Using Exploratory Data Mining to Identify Important Correlates of Nonsuicidal Self-Injury Frequency

Quantifying the Importance of Non-Suicidal Self-Injury Characteristics in Predicting Different Clinical Outcomes: Using Random Forest Model

Existing research on non-suicidal self-injury (NSSI) among adolescents has primarily concentrated on general risk factors, leaving a significant gap in understanding the specific NSSI characteristics that predict diverse psychopathological outcomes. This study aims to address this gap by using Random Forests to discern the significant predictors of different clinical outcomes. The study tracked 348 adolescents (64.7% girls; mean age = 13.31, SD = 0.91) over 6 months. Initially, 46 characteristics of NSSI were evaluated for their potential to predict the repetition of NSSI, as well as depression, anxiety, and suicidal risks at a follow-up (T2). The findings revealed distinct predictors for each psychopathology. Specifically, psychological pain was identified as a significant predictor for depression, anxiety, and suicidal risks, while the perceived effectiveness of NSSI was crucial in forecasting its repetition. These findings imply that it is feasible to identify high-risk individuals by assessing key NSSI characteristics, and also highlight the importance of considering diverse NSSI characteristics when working with self-injurers.

Psychological pain and sociodemographic factors classified suicide attempt and non-suicidal self-injury in adolescents

This study aimed to utilize machine learning to explore the psychological similarities and differences between suicide attempt (SA) and non-suicidal self-injury (NSSI), with a particular focus on the role of psychological pain. A total of 2385 middle school students were recruited using cluster sampling. The random forest algorithm was used with 25 predictors to develop classification models of SA and NSSI, respectively, and to estimate the importance scores of each predictor. Based on these scores and related theories, shared risk factors (control feature set) and distinct risk factors (distinction feature set) were selected and tested to distinguish between NSSI and SA. The machine learning algorithm exhibited fair to good performance in classifying SA history [Area Under Curves (AUCs): 0.65-0.87] and poor performance in classifying NSSI history (AUC: 0.61-0.68). The distinction feature set comprised pain avoidance, family togetherness, and deviant peer affiliation, while the control feature set included pain arousal, painful feelings, and crisis events. The distinction feature set slightly but stably outperformed the control feature set in classifying SA from NSSI. The three-dimensional psychological pain model, especially pain avoidance, might play a dominant role in understanding the similarities and differences between SA and NSSI.

Comparing linear discriminant analysis and supervised learning algorithms for binary classification-A method comparison study

In psychology, linear discriminant analysis (LDA) is the method of choice for two-group classification tasks based on questionnaire data. In this study, we present a comparison of LDA with several supervised learning algorithms. In particular, we examine to what extent the predictive performance of LDA relies on the multivariate normality assumption. As nonparametric alternatives, the linear support vector machine (SVM), classification and regression tree (CART), random forest (RF), probabilistic neural network (PNN), and the ensemble k conditional nearest neighbor (EkCNN) algorithms are applied. Predictive performance is determined using measures of overall performance, discrimination, and calibration, and is compared in two reference data sets as well as in a simulation study. The reference data are Likert-type data, and comprise 5 and 10 predictor variables, respectively. Simulations are based on the reference data and are done for a balanced and an unbalanced scenario in each case. In order to compare the algorithms' performance, data are simulated from multivariate distributions with differing degrees of nonnormality. Results differ depending on the specific performance measure. The main finding is that LDA is always outperformed by RF in the bimodal data with respect to overall performance. Discriminative ability of the RF algorithm is often higher compared to LDA, but its model calibration is usually worse. Still LDA mostly ranges second in cases it is outperformed by another algorithm, or the differences are only marginal. In consequence, we still recommend LDA for this type of application.

Comparing Bayesian Variable Selection to Lasso Approaches for Applications in Psychology

In the current paper, we review existing tools for solving variable selection problems in psychology. Modern regularization methods such as lasso regression have recently been introduced in the field and are incorporated into popular methodologies, such as network analysis. However, several recognized limitations of lasso regularization may limit its suitability for psychological research. In this paper, we compare the properties of lasso approaches used for variable selection to Bayesian variable selection approaches. In particular we highlight advantages of stochastic search variable selection (SSVS), that make it well suited for variable selection applications in psychology. We demonstrate these advantages and contrast SSVS with lasso type penalization in an application to predict depression symptoms in a large sample and an accompanying simulation study. We investigate the effects of sample size, effect size, and patterns of correlation among predictors on rates of correct and false inclusion and bias in the estimates. SSVS as investigated here is reasonably computationally efficient and powerful to detect moderate effects in small sample sizes (or small effects in moderate sample sizes), while protecting against false inclusion and without over-penalizing true effects. We recommend SSVS as a flexible framework that is well-suited for the field, discuss limitations, and suggest directions for future development.

Common, uncommon, and novel applications of random forest in psychological research

Recent reform efforts have pushed toward a better understanding of the distinction between exploratory and confirmatory research, and appropriate use of each. As some utilize more exploratory tools, it may be tempting to employ multiple linear regression models. In this paper, we advocate for the use of random forest (RF) models. RF is able to obtain better predictive performance than traditional regression, while also inherently protecting against overfitting as well as detecting nonlinear effects and interactions among predictors. Given the advantages of RF compared to other statistical procedures, it is a tool commonly used within a plethora of industries, including stock trading, banking, pharmaceuticals, and patient healthcare planning. However, we find RF is used within the field of psychology comparatively less frequently. In the current paper, we advocate for RF as an important statistical tool within the context of behavioral and psychological research. In hopes of increasing the use of RF in the field of psychology, we provide information pertaining to the limitations one might confront in using RF and how to overcome such limitations. Moreover, we discuss various methods for how to optimally utilize RF with psychological data, such as nonparametric modeling, interaction and nonlinearity detection, variable selection, prediction and classification modeling, and assessing parameters of Monte Carlo simulations. Throughout, we illustrate the use of RF with visualization strategies, aimed to make RF models more comprehensible and intuitive.

How to apply variable selection machine learning algorithms with multiply imputed data: A missing discussion

Psychological researchers often use standard linear regression to identify relevant predictors of an outcome of interest, but challenges emerge with incomplete data and growing numbers of candidate predictors. Regularization methods like the LASSO can reduce the risk of overfitting, increase model interpretability, and improve prediction in future samples; however, handling missing data when using regularization-based variable selection methods is complicated. Using listwise deletion or an ad hoc imputation strategy to deal with missing data when using regularization methods can lead to loss of precision, substantial bias, and a reduction in predictive ability. In this tutorial, we describe three approaches for fitting a LASSO when using multiple imputation to handle missing data and illustrate how to implement these approaches in practice with an applied example. We discuss implications of each approach and describe additional research that would help solidify recommendations for best practices. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

NSSI questionnaires revisited: A data mining approach to shorten the NSSI questionnaires

BACKGROUND AND OBJECTIVE

Non-suicidal self-injury (NSSI) is a psychological disorder that the sufferer consciously damages their body tissues, often too severe that requires intensive care medicine. As some individuals hide their NSSI behaviors, other people can only identify them if they catch them while injuring, or via dedicated questionnaires. However, questionnaires are long and tedious to answer, thus the answers might be inconsistent. Hence, in this study for the first time, we abstracted a larger questionnaire (of 662 items in total) to own only 22 items (questions) via data mining techniques. Then, we trained several machine learning algorithms to classify individuals based on their answers into two classes.

METHODS

Data from 277 previously-questioned participants is used in several data mining methods to select features (questions) that highly represent NSSI, then 245 different people were asked to participate in an online test to validate those features via machine learning methods.

RESULTS

The highest accuracy and F1 score of the selected features-via the Genetics algorithm-are 80.0% and 74.8% respectively for a Random Forest algorithm. Cronbach's alpha of the online test (validation on the selected features) is 0.82. Moreover, results suggest that an MLP can classify participants into two classes of NSSI Positive and NSSI Negative with 83.6% accuracy and 83.7% F1-score based on the answers to only 22 questions.

CONCLUSION

While previously psychologists used many combined questionnaires to see whether someone is involved in NSSI, via various data mining methods, the present study showed that only 22 questions are enough to predict if someone is involved or not. Then different machine learning algorithms were utilized to classify participants based on their NSSI behaviors, among which, an MLP with 10 hidden layers had the best performance.

The Link between Sleep Insufficiency and Self-Injury among In-School Adolescents: Findings from a Cross-Sectional Survey of Multi-Type Schools in Huangpu District of Shanghai, China

Both insufficient sleep and self-injury are rising public health issues among middle school students. Understanding their relationship may guide the intervention and policy making to help youths gain a healthy life. Thus, we analysed the data collected from the Shanghai Students Health Risk Behavior Surveillance (2015) in the Huangpu District. Self-injury was self-reported and categorized into ever or never. Sleep duration was classified as sufficient and insufficient according to the Health China 2030 Plan and the National Sleep Foundation's updated sleep duration recommendations. Crude OR and adjusted OR of sleep duration and covariates were estimated for self-injury using the logistic regression models with standard error clustered on school types. Results showed that 8.42% of the participants had conducted self-injury, with girls more than boys and ordinary school students more than key school students. After full adjustment, sleep insufficiency increased the odds of conducting self-injury by approximately two folds (AOR = 2.08, 95%CI = 1.40-3.07). The odds of self-injury were higher among students studying at ordinary schools (AOR = 3.58, 95%CI = 1.25-10.27) or vocational schools (AOR = 2.00, 95%CI = 1.77-2.26), with comparison to those at key schools. Interventions seeking to solve insufficient sleep need to be multifaceted, with consideration of changing the school environment and multiple social contexts, which create stressful burdens for adolescents' development.

Cognition and Non-Suicidal Self-Injury: Exploring Relationships with Psychological Functions

OBJECTIVE

Non-suicidal self-injury (NSSI) is strongly associated with difficulties in emotion regulation, but its relationships with maladaptive cognitive processes are less clear.

METHOD

The current study examined relationships between self-reported NSSI (presence, number of methods, frequency, recency, duration, functions) and negative cognitive processes (rumination, worry, self-criticism, perceived burdensomeness, thwarted belongingness) among 1,357 undergraduates. Cognition variables were submitted to exploratory factor analysis (EFA), and relationships were examined between the resulting factors and NSSI history (among the full sample) and NSSI severity and functions (among those with a history of NSSI).

RESULTS

The EFA derived two higher order cognitive factors: repetitive negative thinking (RNT) and negative self-perception (NSP). Both RNT and NSP were significantly higher among participants with than those without a history of NSSI. Among those with NSSI, NSP, but not RNT, was positively related to lifetime NSSI frequency and number of methods, as well as recency (presence in the past 12 months) and total duration (in years) of NSSI engagement. Moreover, RNT and NSP were positively associated with aggregate intrapersonal (but not interpersonal) functions of NSSI. The two cognitive factors demonstrated differential relationships with the individual intrapersonal NSSI functions.

CONCLUSIONS

Higher order categories of cognitive risk factors may have unique relationships with functions and severity of NSSI, with possible implications for more targeted approaches to risk assessment and intervention.HighlightsNegative thinking and self-perception were higher in people who engage in NSSI.Negative self-perception was associated with greater NSSI severity.Negative thinking and self-perception had different relations to NSSI functions.

Different Ways to Drown Out the Pain: A Meta-Analysis of the Association Between Nonsuicidal Self-Injury and Alcohol Use

OBJECTIVE

There is a significant overlap in the motivations for nonsuicidal self-injury (NSSI) and alcohol use. Moreover, several theories would predict that more frequent alcohol use is likely associated with more NSSI engagement. Still, the size and direction of this association has not been well documented in the literature.

METHOD

To address this gap, the goal of this article was to conduct a meta-analysis of the relation between alcohol use and NSSI.

RESULTS

Across 57 samples and 141,669 participants, we found that there was a significant positive association between NSSI and alcohol use, odds ratio = 1.78, 95% confidence interval [1.53, 2.07], k = 64, m = 52. Moderator analyses found that this effect was stronger for younger samples and samples with more severe alcohol use problems.

CONCLUSIONS

These results help establish a link between NSSI and alcohol use. Implications and future directions for NSSI research and intervention are discussed.HighlightsThere are several reasons to think that NSSI and alcohol use are linked.No reviews or meta-analyses have been conducted.We found a significant and small effect linking greater NSSI with greater alcohol use.

AI enabled suicide prediction tools: a qualitative narrative review

Background: Suicide poses a significant health burden worldwide. In many cases, people at risk of suicide do not engage with their doctor or community due to concerns about stigmatisation and forced medical treatment; worse still, people with mental illness (who form a majority of people who die from suicide) may have poor insight into their mental state, and not self-identify as being at risk. These issues are exacerbated by the fact that doctors have difficulty in identifying those at risk of suicide when they do present to medical services. Advances in artificial intelligence (AI) present opportunities for the development of novel tools for predicting suicide.

Method: We searched Google Scholar and PubMed for articles relating to suicide prediction using artificial intelligence from 2017 onwards.

Conclusions: This paper presents a qualitative narrative review of research focusing on two categories of suicide prediction tools: medical suicide prediction and social suicide prediction. Initial evidence is promising: AI-driven suicide prediction could improve our capacity to identify those at risk of suicide, and, potentially, save lives. Medical suicide prediction may be relatively uncontroversial when it pays respect to ethical and legal principles; however, further research is required to determine the validity of these tools in different contexts. Social suicide prediction offers an exciting opportunity to help identify suicide risk among those who do not engage with traditional health services. Yet, efforts by private companies such as Facebook to use online data for suicide prediction should be the subject of independent review and oversight to confirm safety, effectiveness and ethical permissibility.

Self-criticism longitudinally predicts nonsuicidal self-injury in eating disorders

Dialectical Behavior Therapy (DBT) has long been successfully applied to such behaviors such as nonsuicidal self-injury (NSSI) and more recently, bulimic behaviors. However, it is less clear how patients experiencing these comorbid symptoms may benefit from this treatment modality. Self-criticism, defined as a highly negative attitude towards the self, has been implicated in both EDs and NSSI and is amenable to DBT; thus, further examination of this construct may be beneficial in informing DBT treatment approaches. However, research has only examined these relationships cross-sectionally and no published research has examined self-criticism as a longitudinal predictor of NSSI and ED symptoms. Thus, this study examined self-criticism as a potential driving factor of NSSI in EDs in order to inform treatments, particularly DBT. Data were collected from 92 treatment-seeking adults at ED treatment facilities in the United States. Participants self-reported ED pathology, NSSI engagement, and self-criticism at baseline and a two-month follow-up. A path analysis revealed that self-criticism at baseline was associated with NSSI frequency at follow-up over and above baseline NSSI and ED symptomology. Self-criticism at baseline was not associated with ED pathology at follow-up. Self-criticism longitudinally predicted NSSI, but not ED pathology, in an ED sample. As such, it may be important for clinicians to assess for self-criticism and consider treatments that target both self-criticism and self-injury, like DBT, for this population.