Predicting student outcomes using digital logs of learning behaviors: Review, current standards, and suggestions for future work.

Using traces of behaviors to predict outcomes is useful in varied contexts ranging from buyer behaviors to behaviors collected from smart-home devices. Increasingly, higher education systems have been using Learning Management System (LMS) digital data to capture and understand students' learning and well-being. Researchers in the social sciences are increasingly interested in the potential of using digital log data to predict outcomes and design interventions. Using LMS data for predicting the likelihood of students' success in for-credit college courses provides a useful example of how social scientists can use these techniques on a variety of data types. Here, we provide a primer on how LMS data can be feature-mapped and analyzed to accomplish these goals. We begin with a literature review summarizing current approaches to analyzing LMS data, then discuss ethical issues of privacy when using demographic data and equitable model building. In the second part of the paper, we provide an overview of popular machine learning algorithms and review analytic considerations such as feature generation, assessment of model performance, and sampling techniques. Finally, we conclude with an empirical example demonstrating the ability of LMS data to predict student success, summarizing important features and assessing model performance across different model specifications.

When people estimate their personal intelligence who is overconfident? Who is accurate?

OBJECTIVE: We explore accurate self-knowledge versus overconfidence in personal intelligence-a "broad" intelligence about personality. The theory of personal intelligence proposes that people vary in their ability to understand the traits, goals, plans, and actions of themselves and others. We wondered who accurately knew that they were higher in personal intelligence and who did not, and whether individuals with more accurate estimates were distinguishable from others in their psychological characteristics. METHOD: Three archival data sets were identified that included both self-estimates and objective measures of personal intelligence: The measures were the Self-Estimated Personal Intelligence scale and the Test of Personal Intelligence. RESULTS: People who were over-confident-overestimating their ability-level of personal intelligence-were positive in their outlook and more sociable. People who provided the most accurate self-estimates were higher in verbal and personal intelligences, more open, and more conscientious than others. CONCLUSIONS: People who were accurate about themselves have not been studied before in this context but may, for example, serve as the monitors and thinkers who help keep themselves and others reasonable and on track.

Examining first-generation college student lived experiences with microaggressions and microaffirmations at a predominately White public research university.

OBJECTIVES: Studies have described the experiences of racial microaggressions in everyday life and on college campuses, yet prior research has not explored how microaggressions and microaffirmations are experienced by students who are first in their family to attend college. METHOD: This qualitative investigation of 296 open-ended survey responses described the lived experiences of first-generation college students at a large, public, predominately White institution. RESULTS: Students experienced microaggressions in the form microinsults, microassaults, and microinvalidations, and microaffirmations in the form of microsupports, microcompliments, and microvalidations on campus. CONCLUSIONS: Study implications include the development and implementation of regular positive and intentional communications by faculty, staff, and peers on college campuses. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Estimating HIV Medication Adherence and Persistence: Two Instruments for Clinical and Research Use.

Antiretroviral therapy (ART) requires lifelong daily oral therapy. While patient characteristics associated with suboptimal ART adherence and persistence have been described in cohorts of HIV-infected persons, these factors are poor predictors of individual medication taking behaviors. We aimed to create and test instruments for the estimation of future ART adherence and persistence for clinical and research applications. Following formative work, a battery of 148 items broadly related to HIV infection and treatment was developed and administered to 181 HIV-infected patients. ART adherence and persistence were assessed using electronic monitoring for 3 months. Perceived confidence in medication taking and self-reported barriers to adherence were strongest in predicting non-adherence over time. Barriers to adherence (e.g., affordability, scheduling) were the strongest predictors of non-adherence, as well as 3- and 7-day non-persistence. A ten-item battery for prediction of these outcomes ( www.med.unc.edu/ncaidstraining/adherence/for-providers ) and a 30-item battery reflective of underlying psychological constructs can help identify and study individuals at risk for suboptimal ART adherence and persistence.

Employees High in Personal Intelligence Differ From Their Colleagues in Workplace Perceptions and Behavior.

Personal intelligence (PI) involves the ability to recognize, reason, and use information about personality to understand oneself and other people. Employees in two studies (Ns = 394, 482) completed the Test of Personal Intelligence (TOPI; e.g., Mayer, Panter, & Caruso, 2017a) and assessments of workplace perception and behavior. Higher PI was associated with higher perceived workplace support and lower counterproductive work behavior. These relationships continued to hold after controlling for other key variables. The results indicate the TOPI, although still in research trials, shows promise as a screening device for selecting employees and targeting individuals for training.

Does personal intelligence exist? Evidence from a new ability-based measure.

Personal intelligence has been defined as the ability to reason about personality and personality-relevant information and to use that information to guide one's actions and more generally, one's life. We constructed an initial version of an ability-based measure to test whether personal intelligence can be measured and whether it exists as a unitary intelligence. In 3 studies (N = 241, 308, and 385), we administered this Test of Personal Intelligence (TOPI), composed of 4 sections, to undergraduates along with criterion measures. Results suggested that a personal intelligence can be measured, that it might exist as a unified area of mental abilities, and that it represents psychological qualities that have intriguing predictive aspects.

The culture of time in neuropsychological assessment: exploring the effects of culture-specific time attitudes on timed test performance in Russian and American samples.

Cultural differences in time attitudes and their effect on timed neuropsychological test performance were examined in matched non-clinical samples of 100 Russian and American adult volunteers using 8 tests that were previously reported to be relatively free of cultural bias: Color Trails Test (CTT); Ruff Figural Fluency Test (RFFT); Symbol Digit Modalities Test (SDMT); and Tower of London-Drexel Edition (ToL(Dx)). A measure of time attitudes, the Culture of Time Inventory (COTI-33) was used to assess time attitudes potentially affecting time-limited testing. Americans significantly outscored Russians on CTT, SDMT, and ToL(Dx) (p,.05) while differences in RFFT scores only approached statistical significance. Group differences also emerged in COTI-33 factor scores, which partially mediated differences in performance on CTT-1, SDMT, and ToL(Dx) initiation time, but did not account for the effect of culture on CTT-2. Significant effect of culture was revealed in ratings of familiarity with testing procedures that was negatively related to CTT, ToL(Dx), and SDMT scores. Current findings indicated that attitudes toward time may influence results of time limited testing and suggested that individuals who lack familiarity with timed testing procedures tend to obtain lower scores on timed tests.

An item factor analysis and item response theory-based revision of the Everyday Discrimination Scale.

The Everyday Discrimination Scale (EDS), a widely used measure of daily perceived discrimination, is purported to be unidimensional, to function well among African Americans, and to have adequate construct validity. Two separate studies and data sources were used to examine and cross-validate the psychometric properties of the EDS. In Study 1, an exploratory factor analysis was conducted on a sample of African American law students (N = 589), providing strong evidence of local dependence, or nuisance multidimensionality within the EDS. In Study 2, a separate nationally representative community sample (N = 3,527) was used to model the identified local dependence in an item factor analysis (i.e., bifactor model). Next, item response theory (IRT) calibrations were conducted to obtain item parameters. A five-item, revised-EDS was then tested for gender differential item functioning (in an IRT framework). Based on these analyses, a summed score to IRT-scaled score translation table is provided for the revised-EDS. Our results indicate that the revised-EDS is unidimensional, with minimal differential item functioning, and retains predictive validity consistent with the original scale.

"Looking at Myself in the Future": how mentoring shapes scientific identity for STEM students from underrepresented groups.

BACKGROUND: Mentorship has been well-established in the literature as fostering scientific identity and career pathways for underrepresented minority students in science, technology, engineering, and mathematics (STEM) fields. Mentorship is prioritized by programs that aim to increase diversity and support future leadership in STEM fields, but in-depth understanding of mentorship in these contexts remains limited. Drawing on qualitative interview data, we sought to understand the relationship between mentoring and scientific identity among a diverse sample of 24 students in one such program, in order to inform program development. RESULTS: Qualitative analysis of the data revealed that mentorship, especially research mentorship, was common and played a role in formation of scientific identity. Students with research mentors tended to say they strongly identified as scientists, whereas those who lacked research mentorship varied in their level of scientific identity. In interviews, research-mentored students described mentors as colleagues who gave them opportunities to grow and as examples to look up to. Students valued mentors with whom they identified on the basis of demographic similarity or shared values, as well as those who challenged them in their academic and research endeavors. CONCLUSIONS: Our analysis highlights how different mentoring experiences can contribute to development of future STEM leadership. We discuss implications for practice, including the need for tailored mentoring approaches and research-focused mentoring, and offer several recommendations for research and programming.

Corticosteroids' effect on the height of atopic dermatitis patients: a controlled questionnaire study.

To investigate if children treated with topical corticosteroids have a significantly shorter height than the height of children not treated with corticosteroids and to see if corticosteroids affect the ability for treated children to meet growth potential defined as midparental height. Parents of patients attending the UNC's Dermatology clinic completed the survey. The patient's height and siblings' heights were measured by staff. Parents' heights were self reported as were the child's diagnosis of atopic dermatitis, and duration of use of corticosteroids. The patient's height was standardized using CDC charts. Additionally, the midparental height was calculated and standardized. The difference between present and predicted standardized heights was calculated; 151 surveys yielded data on 83 girls and 63 boys (ages 2-21 yrs). The standing height and the difference in standing height and midparental scores were not significantly different among: (i) children with and without atopic dermatitis; and (ii) children treated and not treated with corticosteroids. The overall height of children examined in this survey who were treated with topical corticosteroids appears to be unaffected.

It matters how and when you ask: self-reported race/ethnicity of incoming law students.

The high-stakes nature of law school testing and admissions puts a premium on the student data presented to admissions committees, such as essays, academic and work history, and student background characteristics including race/ethnicity. 4,472 law school-bound students self-identified their race/ethnicity using (a) a mutually exclusive "choose one" format during registration for the law school admissions test, and (b) an elaborated "check-all-that-apply" format as part of a national survey administered during the first weeks at their chosen law school. Student multiraciality that was masked by the first assessment was associated with self-reported ethnic identity, discrimination experience, intergroup contact, race-related attitudes, academic performance, and trait ratings, as compared to monoracial majority students. A different profile of findings was observed across these constructs when multiracial students were compared to monoracial majority students, to monoracial minority students, and within group. These correlates also predicted the likelihood of changing identification across the two assessment contexts. These findings support the continued study of specific combinations of multiracial groups, fluidity of multiracial identities, and context effects that influence race/ethnicity self-categorizations.

Advancing the Measurement of Personal Intelligence with the Test of Personal Intelligence, Version 5 (TOPI 5).

People use their personal intelligence (PI) to understand personality in themselves and others. In Studies 1 and 2 (Ns = 961 and 548), individuals completed the Test of Personal Intelligence, Version 5 (TOPI 5), which is introduced here. The TOPI 5 is an ability assessment with a broader range of content and more challenging items than earlier test versions. In past research, factor analyses indicated that people employ two distinct but highly correlated abilities to problem-solve in this area. These two-factor models, however, exhibited instabilities and limited applicability between the TOPI 4 and 5 in this research (and as reported in the Supplementary Materials). In Study 3, we successfully test the one-factor models of the TOPI with the present data and archival data sets (Narchival = 19,627). We then use the one-factor models to develop a pair of new test forms: one that is compatible with all the TOPI test versions and another, TOPI 5E, that is better at distinguishing among people scoring in the higher range of performance relative to previous measures.

Factors associated with potentially inappropriate drug utilization in a sample of rural community-dwelling older adults.

BACKGROUND: Many medications present special risks when used by older adults (ie, those aged > or = 65 years) and are considered potentially inappropriate for this population. The Beers criteria are often used to identify such medications. Past research has documented that use of Beers drugs is common among older adults. OBJECTIVE: The aim of this work was to examine factors associated with potentially inappropriate drug use among rural community-dwelling older adults using a conceptual framework adapted from the Andersen-Newman behavioral model of health service use. METHODS: This was a population-based, cross-sectional survey. Data were collected via face-to-face home interviews between 2002 and 2004. Rural community-dwelling older adults residing in a single county in North Carolina were eligible. Potentially inappropriate drug use was operationalized using the Beers criteria. Data concerning predisposing (ie, age, sex, race, education, and marital status), enabling (ie, social support and insurance status), need (ie, disability and history of major depression, hypertension, osteoarthritis, back problems, or other comorbidities), and utilization factors (ie, number of medications used) were collected. RESULTS: Data were gathered from 892 people, with information on medication use available for 800. Two hundred thirteen of these 800 participants (26.6%) used > or = 1 Beers drug. Compared with individuals who used no Beers drugs, those who used > or = 1 Beers drug reported lower levels of social support (odds ratio [OR], 0.94; 95% CI, 0.90-0.99) and higher levels of disability (OR, 1.48; 95% CI, 1.11-1.97), used more medications (OR, 1.07; 95% CI, 1.01-1.13), and were more likely to have a history of major depression (OR, 1.67; 95% CI, 1.05-2.66), hypertension (OR, 1.58; 95% CI, 1.07-2.33), osteoarthritis (OR, 1.58; 95% CI, 1.09-2.29), and back problems (OR, 1.72; 95% CI, 1.19-2.47). CONCLUSION: As suggested by the Andersen-Newman model, the risk of potentially inappropriate drug use is highest among those with the greatest medication needs, as evidenced by poorer health status in this sample of rural community-dwelling older patients.

Assessing tobacco beliefs among youth using item response theory models.

Successful intervention research programs to prevent adolescent smoking require well-chosen, psychometrically sound instruments for assessing smoking prevalence and attitudes. Twelve thousand eight hundred and ten adolescents were surveyed about their smoking beliefs as part of the Teenage Attitudes and Practices Survey project, a prospective cohort study of predictors of smoking initiation among US adolescents. Item response theory (IRT) methods are used to frame a discussion of questions that a researcher might ask when selecting an optimal item set. IRT methods are especially useful for choosing items during instrument development, trait scoring, evaluating item functioning across groups, and creating optimal item subsets for use in specialized applications such as computerized adaptive testing. Data analytic steps for IRT modeling are reviewed for evaluating item quality and differential item functioning across subgroups of gender, age, and smoking status. Implications and challenges in the use of these methods for tobacco onset research and for assessing the developmental trajectories of smoking among youth are discussed.

Do characteristics of HIV/AIDS education and training affect perceived training quality? Lessons from the evaluation of seven projects.

Initial and continuing HIV/AIDS education and training has been a critical way to bring the nation's health providers up to date on emerging developments and approaches. This study reports cross-cutting findings from seven HIV/AIDS education and training projects. Trainers described over 600 training sessions from these projects in terms of their structural characteristics and design elements, while trainees described these sessions on several dimensions related to training quality. Training characteristics were compared to trainee assessments of training quality. Using a decision-tree analytic approach for major training attributes, considerable support emerged for links between training characteristics and perceived quality of the HIV/AIDS training experience. More favorable quality ratings were associated with certain projects, the training setting, the types of trainees served by the training, the intended training impact, discussion of special populations, and training methods involving interactive learning. With increased knowledge regarding how these educational experiences relate to the ways they are perceived and processed, more targeted approaches to training design on HIV/AIDS can be developed.

Risk factors and characteristics of youth living with, or at high risk for, HIV.

Over 8,000 adolescents and young adults (4,111 males; 4,085 females) reported on several HIV-related risk behaviors during enrollment into 10 service demonstration projects targeted to youth living with, or at risk for, HIV. Distinct risk patterns emerged by gender when predicting HIV serostatus (versus unknown serostatus/negative serostatus). Males who had injection drug risk histories, had sex with an HIV positive partner, had sexually transmitted diseases, had sex with males, and/or were homeless had an inflated risk of being HIV positive. Females who engaged in sex with an HIV partner, had sex with an injection drug user, and/or had sexually transmitted diseases, were at the highest HIV risk. For both samples, engaging in sex with women reduced the likelihood of HIV positive status. Very basic information about risk factors obtained at service intake offers important information about HIV status of "high risk" youth presenting for care in community programs, as well as suggests clear risk factors for targeted preventive efforts.

Effects of HIV/AIDS education and training on patient care and provider practices: a cross-cutting evaluation.

Initial and continuing training in HIV/AIDS service provision is a critical way to enable the nation's health providers to use state-of-the-art developments and perspectives. Typically, the efficacy of HIV/AIDS training programs is evaluated using assessments administered to trainees immediately following the training. This study reports cross-cutting findings from telephone interviews conducted with 218 trainees an average of 8 months after training. Long-term training effects are examined in three domains: (a) general perspectives on HIV/AIDS; (b) health care provider service provision; and (c) changes in procedures and operations at the health care system level. The findings show the different ways that the training experience had long-term positive and observable effects in these three domains. In some cases, background characteristics and job positions predicted the specific type of reported training effects. The pattern of results suggests ways in which training methods can be targeted to specific audiences.

Women's steps of change and entry into drug abuse treatment. A multidimensional stages of change model.

The Transtheoretical, or Stages of Change Model, has been applied to the investigation of help-seeking related to a number of addictive behaviors. Overall, the model has shown to be very important in understanding the process of help-seeking. However, substance abuse rarely exists in isolation from other health, mental health, and social problems. The present work extends the original Stages of Change Model by proposing "Steps of Change" as they relate to entry into substance abuse treatment programs for women. Readiness to make life changes in four domains-domestic violence, HIV sexual risk behavior, substance abuse, and mental health-is examined in relation to entry into four substance abuse treatment modalities (12-step, detoxification, outpatient, and residential). The Steps of Change Model hypothesizes that help-seeking behavior of substance-abusing women may reflect a hierarchy of readiness based on the immediacy, or time urgency, of their treatment issues. For example, women in battering relationships may be ready to make changes to reduce their exposure to violence before admitting readiness to seek substance abuse treatment. The Steps of Change Model was examined in a sample of 451 women contacted through a substance abuse treatment-readiness program in Los Angeles, California. A series of logistic regression analyses predict entry into four separate treatment modalities that vary. Results suggest a multidimensional Stages of Change Model that may extend to other populations and to other types of help-seeking behaviors.

Predicting substance abuse among youth with, or at high risk for, HIV.

This article describes data from 4,111 males and 4,085 females participating in 10 HIV/AIDS service demonstration projects. The sample was diverse in age, gender, ethnicity, HIV status, and risk for HIV transmission. Logistic regression was used to determine the attributes that best predict substance abuse. Males who were younger; HIV positive; homeless; involved in the criminal justice system; had a sexually transmitted disease (STD); engaged in survival sex; and participated in risky sex with men, women, and drug injectors were most likely to have a substance abuse history. For females, the same predictors were significant, with the exception of having an STD. Odds ratios as high as 6 to 1 were associated with the predictors. Information about sexual and other risk factors also was highly predictive of substance abuse issues among youth.

Healthcare provider characteristics and perceived confidence from HIV/AIDS education.

This study reports findings from six training projects designed to keep health providers up-to-date on emerging developments and approaches in HIV/AIDS care. Participants were 3,779 individuals who described themselves, their professional background, and their specific experience in the HIV/AIDS field. These characteristics were compared with their self-reported confidence in managing clients, counseling clients, providing services, and the training topics. A repeated-measures design examining level and change of confidence showed little support for links between provider characteristics and confidence due to HIV/AIDS training experience. Thus, knowing a provider's background does not necessarily provide diagnostic information about who might most benefit in improved confidence from HIV/AIDS educational training. These results suggest that HIV/AIDS training programs may be targeted broadly-to a wide range of healthcare providers of diverse backgrounds-with little or no impact on overall levels and changes in provider confidence.