Assessment of Theory of Mind for People with Schizophrenia Questionnaire (AToMS)

1. Abstract

The Assessment of Theory of Mind for People with Schizophrenia (AToMS; Yeh et al., 2023) is a standardized neuropsychological instrument developed to evaluate Theory of Mind (ToM) impairments specifically in individuals diagnosed with schizophrenia. Rooted in contemporary neurodevelopmental and neurobiological paradigms (Abu-Akel & Shamay-Tsoory, 2011; Baron-Cohen, 1995; Hutchins et al., 2012), the instrument operationalizes mentalizing capacity across two foundational neurocognitive domains: Cognitive ToM (inferring the beliefs, intentions, and thoughts of others) and Affective ToM (inferring feelings, emotional states, and affective nuances). The AToMS comprises 16 items that systematically measure eight developmental ToM concepts stratified across basic and advanced complexity levels: first-order false belief, first-order deception, second-order false belief, appearance and reality, banter, hinting, social cues, and faux pas.

Each concept is examined using a paired-item structure consisting of control items and target Theory of Mind assessment items. Items are scored dichotomously (0 = incorrect, 1 = correct), yielding subscale scores ranging from 0 to 8 for both the Cognitive and Affective dimensions, with total scale scores ranging from 0 to 16. Psychometric evaluation in clinical cohorts demonstrates good internal consistency (Cronbach’s alpha = 0.850), exceptional inter-rater reliability (Intraclass Correlation Coefficient [ICC] = 0.997), and high test-retest stability (ICC = 0.901 for the overall scale; ICC = 0.900 for both cognitive and affective subscales) under symptom-stable conditions. Content validity was robustly supported with an overall Content Validity Index (CVI) of 0.91. Construct validity is evidenced by moderate convergent correlations with functional social competence via the Social Skills Performance Assessment (SSPA; r = 0.577, p < 0.001) and divergent validity via negligible correlation with physical motor performance on the 30-second Chair Stand Test (30 CST; r = 0.184, p = 0.081). The AToMS addresses a critical clinical need for an ecologically valid, psychometrically sound, and developmentally graded assessment of social cognitive deficits in psychiatric rehabilitation.

2. Keywords

Theory of Mind, Schizophrenia, Social Cognition, Cognitive ToM, Affective ToM, Neuropsychological Assessment, Psychometrics, Mentalizing, Social Skills, Psychotic Disorders

3. Authors

The Assessment of Theory of Mind for People with Schizophrenia was developed and validated by an interdisciplinary team of researchers in occupational therapy, psychiatry, special education, and clinical psychology:

• Ya-Chin Yeh, Ph.D., OTR: Department of Occupational Therapy, Shu-Zen Junior College of Medicine and Management, Kaohsiung, Taiwan. (Email: [email protected])
• Chi-Fa Hung, M.D.: School of Medicine, College of Medicine, National Sun Yat-sen University; Department of Psychiatry, Kaohsiung Chang Gung Memorial Hospital, Kaohsiung, Taiwan. (Email: [email protected])
• Chung-Ying Lin, Ph.D.: Institute of Allied Health Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan. (Email: [email protected])
• Yuh-Yih Wu, Ph.D.: Department of Special Education, National Kaohsiung Normal University, Kaohsiung, Taiwan. (Email: [email protected])
• Chun-Hong Kuo, M.S., OTR: Academic Office, Kaohsiung Municipal Nan-Zih Special School, Kaohsiung, Taiwan. (Email: [email protected])
• Marc N. Potenza, M.D., Ph.D.: Department of Psychiatry, Child Study Center, and Department of Neuroscience, Yale School of Medicine, New Haven, CT, United States. (Email: [email protected])
• Chun-Hua Cheng, M.S., OTR: Department of Occupational Therapy, Kaohsiung Municipal Kai-Syuan Psychiatric Hospital, Kaohsiung, Taiwan. (Email: [email protected])
• Kuan-Lin Chen, Ph.D., OTR (Corresponding Author): Institute of Allied Health Sciences, College of Medicine, National Cheng Kung University, No. 1, University Road, Tainan 701401, Taiwan. (ORCID: 0000-0001-9996-814X; Email: [email protected])

4. Purpose

Social cognitive impairment represents a core, disabling dimension of schizophrenia that persists independently of positive psychotic symptomatology and profoundly undermines real-world functional recovery, occupational engagement, and independent community living (Green et al., 2015). Among the multidimensional facets of social cognition identified by the National Institute of Mental Health (NIMH) Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) initiative, Theory of Mind (ToM)—defined as the social-cognitive capacity to represent and infer self and others’ unobservable mental states, such as beliefs, desires, epistemic knowledge, and emotional feelings—is widely recognized as the single strongest social cognitive predictor of community functioning and interpersonal success.

Historically, mentalizing abilities in psychiatric populations were evaluated using pediatric developmental tasks (such as classical Sally-Anne first-order change-of-location tasks) or adult instruments transposed from high-functioning autism spectrum disorder paradigms, such as the “Reading the Mind in the Eyes” test or static textual vignettes (e.g., the Hinting Task or Strange Stories). However, systematic reviews and meta-analyses (Yeh et al., 2021) revealed significant methodological constraints in these existing instruments when applied to schizophrenia. Text-heavy tasks frequently impose disproportionate burdens on general neurocognitive domains, including working memory, sustained attention, executive functioning, and verbal reading fluency, resulting in confounded mentalizing scores. Moreover, static vignettes lack the dynamic, temporal, and multimodal characteristics inherent to human nonverbal interaction, leading to poor ecological validity and pronounced floor or ceiling effects.

To overcome these long-standing psychometric limitations, Yeh and colleagues (2023) developed the Assessment of Theory of Mind for People with Schizophrenia (AToMS). The primary purpose of the AToMS is to provide a standardized, objective, and developmentally stratified assessment battery tailored to the specific cognitive profile of schizophrenia. The instrument fulfills three primary aims:

1. Precise Phenotypic Characterization: Disentangling neurocognitive dysfunction from authentic mentalizing failure by incorporating explicit control questions that evaluate basic comprehension, memory, and physical reasoning alongside target ToM inferential questions.
2. Neuroanatomical & Functional Dissociation: Differentiating between Cognitive ToM (inferring cognitive epistemic states, beliefs, and deceptive intents) and Affective ToM (inferring emotional states, empathy, and interpersonal blunders), which rely on partially segregated prefrontal-temporoparietal and frontolimbic neural circuitries.
3. Clinical Staging and Treatment Monitoring: Providing clinical neuropsychologists, occupational therapists, and psychiatric rehabilitation specialists with a sensitive outcome measure capable of detecting granular improvements following targeted social cognitive remediation therapy (SCIT) or pharmacotherapeutic interventions across different stages of chronicity.

5. Psychological Construct

The overarching construct evaluated by the AToMS is Theory of Mind (ToM), or mentalizing capacity. Rather than treating ToM as a unitary cognitive module, the AToMS operationalizes mentalizing as a multicomponent construct structured across two functional neurobiological axes: Cognitive ToM and Affective ToM, further stratified across developmental levels of sociocognitive complexity.

Cognitive Theory of Mind

Cognitive ToM encompasses the epistemic capacity to represent, interpret, and predict human behavior based on inferences regarding what another individual thinks, knows, or intends. In schizophrenia, cognitive mentalizing deficits can manifest either as hypo-ToM (a failure to represent mental states, leading to social isolation and concrete thinking) or hyper-ToM / over-mentalizing (the unwarranted attribution of malevolent intentions, underlying persecutory and paranoid delusions). Cognitive ToM within the AToMS evaluates four primary theoretical concepts:

• First-Order False Belief: The understanding that another person can hold an internal representation of the external world that differs from objective reality and differs from the observer’s own true belief. Individuals must predict a character’s behavior based on the character’s mistaken epistemic state (e.g., searching for an object in the location where they last left it rather than where it was surreptitiously moved).
• First-Order Deception: The cognitive recognition that an agent is intentionally communicating false information or altering an environmental cue to instill a false belief in another agent for tactical or playful purposes. This requires the observer to decouple the deceptive behavior from genuine behavioral intent.
• Second-Order False Belief: The recursive sociocognitive capacity to understand embedded mental states (“John thinks that Mary thinks that…”). This requires maintaining simultaneous, nested representations of two discrete perspectives that conflict with current reality, representing an ontogenetically later developmental milestone typically emerging in middle childhood.
• Appearance and Reality: The distinction between how an object, event, or social behavior superficially appears versus what it essentially or functionally is in reality. In schizophrenia, concrete thinking often compromises this differentiation, resulting in literalistic misinterpretations of social scenes.

Affective Theory of Mind

Affective ToM involves the capacity to identify, infer, and vicariously comprehend what another individual feels, requires understanding the emotional valence of interpersonal interactions, and underpins affective empathy and tactfulness. While cognitive ToM requires cool, symbolic representations of knowledge states, affective ToM demands hot processing of emotional states within complex social contexts. The AToMS assesses four advanced affective concepts:

• Banter: The interpretation of playful, affectionate teasing or ironic remarks that are not intended to harm, distinguishing non-literal friendly teasing from genuine verbal aggression or insult. Individuals with schizophrenia frequently misattribute hostile intent to benign, lighthearted banter due to hyper-intentionality.
• Hinting: The pragmatic inferential ability to decode covert communicative intentions where an interlocutor conveys a desire or request indirectly without explicitly stating it (e.g., stating “It is very chilly in this room” to imply a request to close an open window).
• Social Cues: The fluid perception and synthesis of contextual, non-verbal micro-behaviors—such as posture, gaze direction, vocal cadence, and subtle facial micro-expressions—to ascertain an individual’s true internal emotional experience during ambiguous interactions.
• Faux Pas Detection: The sociocognitive capacity to recognize when a speaker inadvertently says something socially awkward, inappropriate, or hurtful without realizing the negative emotional impact on the listener. Detecting a faux pas requires a sophisticated synthesis of cognitive ToM (knowing that the speaker did not know the remark would be hurtful or did not intend to hurt) and affective ToM (understanding that the listener felt embarrassed, hurt, or offended).

6. Theoretical Framework

The architecture of the AToMS is theoretically grounded in the convergence of two major paradigms in cognitive science and neuropsychiatry: Ontogenetic Developmental Theory and the Neurobiological Dissociation Model of Theory of Mind.

The Neurobiological Model of ToM (Abu-Akel & Shamay-Tsoory)

Contemporary cognitive neuropsychology posits that mentalizing is not mediated by an isolated brain center, but rather emerges from two distinct, interacting large-scale neural networks (Abu-Akel & Shamay-Tsoory, 2011; Shamay-Tsoory et al., 2010):

• The Cognitive Mentalizing Network: Encompasses the dorsolateral prefrontal cortex (DLPFC), the dorsal medial prefrontal cortex (dMPFC), the temporoparietal junction (TPJ), and the precuneus. This network manages the logical, computational, and rule-based processing of perspective taking, epistemic states, and false beliefs. In schizophrenia, DLPFC hypofunction and TPJ aberrancies are frequently associated with impaired cognitive perspective taking and disorganized thought processes.
• The Affective Mentalizing Network: Encompasses the ventromedial prefrontal cortex (vmPFC), the orbitofrontal cortex (OFC), the anterior insula, and the amygdala. This circuitry is responsible for evaluating emotional significance, processing social rewards and punishments, empathy, and detecting social faux pas. In schizophrenia, structural and functional disconnectivity within the frontolimbic system underpins affective flattening, blunted empathy, and severe difficulties in navigating emotional subtleties.

By splitting the evaluation symmetrically into 8 cognitive and 8 affective items, the AToMS directly mirrors this dual-stream neurobiological architecture, allowing clinicians to delineate whether a patient exhibits selective cognitive deficits, selective affective deficits, or global mentalizing impairments.

Ontogenetic and Developmental Hierarchies

The second pillar of the AToMS derives from developmental psychology (Baron-Cohen, 1995; Wellman & Liu, 2004; Hutchins et al., 2012). Theory of Mind is acquired through progressive developmental stages:

1. Early/Basic Mentalizing (Ages 3–5): Acquisition of visual perspective taking, the distinction between mental and physical phenomena (appearance vs. reality), and the understanding of first-order false beliefs and elementary deception.
2. Intermediate Mentalizing (Ages 6–8): Mastering recursive, second-order false beliefs, appreciating that individuals can simultaneously hold conflicting beliefs, and decoding basic indirect linguistic hints.
3. Advanced/Mature Mentalizing (Ages 9–Adulthood): Synthesizing complex linguistic pragmatics, distinguishing ironic humor and friendly banter from hostility, and grasping complex social violations such as faux pas.

The AToMS incorporates this hierarchical complexity directly into its stimulus design. Instead of presenting only high-level tasks that cause floor effects or low-level tasks that yield ceiling effects, the AToMS systematically stratifies items across three progressive tiers of complexity, ensuring sensitivity across acute, subacute, and chronic states of schizophrenia.

7. Validity

The psychometric validation of the AToMS followed stringent international standards for health measurement instruments (COSMIN guidelines; Yeh et al., 2023), confirming robust content, convergent, and divergent validity.

Content Validity

Content validity was evaluated during initial instrument design through a multi-stage expert panel review involving neuropsychiatrists, clinical psychologists, occupational therapists, and psychometricians. Experts rated the relevance, clarity, developmental appropriateness, and clinical representativeness of each scenario and paired question set using Aiken’s Content Validity Index (Aiken’s CVI). The item-level CVIs ranged from 0.80 to 0.93, culminating in an overall scale-level Content Validity Index of 0.91, well exceeding the recognized psychometric threshold of 0.80 for adequate content representation in specialized clinical assessment batteries.

Convergent Validity

To establish convergent construct validity, performance on the AToMS was correlated with an ecologically valid, gold-standard behavioral assessment of interpersonal functioning: the Social Skills Performance Assessment (SSPA; Patterson et al., 2001). The SSPA requires patients to participate in role-play social scenarios (e.g., greeting a new neighbor, negotiating a service request with a landlord), which are scored across domains including conversational skill, social appropriateness, negotiation, and affect. In a well-characterized cohort of patients with schizophrenia, total AToMS scores demonstrated a statistically significant, moderate-to-strong positive correlation with SSPA performance:

r = 0.577, p < 0.001

This empirical correlation demonstrates that laboratory-measured mentalizing competence directly translates to observable functional social skills in real-world communicative exchanges, validating the clinical utility of the AToMS.

Divergent (Discriminant) Validity

Divergent validity was established by comparing AToMS scores against an unrelated physical functional motor capacity metric: the 30-second Chair Stand Test (30 CST; Jones, Rikli, & Beam, 1999), which measures lower-body muscular strength and functional mobility. In the clinical cohort, the correlation between AToMS and the 30 CST was non-significant and weak:

r = 0.184, p = 0.081

This dissociation demonstrates that the AToMS specifically measures social cognitive and mentalizing constructs rather than reflecting non-specific physical fatigue, general motoric speed, or undifferentiated somatic disability.

8. Reliability

The reliability of the AToMS was evaluated across three core psychometric domains: internal consistency, test-retest stability across clinical windows, and inter-rater reproducibility.

Internal Consistency

The internal consistency of the 16-item scale in patients diagnosed with schizophrenia yielded a Cronbach’s alpha coefficient of α = 0.850. This indicates high internal coherence across the eight distinct ToM concepts without excessive item redundancy (which typically manifests at α > 0.90), confirming that the items reliably reflect the underlying latent construct of social cognitive mentalizing.

Test-Retest Reliability

Temporal stability was evaluated using the Intraclass Correlation Coefficient (ICC) across repeated test administrations in a stable patient subsample. The subsample was confirmed to have maintained clinical symptom stability between the two assessment points, with no significant fluctuation in psychotic symptomatology (p = 0.22 on standardized psychiatric symptom ratings). Under these stable clinical conditions, the AToMS exhibited excellent temporal reproducibility:

• Total AToMS Scale: ICC = 0.901 (95% Confidence Interval [CI] = 0.35–0.97)
• Cognitive ToM Subscale: ICC = 0.900 (95% CI = 0.57–0.97)
• Affective ToM Subscale: ICC = 0.900 (95% CI = 0.50–0.97)

These findings substantiate that the AToMS yields dependable, highly stable scores over time in symptom-stable individuals, satisfying a key prerequisite for use in longitudinal psychiatric monitoring and clinical trials.

Inter-Rater Reliability

Because clinical scoring of open social responses can be vulnerable to subjective examiner bias, inter-rater reliability was rigorously evaluated across five independent, blinded clinical raters who scored identical patient assessment sessions. The resulting inter-rater reliability was near-perfect:

ICC = 0.997 (95% CI = 0.994–0.999)

This near-unity ICC confirms that the standardized scoring criteria, explicit rubrics, and paired-control item architecture of the AToMS effectively eliminate rater subjectivity, ensuring objective scoring across different clinical assessors.

9. Factor Analysis

In the primary psychometric evaluation study of the AToMS (Yeh et al., 2023), formal exploratory factor analysis (EFA) or confirmatory factor analysis (CFA) numerical matrices (e.g., individual item factor loadings, root mean square error of approximation [RMSEA], comparative fit index [CFI]) were not reported due to the prioritized clinical focus on developmental item-response calibration and criterion-related psychometric validation within a clinically stabilized psychiatric sample.

However, the latent structural configuration of the AToMS was theoretically derived and constructed a priori around a two-factor orthogonal/oblique model distinguishing Cognitive ToM from Affective ToM. This structural segregation is supported by extensive contemporary psychiatric neuroimaging literature (Abu-Akel & Shamay-Tsoory, 2011; Shamay-Tsoory et al., 2010), which demonstrates that cognitive mentalizing and affective mentalizing map onto distinct prefrontal-temporoparietal sub-networks.

In classical psychometric modeling of mentalizing tasks in schizophrenia, structural investigations typically identify two competing factor structures:

1. A Unidimensional Model: Positing that all social cognitive tasks load on a single generalized “social cognition” factor, which shares substantial variance with general intelligence (g) and neurocognitive executive function.
2. A Two-Dimensional Correlated Model: Positing separate Cognitive ToM and Affective ToM latent factors that correlate moderately (r ≈ 0.40–0.60) but remain distinct in their associations with functional outcomes (e.g., affective ToM correlating more strongly with emotional regulation and empathy, whereas cognitive ToM correlates with paranoia and cognitive insight).

Future large-sample psychometric studies (> 300 participants) are recommended to conduct formal CFA on the AToMS to empirically confirm the structural stability of the two-factor model against a bifactor alternative, and to assess measurement invariance across clinical subtypes (e.g., first-episode psychosis versus chronic institutionalized schizophrenia) and biological sexes.

10. Instrument / Measurement Tool

The Assessment of Theory of Mind for People with Schizophrenia (AToMS) is structured as an objective, standardized performance-based neuropsychological assessment instrument. Below are the structural and administrative specifications:

• Instrument Name: Assessment of Theory of Mind for People with Schizophrenia (AToMS)
• Test Type: Original Clinician-Administered Performance Battery / Questionnaire
• Administration Format: Individual administration via standardized dynamic visual/animated scenarios and structured auditory-verbal prompts
• Target Population: Adults diagnosed with schizophrenia (including young adults [18–29 years], adults in their thirties [30–39 years], and middle-aged adults [40–64 years]; male and female)
• Language: Originally developed and validated in Mandarin Chinese
• Item Count: 16 core items evaluating 8 distinct Theory of Mind concepts
• Concept Distribution:
  • Cognitive ToM (8 items total across 4 concepts): First-order false belief, First-order deception, Second-order false belief, Appearance and reality
  • Affective ToM (8 items total across 4 concepts): Banter, Hinting, Social cues, Faux pas
• Paired Item Structure: Each conceptual scenario features:
1. A Control Question: Assesses factual comprehension, attention, and physical causality to ensure that any failure on the subsequent ToM item is not driven by general cognitive or perceptual breakdown.
2. A Theory of Mind (ToM) Question: Probes the mental state attribution, cognitive inference, or emotional identification.
• Scoring Format & Rules:
  • Each control question is scored 1 point if answered correctly.
  • Each Theory of Mind (ToM) question is scored 1 point if answered correctly.
  • Subscale scores range from 0 to 8 (8 items, each scored 0 or 1).
  • The total AToMS score ranges from 0 to 16.
  • Higher total scores indicate better ToM performance.
• Administration Time: Approximately 20 to 30 minutes

11. Permissions & Fee and Test Year

• Test Publication Year: 2023
• Fee: No fee is required for the clinical or academic research use of the AToMS. The instrument is made available at no financial cost to psychiatric researchers and certified clinical professionals.
• Permissions & Access: While the test is free of licensing fees, the complete stimulus battery, dynamic presentation materials, and examiner’s scoring handbook are proprietary and copyrighted. Interested clinicians and investigators must contact the corresponding author, Dr. Kuan-Lin Chen ([email protected]) or first author Dr. Ya-Chin Yeh ([email protected]) at National Cheng Kung University, Taiwan, to request formal academic permission, administrative materials, and authorized test manuals.

12. References

Abu-Akel, A., & Shamay-Tsoory, S. (2011). Neuroanatomical and neurochemical bases of theory of mind. Neuropsychologia, 49(11), 2971–2984. https://doi.org/10.1016/j.neuropsychologia.2011.07.012

Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. MIT Press. https://doi.org/10.7551/mitpress/4635.001.0001

Green, M. F., Horan, W. P., & Lee, J. (2015). Social cognition in schizophrenia. Nature Reviews Neuroscience, 16(10), 620–631. https://doi.org/10.1038/nrn4005

Hutchins, T. L., Prelock, P. A., & Bonazinga, L. (2012). Psychometric evaluation of the Theory of Mind Inventory (ToMI): A study of typically developing children and children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 42(3), 327–341. https://doi.org/10.1007/s10803-011-1244-7

Jones, C. J., Rikli, R. E., & Beam, W. C. (1999). A 30-s chair-stand test as a measure of lower body strength in community-residing older adults. Research Quarterly for Exercise and Sport, 70(2), 113–119. https://doi.org/10.1080/02701367.1999.10608028

Patterson, T. L., Moscona, S., McKibbin, C. L., Davidson, K., & Jeste, D. V. (2001). Social skills performance assessment among older patients with schizophrenia. Schizophrenia Research, 48(2–3), 351–360. https://doi.org/10.1016/S0920-9964(00)00109-8

Shamay-Tsoory, S. G., Harari, H., Aharon-Peretz, J., & Levkovitz, Y. (2010). The role of the orbitofrontal cortex in affective theory of mind deficits in criminal offenders with psychopathy. Cortex, 46(5), 668–677. https://doi.org/10.1016/j.cortex.2009.04.008

Wellman, H. M., & Liu, D. (2004). Scaling of theory-of-mind tasks. Child Development, 75(2), 523–541. https://doi.org/10.1111/j.1467-8624.2004.00691.x

Yeh, Y.-C., Hung, C.-F., Lin, C.-Y., Wu, Y.-Y., Kuo, C.-H., Potenza, M. N., Cheng, C.-H., & Chen, K.-L. (2023). The animated assessment of theory of mind for people with schizophrenia (AToMS): Development and psychometric evaluation. European Archives of Psychiatry and Clinical Neuroscience, 273(3), 663–677. https://doi.org/10.1007/s00406-022-01498-2

Yeh, Y.-C., Lin, C.-Y., Hung, C.-F., & Chen, K.-L. (2021). A systematic review of theory of mind instruments for use in people with schizophrenia. European Psychiatry, 64(Suppl 1), S482–S483. https://doi.org/10.1192/j.eurpsy.2021.1294

13. Items of the Scale