When Brain Waves Meet Hospital Budgets: A Review of a Pioneering EEG Study on Decision-Making in Hospital Administrative and Financial Units
Dinler, Z. M. (2024). Hastanelerin idari ve mali birimlerinde gorevli personelin kararlarinin elektroensefelografi araciligi ile incelenmesi. Unpublished doctoral dissertation, Sakarya University.
Reviewed by Prof. Dr. M. Nurullah Kurutkan | Duzce University, Faculty of Business, Department of Health Management
April 2026
Introduction: Why This Thesis Matters
Can we peer inside a hospital employee’s brain while they decide whether a medical invoice is correct? Can delta and theta oscillations tell us something useful about how administrative staff navigate the fog of SUT codes, procurement forms, and budget reconciliations? Zeynep Merve Dinler’s doctoral dissertation, supervised by Prof. Dr. Mahmut Akbolat at Sakarya University, dares to ask exactly these questions. In doing so, it becomes what appears to be the first study in the health management literature to apply electroencephalography (EEG) to real-world administrative decision-making in hospitals.
This review examines the thesis through two lenses: its genuine contributions to an underexplored interdisciplinary space, and the methodological frontiers it opens for future investigators. The tone here is deliberately constructive. Interdisciplinary work that bridges health management and cognitive neuroscience deserves encouragement, and the inevitable rough edges of such pioneering efforts are best understood as signposts for the next generation of studies rather than failings of the current one.
What the Thesis Does Well
A Genuinely Novel Interdisciplinary Venture
The neuroscience of decision-making has been well explored in economics (Caroselli et al., 2006), finance (Neo et al., 2020; Toma & Miyakoshi, 2021), and marketing (Roselli et al., 2020). Yet health management, a field where daily decisions directly affect patient safety and institutional solvency, has remained conspicuously absent from this conversation. Dinler’s thesis fills this gap with admirable ambition. The researcher, coming from a health management background, invested approximately seven months in theoretical and practical EEG training at Istanbul Medipol University’s laboratory (February-September 2022) and conducted pilot recordings with ten volunteers before launching the main study. This level of cross-disciplinary commitment is noteworthy and sets a valuable precedent.
A Thoughtfully Designed Mixed-Methods Architecture
The three-phase research design (qualitative interviews, quantitative survey, experimental EEG) follows a coherent internal logic. What makes this design particularly compelling is the organic connection between phases: the qualitative stage identified the specific words and documents that hospital administrative and financial managers actually use in their daily decision-making. These were then woven into the experimental paradigm as stimuli. This is a meaningful departure from many neuroscience studies that rely on generic or abstract stimuli, and it enhances the ecological validity of the entire enterprise.
Twelve managers from six public and six private hospitals across four cities (Ankara, Sakarya, Kocaeli, Istanbul) were interviewed between December 2022 and February 2023. Trustworthiness criteria were systematically addressed: credibility through participant confirmation (with two participants, Y3 and Y7, identifying and correcting insufficiencies), dependability through triangulation with the thesis supervisor and an independent researcher, confirmability through audio recordings, and transferability through purposive sampling. This methodological rigor in the qualitative phase is commendable.
The Standout Finding: Knowing vs. Doing
Perhaps the most valuable empirical contribution of this thesis lies in the contrast between the lexicality paradigm and the application paradigm. When participants processed words related to hospital administrative terminology (the lexicality condition), no significant power spectrum responses emerged. However, when they were presented with actual document examples and asked to identify errors that could cause financial harm to the hospital (the application condition), a theta response appeared in the frontal region.
| Key finding: The distinction between knowing a concept and applying it in practice produces measurably different neurophysiological signatures. This has genuine implications for how we think about training design and competency assessment in healthcare administration. |
This finding resonates with Kahneman’s (2012) dual-process framework and suggests that conceptual knowledge alone does not guarantee effective application. The frontal theta response during the application phase likely reflects the additional cognitive control and working memory demands required when personnel must integrate knowledge with practical judgment. For health management educators, this observation could inform the design of more application-oriented training programs.
Rigorous Ethical and Clinical Screening
The ethical infrastructure of the study meets high standards. Ethical approval was obtained from Sakarya University (September 7, 2022), written informed consent was collected, and participants were screened using the Montreal Cognitive Assessment (MoCA) test, with a mean score of 26.51 (well above the 21-point threshold for cognitive normality in the Turkish adaptation). Inclusion and exclusion criteria were clearly defined: right-hand dominance, absence of neurological or psychiatric diagnosis, no psychotropic medication use, and current employment in hospital administrative or financial units. The preparation process, impedance management (below 10 kohm for surface electrodes, below 5 kohm for ground and reference), and use of BrainVision Recorder with active caps all reflect proper laboratory protocol.
Behavioral Data: Reaction Times Tell a Story
The reaction time analysis across three decision types yielded a highly significant result (F = 136.291, p < 0.001). Intuitive decisions were fastest (1.37 s), rational lexical decisions were intermediate (1.75 s), and rational application decisions were slowest (3.66 s). This gradient is entirely consistent with Kahneman’s prediction that System 2 processes require more time and cognitive effort than System 1 processes. The behavioral data, independent of the EEG findings, provides a solid empirical foundation.
Opportunities for Future Research
Every pioneering study opens as many questions as it answers. The following observations are intended not as criticisms but as a roadmap for the next wave of investigations in this promising interdisciplinary space.
Expanding the Paradigm Design
Adding a Neutral Control Condition
The current design includes three experimental conditions (intuitive lexical, rational lexical, rational application) but no neutral control condition. Future studies could include a simple perceptual task (e.g., identifying the color of a fixation cross) or domain-irrelevant words to establish a cognitive baseline. This would help disentangle decision-specific neural signatures from general task-related brain activation. Several EEG studies in adjacent fields have demonstrated that such control conditions substantially strengthen causal inferences (Bastiaansen et al., 2005; Krause et al., 2006).
Balancing Stimulus Complexity Across Conditions
The three stimulus categories differ substantially in cognitive load: single words for the lexicality conditions versus multi-line document excerpts for the application condition. The dramatic difference in reaction times (1.37 s vs. 3.66 s) may partially reflect stimulus complexity rather than decision-type differences per se. Future designs could equalize stimulus complexity, perhaps by presenting application scenarios as brief phrases rather than full documents, or by adding a word-level application task where participants judge individual terms in context.
Increasing Trial Numbers
Twenty stimuli per category is at the lower boundary for event-related oscillation analysis. The classical lexical decision literature typically recommends 40-80 or more trials per condition to achieve adequate signal-to-noise ratios (Bastiaansen et al., 2005; Krause et al., 2006). Future replications with larger stimulus sets would strengthen the reliability of the oscillatory findings at the individual level.
Strengthening the Analytical Framework
Multiple Comparison Corrections
The current analyses involve numerous statistical tests across 12 brain regions, multiple frequency bands, and multiple time windows, each evaluated at p = 0.05. Modern neuroscience publishing standards increasingly expect correction procedures such as Bonferroni, False Discovery Rate (FDR), or cluster-based permutation tests. Incorporating these corrections in future work would provide stronger confidence that observed effects are not attributable to chance. This is a well-recognized challenge across the EEG field (Larson & Carbine, 2017) and is not unique to this study.
Effect Size Reporting
While some analyses report partial eta-squared values, systematic reporting of effect sizes (Cohen’s d, partial eta-squared) across all comparisons would enhance interpretability. This is particularly important for power calculations in future studies and for meta-analytic integration. The field is moving toward mandatory effect size reporting, and early adoption in this research line would position it well.
Covariate Control in Gender Comparisons
An interesting feature of the sample is that 90.63% of female participants worked in billing units while 66.67% of male participants worked in procurement. This natural occupational distribution, combined with differences in mean age (32.84 vs. 28.21 years) and education level (87.50% vs. 25.00% holding bachelor’s degrees), creates a rich but analytically complex dataset. Future studies could employ ANCOVA or linear mixed models with unit, age, and education as covariates, which would allow researchers to more precisely attribute observed EEG differences to gender-related neurophysiological variation versus occupational experience effects. This distinction has important theoretical implications and could itself become a productive research question: does prolonged exposure to specific administrative tasks reshape neural decision-making patterns?
Broadening the Neurophysiological Window
Including the Alpha Band
The present study focuses on delta (1-3.5 Hz) and theta (4-7 Hz) oscillations, both of which are well-established correlates of attention, memory, and decision-making. However, the alpha band (8-12 Hz) plays a critical role in inhibitory control and semantic processing during decision-making (Klimesch, 1999, 2007). The thesis itself references Razumnikova and Yashanina (2018), who examined alpha responses in rational/irrational thinking. Including alpha (and potentially beta) analyses in future work would provide a more complete picture of the oscillatory dynamics underlying administrative decision-making.
Leveraging the Frontal Midline
The FZ electrode was excluded from the primary analyses because it lies on the midline. However, frontal midline theta (FMT) is one of the most robust and widely replicated neural correlates of cognitive control and conflict monitoring (Cavanagh & Frank, 2014; Nayak et al., 2019). Future studies that retain midline electrodes in the analytical framework could capture this important signal, particularly given the thesis’s central interest in rational decision-making processes.
Connecting Brain Activity to Performance Outcomes
Perhaps the most exciting avenue for future research is connecting EEG signatures to objective performance metrics. The current thesis demonstrates that decision-making produces measurable oscillatory patterns, but does not yet link those patterns to actual decision quality (e.g., error rates, processing accuracy, financial impact of errors). A follow-up study that correlates individual-level EEG profiles with independently measured job performance indicators would be groundbreaking. Such a study would answer the critical question: does the strength of the frontal theta response during the application paradigm predict how many errors a staff member actually makes?
Until this predictive link is established, it would be prudent to frame the current findings as foundational descriptive science rather than as a basis for applied personnel decisions. The thesis’s suggestion that EEG could inform personnel selection is an inspiring long-term vision, but realizing it would require demonstrating individual-level predictive validity, test-retest reliability, and incremental validity over existing assessment tools, all of which represent substantial future research programs in their own right.
Considering Alternative and Complementary Methodologies
Future research in this space could benefit from several complementary approaches. Functional near-infrared spectroscopy (fNIRS) offers portability advantages that would allow data collection in actual hospital settings rather than laboratory environments, directly addressing ecological validity concerns. Multimodal EEG-fNIRS designs could combine EEG’s millisecond-level temporal resolution with fNIRS’s hemodynamic spatial information. On the behavioral side, Situational Judgment Tests (SJTs) and work sample tests have decades of validation evidence in personnel assessment and could serve as important benchmarks against which neurophysiological measures are evaluated.
Conceptual Considerations
The Dual-Process Framework
The thesis builds on Kahneman’s (2012) System 1/System 2 theory and Hamilton et al.’s (2016) rational-intuitive decision-making scale. These two frameworks operate at different levels of analysis: Kahneman describes situational cognitive processes, while Hamilton measures dispositional decision-making styles. The thesis integrates both, which is conceptually ambitious. Future theoretical work could more explicitly articulate how these frameworks complement each other and whether the EEG paradigm measures situational process activation (consistent with Kahneman) or trait-like cognitive dispositions (consistent with Hamilton). This theoretical refinement would sharpen the interpretive power of the findings.
From the Lexical Decision Tradition to Domain-Specific Assessment
The paradigm modifies the classical lexical decision task (which asks “Is this a real word?”) into a domain-specific importance judgment (“Is this term important for hospital administration?”). This is a creative adaptation that enhances ecological relevance. However, it also means that the cognitive processes being measured differ from those in the psycholinguistic tradition. Future publications from this research line could benefit from explicitly positioning the paradigm as a “domain-specific semantic evaluation task” rather than a classical lexical decision task, which would prevent potential confusion when readers compare findings across literatures.
Closing Reflections
Zeynep Merve Dinler’s dissertation represents a genuinely pioneering effort to bring neuroscientific methods into the health management discipline. The study’s mixed-methods architecture is thoughtfully constructed, the qualitative-to-experimental pipeline is well-motivated, and the key finding, that knowing administrative concepts and applying them in practice produce distinguishable neurophysiological signatures, opens a door that future researchers can walk through with more refined tools.
The thesis’s greatest strength lies in its courage to cross disciplinary boundaries. Its greatest opportunity lies in the rich research program it implicitly defines: larger samples, controlled covariates, alpha-band inclusion, performance-linked EEG profiles, and field-based fNIRS applications all represent natural next steps. If this study is the first chapter of a longer story, it is an auspicious beginning.
| Bottom line: This thesis establishes that EEG-based investigation of administrative decision-making in healthcare settings is feasible, produces interpretable results, and reveals a meaningful distinction between conceptual knowledge and applied judgment. The descriptive findings are solid; the applied recommendations are aspirational. The gap between the two defines the research agenda for the next decade. |
Note on References
All references cited in this review (Bastiaansen et al., 2005; Caroselli et al., 2006; Cavanagh & Frank, 2014; Fischer et al., 2013; Hamilton et al., 2016; Kahneman, 2012; Klimesch, 1999, 2007; Krause et al., 2006; Larson & Carbine, 2017; Nayak et al., 2019; Neo et al., 2020; Razumnikova & Yashanina, 2018; Roselli et al., 2020; Toma & Miyakoshi, 2021) are fully cited in the thesis under review. The reader is directed to the thesis bibliography for complete citation details.