The first step in treatment development is identifying the key focus of the work. The need for innovation often becomes clear when there are challenges in treatment. These may be difficulties in treatment acceptability, outcome, or access. For example, treatments that do not address the specific developmental context of YP (Keystone 1), or that too few YP are able to access. These clinical challenges often spark the research process.

To generate a focus for treatment development it is helpful to synthesise clinical observations with the existing literature. Possible sources of observations:

• Clinical reflections
• Recordings of therapy sessions
• Supervision discussions
• Questionnaires

These observations can help us to listen carefully to the key stakeholders (Keystone 2) and give clues to the potential mechanisms at play. In this early phase, it is also important to scope the potential context of any future intervention – in essence considering implementation from the outset.

Combining observations, existing theories and models may lead to a preliminary conceptualisation (Keystone 3), which can be tested and developed further. Here, consider how mechanistic processes may overlap, how they could interact with developmental processes (Keystone 1), or contribute to more severe difficulties.

In summary, clinical observations can help to direct attention to the relevant existing literature, as well as ensuring that treatment innovation has a meaningful target.

To observe the relevant phenomena, qualitative methods can provide rich and complex information by studying a small number of participants, allowing understanding of psychological mechanisms and how they develop (as well as their impact on behaviour, attention, emotions and symptoms).

Case study: Watson et al., 2020

A qualitative study in 34 young people with elevated symptoms of depression (or a diagnosis) revealed a common topic of a lack of positive imagery for future events, in addition to negative memories connected to future events, within the theme of ‘loss of joy or flattening of emotions. Researchers are now targeting this potential mechanism (anhedonia) in imagery-focused interventions for adolescents (e.g. Hutchinson et al., 2024).

Case study: Hewitt et al., 2021

Developmental sensitivities may be indicated when using qualitative methods (Keystone 1). In this qualitative study, young people described feelings of isolation and disconnection from their peers associated with their panic attacks, indicating that the social context (including appraisals of it) affect how young people experience panic attacks. We can use this new connection to inform our conceptual understanding (Braun and Clarke, 2014; Keystone 3), to guide research and ultimately in treatment development. In this example treatment may incorporate work around young people’s thoughts of themselves and others in relation to panic attacks and consider how this could maintain their symptoms.

Interventions may also involve schools, families, and peer networks to help challenge these appraisals.

While qualitative methods can provide very detailed information from participants, allowing nuance that can support quantitative methods and can be transferable to other contexts, it is limited as the findings cannot necessarily be generalised in the traditional sense beyond this group.

More recent approaches to qualitative work include using larger samples of text-based data with lard language models, allowing a wider range of experiences and characteristics to be represented (eg Chiu et al., 2022).

Guideline: COREQ

Numerous tools exist for evaluating the quality of qualitative research (Santiago-Delefosse et al., 2016). The most widely used tool for both researchers and research evaluators in health research is the Consolidated criteria for Reporting Qualitative research: COREQ (Majid & Vanstone, 2018; Walsh et al., 2020). The COREQ (Tong et al., 2007) is a 32-item checklist for qualitative research and is designed to improve quality and promote complete and transparent reporting (Tong et al., 2007).

Guideline: BQQRG

Braun & Clarke (2024) suggest caution in the use of the COREQ as a universal tool for quality. They argue that given the plethora of different orientations and values of qualitative research the COREQ is best suited to post- positivist/realist or ‘small q’ qualitative research. As an alternative for researchers who practice ‘fully’ qualitative, non-positivist, Big Q qualitative research Braun and Clarke developed the Big Q Qualitative Reporting Guidelines (BQQRG: Braun & Clarke, 2025). The aim of the BQQRG is to provide a guide for Big Q qualitative researchers based on the values of qualitative research rather than on a consensus-based framework.

Measurement allows understanding of the scale of a problem, while detecting those in need, informing development of treatments, and tracking outcomes of interventions. Explicit mechanism measures are needed to identify and test if it is possible to modify causal factors contributing to mental health problems.

Important features of measures:

• Capture the relevant phenomena (the target problem or mechanism), by working with people with lived experience of the problem (Keystone 2).
• Have been developed in representative cohorts, including clinical groups in addition to large general population groups, allowing the range of severity to be covered.
• Reliable and valid
• Easy to use across settings (including clinical services, schools and online)
• Can be interpreted by age and gender

Measures may be taken by questionnaires (which could be completed by the individual, a parent or carer, teacher or clinician), assessor-rated interviews, tasks, or observation. The developmental context (Keystone 2) should be considered when shaping how a measure may be developed, validated, and used. This could be in the specific use of language as well as the content of items – for example ensuring that items related to social interactions reflect the importance of the peer context.

Some measures have been developed specifically for YP (e.g., Bird et al., 2020), more often measures have been adapted and/or validated in the target population (e.g., Illingworth et al., 2024).

Case Study: Bird et al 2020

The items for the Bird Checklist of Adolescent Paranoia were developed with people with lived experience of the problem. The measure was then tested with students at a secondary school, as well as young people attending mental health services. This means the measure can be used in both clinical services and wider adolescent contexts – such as identifying those with elevated levels of the difficulty in the school population, which can be useful for screening and prevention  The  psychometric properties of the measure were tested to check for reliability [link to glossary] and validity [link to glossary]. Item response theory was used to assess the properties of each item such that the measure can be applied adaptively – this means fewer items are required to estimate the relevant construct. The measure can then be delivered using computer adaptive tests (CAT) to reduce the burden on participants.

Representative samples, such as those accessed through research-ready schools (e.g., https://wisdom.mhid.org.uk), can be used for identifying patterns which otherwise may not be detected in smaller or more selective clinical samples.

We have developed a community library of measures of mechanisms.

If an appropriate measure cannot be found, development of a measure is the next step, which could be done in one of three ways:

• Adapt an existing measure
• Validate an existing measure in the target population
• Develop a new measure

To develop an appropriate measure, qualitative work can be used to ensure items included are relevant, meaningful, and appropriate to the developmental context of the target group (for example, Boateng et al., 2018)

Methods focus:

Ecological momentary assessment (EMA) involves repeated measure completion in real time while the participant goes about their everyday life – data is often collected on their mobile phone. This reduces bias in recall and provides the opportunity to examine mechanisms in various settings. It is more resource intensive for participants, however it is also more ecologically valid than typical    measure methods.

Large-scale longitudinal datasets can be used to examine mechanisms involved in mental health problems in young people, for example by exploring the scale and frequency of cognitive, emotional, or behavioural processes in the population. These datasets can also be useful for looking at subpopulations, such as those at risk of or currently experiencing mental health problems.

Case Study: WISDOM Research-ready schools

Representative samples, such as those accessed through research-ready schools (e.g., https://wisdom.mhid.org.uk), can be used for identifying patterns which otherwise may not be detected in smaller or more selective clinical samples.

Case Study: Atlas Longitudinal Datasets

Pre-existing youth cohort datasets can also support this kind of research (e.g, https://atlaslongitudinaldatasets.ac.uk).

Longitudinal datasets allow associations to be made between suggested mechanisms and symptoms over time, which can support causal inference (ideally with at least three time-points). They can:

• Allow temporal ordering to be considered.
• Support findings of within-person change, distinguishing this from between-person differences.
• Can allow testing of mediation over time.

However, many large cohort studies include broad symptom measures, general risk factors and nonspecific demographic variables. This means many researchers collect datasets from scratch, in order to use measures of more specific psychological mechanisms (eg attentional bias, safety-seeking behaviours, mental imagery).

Some studies have shown age-related differences in the strength or nature of these associations, suggesting that certain mechanisms may have greater relevance at particular points in development (Keystone 1). This can help us more accurately characterise risk pathways and identify windows of opportunity for intervention.

As mental health problems typically have many causal factors which are often related to each other, there are likely various possible treatment targets.

Definition:

Multifactorial Causation – We assume there are multiple contributory causal factors that interact in a dynamic interplay, and these may vary, at least to some extent, between individuals.

Method focus:

To better understand complexities when multifactorial causation is indicated, a dynamic network perspective may be used, by modelling symptoms and mechanisms as interacting nodes (Borsboom & Cramer, 2013). This may be used to inform the theoretical conceptualization (Keystone 3). Within dynamic networks, highly connected problems indicate valuable treatment targets, allowing treatment to modify common specific and non-specific factors which have so far been overlooked (e.g., Waite et al., 2020).

Clinical staging models indicate that broad spectrum treatments might be particularly valuable in preventative interventions as there are often less differentiated presentations in the early phases of mental health problems (McGorry et al., 2014; Shah et al., 2020).

Network analysis and structural equation modeling (SEM) are advanced statistical techniques ideally suited to this work. These approaches are explicitly built on the recognition of multi-factorial causation and can identify key mechanisms to target in treatment.

Features of network analysis models:

• They can be used to identify the components of systems (nodes).
• They can indicate relations between nodes (links).
• They can inform typology (including the importance of particular nodes; Borsboom et al., 2021; McNally, 2021; Moffa et al., 2017; Pearl, 2009).
• Undirected partial correlation networks make no assumptions regarding the direction of relationships between nodes.
• Directed acyclic graph models (DAGs), built on Bayesian inference, allow assumptions regarding the direction of direct and indirect causal pathways between nodes to be tested.

Case study: Bird et al., 2019

In this example, DAGs have been used to map potential psychological and social factors contributing to paranoia in adolescents (Bird et al., 2019).  In line with the developmental context (Keystone 2), specific factors were found to contribute to the occurrence of inaccurate thoughts of harm from others (paranoia) in adolescents. This included negative affect, peer difficulties – including bullying, and responses to (but not use of) social media.

Method focus:

SEM is a statistical technique used to test hypothesised relationships between variables, including both observed and unobservable constructs (named latent constructs). It combines elements of factor analysis and path analysis, allowing researchers to model complex relationships and assess the fit of their models to observed data. In this example, SEM was used to investigate the contribution of different aspects of social media use such as social comparison, passing time, seeking support, and experiencing hostility (to and from others) to depression in adolescents (Twivy et al., 2025).

Mapping multiple hypothesised causal mechanisms using network analysis or SEM can help identify potentially influential targets within the network of interconnected problems. However, to determine causality, manipulation methods are required.

Experimental methods offer a powerful approach to investigating the causal role of psychological mechanisms in the onset, maintenance, or amelioration of mental health difficulties. These methods are rooted in a causal-interventionist framework (Kendler & Campbell, 2009), which involves deliberately altering the system to observe whether changes in the hypothesised mechanism led to changes in relevant outcomes. This is crucial for isolating mechanisms and establishing causal relationships, providing a robust foundation for the development of targeted interventions.

Firstly, the psychological mechanism of interest (e.g., mental imagery) must be identified, and then an aspect of this should be targeted (e.g., imagery valence, vividness, perspective, or time orientation).

Experimental study design requires consideration of several factors, including:

• How to manipulate the process
• Which control condition to use
• How to account for context and complexity

It is important for researchers to also consider, from the outset, how manipulations can be done in an ecologically valid way that is acceptable and meaningful to the young people of interest. Involving young people and stakeholders (Keystone 3) at this stage can maximise the likelihood of developing meaningful, acceptable procedures that are more likely to translate into therapeutic procedures.

In addition, inclusion of open questions at the end of the testing session to learn about participants’ experience of the procedure, and systematic logging of potential adverse effects (however apparently minor) will all be valuable as therapy processes are developed.

Related to this is the choice of a comparator condition, which could include:

• Passive controls: These participants receive no instructions. A passive control condition allows absolute efficiency of the manipulation to be established.
• Active controls: These participants may be instructed to carry out the opposite behaviour, or an attention placebo may be used. This allows causality to be tested more robustly; it also reduces dependence on sample characteristics, as participants in passive conditions (e.g., no instruction) may engage in widely varying mental processes depending on their symptom profile. Most importantly for the purposes of treatment innovation, it informs development of therapy processes.
• A combination of passive and active controls: this is preferable, but a larger sample size is needed to account for the increased number of conditions, and this might not be feasible.

To establish whether a manipulation has been successful, the vast majority of studies rely on self-report (typically Visual Analogue Scales [VAS] after the manipulation), asking participants to rate the extent to which they followed the instructions or engaged in the particular process during the preceding period. While VAS are quick and easy to include in paradigms, they bring a range of issues, such as:

• Potential biases and subjectivity such as demand characteristics and lack of insight into the particular process
• Measurement error due to common reliance on single item measures
• Inability to accurately capture the temporal variability in how individuals engage with a process over time

While implicit measures overcome some of these issues, there are some concerns associated with them, such as interpretational ambiguity and sensitivity to task conditions. In the absence of a reliable and valid manipulation check, using a combination of implicit and self-report measures may be useful.

When developing experimental paradigms with the aim of treatment relevance, it is essential to select outcome measures carefully – these should go beyond immediate task performance to include proximal and distal indicators of therapeutic change (e.g., changes in affect, behaviour, or symptom expression). Pilot testing can provide insight into feasibility, acceptability and sensitivity of the manipulation and the outcome measures. This allows processes to be refined before they are used in larger studies.

It is highly unlikely that psychological processes operate alone. Instead, processes will act in tandem, reciprocally, to drive symptoms (Everaert et al., 2012; Hirsch et al., 2006). Therefore, there is a tension between the lab and reality, but there are good examples of how we can bring complexity into our experimental paradigms. For instance, multiple mechanisms can be manipulated within one study (AlMoghrabi et al., 2022; Kavallari & Lau, 2022; Platt et al., 2017).

Alternatively, in studies which manipulate one mechanism, the other relevant mechanisms can be measured (e.g., measuring attribution bias when manipulating ruminative thinking in response to stress in the context of depression).

Case study: Leigh et al., 2021

An example of effective mechanism manipulation comes from adolescent social anxiety, in which the manipulation of imagery in social situations (negative vs. benign) was associated with a differential effect on spontaneous use of safety behaviours, with participants using more safety behaviours when holding a negative compared to positive image in mind (Leigh et al., 2021).

In this resource, the term ‘psychological intervention’ is a collective term encompassing the following:

• Prevention: These are strategies aimed at reducing the incidence of mental health problems before they emerge. Prevention can be universal (targeting the whole population), selective (targeting high-risk groups), or indicated (targeting individuals with early signs/symptoms).
• Early intervention: Early interventions are delivered at the earliest stages of a mental health problem to prevent progression to more severe or enduring difficulties. Often, these focus on subthreshold symptoms or recent onset.
• Treatment: Interventions designed to reduce or eliminate symptoms of a diagnosed mental health disorder, including evidence-based therapies (e.g., CBT). Treatment is typically delivered once a full disorder is present.
• Relapse prevention: Interventions aimed at maintaining recovery and preventing recurrence following successful treatment. Relapse prevention may involve ongoing support or booster sessions.
• Recovery support: These are broader interventions that promote functioning, quality of life, and personal recovery (beyond symptom reduction), especially in chronic or recurring conditions.

The experimental paradigm from Step 6 can be used in treatment development (hence intervention development may overlap with experimental testing), by using a causal-interventionist paradigm: randomised controlled trials of treatments focused on a mechanism and measuring change in the primary clinical outcome. Employing a casual-interventionist paradigm – with a specific single focus of the intervention to bring greater confidence in the conclusions drawn from the work – facilitates the testing of the treatment and the underlying theory simultaneously, potentially speeding clinical translation (Freeman, 2011).

Methods focus:

Single-case experimental designs (SCED) are experimental designs which can be used to test the effect of an intervention with a small number of patients (typically one to three) and involve repeated measurements, sequential (potentially randomised) introduction of intervention techniques, and method-specific data analysis such as visual analysis and specific statistics. The essential components of the design framework are:

1. Studying prospectively and intensively a single person or group of individuals over time.
2. Frequent repeated measurement of outcomes and mechanisms.
3. Sequential addition and/or withdrawal of intervention techniques.

The repeated measures allow the individual to act as their own control, by comparing the baseline score (before the intervention is introduced) [phase A] with the intervention phase [phase B]. The methods can be used to evaluate the efficacy of the treatment – the key question of: does it work? It can be used specifically to pilot novel interventions, modify existing treatments, and investigate the active components of an intervention package. SCED can provide an insight into the components of therapy that lead to meaningful change which is necessary to refine the treatment. Regular assessment within treatment also provides the necessary data to track patterns of change and potentially model different trajectories to predict outcome.

Multiple baseline designs can be used to introduce the intervention to different patients, settings, or target behaviours. These can be particularly valuable when there is no expected immediate effect or the treatment effect cannot be ‘withdrawn’, or in psychological therapy when changes in a treatment session may be expected to persist after the application of the technique.

Typically, at least three attempts should be made to demonstrate an intervention effect. This could be at least three phase changes in an introduction/withdrawal [ABA] design, or at least three patients/settings/behaviours in a multiple baseline design.

Given the importance of frequent repeated measurement in SCED careful consideration should be given to the primary outcome measure. Often a combination of standardised and idiographic (individualised and meaningful to the young person) measures are used. Ecological momentary assessment (EMA) methods can also be used in the context of SCED to provide insights into the momentary events in people’s lives and the psychological constructs under investigation.

Randomisation in SCED is used to determine the order or timing of treatment phases, rather than allocation to treatment groups. The baseline phase establishes a trend with which to compare subsequent phases.  SCED analysis typically includes visual inspection of the data.  Statistical tests can also be used. Measures of treatment fidelity and quality will also be important to interpret the findings of treatment evaluation.

Guidelines:

The SCRIBE and ROBINT guidelines provide recommendations on defining the target variable, measurement, and good practice (Tate et al., 2016; Tate et al., 2013). Examples of types of SCED designs (Krasny-Pacini & Evans, 2018; Smith, 2012).

Other treatment development methods include cohort studies, in which an A-B design can be applied to a group of participants with analysis of data by the participant group rather than at the individual level. Feasibility trials are used to gain information regarding the potential viability of completing a full efficacy trial. In this early phase 1 treatment development work, the focus is on generating proof of concept data. Studies are not sufficiently powered to test treatment effects fully. Therefore, analysis should be mainly descriptive or focus on confidence interval estimation, rather than reporting p-values (Lancaster et al., 2004). Robust testing of the treatment effects can be completed through randomised controlled trials (see Step 8).

Randomised controlled trials (RCTs) are the gold standard to evaluate therapeutic efficacy and effectiveness and should be planned and carried out in line with best practice guidelines.

Guidelines:

The CONSORT statement provides guidance for conducting and reporting RCTs (Hopewell et al., 2025).

To strengthen mechanistic insight, RCTs should incorporate measures that assess proposed mechanisms of change, such as validated questionnaires and, where possible, objective or behavioural measures. Including mechanistic assessments allows for testing whether interventions work through the pathways they target, helping refine theory and improve intervention precision (see Step 9).

Further efficiency in translational treatment development may be possible through the use of novel trial designs such as platform trials in which it is possible to test multiple mechanistic treatments simultaneously or in multiphase or sequential designs. Treatment acceptability is often measured by satisfaction and treatment uptake. Yet, few measures of acceptability (e.g., Tarnowski & Simonian, 1992) have been developed, especially for young people.

Throughout treatment development and evaluation, attention should also be given to the therapeutic style and ethos of the approach. This can be done through co-production methods to ensure treatments are meaningful and acceptable to YP, including under-represented groups (Keystone 2).

Method focus:

Qualitative methods can be employed to gain insights into the patient experience of the intervention and explore what works, when, and why (e.g., Waite et al., 2025). This can be used to refine treatment further. Qualitative investigations can also be delivered with treatment deliverers to assess the likely feasibility of future implementation (see Step 10). This can also help identify who is best placed to deliver an intervention – clinician, teacher, peer, or standalone digital resource.

Method focus:

Propensity Score Matching (PSM) is an alternative approach that can be used to reduce selection bias in observational studies when an RCT is not feasible due to ethical, practical, or financial constraints. PSM estimates each participant’s probability of receiving the treatment based on observed covariates, then matches individuals with similar scores to create comparable groups. This mimics some of the benefits of randomisation by balancing covariates between groups. PSM is especially useful when researchers cannot randomly assign participants but still want to draw causal inferences. However, it only accounts for measured confounders and may exclude unmatched cases, reducing sample size.

Intervention studies can offer valuable insights into psychological mechanisms when they are designed with this goal in mind.

Method focus:

Mediation analyses, for example, can test whether changes in a proposed mechanism account for the therapeutic effect, providing evidence for a causal pathway (e.g., Pineda & Dadds, 2013; Idsoe et al., 2019). Beyond mediation, the process of “back-translation” — where specific intervention procedures and their effects are examined in the context of therapy— further contributes to our understanding of mechanisms (Cohen et al., 2023).

Case Study: Leigh et al., 2025

For example, within the context of cognitive therapy for adolescent social anxiety, large effects of the self-focus attention and safety behaviour experiment and video feedback were found on anxiety and self-appraisals, indicating their value in therapy and providing support for the role of self-focus, safety behaviours, and negative imagery in the maintenance of social anxiety (Leigh et al., 2025).

Dismantling studies, which isolate and remove components of a treatment, and factorial RCTs, which systematically vary the presence or absence of different elements, are further powerful methods for identifying active ingredients and their interactions. Moderation analyses can reveal for whom and under what conditions certain mechanisms operate most strongly. In-session data can be analysed to identify trajectories of change. These approaches can highlight individual or contextual differences in how interventions work. Taken together, these approaches can help clarify the psychological processes driving change and guide the refinement of more effective, targeted treatments

Successful implementation of evidence-based treatments is critical for improving outcomes for YP. Yet too often there are delays and barriers, meaning exciting innovations do not reach those who need them. Over the last 20 years the field of implementation science has generated greater understanding and methods to improve the sustainability of new treatment approaches in health care (e.g., Greenhalgh et al., 2004).

A range of models, frameworks, and guidelines have been produced which set out key elements to be considered. Consistent across these are several common principles (e.g., McGinty et al., 2024):

• Firstly, whenever possible integrate implementation throughout the intervention development pathway, for example engaging and listening carefully to key stakeholders (as described in Keystone 2 and Steps 1 and 2). Draw on the range of perspectives and disciplines and embed this understanding in the conceptualisation (Keystone 3).
• Secondly, consider the scale of delivery required and the potential structural barriers to this, including inequities in health problems and access. Recognise and embrace the complexity within the broad range of settings that interventions may be delivered in – for example, clinics, schools, digital (standalone or supported) – and seek and utilise local adaptions to spur further innovations.
• Finally, draw on the variety of methods that can be used throughout the translational pathway, including natural experiments which can complement RCTs.

Implementation then requires evaluation across a number of outcomes, such as acceptability, adoption, appropriateness, feasibility, fidelity, cost, penetration, and sustainability (Proctor et al., 2011) across different settings, and over time. Implementation is ongoing, it must be reviewed and renewed in line with intervention and context changes to ensure YP continue to benefit from effective, acceptable, and accessible interventions.

AlMoghrabi, N., Franken, I. H. A., Mayer, B., & Huijding, J. (2022). A Single-Session Combined Cognitive Bias Modification Training Targeting Attention and Interpretation Biases in Aggression. Behaviour Change, 39(1), 1-20. https://doi.org/10.1017/bec.2021.11

Bird, J. C., Evans, R., Waite, F., Loe, B. S., & Freeman, D. (2019). Adolescent Paranoia: Prevalence, Structure, and Causal Mechanisms. Schizophr Bull, 45(5), 1134-1142. https://doi.org/10.1093/schbul/sby180

Bird, J. C., Loe, B. S., Kirkham, M., Fergusson, E. C., Shearn, C., Stratford, H.,…Freeman, D. (2020). The assessment of paranoia in young people: Item and test properties of the Bird Checklist of Adolescent Paranoia. Schizophr Res, 220, 116-122. https://doi.org/10.1016/j.schres.2020.03.046

Boateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Quiñonez, H. R., & Young, S. L. (2018). Best Practices for Developing and Validating Scales for Health, Social, and Behavioral Research: A Primer. Frontiers in public health, 6, 149. https://doi.org/10.3389/fpubh.2018.00149

Borsboom, D., & Cramer, A. O. (2013). Network analysis: an integrative approach to the structure of psychopathology. Annual review of clinical psychology, 9, 91–121. https://doi.org/10.1146/annurev-clinpsy-050212-185608

Borsboom, D., Deserno, M.K., Rhemtulla, M. et al. Network analysis of multivariate data in psychological science. Nat Rev Methods Primers 1, 58 (2021). https://doi.org/10.1038/s43586-021-00055-w

Braun, V., & Clarke, V. (2014). What can “thematic analysis” offer health and wellbeing researchers? International Journal of Qualitative Studies on Health and Well-being, 9(1), 26152. https://doi.org/10.3402/qhw.v9.26152

Braun, V., & Clarke, V. (2024). How do you solve a problem like COREQ? A critique of Tong et al.’s (2007) Consolidated Criteria for Reporting Qualitative Research. Methods in Psychology, 11  https://doi.org/10.1016/j.metip.2024.100155

Braun, V., & Clarke, V. (2025). Reporting guidelines for qualitative research: a values-based approach, Qualitative Research in Psychology, 22:2, 399-438, doi: 10.1080/14780887.2024.2382244

Chiu, K., Clark, D. M., & Leigh, E. (2022). Characterising Negative Mental Imagery in Adolescent Social Anxiety. Cognitive Therapy and Research. https://doi.org/10.1007/s10608-022-10316-x

Cohen, Z. D., Barnes-Horowitz, N. M., Forbes, C. N., & Craske, M. G. (2023). Measuring the active elements of cognitive-behavioral therapies. Behaviour Research and Therapy, 167, 104364. https://doi.org/https://doi.org/10.1016/j.brat.2023.104364

Everaert, J., Koster, E. H., & Derakshan, N. (2012). The combined cognitive bias hypothesis in depression. Clin Psychol Rev, 32(5), 413-424. https://doi.org/10.1016/j.cpr.2012.04.003

Freeman, D. (2011). Improving cognitive treatments for delusions. Schizophrenia Research, 132(2), 135-139. https://doi.org/https://doi.org/10.1016/j.schres.2011.08.012

Greenhalgh, T., Robert, G., Macfarlane, F., Bate, P., & Kyriakidou, O. (2004). Diffusion of Innovations in Service Organizations: Systematic Review and Recommendations. The Milbank Quarterly, 82(4), 581–629. https://doi.org/10.1111/j.0887-378X.2004.00325.x

Hewitt, O. M., Alice, T., & and Waite, P. (2021). The experience of panic attacks in adolescents: an interpretative phenomenological analysis study. Emotional and Behavioural Difficulties, 26(3), 240-253. https://doi.org/10.1080/13632752.2021.1948742

Hirsch, C. R., Clark, D. M., & Mathews, A. (2006). Imagery and interpretations in social phobia: Support for the combined cognitive biases hypothesis. Behavior Therapy, 37(3), 223-236.

Hopewell, S., Chan, A.-W., Collins, G. S., Hróbjartsson, A., Moher, D., Schulz, K. F.,…Boutron, I. (2025). CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ, 389, e081123. https://doi.org/10.1136/bmj-2024-081123

Hutchinson, T., Lau, J. Y. F., Smith, P., & Pile, V. (2024). Targeting Anhedonia in Adolescents: A Single Case Series of a Positive Imagery-Based Early Intervention. International Journal of Cognitive Therapy, 17(3), 429-465. https://doi.org/10.1007/s41811-024-00202-7

Idsoe, T., Keles, S., Olseth, A.R. et al. Cognitive behavioral treatment for depressed adolescents: results from a cluster randomized controlled trial of a group course. BMC Psychiatry 19, 155 (2019). https://doi.org/10.1186/s12888-019-2134-3

Illingworth, G., Mansfield, K. L., Skripkauskaite, S., Fazel, M., & Waite, F. (2024). Insomnia symptoms in children and adolescents: Screening for sleep problems with the two-item Sleep Condition Indicator (SCI-02). BMC Public Health, 24(1), 2957. https://doi.org/10.1186/s12889-024-20310-5

Kavallari, D., & Lau, J. Y. F. (2022). Testing a Combined Cognitive Bias Hypothesis of Pain and Pain-related Worry in Young People. The Journal of Pain, 23(6), 1082-1091. https://doi.org/https://doi.org/10.1016/j.jpain.2022.01.004

Kendler, K. S., & Campbell, J. (2009). Interventionist causal models in psychiatry: repositioning the mind-body problem. Psychological medicine, 39(6), 881–887. https://doi.org/10.1017/S0033291708004467

Krasny-Pacini, A., & Evans, J. (2018). Single-case experimental designs to assess intervention effectiveness in rehabilitation: A practical guide. Annals of Physical and Rehabilitation Medicine, 61(3), 164-179. https://doi.org/https://doi.org/10.1016/j.rehab.2017.12.002

Lancaster, G. A., Dodd, S., & Williamson, P. R. (2004). Design and analysis of pilot studies: recommendations for good practice. J Eval Clin Pract, 10(2), 307-312. https://doi.org/10.1111/j..2002.384.doc.x

Leigh E, Chiu K, Clark DM (2021) Self-focused attention and safety behaviours maintain social anxiety in adolescents: An experimental study. PLoS ONE 16(2): e0247703. https://doi.org/10.1371/journal.pone.0247703

Leigh, E., Clark, D., & Chiu, K. (2025). Examining two of the ingredients of Cognitive therapy for adolescent social anxiety disorder: Back-translation from a treatment trial. Journal of Behavior Therapy and Experimental Psychiatry, 88, 102020. https://doi.org/https://doi.org/10.1016/j.jbtep.2025.102020

Majid, U., Vanstone, M., 2018. Appraising qualitative research for evidence syntheses: a compendium of quality appraisal tools. Qual. Health Res. 28 (13), 2115–2131. https://doi.org/10.1177/1049732318785358.

McGinty, E. E., Alegria, M., Beidas, R. S., Braithwaite, J., Kola, L., Leslie, D. L., Moise, N., Mueller, B., Pincus, H. A., Shidhaye, R., Simon, K., Singer, S. J., Stuart, E. A., & Eisenberg, M. D. (2024). The Lancet Psychiatry Commission: Transforming mental health implementation research. The Lancet Psychiatry, 11(5), 368–396. https://doi.org/10.1016/S2215-0366(24)00040-3

McGorry, P., Keshavan, M., Goldstone, S., Amminger, P., Allott, K., Berk, M., Lavoie, S., Pantelis, C., Yung, A., Wood, S. and Hickie, I. (2014), Biomarkers and clinical staging in psychiatry. -, 13: 211-223. https://doi.org/10.1002/wps.20144

McNally R. J. (2021). Network Analysis of Psychopathology: Controversies and Challenges. Annual review of clinical psychology, 17, 31–53. https://doi.org/10.1146/annurev-clinpsy-081219-092850

Moffa, G., Catone, G., Kuipers, J., Kuipers, E., Freeman, D., Marwaha, S., Lennox, B. R., Broome, M. R., & Bebbington, P. (2017). Using Directed Acyclic Graphs in Epidemiological Research in Psychosis: An Analysis of the Role of Bullying in Psychosis. Schizophrenia bulletin, 43(6), 1273–1279. https://doi.org/10.1093/schbul/sbx013

Pearl, J. (2009). Causality. Cambridge university press.

Pineda, J., & Dadds, M. R. (2013). Family intervention for adolescents with suicidal behavior: a randomized controlled trial and mediation analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 52(8), 851–862. https://doi.org/10.1016/j.jaac.2013.05.015

Platt, B., M., W. A., Gerd, S.-K., Lina, E., & and Salemink, E. (2017). A review of cognitive biases in youth depression: attention, interpretation and memory. Cognition and Emotion, 31(3), 462-483. https://doi.org/10.1080/02699931.2015.1127215

Polari, A., Lavoie, S., Yuen, H. P., Amminger, P., Berger, G., Chen, E., deHaan, L., Hartmann, J., Markulev, C., Melville, F., Nieman, D., Nordentoft, M., Riecher-Rössler, A., Smesny, S., Stratford, J., Verma, S., Yung, A., McGorry, P., & Nelson, B. (2018). Clinical trajectories in the ultra-high risk for psychosis population. Schizophrenia research, 197, 550–556. https://doi.org/10.1016/j.schres.2018.01.022

Proctor, E., Silmere, H., Raghavan, R., Hovmand, P., Aarons, G., Bunger, A.,…Hensley, M. (2011). Outcomes for Implementation Research: Conceptual Distinctions, Measurement Challenges, and Research Agenda. Administration and Policy in Mental Health and Mental Health Services Research, 38(2), 65-76. https://doi.org/10.1007/s10488-010-0319-7

Santiago-Delefosse, M., Gavin, A., Bruchez, C., Roux, P., Stephen, S.L., 2016. Quality of qualitative research in the health sciences: analysis of the common criteria present in 58 assessment guidelines by expert users. Soc. Sci. Med. 148, 142–151. https://doi. org/10.1016/j.socscimed.2015.11.007.

Shah, J. L., Scott, J., McGorry, P. D., Cross, S. P. M., Keshavan, M. S., Nelson, B., Wood, S. J., Marwaha, S., Yung, A. R., Scott, E. M., Öngür, D., Conus, P., Henry, C., Hickie, I. B., & International Working Group on Transdiagnostic Clinical Staging in Youth Mental Health (2020). Transdiagnostic clinical staging in youth mental health: a first international consensus statement. World psychiatry : official journal of the World Psychiatric Association (WPA), 19(2), 233–242. https://doi.org/10.1002/wps.20745

Smith, J. D. (2012). Single-case experimental designs: a systematic review of published research and current standards. Psychol Methods, 17(4), 510-550. https://doi.org/10.1037/a0029312

Tarnowski, K. J., & Simonian, S. J. (1992). Assessing treatment acceptance: The abbreviated acceptability rating profile. Journal of Behavior Therapy and Experimental Psychiatry, 23(2), 101–106. https://doi.org/10.1016/0005-7916(92)90007-6

Tate, R. L., Perdices, M., Rosenkoetter, U., McDonald, S., Togher, L., Shadish, W.,…Vohra, S. (2016). The Single-Case Reporting Guideline In BEhavioural Interventions (SCRIBE) 2016: Explanation and elaboration. Archives of Scientific Psychology, 4(1), 10-31. https://doi.org/10.1037/arc0000027

Tate, R. L., Perdices, M., Rosenkoetter, U., Wakim, D., Godbee, K., Togher, L., & McDonald, S. (2013). Revision of a method quality rating scale for single-case experimental designs and n-of-1 trials: The 15-item Risk of Bias in N-of-1 Trials (RoBiNT) Scale. Neuropsychological Rehabilitation, 23(5), 619-638. https://doi.org/10.1080/09602011.2013.824383

Tong, A., Sainsbury, P., Craig, J., 2007. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int. J. Qual. Health Care 19 (6), 349–357. https://doi.org/10.1093/intqhc/mzm042.

Twivy, E., Freeman, D., Ciorsdan, A., Sheng, L. B., & Waite, F. (2025). The social media scale for depression in adolescence. International Journal of Adolescence and Youth, 30(1), 2450425. https://doi.org/10.1080/02673843.2025.2450425

Waite, F., Evans, S., Rebello, A., Sharpe, T., Otaiku, J., Iredale, E., Kabir, T., Černis, E., & Freeman, D. (2025). Sleep disruption and its psychological treatment in young people at risk of psychosis: A peer methods qualitative evaluation. The British journal of clinical psychology, 10.1111/bjc.70002. Advance online publication. https://doi.org/10.1111/bjc.70002

Waite, F., Sheaves, B., Isham, L., Reeve, S., & Freeman, D. (2020). Sleep and schizophrenia: From epiphenomenon to treatable causal target. Schizophrenia Research, 221, 44-56. https://doi.org/10.1016/j.schres.2019.11.014

Walsh, S., Jones, M., Bressington, D., McKenna, L., Brown, E., Terhaag, S., Shresta, M., Al-Ghareeb, A., Gray, R., 2020. Adherence to COREQ reporting guidelines for qualitative research: a scientometric study in nursing social science. Int. J. Qual. Methods 19, 1–9. https://doi.org/10.1177/16094096920982145.

Watson, R., Harvey, K., McCabe, C., & Reynolds, S. (2020). Understanding anhedonia: a qualitative study exploring loss of interest and pleasure in adolescent depression. Eur Child Adolesc Psychiatry, 29(4), 489-499. https://doi.org/10.1007/s00787-019-01364-y