, L. A. (2011). A qualitative study of the formation and composition of
social networks among homeless youth. Journal of Research on Adolescence, 21, 802–
817. doi:10.1111/j.1532-7795.2011.00739.x
Tyson, M., Covey, J., & Rosenthal, H. E. S. (2014). Theory of planned behavior interventions for
reducing heterosexual risk behaviors: A meta-analysis. Health Psychology. Advance
online publication. doi:10.1037/hea0000047
Valente, A. M., & Auerswald, C. L. (2013). Gender differences in sexual risk and sexually
transmitted infections correlate with gender differences in social networks among San
Francisco homeless youth. Journal of Adolescent Health, 53, 486–491.
doi:10.1016/j.jadohealth.2013.05.016
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 87
VanderDrift, L. E., Agnew, C. R., Harvey, S. M., & Warren, J. T. (2013). Whose intentions
predict? Power over condom use within heterosexual dyads. Health Psychology, 32,
1038–1046. doi:10.1037/a0030021
Warwick, P., Terry, D., & Gallois, C. (1993). Extending the theory of reasoned action: The role
of health beliefs. In D. J. Terry, C. Gallois, & M. McCamish (Eds.), The theory of
reasoned action: Its application to AIDS-preventive behavior (pp. 117–134). Tarrytown,
NY: Pergamon Press.
Webb, T. L., & Sheeran, P. (2006). Does changing behavioral intentions engender behavior
change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132, 249–
268. doi:10.1037/0033-2909.132.2.249
Wechsberg, W. M., Lam, W. K. K., Zule, W., Hall, G., Middlesteadt, R., & Edwards, J. (2003).
Violence, homelessness, and HIV risk among crack-using African-American women.
Substance Use & Misuse, 38, 669–700. doi:10.1081/JA-120017389
Weinhardt, L. S., & Carey, M. P. (2000). Does alcohol lead to sexual risk behavior? Findings
from event-level research. Annual Review of Sex Research, 11, 125–157.
doi:10.1080/10532528.2000.10559786
Weinhardt, L. S., Carey, M. P., Maisto, S., Carey, K. B., Cohen, M. M., & Wickramasinghe, S.
M. (1998). Reliability of the Timeline Follow-Back sexual behavior interview. Annals of
Behavior Medicine, 20, 25–30. doi:10.1007/BF02893805
Wenzel, S. L., Green, H. D., Jr., Tucker, J. S., Golinelli, D., Kennedy, D. P., Ryan, G., & Zhou,
A. (2009). The social context of homeless women’s alcohol and drug use. Drug and
Alcohol Dependence, 105, 16–23. doi:10.1016/j.drugalcdep.2009.05.026
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 88
Wenzel, S. L., Rhoades, H., Hsu, H.-T., Golinelli, D., Tucker, J. S., Kennedy, D. P., … Ewing, B.
(2012). Behavioral health and social normative influence: Correlates of concurrent sexual
partnering among heterosexually-active homeless men. AIDS and Behavior, 16, 2042–
2050. doi:10.1007/s10461-011-0066-0
Wenzel, S. L., Tucker, J. S., Elliott, M. N., Hambarsoomians, K., Perlman, J., Becker, K., …
Golinelli, D. (2004). Prevalence and co-occurrence of violence, substance use and
disorder, and HIV risk behavior: A comparison of sheltered and low-income housed
women in Los Angeles County. Preventive Medicine, 39, 617–624.
doi:10.1016/j.ypmed.2004.02.027
Westercamp, N., Mattson, C. L., Madonia, M., Moses, S., Agot, K., Ndinya-Achola, J. O., …
Bailey, R. C. (2010). Determinants of consistent condom use vary by partner type among
young men in Kisumu, Kenya: A multi-level data analysis. AIDS and Behavior, 14, 949–
959. doi:10.1007/s10461-008-9458-1
Widman, L., Golin, C. E., Grodensky, C. A., & Suchindran, C. (2013). Do safer sex self-efficacy,
attitudes toward condoms, and HIV transmission risk beliefs differ among men who have
sex with men, heterosexual men, and women living with HIV? AIDS and Behavior, 17,
1873–1882. doi:10.1007/s10461-011-0108-7
Widman, L., Golin, C. E., & Noar, S. M. (2013). When do condom use intentions lead to actions?
Examining the role of sexual communication on safer sexual behavior among people
living with HIV. Journal of Health Psychology, 18, 507–517.
doi:10.1177/1359105312446769
Wingood, G. M., & DiClemente, R. J. (1998). Gender-related correlates and predictors of
consistent condom use among young adult African-American women: A prospective
RISKY SEXUAL BEHAVIOR OF HOMELESS PEOPLE 89
analysis. International Journal of STD & AIDS, 9, 139–145.
doi:10.1258/0956462981921891
Wingood, G. M., & DiClemente, R. J. (2000). Application of the theory of gender and power to
examine HIV-related exposures, risk factors, and effective interventions for women.
Health Education & Behavior, 27, 539–565. doi:10.1177/109019810002700502
Zhang, Z., Zyphur, M. J., & Preacher, K. J. (2009). Testing multilevel mediation using
hierarchical linear models: Problems and solutions. Organizational Research Methods,
12, 695–719. doi:10.1177/1094428108327450

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Social network engagement and HIV risk among homeless former foster youth

PDF

Social network norms and HIV risk behaviors among homeless youth in Los Angeles, California

PDF

Factors and correlates of sexual behaviors among young adults from continuation high schools

PDF

Engaging homeless men and shelter providers to adapt an existing evidence-based HIV prevention intervention

PDF

Couch-surfing among youth experiencing homelessness: an examination of HIV risk

PDF

Homelessness and substance use treatment: using multiple methods to understand risks, consequences, and unmet treatment needs among young adults

PDF

Sex talks: an examination of young Black women's social networks, sexual health communication, and HIV prevention behaviors

PDF

Discrimination at the margins: perceived discrimination and the role of social support in mental health service use for youth experiencing homelessness

PDF

Complicated relationships: the prevalence and correlates of sexual assault, dating violence victimization, and minority stress among sexual minority adolescents throughout the United States

PDF

The algorithm of Black homelessness: a classification tree analysis

PDF

The influence of a humorous intervention on parent-child sexual communication

PDF

A penta-dimensional longitudinal analysis of the predictors of compulsive internet use among adolescents using linear mixed model (LMM)

PDF

The link between maternal depression and adolescent daughters' risk behavior: the mediating and moderating role of family

PDF

A sociocultural and developmental approach to intimate partner violence among a sample of Hispanic emerging adults

PDF

The role of depression symptoms on social information processing and tobacco use among adolescents

PDF

The influence of organizational ethics on job attitudes and government performance

PDF

Transitional housing and wellness center: a holistic approach to decreasing homelessness and mental illness in the Black community

PDF

Community integration of individuals with serious mental illness: a network perspective from India and United States

PDF

Using a structural model of psychopathology to distinguish relations between shared and specific features of psychopathology, smoking, and underlying mechanisms

PDF

My Life, My Wishes, a death education program for community-dwelling Chinese older adults: development and pilot testing

Asset Metadata

Creator Song, Ahyoung (author)

Core Title A closer look at intention to use condoms and risky sexual behaviors among homeless people: application of the theories of reasoned action and planned behavior

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work

Publication Date 08/29/2014

Defense Date 06/06/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag condom use,HIV,homeless,OAI-PMH Harvest,sexual behaviors,theory of planned behavior,theory of reasoned action

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wenzel, Suzanne L. (committee chair), Rice, Eric R. (committee member), Sussman, Steven (committee member)

Creator Email ahyoung.song80@gmail.com,ahyoungs@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-466953

Unique identifier UC11287841

Identifier etd-SongAhyoun-2867.pdf (filename),usctheses-c3-466953 (legacy record id)

Legacy Identifier etd-SongAhyoun-2867.pdf

Dmrecord 466953

Document Type Dissertation

Format application/pdf (imt)

Rights Song, Ahyoung

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA