The development of working alliance in early stages of care from the perspective of patients attending a chiropractic teaching clinic

BACKGROUND

The clinician-patient relationship has consistently been found to predict treatment success in both physical and mental health settings. This relationship has been operationalised in the literature as "Working Alliance," which consists of three key components: patient-clinician agreement on the goals of care, agreement on the tasks required to achieve those goals, and the establishment of a strong bond. While research has demonstrated the impact of working alliance in physical health settings, it often measures working alliance early in patients' care journeys. However, no primary research has investigated how early working alliance develops between patients and chiropractors. Evidence suggests that musculoskeletal practitioners may require further training to feel confident in establishing working alliance. Therefore, this study aims to explore the development of working alliance in the early stages of chiropractic care from the patients' perspective to inform evidence-based practice.

METHODS

Participants for this qualitative study were recruited from a teaching clinic at a specialised healthcare professions training university in the United Kingdom between September 2022 and April 2023. A total of 25 adult patients completed semi-structured interviews during the early stages of their care. The interview transcripts were analysed using Reflexive Thematic Analysis, from a critical realist stance.

RESULTS

The findings highlight that an early working alliance entails the gradual development of patients' confidence in their decision to seek help from trainee chiropractors to alleviate their symptoms. The four themes describe the impact of the clinical context on patients' expectations, the trainee chiropractors' qualities that participants considered important for early working alliance, the role of explanations, and the interplay between pain and early working alliance.

CONCLUSIONS

Establishing an early trainee chiropractor-patient working alliance involves a process of building patients' confidence in the trainee chiropractors' expertise, identifying the correct goals of care, and recognising the value of the proposed treatment plan. Factors shaping this process include the context of the care journey, patients' perceptions of trainee chiropractors' qualities, their bodily sensations, their expectations, their past experiences, and their satisfaction with trainee chiropractors' explanations.

Exploring the performance and explainability of fine-tuned BERT models for neuroradiology protocol assignment

BACKGROUND

Deep learning has demonstrated significant advancements across various domains. However, its implementation in specialized areas, such as medical settings, remains approached with caution. In these high-stake environments, understanding the model's decision-making process is critical. This study assesses the performance of different pretrained Bidirectional Encoder Representations from Transformers (BERT) models and delves into understanding its decision-making within the context of medical image protocol assignment.

METHODS

Four different pre-trained BERT models (BERT, BioBERT, ClinicalBERT, RoBERTa) were fine-tuned for the medical image protocol classification task. Word importance was measured by attributing the classification output to every word using a gradient-based method. Subsequently, a trained radiologist reviewed the resulting word importance scores to assess the model's decision-making process relative to human reasoning.

RESULTS

The BERT model came close to human performance on our test set. The BERT model successfully identified relevant words indicative of the target protocol. Analysis of important words in misclassifications revealed potential systematic errors in the model.

CONCLUSIONS

The BERT model shows promise in medical image protocol assignment by reaching near human level performance and identifying key words effectively. The detection of systematic errors paves the way for further refinements to enhance its safety and utility in clinical settings.

How people reason with counterfactual and causal explanations for Artificial Intelligence decisions in familiar and unfamiliar domains

Few empirical studies have examined how people understand counterfactual explanations for other people's decisions, for example, "if you had asked for a lower amount, your loan application would have been approved". Yet many current Artificial Intelligence (AI) decision support systems rely on counterfactual explanations to improve human understanding and trust. We compared counterfactual explanations to causal ones, i.e., "because you asked for a high amount, your loan application was not approved", for an AI's decisions in a familiar domain (alcohol and driving) and an unfamiliar one (chemical safety) in four experiments (n = 731). Participants were shown inputs to an AI system, its decisions, and an explanation for each decision; they attempted to predict the AI's decisions, or to make their own decisions. Participants judged counterfactual explanations more helpful than causal ones, but counterfactuals did not improve the accuracy of their predictions of the AI's decisions more than causals (Experiment 1). However, counterfactuals improved the accuracy of participants' own decisions more than causals (Experiment 2). When the AI's decisions were correct (Experiments 1 and 2), participants considered explanations more helpful and made more accurate judgements in the familiar domain than in the unfamiliar one; but when the AI's decisions were incorrect, they considered explanations less helpful and made fewer accurate judgements in the familiar domain than the unfamiliar one, whether they predicted the AI's decisions (Experiment 3a) or made their own decisions (Experiment 3b). The results corroborate the proposal that counterfactuals provide richer information than causals, because their mental representation includes more possibilities.

Pediatrician Explanations of Pediatric Pain in Clinical Settings: A Delicate Craft

Explaining chronic pain to children and families can be challenging, particularly in the absence of an obvious physiologically identifiable cause for the child's pain. In addition to medical intervention, children and families may expect clinicians to provide clarity around the cause of pain. Such explanations are often provided by clinicians who have not received formal pain training. This qualitative study sought to explore the following question: What do pediatricians consider to be important when providing pain explanations to children and their parents? Using semistructured interview methods, 16 UK pediatricians were interviewed regarding their perceptions of explaining chronic pain to children and families in clinical settings. Data were analyzed using inductive reflexive thematic analysis. Analyses generated 3 themes: 1) timing of the explanation, 2) casting a wider net, and 3) tailoring of the narrative. Study findings demonstrated the need for pediatricians to skilfully interpret where children and families are in their pain journey and deliver an appropriate and adaptable explanation relating to individual needs. Analyses identified the importance of providing a pain explanation that could be repeated and understood by others outside the consultation room, to enable children and families to accept the explanation. PERSPECTIVE: Study findings identify the importance of language in addition to familial and broader factors that may influence the provision and adoption of chronic pain explanations provided by pediatricians to children and families. Improving pain explanation provision may influence treatment engagement for children and their parents, subsequently impacting pain related outcomes.

Explanations in the wild

Why do some explanations strike people as highly satisfying while others, seemingly equally accurate, satisfy them less? We asked laypeople to generate and rate thousands of open-ended explanations in response to 'Why?' questions spanning multiple domains, and analyzed the properties of these explanations to discover (1) what kinds of features are associated with greater explanation quality; (2) whether people can tell how good their explanations are; and (3) which cognitive traits predict the ability to generate good explanations. Our results support a pluralistic view of explanation, where satisfaction is best predicted by either functional or mechanistic content. Respondents were better able to judge how accurate their explanations were than how satisfying they were to others. Insight problem solving ability was the cognitive ability most strongly associated with the generation of satisfying explanations.

Co-learner presence and praise alters the effects of learner-generated explanation on learning from video lectures

Learning from video lectures is becoming a prevalent learning activity in formal and informal settings. However, relatively little research has been carried out on the interactions of learning strategies and social environment in learning from video lectures. The present study addresses this gap by examining whether learner-generated explanations and co-learner presence with or without nonverbal praise independently and interactively affected learning from a self-paced video lecture about infectious diseases. University students were randomized into viewing either the video with instructor-generated explanations or the same video but generating explanations themselves. Outcomes were assessed by the quality of explanations, learning performance, mental effort, attention allocation, and behavioral patterns. Between-group comparisons showed that, in the absence of a peer co-learner, learning performance was similar in both the instructor-generated and learner-generated explanation groups. However, in the presence of a peer, learner-generated explanation facilitated learning performance. Furthermore, learner-generated explanation in the presence of a co-learner also reduced learners' mental effort and primed more behaviors related to self-regulation and monitoring. The results lead to the following strong recommendation for educational practice when using video lectures: if students learn by generating their own explanations in the presence of a co-learner, they will show better learning performance even though the learning is not necessarily more demanding, and will engage in more behaviors related to explanation adjustment and self-regulation.

Interpretable classifiers for prediction of disability trajectories using a nationwide longitudinal database

OBJECTIVES

To explore the heterogeneous disability trajectories and construct explainable machine learning models for effective prediction of long-term disability trajectories and understanding the mechanisms of predictions among the elderly Chinese at community level.

METHODS

This study retrospectively collected data from the Chinese Longitudinal Healthy Longevity and Happy Family Study between 2002 and 2018. A total of 4149 subjects aged 65 + in 2002 with completed activities of daily living (ADL) information for at least three waves were included. The mixed growth model was used to identify disability trajectories, and five machine learning models were further established to predict disability trajectories using epidemiological variables. An explainable approach was deployed to understand the model's decisions.

RESULTS

Three distinct disability trajectories, including normal class (77.3%), progressive class (15.5%), and high-onset class (7.2%), were identified for three-class prediction. The latter two were further merged into abnormal class, accompanied by normal class for two-class prediction. Machine learning, especially random forest and extreme gradient boosting achieved good performance in both two tasks. ADL, age, leisure activity, cognitive function, and blood pressure were key predictors.

CONCLUSION

The findings suggest that machine learning showed good performance and maybe of additional value in analyzing quality indicators in predicting disability trajectories, thereby providing basis to personalize intervention measures.

Conviction Narrative Theory: A Theory of Choice Under Radical Uncertainty

Conviction Narrative Theory (CNT) is a theory of choice under radical uncertainty-situations where outcomes cannot be enumerated and probabilities cannot be assigned. Whereas most theories of choice assume that people rely on (potentially biased) probabilistic judgments, such theories cannot account for adaptive decision-making when probabilities cannot be assigned. CNT proposes that people use narratives-structured representations of causal, temporal, analogical, and valence relationships-rather than probabilities, as the currency of thought that unifies our sense-making and decision-making faculties. According to CNT, narratives arise from the interplay between individual cognition and the social environment, with reasoners adopting a narrative that feels 'right' to explain the available data; using that narrative to imagine plausible futures; and affectively evaluating those imagined futures to make a choice. Evidence from many areas of the cognitive, behavioral, and social sciences supports this basic model, including lab experiments, interview studies, and econometric analyses. We propose 12 principles to explain how the mental representations (narratives) interact with four inter-related processes (explanation, simulation, affective evaluation, communication), examining the theoretical and empirical basis for each. We conclude by discussing how CNT can provide a common vocabulary for researchers studying everyday choices across areas of the decision sciences.

ChemInformatics Model Explorer (CIME): exploratory analysis of chemical model explanations

The introduction of machine learning to small molecule research- an inherently multidisciplinary field in which chemists and data scientists combine their expertise and collaborate - has been vital to making screening processes more efficient. In recent years, numerous models that predict pharmacokinetic properties or bioactivity have been published, and these are used on a daily basis by chemists to make decisions and prioritize ideas. The emerging field of explainable artificial intelligence is opening up new possibilities for understanding the reasoning that underlies a model. In small molecule research, this means relating contributions of substructures of compounds to their predicted properties, which in turn also allows the areas of the compounds that have the greatest influence on the outcome to be identified. However, there is no interactive visualization tool that facilitates such interdisciplinary collaborations towards interpretability of machine learning models for small molecules. To fill this gap, we present CIME (ChemInformatics Model Explorer), an interactive web-based system that allows users to inspect chemical data sets, visualize model explanations, compare interpretability techniques, and explore subgroups of compounds. The tool is model-agnostic and can be run on a server or a workstation.

Cognitive style predicts how people explain mental magic tricks

Magic tricks are deceiving, yet we can readily generate an explanation for a trick that we do not fully understand. In three experiments, we show that the way people explain a mental magic trick depends on their individual cognitive style. Analytical thinkers tend to generate explanations that appeal to rationality, such as using physical props to accomplish an effect. In contrast, intuitive thinkers are more likely to generate irrational explanations that accord with the magician's provided backstory, such as using subliminal cues to guide a spectator's choices. We observe this effect when measuring a participant's cognitive style using the Cognitive Reflection Test, and also when manipulating a participant's cognitive style using a simple narrative prompt.

Benefits of Writing an Explanation During Pauses in Multimedia Lessons

Generative learning theory posits that learners engage more deeply and produce better learning outcomes when they engage in selecting, organizing, and integrating processes during learning. The present experiments examine whether the generative learning activity of generating explanations can be extended to online multimedia lessons and whether prompts to engage in this generative learning activity work better than more passive instruction. Across three experiments, college students learned about greenhouse gasses from a 4-part online lesson involving captioned animations and subsequently took a posttest. After each part, learners were asked to generate an explanation (write-an-explanation), write an explanation using provided terms (write-a-focused-explanation), rewrite a provided explanation (rewrite-an-explanation), read a provided explanation (read-an-explanation), or simply move on to the next part (no-activity). Overall, students in the write-an-explanation group (Experiments 2 and 3), write-a-focused-explanation group (Experiment 2), and rewrite-an-explanation group (Experiment 3) performed significantly better on a delayed posttest than the no-activity group, but the groups did not differ significantly on an immediate posttest (Experiment 1). These results are consistent with generative learning theory and help identify generative learning strategies that improve online multimedia learning, thereby priming active learning with passive media.

Transparency as design publicity: explaining and justifying inscrutable algorithms

In this paper we argue that transparency of machine learning algorithms, just as explanation, can be defined at different levels of abstraction. We criticize recent attempts to identify the explanation of black box algorithms with making their decisions (post-hoc) interpretable, focusing our discussion on counterfactual explanations. These approaches to explanation simplify the real nature of the black boxes and risk misleading the public about the normative features of a model. We propose a new form of algorithmic transparency, that consists in explaining algorithms as an intentional product, that serves a particular goal, or multiple goals (Daniel Dennet's design stance) in a given domain of applicability, and that provides a measure of the extent to which such a goal is achieved, and evidence about the way that measure has been reached. We call such idea of algorithmic transparency "design publicity." We argue that design publicity can be more easily linked with the justification of the use and of the design of the algorithm, and of each individual decision following from it. In comparison to post-hoc explanations of individual algorithmic decisions, design publicity meets a different demand (the demand for impersonal justification) of the explainee. Finally, we argue that when models that pursue justifiable goals (which may include fairness as avoidance of bias towards specific groups) to a justifiable degree are used consistently, the resulting decisions are all justified even if some of them are (unavoidably) based on incorrect predictions. For this argument, we rely on John Rawls's idea of procedural justice applied to algorithms conceived as institutions.

What is and what never should have been: Children's causal and counterfactual judgments about the same events

Substantial research with adults has characterized the contents of individuals' counterfactual thoughts. In contrast, little is known about the types of events children invoke in their counterfactual thoughts and how they compare with their causal ascriptions. In the current study, we asked children open-ended counterfactual and causal questions about events in which a character's action enabled a force of nature to cause a minor mishap. Children aged 3.5-8 years (N = 160) tended to invoke characters' actions in their counterfactual judgments to explain how an event could have been prevented (e.g., "She should have closed the window") and tended to invoke forces of nature in their causal judgments (e.g., "The rain got it wet"). Younger children were also significantly more likely than older children to invoke forces of nature in their counterfactuals (e.g., "It shouldn't have rained"). These results indicate that, similar to reasoning patterns found in adults, children tend to focus on controllable enabling conditions when reasoning counterfactually, but the results also point to some developmental differences. The developmental similarities suggest that counterfactual reasoning may serve a similar function from middle childhood through adulthood.

Does expertise moderate the seductive allure of reductive explanations?

Non-experts are unduly attracted to explanations of scientific phenomena that contain irrelevant reductive language (e.g., explanations of biological phenomena that mention chemistry; Hopkins, Weisberg, & Taylor, 2016). To determine if expertise would reduce this reasoning error, the current study recruited individuals with graduate-level training in six scientific fields and in philosophy (N = 580) and asked them to judge explanations for phenomena from those fields. Like the novices in Hopkins et al. (2016), scientists' ratings of bad explanations were influenced by reductive information when viewing phenomena from outside their field of expertise, but they were less likely to show this bias when reasoning about their own field. Higher levels of educational attainment did improve detection of bad explanations. These results indicate that advanced training in science or logic can lead to more accurate reasoning about explanations, but does not mitigate the reductive allure effect.

Teaching Problem Explanations Using Instructive Feedback

Instructive feedback (IF) is a procedure in which extra information is presented to a participant during the consequence portion of instruction for other skills. Previous research has demonstrated that participants with intellectual disabilities may acquire a portion of non-targeted skills (secondary targets) without explicit instruction when extra information is presented. Previous research has demonstrated that IF has resulted in more efficient instruction for participants with disabilities as a whole. However, few studies have focused on participants with autism spectrum disorders (ASDs). Additionally, the measures of secondary target acquisition in past research have focused solely on discrete responses (e.g., one-word utterances). The current investigation extended the IF literature related to participants with ASD by including longer verbal responses as secondary targets and assessing maintenance for both primary and secondary targets. Across three participants, IF resulted in the acquisition of at least a portion of secondary targets without explicit teaching. For two participants, additional instruction was required before IF resulted in acquisition of secondary targets. Across all three participants, gains observed for both primary and secondary targets in intervention were maintained.

"Because It's Hers": When Preschoolers Use Ownership in Their Explanations

Young children show competence in reasoning about how ownership affects object use. In the present experiments, we investigate how influential ownership is for young children by examining their explanations. In three experiments, we asked 3- to 5-year-olds (N = 323) to explain why it was acceptable (Experiments 1-3) or unacceptable (Experiment 2 and 3) for a person to use an object. In Experiments 1 and 2, older preschoolers referenced ownership more than alternative considerations when explaining why it was acceptable or unacceptable for a person to use an object, even though ownership was not mentioned to them. In Experiment 3, ownership was mentioned to children. Here, younger preschoolers frequently referenced ownership when explaining unacceptability of using an object, but not when explaining why using it was acceptable. These findings suggest that ownership is influential in preschoolers' explanations about the acceptability of using objects, but that the scope of its influence increases with age.

Improving GRADE evidence tables part 3: detailed guidance for explanatory footnotes supports creating and understanding GRADE certainty in the evidence judgments

BACKGROUND

The Grading of Recommendations Assessment, Development and Evaluation (GRADE) is widely used and reliable and accurate for assessing the certainty in the body of health evidence. The GRADE working group has provided detailed guidance for assessing the certainty in the body of evidence in systematic reviews and health technology assessments (HTAs) and how to grade the strength of health recommendations. However, there is limited advice regarding how to maximize transparency of these judgments, in particular through explanatory footnotes or explanations in Summary of Findings tables and Evidence Profiles (GRADE evidence tables).

METHODS

We conducted this study to define the essential attributes of useful explanations and to develop specific guidance for explanations associated with GRADE evidence tables. We used a sample of explanations according to their complexity, type of judgment involved, and appropriateness from a database of published GRADE evidence tables in Cochrane reviews and World Health Organization guidelines. We used an iterative process and group consensus to determine the attributes and develop guidance.

RESULTS

Explanations in GRADE evidence tables should be concise, informative, relevant, easy to understand, and accurate. We provide general and domain-specific guidance to assist authors with achieving these desirable attributes in their explanations associated with GRADE evidence tables.

CONCLUSIONS

Adhering to the general and GRADE domain-specific guidance should improve the quality of explanations associated with GRADE evidence tables, assist authors of systematic reviews, HTA reports, or guidelines with information that they can use in other parts of their evidence synthesis. This guidance will also support editorial evaluation of evidence syntheses using GRADE and provide a minimum quality standard of judgments across tables.

Building theory-based concepts: four-year-olds preferentially seek explanations for features of kinds

Is the structure of human concepts continuous across development, or does it undergo qualitative transformations? Extensive evidence with adults has demonstrated that they are motivated to understand why categories have the features they do. To investigate whether young children display a similar motivation-an issue that bears on the question of continuity vs. transformation in conceptual structure-we conducted three studies involving 4-year-olds (N=90) and adults (N=124). Experiments 1 and 2 suggested that 4-year-olds indeed display a strong motivation to explain why categories have the features they do. Specifically, when provided with the option of asking "why?" about features of novel categories vs. features of individuals from other novel categories, children preferred to ask "why?" about the category features. Moreover, children's explanatory preference was specific to facts about categories per se and did not extend to facts that were merely presented in the context of multiple category instances. Experiment 3 also ruled out the possibility that the category facts were preferred because these facts were more surprising. In sum, these three studies reveal an early-emerging motivation to make sense of the categories encountered in the world and, more generally, speak to the richness of children's conceptual representations.