Advancing clinical psychology with improved doctoral training in science

Abstract

Clinical psychology is a young science. Psychological suffering remains extensive, with core scientific issues the subject of ongoing study and debate. Clinical practice depends on the research base for elucidating etiology, systematic improvements in provision of mental health care, improvements in public health, and credibility with clients and the larger public. Science-oriented training programs endeavor to develop students into independent scientists, clinicians, and/or administrators who produce knowledge and translate it to improve clinical care. Existing training standards accommodate programs with a heavy clinical focus and little to no research training. To accelerate scientific progress, these standards must also accommodate increased scientific and research training by reducing generalist requirements and increasing curricular flexibility for students. On net, these changes would allow clinical psychologists more opportunities to work at the level of the clinic, system, policy, or laboratory; and thereby accelerate scientific discovery and translation to understand and reduce mental health burdens in our society, all while existing programs, including practitioner-oriented programs, may continue to flourish.

A standard and formal training model for clinical psychology in the United States has been in place for over 70 years, since the first training programs were accredited in 1948. Slowly but surely, clinical psychology has been influenced by advances in empirically-based assessment and intervention. Research standards have been developed to adjudicate effective from ineffective techniques, and clinical psychologists are keen to use what works. However, it is difficult to discern whether clinical psychology has made a substantial impact on the overall mental health burden (Furukawa & Kessler, 2019; Greenberg et al., 2003; Kazdin & Blase, 2011; Kessler, Berglund, et al., 2005; Kessler, Demler, et al., 2005). Contrast this with marks of progress in other applied sciences over the same period: vaccines inoculate, crop yields feed billions, human organs are grown, and videoconferences are joined by smartphone. These are very real and obvious objective markers of technological progress and the scientific knowledge underlying those successes. To what obvious objective success does clinical psychology point? We have certainly learned something about the nature and origins of human individual differences and psychological suffering but human behavior is exceedingly complex, and our ability to apply our limited knowledge has not had an obvious and overwhelming impact on the most basic issues that face our profession: the understanding, assessment, amelioration, and prevention of mental illness. With rates of psychopathology seeming to rise across industrialized economies (Rehm & Shield, 2019), there is an increasingly urgent need to address these challenges that plague our communities. We contend that one important way forward is to improve the quality and quantity of clinical psychology research through improved research training, while still nurturing the connection between the laboratory and the clinic. Our proposal is not entirely new, but rather borrows from past proposals and structures, adapting them to meet current challenges in applied science training in clinical psychology.

In research-oriented clinical psychology training programs with which we are familiar, each new student is tasked with a first-year research project. For many of them, it is the first time they have shepherded a research project of any significance from beginning to end. Year after year, these new students take on increasingly demanding projects filled with big theoretical ideas, complex technologies, large datasets, and analytic strategies that were not possible even five or ten years ago. The first-year project is a fundamental milestone for these students and reflects a core tenant that the primary goal of research-oriented programs is to train students to be highly competent in clinical psychology research. Yet, the conduct of impactful original research is minimally weighted in accreditation standards of clinical psychology training programs. Moreover, clinical students who wish to pursue research careers are at a fundamental disadvantage to students in pure research training programs as they seek to maximize their research knowledge base while pursuing other more applied competencies.

We propose that training standards, regardless of accrediting body, would do well to accommodate programs that are focused on both the generation of that knowledge and its application to clinical practice. The proposed program structure could be entertained by either current accrediting body, given that APA CoA has long had a broad scope including programs ranging from research-intensive to practice-intensive, and PCSAS specifically promotes scientific training. The most comprehensive formulation of this thesis as it relates to clinical psychology training is from Baker and colleagues (Baker, McFall, & Shoham, 2008), and was previously argued polemically by McFall (1991, 2000) and Meehl (1972). These authors were critical of existing clinical training models yet aspired to a bright future, arguing that some models should center scientific training in clinical psychology and that this would improve the field, public health, and the lives of those who suffer from mental illness. Indeed, each of these scholarly efforts contributed to a movement toward “clinical science” programs versus more traditional “clinical psychology” programs. We recapitulate some of these arguments here, updating them with advances, or lack thereof, in clinical psychology over the past decade. At the same time, our proposal is new and more concrete. It goes further in the explicit argument for more intensive training in research and acknowledges more directly that if clinical psychology as a science is to more rapidly advance, an increased emphasis on research training must come at the expense of applied skills training, especially in the context of a fixed term (5 or 6 years) program.

We do not claim that all clinical psychology training programs must or would even do well to emphasize research training at the expense of applied clinical training, but simply that the norms of our field and accrediting bodies should allow programs transparently to do so. Some students in research-intensive training programs may need extensive applied clinical skills training to inform their research activities, others may need less. For example, students conducting epidemiological research on the link between maladaptive personality and health outcomes may benefit less from training in the administration of neuropsychological tests or provision of psychotherapy, freeing their time for advanced training in their research domain. On the other hand, students conducting research with client populations, including the development of neuropsychological tests, or clinical intervention trials, may benefit greatly from increased client contact, acquiring substantial hands-on skills in test administration or intervention. In a program such as we envision, the student would make such training decisions under the supervision of their mentors and the program’s training faculty, within some basic guardrails that require sufficient exposure to standard assessment and intervention curriculum and client contact. Flexibility of training, within expanded yet reasonable boundaries, is a key goal.

Below, we a) argue that clinical psychologists with strong scientific training are needed now more than ever for future growth of applied knowledge and practice, b) briefly describe current training models in clinical psychology, c) outline principles to guide the development of research-oriented training models and d) explore the practical impact of such principles.

The need for clinical psychologists who are scientists

Opportunities: Improved scientific training to advance our applied science

As a field we have made many advances since the first clinical psychology programs were accredited in 1948. To name a few, we contributed to the development of psychopathology and personality assessment and diagnosis; created and disseminated a variety of moderately efficacious and efficient psychotherapies, discarding many with dubious evidentiary bases; begun to characterize the developmental course and general etiological sources of many pathological behaviors; learned a great deal about the psychological functions of neural aberrancies associated with mental illness; and described the continuum of individual differences in psychopathology. At the same time, our science remains quite limited, something that has been discussed many times previously (Baker et al., 2008; Berenbaum et al., 2021; Dodge et al., 2024; Yee et al., 2025), but major research opportunities are reiterated here to make the argument that scientific training is needed now more than ever.

Scientific opportunity 1.

Formal diagnostic systems remain young (Kotov et al., 2020; Krueger et al., 2018). Few putative disorders are categorical or etiologically distinct (Lee et al., 2019; Smoller et al., 2019; Waldman, Poore, Luningham, & Yang, 2020), in most cases mapping inadequately on to reliable neural markers (First et al., 2018) and specific treatments (Cuijpers, Reijnders, & Huibers, 2019). At the same time, promising alternative conceptualizations are based largely on complex partial regressions (Cramer, Waldorp, van der Maas, & Borsboom, 2010) or factor analysis of covariance matrices of comorbid disorders/symptoms (Kotov et al., 2017) from observational studies of unrelated individuals. Efforts to understand the structure of maladaptive behavior will benefit from improved study design and associations with external criteria that will deepen our knowledge of more specific etiological mechanisms (Vaidyanathan, Vrieze, & Iacono, 2015; Wittchen & Beesdo-Baum, 2018). These developments, in turn, will drive innovation in intervention, because a better understanding of the empirical nature of mental disorder can frame more integrated intervention efforts (Ruggero et al., 2019).

Scientific opportunity 2.

Many psychosocial therapies work somewhat better than placebo controls (Leichsenring, Steinert, Rabung, & Ioannidis, 2022), especially exposure-based therapies for fear conditions, but available evidence indicates that effective delivery of such therapies does not require or even benefit from a doctorate (Berman & Norton, 1985; Christensen & Jacobson, 1994; Durlak, 1979; Mason, Grey, & Veale, 2016; Montgomery, Kunik, Wilson, Stanley, & Weiss, 2010; Tracey, Wampold, Lichtenberg, & Goodyear, 2014). While the interventions have modest effects, it is unclear why they work, or whether one works better than another depending on patient or provider characteristics (Cuijpers et al., 2019; Cuijpers et al., 2013; Kazdin, 2007, 2008, 2009). For example, the matching of treatments to client characteristics rarely has strong scientific bases (Barlow et al., 2017; Cuijpers et al., 2013; Huey, Tilley, Jones, & Smith, 2014; Marcus, O’Connell, Norris, & Sawaqdeh, 2014). Even if today’s gold-standard interventions were highly effective, they cannot be implemented at the scale needed to meet mental health needs of the population in most regions of the world, including the United States. Under the current delivery paradigm, there simply are not and will not be enough professionals to manage care for or conduct psychotherapy with all who would obviously benefit (Kazdin & Blase, 2011).

Scientific opportunity 3.

The interplay of genetic and environmental etiology remains entangled, and common research designs in clinical psychology cannot disentangle them. Psychopathology is moderately heritable and highly polygenic, such that individual genetic variants rarely have large effects. When they do, the genetic variant itself is rare or physically large in the DNA (e.g., large deletions), imprecise in mechanism of action and with variable expressivity (e.g., Glassford et al., 2016). Genetic etiology is therefore typically general rather than specific – characterized by some percentage of heritability or genome-wide risk score predicting a small proportion of variance in maladaptive behavior - with limited clinical utility at present (Martin et al., 2019). Despite terms like “social determinants,” our knowledge of environmental causes is similarly limited. While some environmental influences are obvious, such as extensive exposure to heavy metals, certain brain lesions, or severe neglect, these represent extremes of human experience and account for a limited fraction of behavioral variation in the larger population. Common sorts of environmental experiences, including most adverse life events, peer influences, and parenting styles, are weakly associated with later health and functioning (e.g., correlations <.2; Flaherty et al., 2013; Reuben et al., 2016). To greatly complicate things, such experiences are themselves heritable (Jaffee & Price, 2007; Kendler & Baker, 2007; Scarr & McCartney, 1983; von Stumm & d’Apice, 2021), thus limiting conclusions from most observational studies about environmental influences per se. Experimental designs can improve causal inference but practical and ethical limitations quickly limit external validity. As a result, strong theories that make causal predictions from environmental exposures to clinically-relevant behavior are few and far between (Eronen & Bringmann, 2021; Meehl, 1970, 1978; Rutter, 2007), limiting translational potential.

Scientific opportunity 4.

Even ignoring the high bar of etiological explanation, prediction of clinical phenomena is challenging, such that clinical psychologists routinely are unequipped to make even moderately precise predictions for complex clinical outcomes such as response to treatment, relapse, or recidivism (Grove, Zald, Lebow, Snitz, & Nelson, 2000; Salganik et al., 2020). What might seem like a simple clinical question can have few good answers anchored in solid data. Client A has presenting complaint X and clinical history Y in their chart. What assessments Z should be administered, and how should they be integrated, to measure the most relevant constructs in the most efficient manner to make important care decisions? What prior probabilities and measurement errors related to X, Y, and Z are relevant to any given clinical conclusion (Meehl & Rosen, 1955)? This involves issues of incremental test validity, which is often unknown for any two measures as they relate to important clinical outcomes, let alone for entire batteries of tests and semi-structured interviews that typify psychological assessment (Hunsley, 2003). We can be confident that any integration will be done “clinically” in the head of the clinician (Vrieze & Grove, 2009) despite the limitations of the human mind for such tasks (Kahneman, 2011; Meehl, 1954). This further complicates the incremental validity question as one of the human judge (Lima et al., 2005) rather than of statistically combined empirical relationships, despite the known superiority of the latter (Grove et al., 2000; Meehl, 1954).

Each of the examples above impinges on the conduct of clinical psychology, and the science related to these issues has limited explanatory and predictive power. To advance, we need clinical psychologists who will lead efforts to generate knowledge, including how to apply scientific findings to clinical problems, thus improving our ability to assess and treat individuals suffering from maladaptive behavior and psychological distress (Baker et al., 2008). And we need programs capable of training these individuals.

Current training models

Training models in clinical psychology are defined in part by how they balance research training versus the acquisition of applied skills such as assessment and intervention. In the past, there were three models representing different balances, and these were explicitly acknowledged by the major accrediting body of clinical psychology, the Commission on Accreditation (CoA) of the American Psychological Association (APA). In order from least to most research training, these were scholar-practitioner models (little to no research), scientist-practitioner, and clinical science (lots of original research). Regardless of research emphasis, CoA then and now requires all programs to meet basic accreditation standards to produce minimally competent clinical psychologists, to ensure not only that the education students receive is consistent with a set of professional standards, but also of course to protect the recipients of clinical services.

The CoA standards are explicitly generalist. They state (Commission on Accreditation, 2015b) that accredited programs are “doctoral programs in psychology that provide broad and general training” (p. 2) and doctoral education should be a “broad and general preparation for practice” (p. 3, italics ours). Requirements include assurance of training in “discipline-specific knowledge” (p. 4) encompassing 10 categories within four domains, including content areas in “scientific psychology” including affective, biological, cognitive, developmental, and social aspects of behavior (Commission on Accreditation, 2015a), essentially the same domains described in the very first APA accreditation standards in 1947 (Hilgard et al., 1947). To ensure the generalist flavor, training experiences that satisfy these domains cannot be focused on issues directly related to clinical psychology. For example, training in neuropsychology or pharmacology cannot alone fulfill breadth requirements in biological aspects of behavior, because they are too related to clinical psychology. In addition to these 10 discipline-specific knowledge requirements, CoA also requires competence in nine profession-wide competencies. All told, across all 19 domains of training, only one domain is referred to as “research”. The remaining 18 can be achieved with coursework and direct clinical care, with no research training required. Even the research domain requirements can be met without conducting original research beyond a simple literature review. Someone can become a clinical psychologist without engaging in the demanding original research efforts that typify impactful scientific contributions. Such individuals not only are not trained to conduct such research but consequently may have difficulty critiquing such research, thereby impairing efforts to translate scientific knowledge to clinical practice.

CoA-accredited programs must be able to demonstrate that each student is minimally competent in each of the 19 domains. Yet, cutting-edge student research programs are specialized, and by their nature will very rarely if ever satisfy all or even many domain-specific knowledge or profession-wide competency requirements. Therefore, to ensure that every student achieves the minimal competence standard, clinical training programs force relevant content into highly standardized experiences such as required courses and practica, thus ensuring that every single student becomes adequately and demonstrably (minimally) trained. The practical consequences include large, required course loads for all clinical students that are much larger than for non-clinical psychology students, with limited flexibility in course choice. There simply is not room to take advanced courses in statistics, psychometrics, clinical trials, neuroanatomy, or genetics. Students who are eager to create new psychological tests, understand brain-behavior relationships, implement novel intervention trials, or understand the biological basis of disorders are limited relative to other students from other disciplines who will have no formal training in clinical psychology or client experience in a health care system.

In addition to the broad coursework (affective aspects of behavior, ethics, etc.), clinical students spend about half their graduate training learning applied clinical skills. This training is embedded in core coursework and practicum/internship through direct client contact in clinical settings. The median clinical psychology student, according to data from the Association of Psychology Postdoctoral and Internship Centers (APPIC), spends ~800 hours of “face-to-face” activity with clients during practica (Keilen, 2018). For every one of these hours, we estimate a minimum of one more hour (likely two) is spent on a combination of supervision, report writing, case conferences, and other support hours, resulting in a median of 1600–2400 hours of applied clinical training related to client care. (APPIC does not report surveys of face-to-face hours versus support hours.) Students then complete an additional independent predoctoral year-long internship devoted to direct client care. All told, research-oriented students typically spend around 1500 (core coursework) + 2000 (practicum) + 2000 (internship) = 5500 hours, or nearly 3 years acquiring skills in hands-on client care. All the while, they are also engaging in scientific activities, such as statistical training and the mastery of research design and execution, necessary to complete the doctorate.

All else being equal, the typical level of applied training under current standards would seem to be acceptable and useful for individuals training to engage in direct patient care, which may include Clinical Psychologists trained under present-day models as well as what we term “Direct Service Practitioners”; more on this later. Of course, all else is never equal, and in the context of a five- or six-year training program acquisition of applied clinical skills comes at the expense of research and other domain expertise. It is a zero-sum game.

Training hours are only one piece of the problem. Curricular offerings are also more limited for clinical psychology students. For non-clinical psychology PhD students, the curriculum is highly flexible and tailored to each student’s research and scholarly interests. In clinical, they are largely the same for every student and every APA-accredited clinical psychology program, unchanged over decades: Introductory applied statistics, psychopathology, ethics, and two semesters each of assessment and intervention. Other content including culture, diversity, history, systems, and supervision are infused into these existing courses, or have courses of their own. Where can clinical students find the time to learn about advances in genetics, multivariate statistics, scientific computing, experimental psychopathology, clinical trials, or cognitive/affective neuroscience, or to extend and deepen their expertise in multi-culturalism and diversity, all while maintaining some semblance of a life outside of training? These additional domains and related skills are central to the conduct of cutting-edge research. It is unrealistic to expect that the typical trainee will have either the time or the motivation to achieve excellence in both applied clinical and research domains.

The need to encourage research specialization

The golden age of clinical psychology has been said to follow World War II, when it grew quickly to respond to mental health needs after the war. Applied clinical training programs were funded by the Veteran’s Administration and National Institutes of Health, with the express purpose of training practitioners. APA’s CoA was created to regulate this growing field, and the influential “scientist-practitioner” Boulder model was created in the late 1940s. While widely endorsed, cracks showed immediately and have grown in the following 80 years. On the one hand, questions arose about the need for research training at all, especially in the training of “front-line” practitioners to meet the growing mental health needs of the populace. Indeed, there is now the PsyD degree specifically for professional clinical psychology training that requires no original research at all, and the current CoA standards accommodate this model. On the other hand, research-intensive training programs bristled at restrictions on training models and academic freedom. This intensified to the point that new scientific associations were formed in the 80s and 90s to emphasize science, and subsequently the Psychological Clinical Science Accreditation System (PCSAS) was recently formed to accredit scientifically oriented clinical psychology training programs (for a comprehensive historical review see Callahan and Braggs (2022)).

While post-World War II may have been one golden age, right now is surely the most exciting time to study human behavior. Our theories are limited but have advanced, and we discarded some failed or unworkable ones. Technologies available now were unfathomable to prior generations and are getting better. Brains are scanned. Genomes sequenced. Wireless sensors including smartphones can monitor behavior in real time, offering just-in-time behavioral interventions. Analyses of these high dimensional data can be conducted within high performance computing environments – computer systems and programming languages that are foreign to the typical clinical student. On top of these technological advances, new standards are being adopted to engage in open science practices. Studies are ideally preregistered and data made widely available to enable replication and help ensure growth of knowledge. Each of these research practices requires specialized knowledge and training. If clinical psychology cannot train students to do this work, other disciplines will, to the detriment of our science, our intervention efforts, and other public health applications.

We need to ensure the next generation of clinical psychological scientists knows how to combine theory, measurement, study design, and analytics with responsible inference and transparent reporting to get traction on longstanding questions in clinical psychology. Under a fixed budget of student time and effort, this goal is in tension with the generalist and applied training paradigm espoused by CoA in 1947 (Hilgard et al., 1947) and continuing to the present day. It is unclear whether it is consistent with PCSAS standards, where highly abstract standards focus largely on career outcomes. We argue that the generalist requirement must be abandoned, and interested programs should be allowed, encouraged even, to expand time and opportunities for research training, even if at the expense of applied training. Of course, some programs will emphasize applied training at the expense of research; those programs already exist and are accredited.

How have accreditation standards kept up with the rapid growth of psychology and increasing need for specialization? A stunning fact is that clinical training looks little different than it did decades ago, when McFall wrote his Manifesto (McFall, 1991), or even nearly 60 years ago when the APA’s CoA was established. The structure is one-size-fits-all with little room for electives to complement a research program, lots of applied training with direct client contact, and increasing paperwork and documentation of “minimal competence”, a standard that can be set so low that “competence” could justifiably be replaced with “exposure”.

Pushing the boundaries of clinical psychology training

The major constraints in clinical psychology training are set out by accreditation standards, but those may be changing. The CoA can evolve over time and a new accreditation standard, the PCSAS, may represent a different or even competing model and set of principles. The distinction between these two systems remains somewhat murky, as all but one PCSAS accredited program is simultaneously currently accredited by the CoA, thus meeting both sets of partially overlapping standards. This is changing as we write, as the clinical psychology program at the University of California Berkeley has sent their first cohort of clinical psychology PhD students to internship under a PCSAS-only accredited program (congratulations to them, they all matched; personal communication, Keanan Joyner, April 16 2025); those students will not graduate from a CoA-accredited program. Momentum is growing and several other prominent programs have made public statements that they will follow suit (Stony Brook, Washington University, UCLA).

Regardless of the accreditation system, programs have some latitude in how they run their training program, and accreditation systems can and hopefully will change to reflect new realities in science and practice, including those described herein. We espouse five principles to inform the nature of training in our discipline and inform accreditation standards, whether from CoA, PCSAS, or another yet to be founded, and whether applied to clinical science, scientist-practitioner, or adjacent training models.

1. Science is foundational to Clinical Psychology.

Clinical psychology is an applied science. Clinical psychology training benefits from freedom and flexibility to develop curricula that optimally trains students to conduct independent and impactful clinical research.

2. Freedom during training to specialize in research is crucial to the future of clinical psychology as a science.

To produce budding independent investigators and science-oriented practitioners, clinical training programs must emphasize intensive and extensive conduct of rigorous original research and deliver training directly relevant to this goal. Given the specialized nature of much current research, trainees must be allowed flexibility to devote considerable time and effort to their chosen area of research. The broad generalist model in place since the 1940s must be replaced. Time must be freed to allow laboratory work and the formation of a supporting program of elective coursework, which will and should be different across clinical students within and across training programs. Trainees should have the latitude to have secondary specializations in measurement, statistics, genetics, or neuroscience, to name a few.

3. Applied training is still critical for science-oriented programs, but less is necessary to accommodate increased training in research.

Clinical psychology training must include applied training and direct client contact. These are essential to understand the nature of health care delivery on the front lines and to provide first-hand experience with clinical phenomena including symptoms and signs, impairment and suffering, how to help, and the care systems that coordinate that help. Applied experiences provide perspective regarding the clinical presentation of psychopathology and form the basis for communication and understanding between the clinical researcher and clinical practitioner, as well as the practical limits of translation from bench to bedside. At the same time, students must be able to limit the extent of applied experience to avoid interference with principles 1 and 2 above. Concrete suggestions follow in the next section.

4. Require proficiency in some domains of choice and tolerate limited or even no formal exposure to others.

“Minimal competence”, the standard in APA CoA accreditation, can be a recipe for mediocrity but is also a recognition that clinical psychologists simply cannot become even modestly competent, let alone proficient, in all domains within clinical psychology. There is not the time nor resources to do so. In training, we must recognize this fact and tolerate limited exposure to some domains, freeing time for mastery of others. This is already happening in scholar-practitioner-oriented programs. Such programs acknowledge that they will not provide more than the barest scientific training. In parallel fashion, students in other programs will benefit from reductions in applied clinical training, such that they can focus on research.

Flexibility is key. We are not arguing that all or even most programs must espouse the principles above. While we believe these principles are of paramount importance, we recognize that others will disagree. Therefore, we merely contend that programs should adopt these principles, if they wish, and be supported in so doing by accrediting bodies.

Practical implications for graduate training programs

These principles have practical implications that could be acted upon in the near term. Required coursework could be substantially reduced, with much less emphasis on practical aspects of assessment and intervention, and a more limited set of generalist requirements. Direct client contact during graduate training could be cut even further, mostly through shifting the year-long predoctoral internship to one that is postdoctoral (Berenbaum et al., 2021). Students would receive substantial applied exposure during graduate school to inform clinical questions and understand the provision of psychological services, but minimal competence may not be achieved in all applied domains until internship or beyond.

To illustrate this proposed model, we show the typical balance of coursework, research, and client contact over five years for the average Minnesota student in Supplementary Tables 1 and 2, which compares the existing Minnesota curriculum with an example proposed curriculum (all under the assumption of a 40-hour work week). Currently, the first year is dominated by coursework, most of which is applied, with little time for research activities. As the years go on, time allocated to research increases until a maximum of about 65% research activity is reached in Year 4, after which it plummets to ~0 during internship. Our proposed model would conceivably cut required coursework but increase elective coursework to accommodate the need for student research specialization. Applied clinical experiences would be reduced as well, resulting in an increase of 1300 – 3000 hours for research activities over the course of a five or six year program.

A helpful reviewer noted that a complication regarding practicum hours is that numbers of hours are not dictated by accreditors but rather reflect norms and implicit standards of the field. Despite evidence to the contrary (Callahan et al., 2025), students clamor for many hours in part because they think it will make them more competitive for internship. Some internships – including those that advertise themselves as research or science-oriented – set minimum hour requirements for applicants. Minimums might be 800 face-to-face hours, 1000, or even more. It would be inadvisable, in our opinion, for accreditation standards to enforce minimum or maximum hours requirements. Short of cultural change within the field, tools remain. The number of hours to prepare for internship is of course quasi-arbitrary. Surely it must be higher than a couple hundred. Surely at some point there are diminishing returns. Accreditation standards may provide rules of thumb, or ranges, to serve as soft anchors for accredited programs. If that is a bridge too far, standards could communicate greater license for programs to test the waters. For example, site visitors tend to exert greater scrutiny on programs when graduation or match rates are not perfect. The lower the rate, the more scrutiny is applied, as low rates are taken as evidence (perhaps correctly) that programs are not meeting their commitments to students. Reducing such scrutiny for programs that experiment with constraints on hours in a fashion transparent to admitted students, may free training programs from unhelpful extra scrutiny.

Under this proposed model, the predoctoral internship is removed and becomes postdoctoral, resulting in five years of graduate school followed by one year of internship. Some students may use this last year of graduate training for research; whereas others may allocate extra time during graduate school to client hours, or take additional applied coursework to pursue a specialization in applied clinical skills, such as neuropsychology or psychotherapy. The proposal here does not prohibit any activities. It does, however, create room and flexibility for the research specialization necessary to work at the leading edge of clinical psychology research.

Aside from the major practical issues that would need to be worked out, there are limited training-based reasons to require predoctoral internships, a point that is perhaps obvious when one considers that internships (followed by residencies) in medicine are postdoctoral. A postdoctoral internship is not a new suggestion (Berenbaum et al., 2021). There was a major movement in the mid to late ‘90s to make the change, supported by dominant organizations in clinical psychology, but the APPIC organization that manages internship placement declined to act (Boggs & Douce, 2000; Robiner & Grove, 2001). Even if postdoctoral internships were adopted, they undoubtedly would remain a normative aspect of the clinical psychologist’s training. An internship experience would almost certainly still be required for licensure; that, and the nature of the experience would, of course, be up to the state rules that govern such activities. An internship may even be required in job ads for academic faculty or research scientist positions as determined by the hiring organization.

Our proposal would not obviate research-oriented postdoctoral placements, which have become ensconced in the biomedical sciences including clinical psychological science, although it would make them less common. Under our proposed model, students who pursue them will be more prepared, more productive, and better positioned to leave postdoctoral training sooner. Ultimately, these junior scientists will be more prepared for the research job market and poised to make new and important discoveries in the etiology, assessment, and treatment of clinical phenomena.

Combined, these changes would give graduate training programs the flexibility to create a curriculum tailored to a given clinical students’ research area and professional interests, while still teaching them the basics of clinical assessment, intervention, and ethics. This would leave open the possibility of post-graduate training in clinical practice (e.g., internship, clinical postdocs) in support of desired careers that include some level of client care. This proposed framework would require graduate programs to define clear training emphases (more practice; less practice, etc.), advertise those emphases to incoming students, and proceed accordingly in a way that permits students to choose the type of career path that best suits their interests. The consequences of adopting programs like that which we promote here are uncertain. While the clinical psychology workforce as a whole would have all relevant expertise, under our proposal individuals within that workforce would be more specialized, and in those individuals may be sundered the research/clinical connection that clinical psychology holds dear. How this might look we turn to next.

Adapting training to a shifting landscape of clinical psychology care

Clinical psychologists who primarily work directly with clients spend a great deal of their professional effort engaging with those clients one-on-one. Whether in psychotherapy or assessment, the psychologist routinely administers tests, conducts interviews, and provides psychotherapy. Given the enormous demand for mental health services, it is imperative that we consider whether these tasks must forever be uniformly delivered by doctoral-level psychologists. The PhD is a research degree. Does one require research training to competently provide mental health services?

The administration of objective psychological tests is described in detail by test manuals and valid administrations are routinely made by non-doctoral-level personnel. Case in point, neuropsychologists typically do not administer test batteries, technicians do. The neuropsychologist interprets results, renders diagnoses, prognoses, and proposes interventions. Clinical interviews are similar. Structured interviews resemble objective tests, with prescribed questions and coding documentation that do not require doctorate degrees. In fact, research studies that validate such interviews rarely employ psychologists with doctorates; it is cost-prohibitive to do so, and the same is true in the clinic.

Therapy has historically required extensive formal training (e.g., psychodynamic interventions) but modern psychosocial interventions have become more like objective tests – they require skill and experience to do well but are often standardized/manualized and can be administered by other professionals. Indeed, some novel interventions consist of algorithms embedded in software applications, with early hints of efficacy (Torous et al., 2021) and considerable hype due to large language models. Conclusive research on the comparative effectiveness of these different approaches remains to be done.

To increase efficiency and to serve more people, one might consider a simple model such as that illustrated in Figure 1, which is roughly already being implemented in a number of health care organizations. Here, a significant proportion of direct client assessment and care is accomplished not by clinical psychologists, but by what we refer to as Direct Service Practitioners (DSPs; in blue in the Figure). These are individuals with primarily skills-based training who may have Master’s, Bachelor’s, or other terminal degrees, potentially along with other focused training and relevant credentialing. They are qualified to administer tests, provide therapy, and work with other providers to ensure resources are in place for the clients’ success. DSPs would work individually or in teams under the supervision of doctoral-level Clinical Psychologists who focus on practice (orange in the Figure). These clinical psychologists work with the client and the DSP to determine the course of care. The psychologist would integrate test results to arrive at care decisions including provision of therapy, which could be carried out by the Direct Service Practitioners. Complex or refractory cases may be handled directly by the Psychologist, but otherwise care would be administered by DSPs.

Figure 1. An aspirational model of the profession of clinical psychology.

Professionals are grouped into three buckets, two of which will typically be doctoral level clinical psychologists. The buckets are not tiered or hierarchical, but rather represent distinct areas of emphasis, including the care of clients, the management of that care, and the generation of knowledge related to that care. Both buckets of clinical psychologists are eligible for licensure. However, if internships became postdoctoral, then we assume only those psychologists who complete a postdoctoral clinical internship (which we envision would be analogous to current predoctoral internships) would be license eligible.

The Clinical Psychologist who focuses on practice would receive training in clinical practice and research in graduate school and possibly beyond, such that they are able to critically evaluate, interpret, and apply scientific knowledge to the care of any given client, the development and management of their team, clinics, health care systems, and policy. This training may be indistinguishable from current accredited training models, such as scientist-practitioner or clinical science models. These Clinical Psychologists may formulate clinic operations and collect and analyze data to evaluate clinic outcomes, making changes consistent with local conditions and the scientific literature.

In a third group are Clinical Psychologists who focus more on science (in green in the Figure) who would be responsible for producing basic and applied scientific knowledge in the academic, industry, and clinical setting, more general policy across health care institutions, and working with other Clinical Psychologists to develop and administer systems of care. These individuals would receive applied clinical training and extensive specialized training in the conduct of research, both during the graduate training years and beyond, such that they would be able to engage in original independent programs of research and expertly evaluate existing research findings and their application. They would be eligible for licensure and may see clients, but their training would be more focused on the critical understanding of scientific issues in clinical psychology. It is important to emphasize that these two groups of clinical psychologists are not mutually exclusive nor are they hierarchically pitted against one another. Each would receive training focused on different aspects of clinical psychology to make it easier for them to specialize in one versus the other. Ideally, they would be able to work side-by-side, each contributing to the success of the other. Some individuals may even move seamlessly between these two groups, contributing to one or the other during their careers.

Summary

What makes psychology distinct from other mental health professions is not the provision of direct client care and healing in face-to-face encounters with clients, but our ability to propagate the empirical and theoretical bases of our field to understand the nature of psychopathology, its etiology, its assessment, and how to prevent or intervene accordingly.

Current training models make it difficult to obtain the training necessary to contribute to the generation of impactful scientific knowledge and to avoid stagnation. We encourage accreditation systems and clinical psychology training programs to consider more heavily focusing on advancing knowledge and ameliorating psychological suffering through rigorous applied science. Here we present one model – rooted in the principles described above – and strongly influenced by the research programs of the present authors and their students. Other programs will develop other models based on their research, clinical, and training traditions and aspirations, but only if accreditation systems permit greater flexibility than in the past.

Public significance:

Clinical psychology is an applied science yet science training during graduate school is constrained under present models. This manuscript proposes a set of principles and a training structure that would improve science training.

Biographies

Scott Vrieze received his PhD in clinical psychology from the University of Minnesota, interned at the Minneapolis VA Medical Center, and completed postdoctoral training in biostatistics at the University of Michigan. He is a clinical psychologist and behavioral geneticist working to understand the nature, origins, and consequences of addiction and psychopathology. To disentangle the influences of inheritance and human experience he discovers and characterizes genes that affect risk for these behaviors, and closely follows multi-generational families with twins and adoptees. He is currently Professor of Psychology and Director of Clinical Training of the APA- and PCSAS-accredited Clinical Science and Psychopathology Research program at the University of Minnesota.

Maxwell Elliott received his PhD in clinical psychology from Duke University, interned at McLean Hospital, and completed postdoctoral training in psychology at Harvard University. He is a clinical scientist and neuroscientist working to understand individual differences in cognitive and brain aging, specifically characterizing why some individuals remain resilient to age-related decline while others are vulnerable to neurodegenerative disease. To address these questions, Dr. Elliott utilizes longitudinal methods that combine high-precision neuroimaging techniques (e.g., cluster scanning), novel biological aging clocks (e.g., DunedinPACNI), and deep phenotyping of cognitive aging and psychosocial exposures. He is currently an Assistant Professor in the Clinical Science and Psychopathology Research program within the Department of Psychology at the University of Minnesota.

William Iacono is Regents Professor and Distinguished McKnight University Professor Emeritus of Psychology, Psychiatry, Neuroscience, and Law at the University of Minnesota. Trained as a clinical psychologist, psychophysiologist, and behavior geneticist, he has employed family, adoption, and twin study designs to investigate the neurobehavioral development and etiology of schizophrenia and common mental disorders, including substance use, antisocial, and major depressive disorders. As a forensic psychologist, his research has also focused on the validity of applied lie detection techniques.

Bonnie Klimes Dougan received her PhD in clinical psychology from Florida State University, interned at Duke Medical Center, and completed postdoctoral training in developmental psychopathology at the National Institute of Mental Health. She is a scholar and clinician whose body of work seeks to enhance understanding and alleviate distress, depression, and despair (including self-harm and suicide) in adolescents. Using a biopsychosocial perspective, she focuses on elucidating relevant neural and physiological mechanisms of threat as they relate to psychopathology, and more recently has focused on marginalized populations at highest risk for suicide (Native American and LGBTQ+ youth). She is currently Professor and Associate Director of Practicum and Internship Placement of the APA- and PCSAS-accredited Clinical Science and Psychopathology Research program at the University of Minnesota.

Robert Krueger received his PhD in clinical psychology from the University of Wisconsin-Madison, and interned at Brown. His research interests center around psychiatric classification, personality and personality disorders, health, aging, and genetics. He currently serves as Editor for the Journal of Personality Disorders. He is a McKnight Distinguished University Professor and former Director of Clinical Training of the APA- and PCSAS-accredited Clinical Science and Psychopathology Research program at the University of Minnesota.

Shmuel Lissek is Associate Professor in the Department of Psychology at the University of Minnesota and the past director of Practicum and Internship Placement in the Clinical Science and Psychopathology Research PhD program at the University of Minnesota. He received his PhD in clinical psychology from St. John’s University, interned at Interfaith Medical Center, and completed post doctoral training in affective neuroscience at the NIMH. His research applies psychophysiology, neuroimaging, and fear conditioning methodologies to elucidate the brain basis and clinical relevance of fear-generalization and anxiety-related decision making.

Monica Luciana received her PhD in clinical psychology from the University of Minnesota, interned at the Minneapolis VA Medical Center, and completed postdoctoral training in child development at the University of Minnesota. She is a clinical scientist and developmental neuroscientist working to describe patterns of neurodevelopment in the adolescent period, associations between neurodevelopmental mechanisms and behavior, and how substance use impacts neurodevelopmental trajectories. To address these questions, Dr. Luciana utilizes longitudinal methods that include neuroimaging, assessments of cognition with a particular emphasis on executive functions, and assessments of psychopathology. She is currently Distinguished McKnight University Professor of Psychology within the Clinical Science and Psychopathology Research program at the University of Minnesota and as well as Director of Graduate Studies in the Department of Psychology.

Angus W. MacDonald, III received his PhD in clinical psychology from the University of Pittsburgh, where he also completed an internship and post-doctoral fellowship at Western Psychiatric Institute & Clinic. He is a clinical neuroscientist who employs experimental cognitive psychological perspectives and tools—including neuroimaging, behavioral genetics, and neuromodulation—to understand and ameliorate the symptoms of psychosis, addiction, and other forms of mental illness. He is currently the Distinguished McKnight University Professor at the University of Minnesota, serves on the National Advisory Mental Health Council and has served as the Director of Clinical Training and editor for the Journal of Psychopathology and Clinical Science.

Whitney Ringwald received her PhD in clinical psychology from the University of Pittsburgh and interned at Stony Brook University. She is a clinical psychologist whose research aims to understand behavioral mechanisms that maintain psychopathology in daily life. To capture the complex, contextual nature of these mechanisms, she uses ambulatory assessment methods integrating data from multiple modalities including self-reports, natural language, video, and smartphone sensors. She is currently Assistant Professor of Psychology of the APA- and PCSAS-accredited Clinical Science and Psychopathology Research Program at the University of Minnesota.

Wayne Siegel received his Ph.D. in counseling psychology from the University of Iowa and completed his psychology internship at the Veterans Affairs / Vanderbilt University Consortium in 1993. He began his career with the Minneapolis VA immediately after completing internship, and provided clinical services on the substance abuse treatment team. In 2001, he assumed the role of training director, overseeing training at the practicum, internship, and postdoctoral levels. Shortly thereafter, he was given the additional responsibilities of psychology section supervisor and manager of the Minneapolis VA Medical Center’s Academic Excellence Council, which is responsible for fostering the research and education aspects of VA’s mission. Dr. Siegel has also held various national service roles, including Associate Editor for the TEPP (Training and Education in Professional Psychology) journal, Chair of the VA Psychology Training Council, the APA’s Commission on Accreditation, APPIC’s Board of Directors, Chair of APPIC’s Postdoctoral Committee, the Board of Directors for the Academy of Psychological Clinical Science, and the Board of Directors for the American Board of Clinical Psychology. Dr. Siegel retired from the VA in 2024 and now maintains a private practice focused on evidence-based treatment and consulting in areas related to psychology education and training.

Irwin Waldman is a clinical psychologist and behavior geneticist who completed his undergraduate degree in Human Development and Family Studies at Cornell University, attended graduate school at the University of Waterloo earning a Ph.D. in Clinical Psychology, and completed a Postdoctoral Fellowship in Behavioral Genetics at the University of Minnesota. Dr. Waldman’s overarching interests are in developmental psychopathology and behavior genetics. He works at the intersection of psychology, statistics, genetics, and biology to understand the classification, causes, development, and biological underpinnings of psychopathology and relevant personality and neurocognitive traits across the lifespan. He was President of the Behavior Genetics Association from 2010–11 and has been a faculty member in the Psychology Department at Emory University since the fall of 1991.

Sylia Wilson received her BA from Macalester College and her MS and PhD in Clinical Psychology from Northwestern University. She is a developmental psychopathologist whose research focuses on understanding the origins of depression, substance use, and related forms of psychopathology across development. Her work integrates longitudinal, genetically informed designs, including family, twin, and adoption studies, with multimethod assessments of behavior, cognition, and brain structure and function to examine how early environmental and familial experiences shapes neurodevelopment trajectories. She is currently Associate Professor at the Institute of Child Development at the University of Minnesota.