Reading comprehension is a dynamic cognitive process governed by numerous factors; however, previous research has examined only a few factors at a time, overlooking their dynamic interactions. Additionally, most research has focused on isolated sentences, limiting the generalizability to real-world reading. Thus, this study characterizes the processing of full stories and uses hypothesis-free modeling to explore a broad space of linguistic factors. I analyzed self-paced reading times (SPRTs), recorded as 181 participants read up to 10 stories word by word, pressing a button to advance. The stories look like real-world texts, but manipulate comprehension difficulty and are thus ideal for characterizing language processing. Using Bayesian Non-negative Matrix Factorization, I deconstructed the time-series of SPRTs from half the stories into factors that capture variance in reading times and represent distinct cognitive processes. I then determined which of over 100 known psycholinguistic variables contributed to each factor via a regularized regression. To validate my findings, I determined how well those factors predict the SPRTs from the remaining stories. This study will identify well-documented factors influencing reading times, such as word frequency, but also novel and complex interactions between sentence structure and meaning. The findings can help to establish what cognitive processes shape real-world comprehension as it unfolds, informing future work in educational practices, accessibility tools, and cognitive models.

Moderate-to-severe premenstrual syndrome (PMS) symptoms are associated with an increased risk of developing alcohol use disorder (AUD) in females. This study analyzes the relationship between AUD and premenstrual symptom severity in females. We hypothesize that females with AUD will report greater premenstrual pain and negative affect severity than healthy controls. This secondary analysis included 45 premenopausal females (ages 18 to 45), with 13 females using hormonal contraceptives. Participants were clinically assessed for AUDs, with 22 participants meeting the criteria for current moderate-to-severe AUD and 23 matched controls with no AUD. Premenstrual pain and negative affect symptom severity was self-reported via subscales of the Shortened Premenstrual Assessment Form. We used independent t-tests to compare group (AUD vs. control) differences in premenstrual pain and negative affect severity. Females with AUD reported greater premenstrual pain (t(37) = 3.13,p = 0.003) and negative affect severity (t(43) = 2.13, p=0.039 ) than healthy controls. Our results suggest that females with AUD report greater premenstrual pain and negative affect severity than health controls. These findings highlight the positive association between heavy alcohol use and premenstrual symptom severity. Significance: The significance of the results is that premenstrual symptomatology is associated with AUD. Future research should examine premenstrual pain and negative affect symptoms as determinants of alcohol use in females.

Previous research shows that caregiver’s use of emotion words helps children develop emotion categories and emotional regulation skills. Emotional vocabulary is necessary for children to understand their own and others’ emotions. However, cultural norms affect how caregivers talk to boys versus girls. Previous studies have shown that mothers are more likely to mention feelings of sadness around daughters than sons. Since language impacts children’s ability to understand and regulate emotions, differences in caregiver’s use of emotion language by genders could contribute to differences in how men and women express their emotions later in life. Using the transcribed conversations between children and their caregivers from the CHILDES database, this study will explore whether parents use emotion words differently with boys and girls aged 2-7 during free play. My results show that the emotion words heard by children increase with age, but there is no statistically significant difference in the number of emotion words boys and girls hear in any age group.The results contribute to our understanding of children’s early language environment and demonstrates that parents use emotion words similarly for boys and girls.

Posttraumatic stress disorder (PTSD) and depression are prevalent in survivors of strokes and transient ischemic attacks (TIA), but the degree to which people seek treatment is unclear. We investigated the frequency of probable depression and PTSD in patients one month after stroke/TIA hospitalization, along with associations with self-reported counseling and medication treatment. A sample of 77 adults (mean age=61.3[SD=16.4]; 53.2% female; 37.7% Hispanic) hospitalized after evaluation for stroke/TIA were enrolled in the REACH Stroke study. Demographic information was collected in-hospital, and NIH Stroke Severity Scale scores were extracted from medical records. During a 1-month follow-up, participants reported their depressive symptoms and PTSD symptoms, and whether they sought counselling and/or medications to cope with their emotional experience in the past month. Approximately 1 month after the stroke/TIA event, 15.6% (n=12) of participants met for probable PTSD and/or depression, and 23.3% (n=18) reported using counseling or medications to cope. The percentage of participants with probable PTSD and/or depression who accessed counseling/medication (50%) was significantly greater than that for those without probable PTSD and depression (15.6%, Fisher’s exact test p=.03). Our study suggests that stroke/TIA patients are much more likely to seek treatment when they experience likely clinically significant levels of PTSD and/or depressive symptoms, but those with symptoms below clinical threshold still access services often.

Early-life environments have been found to exert a lasting influence on health and well-being into adulthood. Past research finds that early-life adversity predicts poor mental health outcomes, impacting vital social and emotional processes. We first investigated if cross-sectional associations between childhood unpredictability (CU) and perceived stress (PS) in adulthood were mediated by emotional regulation and perceived emotional support in a sample of 438 adults participating in a RCT designed to reduce PS levels. We found that higher levels of CU predicted both poor emotional regulation and low levels of perceived emotional support, both of which, in turn, predicted high levels of PS. Next, we investigated how these mediators were associated with change in PS levels from baseline to post-intervention. We found that emotion regulation skills, but not perceived emotional support, mediated the association between CU and PS levels. These results suggest that higher-risk individuals require rigorous treatment approaches that address the psychosocial and behavioral consequences of early-life stress to thrive.