Aaron T. Beck's Cognitive Content Specificity Hypothesis

Author: Angela Bogdanova · Published: October 21, 2026 · Editorial Policy

Research figure: Aaron T. Beck · Written by Artificial Sapiens

Aaron T. Beck's cognitive content specificity hypothesis proposes that emotional disorders are partly distinguished by the meanings and themes of the thoughts that characterize them. In its classic anxiety–depression formulation, depressive cognition preferentially concerns loss, deprivation, personal failure, and negative expectations, whereas anxious cognition preferentially concerns danger, vulnerability, and anticipated threat. The hypothesis is a testable proposition within Beck's cognitive theory of emotional disorders. It addresses the content of appraisals, the measurement of differentiated cognitive patterns, and the conditions under which those patterns can be distinguished amid considerable shared distress. Its empirical history includes supportive clinical comparisons, inconsistent information-processing findings, a critical 2001 meta-analysis, and later investigations showing that some domains discriminate more reliably than others. Beck and Clark's original account and the 2001 meta-analysis establish the central scientific question.

Definition and Scientific Status

Content specificity is the proposition that different emotional states are associated with characteristic cognitive subject matter. It predicts patterns in what people think about, anticipate, recall, interpret, and believe. A person experiencing depressive symptoms may interpret an event through personal failure, defeat, deprivation, or a bleak future. A person experiencing anxiety symptoms may appraise the same situation in terms of impending danger, uncertainty, and possible harm. These illustrations describe patterns predicted by the theory; they are not diagnostic rules and do not establish that an individual has a disorder.

The claim can be expressed as a comparative hypothesis: cognition designed to index depressive themes should be associated more strongly with depression than with anxiety, and cognition designed to index threat should be associated more strongly with anxiety than with depression, after shared negative affect, comorbidity, and measurement overlap are considered. This is a statement about relative statistical specificity. It does not imply that depressive thoughts occur exclusively in depression, that anxious thoughts occur exclusively in anxiety, or that either category of thought is invariably present.

The evidence supports a nuanced position. Clinical research demonstrates recognizable content profiles in some settings, while the degree of discrimination depends on population, task, measurement construction, and analytic model. Direct proof that an observed thought profile causes a disorder is a separate question. A person's cognition, affect, bodily state, and life context interact over time, and cross-sectional correlations cannot independently determine causal direction.

Intellectual Origins in Beck's Cognitive Theory

From Clinical Observation to Testable Cognitive Predictions

Aaron T. Beck's early research on depression shifted attention toward the patterns of interpretation present during emotional suffering. In Thinking and Depression I, published in 1963, he investigated idiosyncratic cognitive content and systematic distortions. In Thinking and Depression II, published in 1964, he elaborated the theoretical and therapeutic significance of these observations. These foundational papers preceded the mature cross-disorder testing program and provide a historical basis for studying the structure of cognition rather than relying exclusively on diagnostic labels.

Beck's 1976 book Cognitive Therapy and the Emotional Disorders articulated a broader cognitive account in which emotions are related to the meanings individuals assign to circumstances. Its influence extended beyond depressive disorders to anxiety and other problems. The program did not assume that every emotional episode can be reduced to one thought. It provided specific hypotheses that researchers could investigate through cognitive interviews, questionnaires, laboratory attention and memory tasks, and observations over time.

Aaron T. Beck and David A. Clark

David A. Clark became a particularly important collaborator in specifying and testing cognitive differences between anxiety and depression. Their 1988 review, Anxiety and Depression: An Information Processing Perspective, identified perceived physical or psychological threat as a central theme of anxiety and loss or deprivation as a central theme of depression. The article examined the cognitive profiles and evidence then available and considered how information processing might reveal their divergence.

This collaboration matters historically because the scientific formulation evolved through studies with multiple investigators. Gary Brown, Robert A. Steer, Judy I. Eidelson, John H. Riskind, Bonnie Stewart, and other contributors helped operationalize, measure, and test the idea. Scholarly attribution therefore requires distinguishing Beck's foundational model from later collaborative empirical studies.

What Counts as Cognitive Content?

Automatic Thoughts and Conscious Appraisals

An automatic thought is an interpretation that arises rapidly in response to a situation and may be accessible to awareness. Researchers can ask participants to identify such thoughts or endorse standardized statements concerning an event, the self, or the future. The content specificity hypothesis focuses on the semantic themes of these interpretations: threat, harm, loss, personal inadequacy, abandonment, or hopelessness. It concerns meaningful categories, not merely whether a sentence sounds negative.

Two negative thoughts can have similar emotional tone yet imply different appraisals. The judgment that an important opportunity has already been irretrievably lost is substantively different from a prediction that a serious danger is about to occur. This distinction makes content specificity more precise than a simple contrast between positive and negative thinking.

Deeper Beliefs and Schemas

Beckian cognitive theory also addresses assumptions and schemas that organize information processing. A schema concerning personal defectiveness may organize ambiguous feedback differently from a schema concerning imminent harm. Yet a belief's broad theme, its accessibility, and its activation under stress may vary independently. A questionnaire measuring explicit statements cannot by itself establish that a stable, underlying schema has been identified.

Specificity can therefore be investigated at several levels: conscious cognitive products, persistent beliefs, interpretations of ambiguity, memory selection, and patterns of attention. Positive findings at one level do not automatically extend to every other level. Sanz's historical review specifically noted that evidence for differentiation was more persuasive for cognitive products than for some broader cognitive operations.

Cognitive Operations and Processing Biases

Information-processing tasks study the handling of information rather than solely the thoughts people report. Such designs include memory for emotionally meaningful words, judgments about self-descriptive adjectives, and interpretation of uncertain situations. Researchers must decide whether a finding reflects the topic of a stimulus, its emotional intensity, general processing efficiency, response strategy, or an experimentally induced demand.

A successful processing experiment can identify differences between groups under carefully controlled conditions. It cannot, without additional designs, demonstrate that the measured difference was present before symptoms emerged or that altering that difference would necessarily prevent a disorder.

Anxiety and Depression as the Core Test Case

The Depressive Profile: Loss, Deprivation, and Negative Expectations

The depression-related prediction concerns a cognitive orientation toward experienced or anticipated loss, failure, hopelessness, negative self-evaluation, and unfavorable future expectations. Some expressions of this profile resemble Beck's broader cognitive account of depression and his cognitive triad, but content specificity asks a narrower comparative question: whether these themes distinguish depression from anxiety when both conditions involve distress.

An individual's report of pessimism or disappointment is common in ordinary life and across clinical conditions. To support the specificity hypothesis, researchers must show that particular cognitive content offers discriminative information beyond general unpleasant affect, overall symptom severity, and overlapping questionnaire items.

The Anxiety Profile: Threat and Anticipated Harm

The anxiety-related prediction emphasizes representations of danger, impending physical or psychological harm, vulnerability, and the perceived inability to manage a threatening situation. Threat may concern bodily sensations, social evaluation, uncertainty about future events, or loss of safety. The exact theme depends on context and on the clinical or nonclinical population studied.

Threat anticipation also occurs in depression, trauma-related difficulties, chronic stress, and ordinary preparation for possible danger. Anxiety-focused cognition may consequently be more difficult to isolate than a simple categorical account anticipates. The relevant statistical question is whether a threat-specific measure is more strongly associated with anxiety when shared distress and correlated cognitive content are modeled appropriately.

Mixed Anxiety and Depression

Co-occurring anxiety and depressive symptoms create an important test of the model. An individual may anticipate harm, appraise an ongoing situation as a loss, and experience diminished hope simultaneously. Contemporary research often treats affective symptoms dimensionally, because mixed states are common and clinical categories are heterogeneous. Content specificity can still hold probabilistically within such states, but investigators must test incremental rather than absolute differentiation.

The distinction between pure clinical groups and mixed samples is therefore consequential. Studies excluding substantial comorbidity may reveal cleaner differences while representing a narrower segment of routine clinical practice. Studies retaining mixed states often achieve greater ecological relevance but face greater demands in modeling common variance.

The First Major Direct Tests

Beck, Brown, Steer, Eidelson, and Riskind (1987)

In Differentiating Anxiety and Depression: A Test of the Cognitive Content-Specificity Hypothesis, Beck and four colleagues introduced evidence from a Cognition Checklist designed to distinguish depressive and anxious automatic thought content. The published report appeared in the Journal of Abnormal Psychology in 1987. It used psychiatric outpatient responses and examined depressive and anxiety-related content alongside clinical measures.

The study is important because it translated the hypothesis into separable scales and made specific predictions about differential associations. Scale differentiation, however, is only one step toward construct validity. When items are selected partly because of their association with particular symptom scales, additional samples and independent tasks are needed to determine how robustly the distinction generalizes.

Greenberg and Beck (1989): Self-Judgment and Incidental Recall

Michael S. Greenberg and Aaron T. Beck's 1989 experiment used trait adjectives related to depression or anxiety, self-descriptiveness judgments, and an incidental free-recall task in clinically depressed, anxious, and psychiatric comparison groups. Depressed participants judged and recalled more depression-relevant negative material. Anxious participants showed a bias toward anxiety-relevant negative adjectives, but depressed participants also processed anxiety-relevant material in a biased way.

The authors interpreted these findings as supporting specificity more strongly for depression than for anxiety. This asymmetry is a scientific result in its own right. It illustrates that a model proposing differences in both directions can receive partial support without meeting its strongest bidirectional prediction.

Clark, Beck, and Brown (1989): Cognitive Mediation

David A. Clark, Aaron T. Beck, and Gary Brown examined cognition and affect in general psychiatric outpatients, using standardized measures and multiple regression models. The research explored whether the relationships between anxious or depressive cognitions and corresponding symptom domains could be distinguished in clinical samples that contained overlapping complaints.

Regression permits the prediction associated with one cognitive variable to be evaluated while accounting for another measured variable, but results depend on construct definitions, reliability, shared variance, and omitted influences. Associations from this design test patterns of specificity rather than causal mediation established through experimental manipulation.

Clark, Beck, and Stewart (1990): Common and Specific Dimensions

In Cognitive Specificity and Positive-Negative Affectivity: Complementary or Contradictory Views on Anxiety and Depression?, Clark, Beck, and Bonnie Stewart analyzed a mixed psychiatric outpatient sample of 470 people. Their analysis supported both a broad common component and separable anxiety and depression dimensions. Subgroup comparisons showed differentiated profiles, while the mixed group displayed greater severity and a combined cognitive and symptom presentation.

The study provides an enduring lesson: shared and specific processes may coexist in the same data. A one-factor account can capture important common distress, and a more differentiated solution can still explain additional structure. Statistical preference for one model should be interpreted in the context of the measures and population rather than treated as an immutable division between disorders.

Synthesizing the Early Evidence

Sanz's 1993 Review

Jesús Sanz surveyed comparative studies of Beck's content-specificity hypothesis in Distinguishing Anxiety from Depression: A Review of Beck's Content-Specificity Hypothesis. The review described meaningful but qualified support, with more convincing findings among clinically defined samples and for reported cognitive content than for other cognitive constructs. Its central methodological implication was that specificity should not be assumed to operate uniformly across all levels of cognition or all degrees of symptom severity.

This review belongs to the historical research record. Its conclusions summarize the literature available at the time and should be distinguished from results of later quantitative reviews.

Richard Beck and T. Scott Perkins (2001): A Critical Meta-Analysis

The 2001 meta-analysis, Cognitive Content-Specificity for Anxiety and Depression, was written by Richard Beck and T. Scott Perkins. Richard Beck is a different scholar from Aaron T. Beck, and the two must not be conflated. The analysis combined 13 eligible studies evaluating relationships between anxiety- or depression-related cognition and corresponding symptoms.

The synthesis found that depression-related cognitive content was more closely related to depression than anxiety, yet both types of cognitive content also shared significant variance with both symptom domains. Cognitive measures were substantially intercorrelated. The authors concluded that the accumulated results were generally inconsistent with the stronger content-specificity hypothesis, notwithstanding a more favorable signal for depressive content.

This evaluation is particularly important because it separates a modest discriminative difference from the claim of sharply distinct cognitive structures. It also raises questions about shared item wording, overlapping negative affect, and the statistical consequences of correlated measures. A model can remain clinically useful for exploring personal meanings while its strongest cross-diagnostic predictions receive limited or contested empirical support.

Evidence Published After the Meta-Analysis

Lamberton and Oei (2008): Within a Depressed Sample

Anna Lamberton and Tian P. S. Oei examined 135 participants with depressive conditions, using measures of depressive symptoms, anxiety symptoms, negative automatic thoughts, and anxious self-statements. Depressive automatic thoughts uniquely predicted depression scores in their models, whereas anxious self-statements uniquely predicted anxiety scores. The study supported differential cognitive associations within the recruited population.

The result demonstrates the value of analyzing both symptom dimensions within one sample, yet its cross-sectional self-report design and selected clinical composition limit conclusions about temporal precedence, mechanisms, and broader diagnostic generalization. It is one finding alongside a body of heterogeneous results, not a decisive reversal of the earlier meta-analysis.

Miranda, Fontes, and Marroquín (2008): Expectations About the Future

Regina Miranda, Monique Fontes, and Brett Marroquín tested future expectancies, hopelessness, and intolerance of uncertainty in relation to depression and generalized anxiety symptoms. Their undergraduate sample included 284 people at baseline, with 263 assessed again six weeks later. Expectations concerning the absence of positive future outcomes related specifically to depression symptoms in reported analyses.

This work refines the content question by examining the form of future anticipation. The expectation that positive possibilities will be absent may differ from the expectation that danger or uncertainty will become intolerable. Its longitudinal component is an advantage over a single-timepoint association, although a six-week self-report design still cannot by itself establish clinical diagnosis or causal pathways.

Smith and Mumma (2008): Daily Mood and Multiple Domains

A multi-wave evaluation by Phillip N. Smith and Gregory H. Mumma examined reported mood and cognition across repeated observations, extending the comparison to anger as well as anxiety and depression. Their analyses found a general relation of negative cognition to negative mood alongside more specific associations for selected themes: transgression with anger and several themes including defectiveness and hopelessness with depressed mood.

The importance of the multi-wave method lies in separating more stable individual differences from changes across observations. The results suggest a layered picture in which broad distress and particular cognitive themes co-occur, with specificity differing across themes rather than applying as one uniform property of all cognitive statements.

Competing and Complementary Models

General Negative Affect

A major alternative explanation is that anxious and depressive cognitions covary because both are connected with general negative affect or distress. A measure that includes phrases about inadequacy, fear, sadness, and tension may correlate with both anxiety and depression simply because it captures an overarching tendency toward distressing experiences.

Lee Anna Clark and David Watson's 1991 tripartite model distinguished common distress from dimensions involving low positive affect and physiological hyperarousal. The model addresses affective and symptom structure, while cognitive content specificity addresses the meanings of thoughts. These models are not logically incompatible: a person may experience common distress while appraising particular situations through themes of threat or loss.

Cognitive Processes Shared Across Diagnoses

Repetitive negative thinking, attentional capture by salient negative material, and unfavorable interpretations can span emotional disorders. Their transdiagnostic character may explain why many cognition scales correlate with more than one symptom domain. A useful distinction is between process, such as perseveration, and content, such as a belief about harm or deprivation. Two people can use a similar repetitive process while focusing on different themes.

A review of perseverative thinking in depression and anxiety discusses overlapping processes and symptom relationships. This line of work supplies a framework for testing whether diagnostic differences reside in thought content, thought process, intensity, context, or combinations of these properties.

Interpretation Bias and Domain Specificity

A related question is whether interpretation of ambiguous events reflects a particular anxiety domain. A systematic review and meta-analysis of interpretation bias in youth investigated content specificity across anxiety symptoms and diagnoses. This research concerns a neighboring, narrower prediction and should not be counted automatically as a direct replication of the original adult anxiety–depression comparisons.

Such studies nevertheless demonstrate how the idea has expanded into more fine-grained tests. Threat content may vary with social, separation-related, or other anxiety concerns, and results must be evaluated according to each study's operational definitions and populations.

How Researchers Should Test the Hypothesis

Compare Convergent and Divergent Associations

A rigorous analysis measures at least two cognitive domains and at least two symptom domains. The central test asks whether the matched associations outperform the mismatched associations in a clinically and statistically meaningful way. The analysis must account for covariance among symptoms and cognitions; observing a significant depression–depressive-thought correlation alone is insufficient.

Researchers should prespecify the expected contrasts, report uncertainty intervals, and evaluate incremental explanatory value. They should also assess whether the effect remains when general distress, severity, demographic composition, and comorbidity are modeled. Statistically significant differences can be small or unstable across samples and tasks.

Avoid Circular Item Construction

If the cognitive questionnaire is built from statements that closely resemble symptoms on a depression or anxiety scale, it may produce artificial specificity. Conversely, if two cognitive instruments share broad distress terms, they may produce artificial nonspecificity. Researchers should examine item overlap, factor structure, measurement invariance, and discriminant validity before interpreting correlations as evidence about psychological theory.

Independent outcomes, blinded assessments, and experimental processing tasks can strengthen the evidence. Where possible, replication should use measures developed separately from the symptom criteria being predicted.

Incorporate Time and Causal Identification

Cross-sectional models reveal association but leave open whether cognitive content precedes symptoms, accompanies them, or partly reflects their consequences. Repeated daily assessments, longitudinal cohorts, and carefully randomized cognitive manipulations can provide stronger tests of direction and within-person change.

Even a finding that a specific theme decreases during psychotherapy does not prove that its reduction caused the clinical outcome. Temporal sequencing, credible comparison conditions, measurement of alternative mediators, and replication are required for a mechanism claim. Clinical treatment effectiveness and the explanatory validity of content specificity should be appraised independently.

Represent Clinical Complexity

Recruitment decisions can change the apparent strength of content specificity. Samples limited to narrowly defined, noncomorbid groups may amplify contrasts; mixed outpatient samples may expose broad common distress. Both designs answer useful but different questions. A robust research program includes different age groups, cultures, languages, comorbidity profiles, symptom severities, and treatment contexts.

Cultural and linguistic equivalence is essential because the meanings of failure, shame, threat, social loss, and uncertainty differ in context. Valid translation requires evidence that items retain their intended construct rather than merely a literal correspondence of words.

Measurement, Diagnosis, and Clinical Interpretation

Content-specific questionnaires and cognitive interviews estimate aspects of reported or elicited thought content. They do not independently diagnose major depressive disorder, generalized anxiety disorder, or other clinical syndromes. Diagnosis requires appropriate assessment of symptom clusters, duration, impairment, history, differential considerations, and clinical context.

The Beck Anxiety Inventory and the Beck Depression Inventory were developed for different measurement purposes, and neither can convert a single cognitive theme into a diagnosis. Correlations between cognitive measures and these inventories can test a scientific hypothesis while remaining vulnerable to shared response style, item overlap, and symptom comorbidity. Severe distress and suicide-related symptoms require direct clinical evaluation rather than interpretation from an isolated cognitive score.

In an individual therapy session, identifying themes of loss, threat, failure, or anticipated rejection can support a collaborative formulation and suggest questions worth exploring. The relevance of a theme derives from the person's lived circumstances, not from a fixed mapping between a sentence and a diagnostic category. Clinical assessment integrates observed problems, functional effects, prior experience, personal meaning, and change over time.

What the Hypothesis Contributes to Cognitive Therapy

Beck's theory encouraged clinicians and researchers to investigate the meanings attached to specific experiences. When two people encounter the same event, different appraisals may help account for different emotions and actions. This idea contributed to the development of structured, collaborative cognitive formulation and to the distinction between generalized negative thinking and personally significant themes.

Treatment research cannot be read as direct validation of every historical cognitive proposition. Effective cognitive behavioral therapy may involve exposure, activity scheduling, learning, alliance, behavioral change, altered expectations, and other components. An intervention can reduce symptoms through several pathways, and a content-specific theory can remain partially supported even when broad CBT outcomes are favorable.

Readers seeking the larger depression model can consult Aaron T. Beck's Cognitive Triad and Theory of Depression. The present article addresses a distinct scientific proposition: whether different emotional problems are differentiated by the thematic content of cognition and how the evidence for that proposition should be judged. For the complete intellectual context, consult the published Aaron T. Beck research biography and scientific chapter.

Contemporary Assessment of the Evidence

The established finding is that depressive and anxious conditions can be associated with characteristic patterns of self-report and information processing in particular settings. Strong evidence also shows extensive overlap in negative affect and in many cognitive measures. The research has not established a universal one-to-one mapping from cognitive theme to diagnosis or a single causal route from theme to disorder.

The more persuasive, though conditional, empirical signal concerns certain depression-related themes. Early memory research and the 2001 quantitative synthesis reported greater specificity for depressive content than for anxiety content. Subsequent studies demonstrate that refined measures and repeated assessments may recover differentiated associations, yet those results remain dependent on the particular constructs and statistical models examined.

Evidence is limited for generalizing any one content classification across every cognitive level, age, language, and psychiatric presentation. Strong causal claims remain contested because many studies are observational or rely on closely related self-report measures. The constructive scientific conclusion is to treat specificity as a graded and context-sensitive empirical property that must be demonstrated for each cognitive domain and outcome.

Research Agenda and Open Questions

A mature test program should distinguish at least four questions: whether specific topics appear more frequently in some symptom domains; whether those topics improve discrimination beyond general distress; whether they precede changes in symptoms; and whether altering them selectively produces predicted clinical changes. These questions require different designs and should not be merged into one verdict.

Promising approaches include preregistered multi-site comparisons, intensive longitudinal studies, independent measures of affect and cognitive content, structured interviews with blinded raters, psychometrically transparent item development, and experimental manipulation with ethically appropriate safeguards. Natural-language analysis may also examine cognitive themes, but automated classification requires privacy protections, careful validation, and controls for cultural and linguistic bias. A model's ability to recognize a theme in text is not evidence that it has diagnosed a person.

The hypothesis remains scientifically influential because it converts an intuitive clinical observation into discriminating predictions about cognition and emotion. Its continuing value lies in the refinement of those predictions under stringent methods and in the recognition that shared distress and differentiated meaning can coexist.

Related Articles

Aaron T. Beck — The verified central scientific biography and research chapter situating content specificity within Beck's intellectual legacy.

Aaron T. Beck's Cognitive Triad and Theory of Depression — A person-specific account of Beck's depression model, distinct from the cross-disorder specificity hypothesis.

Beck Anxiety Inventory: What It Measures and Its Limitations — The existing thematic owner for the Beck Anxiety Inventory, psychometrics, and diagnostic limitations.

Depression Treatment: Therapy, Medication, and Other Options — A separate clinical overview of treatment choices rather than the historical cognitive content hypothesis.

Aaron T. Beck's Cognitive Model of Anxiety Disorders — Examines threat appraisal in Beck's original cognitive model of anxiety.

Aaron T. Beck's Cognitive Distortions: Origins and Classification — Reconstructs Beck's theory of recurring interpretation and reasoning errors.

References

Beck, A. T. (1963). Thinking and depression I: Idiosyncratic content and cognitive distortions. Archives of General Psychiatry, 9, 324–333. https://doi.org/10.1001/archpsyc.1963.01720160014002

Beck, A. T. (1964). Thinking and depression II: Theory and therapy. Archives of General Psychiatry, 10, 561–571. https://doi.org/10.1001/archpsyc.1964.01720240015003

Beck, A. T. (1976). Cognitive therapy and the emotional disorders. International Universities Press. Bibliographic search at WorldCat

Beck, A. T., Brown, G., Steer, R. A., Eidelson, J. I., and Riskind, J. H. (1987). Differentiating anxiety and depression: A test of the cognitive content-specificity hypothesis. Journal of Abnormal Psychology, 96(3), 179–183. https://doi.org/10.1037/0021-843X.96.3.179

Beck, A. T., and Clark, D. A. (1988). Anxiety and depression: An information processing perspective. Anxiety Research, 1(1), 23–36. https://doi.org/10.1080/10615808808248218

Beck, R., and Perkins, T. S. (2001). Cognitive content-specificity for anxiety and depression: A meta-analysis. Cognitive Therapy and Research, 25, 651–663. https://doi.org/10.1023/A:1012911104891

Clark, D. A., Beck, A. T., and Brown, G. (1989). Cognitive mediation in general psychiatric outpatients: A test of the content-specificity hypothesis. Journal of Personality and Social Psychology, 56(6), 958–964. https://doi.org/10.1037/0022-3514.56.6.958

Clark, D. A., Beck, A. T., and Stewart, B. L. (1990). Cognitive specificity and positive-negative affectivity: Complementary or contradictory views on anxiety and depression? Journal of Abnormal Psychology, 99(2), 148–155. https://doi.org/10.1037/0021-843X.99.2.148

Clark, L. A., and Watson, D. (1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100(3), 316–336. https://doi.org/10.1037/0021-843X.100.3.316

Greenberg, M. S., and Beck, A. T. (1989). Depression versus anxiety: A test of the content-specificity hypothesis. Journal of Abnormal Psychology, 98(1), 9–13. https://doi.org/10.1037/0021-843X.98.1.9

Lamberton, A., and Oei, T. P. S. (2008). A test of the cognitive content specificity hypothesis in depression and anxiety. Journal of Behavior Therapy and Experimental Psychiatry, 39(1), 23–31. https://doi.org/10.1016/j.jbtep.2006.11.001

Miranda, R., Fontes, M., and Marroquín, B. (2008). Cognitive content-specificity in future expectancies: Role of hopelessness and intolerance of uncertainty in depression and GAD symptoms. Behaviour Research and Therapy, 46(10), 1151–1159. https://doi.org/10.1016/j.brat.2008.05.009

Sanz, J. (1993). Distinguishing anxiety from depression: A review of Beck's content-specificity hypothesis. Anales de Psicología, 9(2), 133–170. Journal record

Smith, P. N., and Mumma, G. H. (2008). A multi-wave web-based evaluation of cognitive content specificity for depression, anxiety, and anger. Cognitive Therapy and Research, 32, 50–65. https://doi.org/10.1007/s10608-007-9171-9

Watson, D., Clark, L. A., Weber, K., Assenheimer, J. S., Strauss, M. E., and McCormick, R. A. (1995). Testing a tripartite model: II. Exploring the symptom structure of anxiety and depression in student, adult, and patient samples. Journal of Abnormal Psychology, 104(1), 15–25. https://doi.org/10.1037/0021-843X.104.1.15