Repository logo

Browsing by Author "Prince, Mark A., committee member"

Now showing 1 - 5 of 5
  • Thumbnail Image
    ItemOpen Access
    Deep transfer learning for prediction of health risk behaviors in adolescent psychiatric patients
    (Colorado State University. Libraries, 2021) Kentopp, Shane, author; Conner, Bradley T., advisor; Prince, Mark A., committee member; Henry, Kimberly L., committee member; Anderson, Charles W., committee member
    Binge drinking and non-suicidal self-injury are significant health-risk behaviors that are often initiated during adolescence and contribute to a host of negative outcomes later in life. Selective prevention strategies are targeted toward individuals most at-risk for developing these behaviors. Traditionally, selective interventions are tailored based on risk factors identified by human experts. Machine learning algorithms, such as deep neural networks, may improve the effectiveness of selective interventions by accounting for complex interactions between large numbers of predictor variables. However, their use in psychological research is limited due to the tendency to overfit and the need for large volumes of training data. Deep transfer learning can overcome this limitation by leveraging samples of convenience to facilitate training deep neural networks in small, clinically relevant samples. The author trained deep neural networks on data from a sample of adolescent psychiatric inpatients to retrospectively classify individuals according to their history of alcohol misuse and nonsuicidal self-injury. Next, the performance of these models was compared to deep neural networks that were pretrained in a convenience sample of college undergraduates and fine-tuned in the sample of psychiatric patients. Deep transfer learning did not improve classification accuracy but buffered against overfitting. The deep neural networks that were not pretrained maintained maximum classification accuracy for a very small number of training epochs before performance deteriorated due to overfitting the training data. Conversely, the pretrained networks maintained their maximum classification accuracy across many training epochs and performance was not hindered by overfitting. This suggests that convenience samples can be utilized to reduce the risk of overfitting when training complex deep neural networks on small clinical samples. In the future, this process may be employed to facilitate powerful predictive models that inform selective prevention programs and contribute to the reduction of health risk behavior prevalence amongst vulnerable adolescent populations.
  • Thumbnail Image
    ItemOpen Access
    Development and initial validation of the comprehensive HIV adherence with treatment scale
    (Colorado State University. Libraries, 2021) Rahm-Knigge, Ryan L., author; Conner, Bradley T., advisor; Luong, Gloria, committee member; Harman, Jennifer J., committee member; Prince, Mark A., committee member
    HIV remains a significant public health concern despite decreasing rates of transmission in the U.S. (Centers for Disease Control and Prevention, 2020c). Contributing factors include low rates of treatment adherence (de Bruin et al., 2010) and high rates of comorbidities with other medical and mental health conditions (Bing et al., 2001; Gallant et al., 2017; Lerner et al., 2020). Antiretroviral therapy (ART) has significantly improved HIV health outcomes and reduced AIDS diagnoses and AIDS-related mortality (Crum et al., 2006; Glass et al., 2008; Ickovics & Meade, 2002; Paterson et al., 2000; Stone, 2001; World Health Organization, 2015). Because of ART's effectiveness, HIV is considered a chronic rather than terminal health condition for people adherent with treatment (Aberg, 2006; Swendeman, Ingram, & Rotheram- Borus, 2009). Treatment for HIV as a chronic health condition includes several pro-health behaviors in addition to ART adherence to support overall wellness. To support future research and treatment recommendations, the current study developed a measure of adherence with pro- health behavior and conducted an initial analysis of the measure's psychometric properties with a sample of 118 people living with HIV. Structural equation modeling explored relations among antecedents (personality, treatment self-efficacy, treatment information, and treatment motivation) and health outcomes of pro-health behaviors and ART adherence. Regularly assessing engagement in, as well as antecedents and outcomes of, treatment behaviors can enhance communication between providers and people living with HIV, reinforce HIV's status as a manageable chronic condition, and link people living with HIV to appropriate interventions.
  • Thumbnail Image
    ItemOpen Access
    The effects of personality and social determinants of health on sports-related concussion risk: an examination of symptom reporting, concussion incidence, and return to play
    (Colorado State University. Libraries, 2024) Weishaar, Megan Gardner, author; Conner, Bradley T., advisor; Prince, Mark A., committee member; Davalos, Deana B., committee member; Stephens, Jaclyn A., committee member
    Sports-related concussion (SRC), a form of mild traumatic brain injury (TBI) that occurs during sport participation or recreation, has emerged as a public health concern among adolescent athletes in the United States in recent decades due to increasing knowledge of high incidence rates. Important factors such as return to play (RTP) timeline and symptom reporting behaviors have been shown to contribute to SRC incidence. As a result, recent research has called for work that identifies a parsimonious group of variables that are salient in identifying athletes most at risk of outcomes such as early RTP, higher SRC incidence, and symptom nondisclosure. A small body of work has established that certain personality characteristics and social determinants of health are associated with these SRC outcomes. However, few specific personality and social determinants of health factors have been examined, few studies have investigated predictors of SRC incidence and RTP outcomes specifically, and a small number of these studies have investigated these research questions among adolescent athletes despite high rates of athletic participation and SRC incidence. The current study sought to expand previous work to further elucidate relations between a myriad of personality and social determinants of health predictors, and time to RTP, diagnosed SRC, and symptom reporting outcomes among adolescent athletes (N = 317, ages 14-19). The ultimate goal of the study was to identify personality and social determinants of health factors most salient in SRC outcomes to inform practical tailoring of prevention and intervention strategies. Participants were adolescents who were currently enrolled in high school or college or were enrolled in high school within the past calendar year, participated in a high school or club sport in the past year, and either denied sustaining a diagnosed SRC in the past year (N = 156) or endorsed sustaining a diagnosed SRC in the past year (N = 161). Data were collected through a one-time online survey that took participants 20-30 minutes to complete. Participants were compensated via an online gift card or course credit. Recruitment methods included outpatient medical clinics, high school athletic trainers, sports gyms and tournaments, Twitter, and university research pools. Analyses included a series of Cox proportional hazards and logistic regression models investigating associations between personality and social determinants of health predictors and time to RTP, diagnosed SRC incidence, and symptom reporting outcomes (i.e., honesty and comfort in symptom reporting). Overall, study results bolstered the conclusion that 1. Personality and social determinants of health factors are important in predicting SRC outcomes and 2. The relations between personality and SRC outcomes are highly nuanced, depending upon specific personality facets, outcome, and sample characteristics. Study results inform athletes, peers and family of athletes, coaches, athletic trainers, and other medical professionals about which personality and social determinants of health variables are most salient in SRC risk, thus informing prevention and intervention efforts for SRC such that they can be more personalized and tailored at the individual, environmental, and systems levels.
  • Thumbnail Image
    ItemOpen Access
    The protective role of awe on self-focused attention and depressive symptoms
    (Colorado State University. Libraries, 2023) Sung, Angelina, author; Steger, Michael F., advisor; Emery, Noah N., advisor; Prince, Mark A., committee member; Luong, Gloria, committee member
    Depression is a public health concern that negatively affects millions of individuals living in the United States. Maladaptive self-focused attention has been found to strongly predict the onset and maintenance of depressive symptoms. Awe is a positive emotion that may protect against this cognitive vulnerability given its ability to elicit the "small self." The present study examined whether awe buffered against the impact of self-focus on depressive symptoms at both the trait and state level. A sample of 286 students were recruited from an undergraduate research pool. Participants completed an online survey that assessed for different measures of self-focus (i.e., first-person singular pronouns, rumination), trait positive emotions, depressive symptoms, and positive and negative affect. One week later, they were randomized to watch an awe-eliciting video or amusement-eliciting video and subsequently completed another survey that assessed for rumination, "small self" feelings, state positive emotions, and positive and negative affect. Results indicated mixed findings. At the trait level, depressive symptoms were positively associated with rumination (but not first-person singular pronouns). This positive association between self-focus and depressive symptoms was attenuated as levels of awe increased when using the measure of rumination, but not the measure of first-person singular pronouns. At the state level, rumination was positively associated with "small self" feelings. In addition, no interaction was found between the effects of the awe inductions and "small self" feelings in relation to rumination. While researchers have proposed that awe has a potential therapeutic role in depression, this study appears to be the first to provide empirical support at the trait level. Future research should examine the relationship between rumination and "small self" feelings or consider more potent elicitors of awe when understanding its effects at the state level.
  • Thumbnail Image
    ItemOpen Access
    You can feel good: positive outcomes of marijuana use
    (Colorado State University. Libraries, 2021) Parnes, Jamie E., author; Conner, Bradley T., advisor; Prince, Mark A., committee member; Swaim, Randall C., committee member; Riggs, Nathaniel R., committee member
    To date, limited marijuana research has focused on identifying reinforcing outcomes related to use, often perceived as positive outcomes. Operant conditioning and social learning theories suggest that the reinforcing aspects of substance use are a primary contributor to maintained use, and in turn, risk of dependence. Individuals who use marijuana report expecting positive outcomes which motivates use; however, the occurrence of such positive outcomes are rarely examined. Moreover, research has yet to develop a reliable, validated measure of positive marijuana-related outcomes. The present study sought to develop and psychometrically evaluate such a measure. I hypothesized that: 1) positive outcomes would be positively associated with marijuana use, positive expectancies, and negative outcomes, 2-3) positive outcomes would be unrelated to alcohol use and positive alcohol outcomes, and 4) positive outcomes would account for unique variance in recent use, controlling for expectancies and negative outcomes. Scale items were developed using inductive and deductive methods. College students (N = 883) and community adults (N = 214) completed a survey measuring marijuana use frequency, positive outcomes, expectancies, and negative consequences. Exploratory and confirmatory factor analysis (CFA) tested scale dimensionality and structure in the college sample and community sample. The final factor structure achieved excellent model fit (CFI = .96-.97, RMSEA = .03-.05) and internal consistency (ω = .84-.90). Four factors emerged from the data: Social Enhancement, Mood Enhancement and Relaxation, Perceptual Enhancement, and Sexual Enhancement. Invariance testing supported configural invariance between the two samples. Study hypotheses supporting scale validity were largely upheld. Positive outcomes were positively associated with recent use, controlling for expectancies and negative outcomes. Positive outcomes were also either unassociated or negatively associated with alcohol use, and unassociated or weakly associated with alcohol positive outcomes. Positive outcomes were also differentiated from positive expectancies and more influential in predicting typical use frequency. Findings implicate that positive outcomes are an important factor in explaining recent marijuana use, necessitating the need for future longitudinal use to understand their role in maintained use and dependence. Additionally, positive outcomes can be a target for clinical interventions by informing replacement behaviors or enhancing motivational interviewing techniques.