Psychotherapeutic Moderators of Far Transfer in Working Memory Training: A Review of Empirical Evidence and Mechanisms

Psychotherapy emerges as a promising yet underexplored modulator in working memory (WM) interventions, particularly for enhancing far transfer effects to real-world cognitive functioning and psychopathology outcomes (Ansari & Kemperman, 2015; Stevenson et al., 2020). Scientific literature highlights its indirect role through mechanisms like motivation, therapeutic alliance, and neuroplasticity, though direct integrations remain rare (Schwaighofer et al., 2015). This article builds on these ideas using evidence from peer-reviewed journal articles, covering the evidence, mechanisms, psychotherapy types, and next steps in research.

Challenges in WM Transfer Effects

Far transfer from WM training, generalisation to dissimilar tasks like fluid intelligence or daily functioning, remains elusive, with meta-analyses reporting effect sizes of g = 0.14 to 0.24 post-training, often nullifying after delays (Melby-Lervåg et al., 2016; Schwaighofer et al., 2015). For instance, Melby-Lervåg et al. (2016) analysed 87 randomised trials (n > 8,000) and found reliable near transfer (g = 0.72 for verbal WM, g = 0.37 for visuospatial), but far transfer to reasoning or inhibition averaged g = 0.16, attributed to task-specific practice rather than capacity enhancement. Au et al. (2015) countered with g = 0.24 for fluid intelligence in adaptive training, yet critics note publication bias and inactive controls inflate effects. In clinical contexts, such as ADHD, Klingberg et al. (2005) showed near transfer in children via Cogmed training, but far transfer to academics or symptoms was inconsistent across meta-analyses like Cortese et al. (2015; n = 1,743; SMD = -0.10 for ADHD core symptoms). These nulls stem from motivational decay, poor engagement, and unaddressed psychological barriers, setting the stage for psychotherapy as a facilitator (Jaeggi et al., 2014).

Psychotherapy’s Therapeutic Potential in WM Interventions

Psychotherapy addresses WM deficits central to disorders like ADHD, depression, and schizophrenia, where impairments link to prefrontal dysfunction and predict functional outcomes (Ansari & Kemperman, 2015; Subramaniam et al., 2014). Ansari and Kemperman (2015) reviewed evidence that WM training induces prefrontal-parietal changes (Buschkuehl et al., 2012), mirroring psychotherapy’s neuroplastic effects, but clinical transfer requires motivational scaffolding. In schizophrenia, Subramaniam et al. (2014) reported sustained WM gains (6 months post-16-week training) alongside prefrontal efficiency improvements, suggesting supplementary psychotherapy could amplify this via adherence. For depression, WM biases toward negative rumination resist training alone (Wanmaker et al., 2015; no effects on rumination/anxiety post-6 days), as low intrinsic motivation, hallmarked by anhedonia, undermines effortful practice. Jaeggi et al. (2014) emphasised self-perceived deficits and motivation as transfer mediators, proposing psychotherapy to reframe these. Beck et al. (2010) integrated coach-supported training in ADHD youth, yielding symptom reductions (inattention decreased), akin to behavioural therapy elements. Overall, psychotherapy’s role lies in boosting compliance and expectancy, with preliminary evidence from hybrid protocols showing promise over isolated training (Ansari & Kemperman, 2015).

Specific Psychotherapy Types and Approaches

Direct studies on psychotherapy types in WM transfer are sparse, but proxies emerge from motivational and alliance-focused interventions. Motivational interviewing (MI), a client-centred approach emphasising discrepancy resolution and change talk, enhances engagement in cognitive training; Stevens et al. (2015) used MI-infused coaching in ADHD adolescents, normalising frontoparietal activation versus standard Cogmed. MI’s transtheoretical model boosts intrinsic motivation, critical as external incentives predict WM gains but fade without internalisation (Jaeggi et al., 2014). Cognitive-behavioural therapy (CBT) elements appear in strategy training; Chan et al. (2019) trained semantic categorisation and rehearsal, CBT-like metacognitive strategies, in 6 to 9-year-olds, yielding far transfer to novel problem-solving (all training groups outperformed controls; performance index p < 0.05), outperforming core-based methods. Semantic grouping (e.g., “things that fly”) mirrors CBT reframing for efficient encoding, with rehearsal akin to behavioural repetition. Supervision/therapeutic alliance, from psychodynamic roots but operationalised in training, moderates transfer: Schwaighofer et al. (2015) meta-analysed 35 studies, finding supervised training doubled far-transfer odds (g = 0.33 vs. unsupervised g = 0.12). No pure psychodynamic or humanistic trials exist, but alliance research (e.g., Horvath et al., 2011 meta-analysis) implies relational support enhances adherence, transferable to WM via expectancy effects. In depression, CBT-WM hybrids could counter interference (Joormann & Gotlib, 2008), though untested. Hubacher et al. (2013) piloted 4-week WM in schizophrenia with supportive counselling, improving verbal/visual memory, suggesting alliance-driven approaches.

Moderating Mechanisms and Neurocognitive Links

Psychotherapy moderates via psychological (motivation, expectancy) and neural pathways. Motivation mediates: external coaching (behavioural therapy proxy) predicted gains in ADHD (Stevenson et al., 2020), but intrinsic factors like self-efficacy, fostered by MI/CBT, drive transfer. Neuroplastically, psychotherapy upregulates BDNF and prefrontal connectivity (e.g., psychodynamic therapy alters default mode; see reviews), synergising WM-induced changes (Li et al., 2015 meta-analysis: schizophrenia brain activation shifts). Strategy training (Chan et al., 2019) exemplifies: rehearsal reduces decay (active maintenance per Camos & Barrouillet, 2011), semantic chunking leverages LTM networks (semantic > perceptual grouping; Schelble et al., 2012), both psychotherapy-teachable via metacognitive therapy. Supervised formats ensure alliance, buffering dropout (30 to 50% in unsupervised WM; Schwaighofer et al., 2015).

Gaps and Future Directions

Literature gaps include no RCTs directly randomising psychotherapy + WM versus WM-alone for far transfer; most evidence infers from moderators (motivation/supervision; Ansari & Kemperman, 2015; Stevenson et al., 2020). Clinical populations (e.g., depression) show least transfer due to amotivation, warranting trials. Future designs: integrate MI/CBT with adaptive WM (e.g., 5 weeks, n > 100/group), measure alliance (WAI scale), and far-transfer batteries (e.g., fluid intelligence composites). Longitudinal neuroimaging could validate synergies (Cramer et al., 2017).