Twenty Statement Test (TST)

Abstract

The Twenty Statement Test (TST), also historically recognized as the Self-Attitudes Test, was designed by sociologists Manford H. Kuhn and Thomas S. McPartland in 1954 to operationalize the empirical study of the self-concept within the tradition of symbolic interactionism. The instrument consists of an open-ended, semi-structured self-report protocol in which respondents are presented with twenty numbered blank lines and asked to write twenty distinct answers to the fundamental interrogative prompt: “Who am I?” Developed as the cornerstone methodological tool of the Iowa School of symbolic interactionism, the TST bypasses the constraints of forced-choice Likert scales, allowing individuals to articulate the salient hierarchies of their identity using their own phenomenological categories. Over seven decades of psychometric inquiry, diverse content-coding frameworks have been established for the instrument. The original taxonomy developed by Kuhn and McPartland distinguished between consensual (socially defined, institutional roles and structural statuses verifiable by objective consensus) and subconsensual (individualistic traits, affective dispositions, and idiosyncratic characterizations requiring direct subjective interpretation) self-attitudes. Subsequent psychometric typologies, most notably Chad Gordon’s four-mode coding system, classify responses into Physical/Bodily (A-mode), Social/Relational (B-mode), Reflective/Psychological (C-mode), and Oceanic/Transcendental (D-mode) self-orientations. Cross-cultural research has extensively utilized the TST to quantify differences in independent versus interdependent self-construals, demonstrating robust criterion-related and construct validity across dozens of cultural groups. While classical internal consistency metrics such as Cronbach’s alpha are structurally inapplicable to open-ended generative qualitative inventories, the instrument exhibits substantial psychometric integrity, consistently demonstrating high inter-rater agreement (Krippendorff’s alpha and Cohen’s kappa regularly exceeding .85) and acceptable test-retest reliability across short- and moderate-term intervals. This article provides a comprehensive evaluation of the TST, delineating its theoretical foundation, psychometric operationalization, structural validation, and contemporary utility across cultural, social, and clinical psychology.

Keywords

Twenty Statement Test, TST, Self-Attitudes Test, self-concept, symbolic interactionism, Iowa School, consensual self, subconsensual self, independent self-construal, interdependent self-construal, identity structure, open-ended psychometrics

Authors

The Twenty Statement Test was originally formulated and empirically validated by Manford H. Kuhn (1911–1963) and Thomas S. McPartland (1918–1978) in 1954 at the Department of Sociology and Anthropology, State University of Iowa (now the University of Iowa), Iowa City, Iowa, United States.

Manford H. Kuhn served as a prominent professor of sociology at the University of Iowa and was the primary architect of the empirical wing of symbolic interactionism, commonly known as the “Iowa School.” Kuhn sought to translate the philosophical theories of George Herbert Mead into operationalized, quantitative, and empirically verifiable behavioral science. Thomas S. McPartland completed his doctoral dissertation under Kuhn’s direction, titled The Self and Social Structure: An Empirical Approach (1953), which provided the initial empirical foundation for the 1954 seminal validation publication in the American Sociological Review. Later major theoretical and structural coding elaborations were advanced by sociologists such as Chad Gordon (Harvard University) and cross-cultural psychologists including Harry C. Triandis (University of Illinois at Urbana-Champaign) and Shinobu Kitayama (University of Michigan).

Purpose

The primary purpose of the Twenty Statement Test is to elicit, systematically assess, and quantify the global self-concept, self-attitudes, and the hierarchical organization of identity as conceptualized by the individual. In the early 1950s, the operationalization of the “self” in psychology and sociology was hampered by methodological bifurcation: one paradigm relied upon impressionistic, non-replicable psychoanalytic interviews, whereas the alternative relied upon rigid, normative, closed-ended rating inventories that superimposed arbitrary structural categories onto respondents’ internal schemas. Kuhn and McPartland designed the TST to solve this dilemma by providing a completely open-ended elicitation format paired with standardized, objective, and reproducible content-analytic coding schemes.

In empirical research, the TST is employed to achieve several discrete objectives:

• Assessing Identity Salience and Structural Hierarchy: By examining the sequential order in which statements are produced, investigators determine which identities (e.g., gender, profession, ethnicity, family role, personal attributes) occupy superordinate positions within the respondent’s phenomenological self-hierarchy. Items produced earliest are theorized to possess the highest accessibility and functional salience in guiding social behavior.
• Evaluating Societal and Cultural Construals of Self: The instrument is one of the most widely deployed tools in cross-cultural psychology for evaluating the balance between individualistic (independent, idiocentric) self-definitions and collectivist (interdependent, allocentric) self-definitions across diverse cultural, regional, and ethnic demographics.
• Investigating Developmental and Role-Socialization Trajectories: The TST is used to capture shifts in self-attitudes across developmental epochs (e.g., transition from childhood to adolescence, entry into professional training programs, occupational retirement) as internalized social structures reshape self-understanding.
• Clinical and Counseling Diagnostic Assessments: In applied clinical frameworks, the TST reveals patterns of self-fragmentation, excessive self-criticism, role diffusion, identity foreclosure, or pervasive existential detachment (manifested via disproportionate “Oceanic” self-statements) in individuals presenting with personality disorders, mood disturbances, or dissociative processes.

Psychological Construct

The core psychological construct measured by the TST is the self-attitude or the multi-faceted self-concept. In Kuhn and McPartland’s formulation, an attitude is a disposition to act, and the self-concept represents a constellation of attitudes directed toward one’s own person as an object. Because the human organism can interact reflexively with itself through linguistic symbols, the self consists of internalized social definitions, categories, and behavioral expectations. Over decades of empirical evolution, psychometricians have established two primary theoretical dimensions to organize the self-statements generated on the TST: the Consensual versus Subconsensual Continuum and the Four-Mode Identity Taxonomy (Gordon’s Typology).

1. The Consensual versus Subconsensual Dimension

Kuhn and McPartland (1954) originally developed an operational dichotomy based upon the degree to which an identity claim relies upon public, institutional verification versus private, subjective attribution:

• Consensual Self-Attitudes: Statements that refer to groups, statuses, roles, or demographic classes whose boundaries are socially recognized, formalized, and publicly verifiable. These statements define selfhood by reference to structural positions that imply direct normative expectations and behavioral limits. Examples include: “I am a medical student,” “I am a mother,” “I am a Catholic,” “I am a resident of Chicago.” Consensual statements require no qualification or idiosyncratic interpretation to establish consensus between the actor and an external observer.
• Subconsensual Self-Attitudes: Statements that describe personal traits, states, styles of behavior, idiosyncratic abilities, or affective tendencies that cannot be externally verified without positional qualification or subjective interpretation. These responses denote individual traits divorced from formalized structural roles. Examples include: “I am ambitious,” “I am often anxious,” “I am a good conversationalist,” “I am tired.”

2. Gordon’s Four-Mode Identity System (A-B-C-D Modes)

Chad Gordon (1968) expanded the categorization of TST statements into four comprehensive, mutually exclusive modes that capture the full spectrum of self-representation:

• A-Mode (Physical and Bodily Self): Statements denoting physical attributes, biological characteristics, bodily awareness, physical capabilities, or geographical/spatial location. Examples: “I am tall,” “I am a woman,” “I am blue-eyed,” “I have dark hair.”
• B-Mode (Social and Relational Self): Statements referencing social roles, institutional statuses, memberships, family lineages, and group affiliations (largely corresponding to Kuhn and McPartland’s consensual dimension). Examples: “I am a brother,” “I am a teacher,” “I am a member of the labor union.”
• C-Mode (Reflective and Attributive Self): Statements identifying psychological traits, moral values, habitual emotional patterns, cognitive styles, and personal interests independent of specific social frameworks. Examples: “I am conscientious,” “I am optimistic,” “I am interested in philosophy,” “I love classic cinema.”
• D-Mode (Oceanic and Transcendental Self): Highly generalized, abstract, philosophical, or cosmic statements that do not tie the individual to specific physical, social, or psychological characteristics. Examples: “I am a living consciousness,” “I am a speck of dust in the universe,” “I am part of all humanity,” “I am nothing.”

3. Independent versus Interdependent Self-Construal

In modern cross-cultural psychometrics (e.g., Markus & Kitayama, 1991; Cousins, 1989), the TST is coded for Independent (Idiocentric) versus Interdependent (Allocentric) self-orientations. Independent statements emphasize internal, autonomous attributes, self-contained goals, and uniqueness. Interdependent statements emphasize social context, interpersonal relationships, kinship obligations, and situational social harmony.

Theoretical Framework

The Twenty Statement Test is historically rooted in symbolic interactionism, a major sociological and social-psychological paradigm developed through the work of George Herbert Mead and Charles Horton Cooley. Cooley introduced the construct of the looking-glass self, postulating that individuals develop their self-conceptions by perceiving and interpreting the reactions of others toward them in social interaction. Mead elaborated this theoretical architecture in Mind, Self, and Society (1934), bifurcating the self into two reflexive phases:

• The “I”: The spontaneous, impulsive, unorganized, and acting subject that initiates novel responses to the environment.
• The “Me”: The organized set of attitudes of others that the individual assumes; the internalized perspective of the “generalized other” representing social norms, institutional expectations, and linguistic conventions.

During the mid-twentieth century, symbolic interactionism fragmented into two distinct methodological schools: the Chicago School, championed by Herbert Blumer, and the Iowa School, led by Manford H. Kuhn. Blumer argued that human action is indeterminate, fluid, and interpretive, asserting that the self cannot be captured through standardized psychometric tools or operational variables. Kuhn radically diverged from Blumer’s qualitative non-positivism. Kuhn maintained that the “Me”—the internalized, structural aspect of the self—is sufficiently stable, organized, and cognitive to be measured empirically through standardized self-attitudes.

Kuhn proposed that an individual’s behavior is determined by the specific self-definitions they hold. These definitions are organized hierarchically: an individual with an identity structure dominated by professional role designations will act differently in ambiguous social situations than one whose identity structure is anchored in familial or psychological categories. The TST was engineered precisely as an operational probe of the “Me,” prompting the individual to verbalize their internalized social classifications under non-directive conditions.

Validity

The validity of the Twenty Statement Test has been extensively investigated across sociological, personality, and cross-cultural domains using diverse empirical strategies.

Construct and Structural Validity

Construct validity for the TST was initially confirmed by Kuhn and McPartland (1954) through the demonstration of a strict Guttman-like scaling order. They hypothesized that individuals generate consensual responses before subconsensual responses. Analyzing protocols across multiple university and occupational samples, Kuhn and McPartland found that if a respondent utilized a subconsensual response early in the sequence, they rarely returned to consensual responses; conversely, the emergence of consensual responses after a sequence of subconsensual statements was statistically rare. The coefficient of reproducibility for this sequential ordering exceeded .90 across their original samples, confirming that the self-concept operates as a structured, hierarchical cognitive schema rather than an unstructured associative collection of labels.

Criterion-Related and Predictive Validity

The TST exhibits notable predictive utility regarding social roles and reference group dynamics. Kuhn (1960) administered the instrument across diverse age cohorts and professional tracks (e.g., student nurses undergoing professional training). As students progressed through their clinical education, the proportion and salience rank of professional self-identifications (e.g., “I am a nurse”) increased systematically, demonstrating the test’s sensitivity to anticipatory socialization and professional identity consolidation. In clinical settings, McPartland and associates showed that psychiatric inpatients displaying profound social withdrawal produced significantly fewer consensual identities and an elevated proportion of D-mode (Oceanic/transcendental) statements, reflecting alienation from institutional social structures.

Cross-Cultural and Convergent Validity

Cross-cultural investigations have yielded substantial evidence for the convergent and discriminant validity of the TST. In a landmark investigation, Cousins (1989) contrasted Japanese and American university students using both the standard TST and a modified contextualized TST. On the standard, open-ended TST, American respondents produced significantly more psychological trait attributes (C-mode, subconsensual statements), whereas Japanese respondents generated significantly more social roles and situational identities (B-mode, consensual statements), converging with theoretical models of Western individualism and East Asian collectivism (Markus & Kitayama, 1991). When situational contexts were explicitly specified in the prompts (e.g., “Who am I at school?” “Who am I with friends?”), Japanese students increased their trait descriptions, demonstrating that the apparent absence of traits on the standard TST reflected a normative cultural restraint against decontextualized self-descriptions rather than a cognitive deficit in trait conceptualization.

Reliability

Because the TST relies upon an open-ended, free-response methodology rather than closed-ended, multi-item continuous scales, traditional internal consistency estimates (such as Cronbach’s coefficient alpha or split-half reliability) cannot be directly computed across items. Instead, psychometric reliability is established via inter-coder (inter-rater) agreement and test-retest stability.

Inter-Coder Reliability

Inter-rater reliability assesses whether independent judges classify statements into respective categories (e.g., consensual vs. subconsensual, or the A, B, C, D modes) with statistical consistency. In their initial 1954 validation study, Kuhn and McPartland reported percent agreement between independent coders classifying consensual versus subconsensual statements exceeding 95%. Subsequent investigations adopting Gordon’s four-mode taxonomy have consistently demonstrated Cohen’s kappa coefficients (κ) ranging from .82 to .94 and Krippendorff’s alpha coefficients (α) exceeding .85 across diverse cross-cultural coders. The high degree of concordance is maintained provided that coders receive standardized training on explicit scoring dictionaries.

Test-Retest Reliability

Temporal stability metrics assess whether an individual’s self-attitude profile remains consistent over time. Test-retest evaluations across short intervals (e.g., one to two weeks) have yielded Pearson correlation coefficients for category proportions (e.g., total percentage of B-mode responses) ranging between r = .70 and r = .85. Stability coefficients across longer intervals (e.g., 6 months to 2 years) decline systematically (r = .45 to .62), which accords with the symbolic interactionist premise that while core aspects of the self-concept exhibit structural stability, self-attitudes remain dynamically responsive to shifts in social milieu, developmental stage, and institutional roles.

Factor Analysis and Structural Properties

Because the TST generates nominal categorical data (linguistic statements), classical exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) based on linear product-moment correlation matrices cannot be performed on individual items in the manner of standard Likert inventories. However, structural and factor-analytic properties have been evaluated through Multidimensional Scaling (MDS), Correspondence Analysis, and factor analysis of the aggregated coding categories.

Factor Structure of Proportional Coding Categories

When researchers aggregate responses into standardized proportions (e.g., percentages of responses allocated to Physical, Social, Reflective, and Oceanic categories, or subdivisions thereof) across large demographic cohorts, factor analyses consistently reveal a clear two-factor or four-factor structural solution:

• The Allocentric / Structural Factor: Characterized by high positive loadings for institutional roles, kinship statuses, and religious/ethnic memberships, paired with negative loadings for decontextualized personal traits.
• The Idiocentric / Individuated Factor: Characterized by robust positive loadings for autonomous personality attributes, subjective preferences, and personal competencies, reflecting an individualized locus of selfhood.

Multidimensional Scaling and Sequential Analyses

Multidimensional scaling studies investigating the semantic proximity of statement categories confirm that respondents cognitively organize the task along two primary dimensions: Social Embeddedness vs. Autonomy, and Concrete/Observable Attributes vs. Abstract/Existential Interpretations. Furthermore, hierarchical structural analysis confirms Kuhn and McPartland’s original discovery regarding statement sequencing: consensual statements display elevated mean rank positions (mean position: 1–7), while subconsensual and abstract oceanic statements emerge predominantly in the lower tier of the sequence (mean position: 12–20).

Instrument / Measurement Tool

The Twenty Statement Test is formatted as a semi-structured, free-response qualitative and quantitative measurement tool. The structural and operational parameters of the instrument are outlined below:

• Test Type: Open-ended self-report identity inventory.
• Administration Format: Paper-and-pencil or computer-assisted self-administered protocol. Can be delivered in group or individual settings.
• Standard Administration Time: Typically 10 to 15 minutes. In traditional administration, an explicit time limit of 12 minutes is frequently imposed to encourage spontaneous responding and prevent over-intellectualization.
• Total Items: 20 numbered blank lines.
• Prompt Instructions: Respondents are presented with instructions requesting twenty different answers to the single question “Who am I?” placed on twenty numbered blank lines. The instructions explicitly direct respondents to answer as if answering to themselves, not to an external audience, and to write statements in the order they occur without concern for internal consistency or logical sequence.
• Response Format: Open generative natural language (free-text string for each numbered line).
• Primary Scoring Rules:
  • Kuhn-McPartland Consensual/Subconsensual System: Each valid statement is coded as either Consensual (C) or Subconsensual (S). The total number of Consensual statements (range 0–20) is summed to establish the Consensual Score. A “Locus Score” can also be derived based on the point of transition where subconsensual statements permanently supersede consensual statements.
  • Gordon A-B-C-D System: Each statement is assigned to one of four mutually exclusive modes: Mode A (Physical), Mode B (Social), Mode C (Reflective), or Mode D (Oceanic). Scores are expressed as absolute frequencies and percentage proportions of the total completed responses.
  • Salience/Rank Order Scoring: Weights are assigned to statements based on sequential order (e.g., statement 1 receives a weight of 20, statement 20 receives a weight of 1) to determine the psychological accessibility of specific identity categories.

Permissions, Fee, and Test Year

The Twenty Statement Test was initially published in 1954 by Manford H. Kuhn and Thomas S. McPartland in the American Sociological Review. The instrument was developed under academic research auspices and has never been commercialized or placed under proprietary copyright restrictions. As an instrument first published in 1954 within a standard scientific journal without individual trademark or commercial licensing restrictions, the core prompt and its variations reside within the public domain for academic, clinical, and scientific research purposes.

Researchers and practitioners may reproduce and administer the original prompt and its scoring protocols without paying licensing fees or securing formal institutional authorization, provided that appropriate scholarly attribution is accorded to Kuhn and McPartland (1954) and subsequent coding authors (e.g., Gordon, 1968). Standard ethical research standards regarding informed consent, voluntary participation, and respondent confidentiality apply during all administrations.

References

• Cooley, C. H. (1902). Human nature and the social order. Charles Scribner’s Sons. https://archive.org/details/humannaturesocia00cooluoft
• Cousins, S. D. (1989). Culture and self-perception in Japan and the United States. Journal of Personality and Social Psychology, 56(1), 124–131. https://doi.org/10.1037/0022-3514.56.1.124
• Gordon, C. (1968). Self-conceptions: Configurations of content. In C. Gordon & K. J. Gergen (Eds.), The self in social interaction (Vol. 1, pp. 115–136). John Wiley & Sons.
• Hartley, W. S. (1970). Manual for the Twenty Statements Test. Department of Psychiatry, Kansas City General Hospital and Medical Center.
• Kuhn, M. H. (1960). Self-attitudes by age, sex, and professional training. The Sociological Quarterly, 1(1), 39–54. https://doi.org/10.1111/j.1533-8525.1960.tb01450.x
• Kuhn, M. H., & McPartland, T. S. (1954). An empirical investigation of self-attitudes. American Sociological Review, 19(1), 68–76. https://doi.org/10.2307/2088175
• Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. https://doi.org/10.1037/0033-295X.98.2.224
• McPartland, T. S. (1953). The self and social structure: An empirical approach [Unpublished doctoral dissertation]. State University of Iowa.
• Mead, G. H. (1934). Mind, self, and society from the standpoint of a social behaviorist. University of Chicago Press.
• Robinson, J. P., & Shaver, P. R. (1969). Measures of political attitudes. Survey Research Center, Institute for Social Research, University of Michigan. https://www.psc.isr.umich.edu/dis/infoserv/isrpub/pdf/Measuresofsocialpsychologicalattitudes_2928_.PDF
• Schwirian, K. P. (1964). Variation in structure of the Kuhn-McPartland twenty statements and related response differences. The Sociological Quarterly, 5(1), 47–59. https://doi.org/10.1111/j.1533-8525.1964.tb02247.x
• Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96(3), 506–520. https://doi.org/10.1037/0033-295X.96.3.506

13. Items of the Scale (Questionnaire)