Background Recent medical education literature pertaining to professional identity development fails to reflect the impact social media has on professional identity theory. Social media is transforming the field of medicine, as the web-based medium is now an avenue for professional development and socialization for medical students and residents. Research regarding identity development in social media has been primarily confined to electronic professionalism through best practice guidelines. However, this neglects other potential aspects pertinent to digital identity that have not yet been explored. Objective This study aims to define the properties and development of the digital self and its interactions with the current professional identity development theory. Methods A qualitative study was conducted using thematic analysis. A total of 17 participants who are social media education and knowledge translation experts were interviewed. The initial participants were from emergency medicine, and a snowball sampling method was used following their respective web-based semistructured interviews to enable global recruitment of other participants from interprofessional disciplines. The research team consisted of a diverse group of researchers including one current social media knowledge translation physician clinician educator, one postdoctoral researcher who is regularly engaged in social media knowledge translation, and 3 nonphysician research assistants who are not social media users. Half of the team conducted the initial coding and analysis, whereas the other 2 investigators audited the procedures followed. Results A total of 4 themes were identified that pertain to digital identity. In the first theme, origins of initial digital identity formation were found to be derived from perceived needs in professional roles (eg, as a medical student or resident). The second theme consisted of the cultivation of digital identity, in which digital identity was developed parallel to professional identity. The third theme that emerged was the management between the professional and personal components of digital identity. Participants initially preferred keeping these components completely separate; however, attempts to do so were inadequate while the integration of both components provided benefits. The fourth theme was the management of real-life identity and digital identity. Participants preferred real-life identity to be wholly represented on the web. Instances of misalignment resulted in identity conflict, compromising one of the identities. Conclusions Social media introduces new features to professional identity in the digital world. The formation of digital identity, its development, and reconciliation with other identities were features captured in our analysis. The virtual component of professional identity must not be neglected but instead further explored, as educational institutions continue to give more importance to navigating professional identity development.

Introduction There still remains a gap between those who conduct science and those who engage in educating others about health sciences through various forms of social media. Few empirical studies have sought to define useful practices for engaging in social media for academic use in the health professions. Given the increasing importance of these platforms, we sought to define good practices and potential pitfalls with help of those respected for their work in this new field. Methods We conducted a qualitative study, guided by constructivist grounded theory principles, of 17 Electronic supplementary material The online version of this article (

Background Residents receive a numeric performance rating (eg, 1-7 scoring scale) along with a narrative (ie, qualitative) feedback based on their performance in each workplace-based assessment (WBA). Aggregated qualitative data from WBA can be overwhelming to process and fairly adjudicate as part of a global decision about learner competence. Current approaches with qualitative data require a human rater to maintain attention and appropriately weigh various data inputs within the constraints of working memory before rendering a global judgment of performance. Objective This study explores natural language processing (NLP) and machine learning (ML) applications for identifying trainees at risk using a large WBA narrative comment data set associated with numerical ratings. Methods NLP was performed retrospectively on a complete data set of narrative comments (ie, text-based feedback to residents based on their performance on a task) derived from WBAs completed by faculty members from multiple hospitals associated with a single, large, residency program at McMaster University, Canada. Narrative comments were vectorized to quantitative ratings using the bag-of-n-grams technique with 3 input types: unigram, bigrams, and trigrams. Supervised ML models using linear regression were trained with the quantitative ratings, performed binary classification, and output a prediction of whether a resident fell into the category of at risk or not at risk. Sensitivity, specificity, and accuracy metrics are reported. Results The database comprised 7199 unique direct observation assessments, containing both narrative comments and a rating between 3 and 7 in imbalanced distribution (scores 3-5: 726 ratings; and scores 6-7: 4871 ratings). A total of 141 unique raters from 5 different hospitals and 45 unique residents participated over the course of 5 academic years. When comparing the 3 different input types for diagnosing if a trainee would be rated low (ie, 1-5) or high (ie, 6 or 7), our accuracy for trigrams was 87%, bigrams 86%, and unigrams 82%. We also found that all 3 input types had better prediction accuracy when using a bimodal cut (eg, lower or higher) compared with predicting performance along the full 7-point rating scale (50%-52%). Conclusions The ML models can accurately identify underperforming residents via narrative comments provided for WBAs. The words generated in WBAs can be a worthy data set to augment human decisions for educators tasked with processing large volumes of narrative assessments.

Introduction: As academia begins to incorporate modern communication technologies into its scholarly structures, there are both enablers and barriers which foster academics’ uptake of these innovations. Those who are early adopters of academic social media - whether it be for education, research-related networking, or knowledge translation - may therefore be best positioned to highlight both enablers and barriers within their work environments. Methods: The authors conducted a constructivist grounded theory study to discern what prominent practitioners of academic social media (e.g. Twitter) have encountered in their careers. Participants were recruited via a snowball sampling technique and invited to participate in semi-structured interviews. Three investigators engaged in constant comparative analysis of incoming transcripts. To enhance rigour, we conducted an audit of the analysis and a participant member check. Results: Seventeen emerging influencers in the field of academic social media were recruited. After axial coding, the 30 enablers and 21 barriers to academic social media use were mapped to three spheres of influence: personal, institutional, and virtual. The investigators propose a framework that organizes these enablers and barriers around a tipping point where sustainability becomes possible. Conclusions: Multiple enablers and barriers were described to influence social media users within academic medicine. By organizing these facets into a personal, institutional, and virtual framework along a spectrum, we can begin to understand the underlying structures that potentiate the academic ecosystems in which social media and similar innovations may flourish.

BACKGROUND Residents receive a numeric performance rating (e.g., 1-7 scoring scale) along with a narrative (i.e., qualitative) feedback based on their performance in each workplace-based assessment (WBA). Aggregated qualitative data from WBA can be overwhelming to process and fairly adjudicate as part of a global decision about learner competence. Current approaches with qualitative data require a human rater to maintain attention and appropriately weigh various data inputs within the constraints of working memory before rendering a global judgment of performance. OBJECTIVE This study evaluates the accuracy of a decision support system for raters using natural language processing (NLP) and machine learning (ML). METHODS NLP was performed retrospectively on a complete dataset of narrative comments (i.e., text-based feedback to residents based on their performance on a task) derived from WBAs completed by faculty members from multiple hospitals associated with a single, large, residency program at McMaster University, Canada. Narrative comments were vectorized to quantitative ratings using bag-of-n-grams technique with three input types: unigram, bigrams, and trigrams. Supervised machine learning models using linear regression were trained for two outputs using the original ratings and dichotomized ratings (at risk or not). Sensitivity, specificity, and accuracy metrics are reported. RESULTS The database consisted of 7,199 unique direct observation assessments, containing both narrative comments and a 3 to 7 rating in imbalanced distribution (3-5: 726, and 6-7: 4,871 ratings). Total of 141 unique raters from five different hospitals and 45 unique residents participated over the course of five academic years. When comparing the three different input types for diagnosing if a trainee would be rated low (i.e., 1-5) or high (i.e., 6 or 7), our accuracy for trigrams was (87%), bigrams (86%), and unigrams (82%). We also found that all three input types had better prediction accuracy when using a bimodal cut (e.g., lower or higher) compared to predicting performance along the full 7-scale (50-52%). CONCLUSIONS The ML models can accurately identify underperforming residents via narrative comments provided for work-based assessments. The words generated in WBAs can be a worthy dataset to augment human decisions for educators tasked with processing large volumes of narrative assessments. CLINICALTRIAL N/A

BACKGROUND Recent medical education literature surrounding professional identity development fail to reflect the impact social media has on professional identity theory. Social media is transforming the field of medicine, as the digital space is now an avenue for professional development and socialization for medical students and residents. Research regarding identity development in the digital space has been primarily confined to e-professionalism through best practices guidelines. This prior work, however, neglects other potential aspects pertinent to digital identity that have not yet been explored. OBJECTIVE We sought to define the properties and development of the digital self, and its interactions with current professional identity development theory. METHODS A qualitative study was conducted using thematic analysis. Seventeen participants who are social media education and knowledge translation experts were interviewed. A snowball sampling method was used following their respective semi-structured interviews. The research team consisted of a diverse group of researchers including one current social media knowledge translation physician clinician educator, one postdoctoral fellow immersed in social media and 3 non-physician research assistants who are not social media users. Half of the team conducted the initial coding and analysis, while the other two investigators conducted an audit of the procedure. RESULTS Four themes surrounding digital identity emerged from our analysis. First, origins of initial digital identity formation were found to be derived from perceived needs in professional roles (e.g. as a medical student or resident). The second theme consisted of the cultivation of digital identity, in which digital identity developed parallel to professional identity development. The third theme that emerged from our analysis was the management between the professional and personal components of digital identity. Participants initially preferred keeping these two completely separate; however, attempts to do so were inadequate while integration of both provided benefits. The last theme conveyed in the analysis was the management of in-real-life identity and digital identity. Our participants preferred in-real-life identity to be wholly represented online. Instances of misalignment resulted in identity conflict, compromising one of the identities. CONCLUSIONS Social media introduces new features to professional identity in the digital space. The formation of digital identity, its development, and its reconciliation with other identities were features captured in our analysis. With the high importance of navigating professional identity development placed by educational institutions, the virtual component must not be neglected and, instead, further explored.