Attitude Scale

Abstract

The Attitude Scale, constructed by Jaques Kaswan, Morris Wasman, and Lawrence Z. Freedman in 1960, is an extensive 205-item psychometric assessment developed within the methodological traditions of Thurstone scaling. Engineered specifically to examine conscious evaluative orientations toward primary social figures and environmental objects, the instrument was conceived as a central component of an empirical investigation testing the construct validity and behavioral correlates of the Rosenzweig Picture-Frustration Study. The assessment systematically investigates interpersonal attitudes across 45 discrete five-item continua, where each continuum captures a spectrum ranging from absolute approval and idealized affection to maximum disapproval, hostility, and rejection. Administered via a dichotomous “Yes/No” self-report format, the scale evaluates respondent endorsement patterns to classify attitudes into positive, negative, or ambivalent configurations. Psychometric validation was originally conducted on an institutionalized cohort of adult male state prison inmates in the United States, linking directional attitudes toward parental figures, peers, and institutional authorities to overt versus covert aggressive behaviors. Reliability of the underlying scaling structure was empirically established through an item-sorting paradigm with an independent, naïve rater, demonstrating a minimal discrepancy rate of 3.6% against the predetermined scaling order. Although modern factor-analytic validation was not performed in the original 1960 investigation, the instrument represents an important historical milestone in mid-twentieth-century criminological assessment, attitude measurement, and the empirical differentiation between projective extrapunitiveness and direct, conscious self-report attitudes.

Keywords

Attitude Scale, Thurstone Scaling, Extrapunitiveness, Rosenzweig Picture-Frustration Study, Aggression Assessment, Inmate Evaluation, Interpersonal Attitudes, Psychometrics, Test Construction, Relational Aggression, Evaluative Continua, Criminological Psychology

Authors

The Attitude Scale was developed through a collaborative research initiative between psychologists and psychiatrists at the University of California and Yale University:

• Jaques W. Kaswan, Ph.D. — Affiliated with the Department of Psychology at the University of California. Dr. Kaswan’s scholarly work focused on clinical assessment, family interaction dynamics, perceptual processes in psychopathology, and the empirical measurement of interpersonal attitudes.
• Morris Wasman, Ph.D. — Affiliated with the Department of Psychiatry at the Yale University School of Medicine. Dr. Wasman conducted extensive experimental research in psychopathology, behavioral responses to frustration, neurophysiology, and psychodynamic assessment methodologies.
• Lawrence Zelic Freedman, M.D. — Foundations’ Fund Research Professor of Psychiatry at the Yale University School of Medicine and later the University of Chicago. Dr. Freedman was an internationally recognized forensic psychiatrist, psychoanalyst, and criminologist who investigated the biological, sociological, and psychological determinants of human aggression, criminal behavior, and social deviance.

Purpose

The primary purpose of the Attitude Scale (Kaswan, Wasman, & Freedman, 1960) was to establish a rigorous, standardized inventory capable of mapping an individual’s explicit, conscious attitudes toward primary socialization agents, institutional authority figures, and significant environmental objects. Developed during an era when projective techniques dominated personality assessment, the scale was created to address a pressing clinical and empirical question: to what degree do projective measurements of aggression—specifically the direction and type of aggression captured by Saul Rosenzweig’s Picture-Frustration (P-F) Study—converge with, diverge from, or predict direct verbalized attitudes and overt behavioral manifestations of hostility?

In clinical and correctional contexts, assessing attitudes toward authority, family, and peers is fundamental to understanding an individual’s psychological adjustment, capacity for institutional compliance, and propensity toward interpersonal conflict. The authors sought to construct an objective measurement device that circumvented the arbitrary scoring boundaries common to unstructured self-reports while avoiding the high inference requirements of projective protocols. By structuring the assessment around 45 distinct evaluative continua, the scale was designed to provide clinicians and researchers with an objective psychometric profile that could identify:

• Directionality of Affect: Quantifying whether an individual maintains predominantly positive (approving), negative (hostile/disapproving), or ambivalent attitudes toward critical relational targets (e.g., father, mother, institutional rules, companions).
• Psychological Ambivalence: Detecting internal conflicts manifested when an examinee simultaneously endorses mutually incompatible statements across a continuum, signifying emotional conflict or cognitive dissonance regarding significant persons.
• Defensive Distortion and Dissociation: Comparing conscious, verbal self-descriptions with projective measures of extrapunitive, intropunitive, or impunitive blame, thereby uncovering defensive structures such as reaction formation, denial, or externalization.
• Behavioral Prediction in Institutional Settings: Differentiating individuals whose verbalized attitudes align with overt behavioral infractions from those whose verbal reports disguise underlying destructive impulses.

Beyond its diagnostic utility within forensic institutions, the Attitude Scale provided a template for the empirical analysis of social-evaluative constructs. It offered a standardized method to quantify social perceptions across multiple strata of human relationships, bridging psychodynamic theories of interpersonal defense with the quantitative rigor of experimental psychometrics.

Psychological Construct

The core construct evaluated by the Attitude Scale is interpersonal and institutional attitude, conceptualized as a relatively enduring organization of beliefs, feelings, and behavioral predispositions toward specific individuals, social roles, and systemic structures. Rather than treating attitudes as monolithic, unidimensional traits, Kaswan, Wasman, and Freedman (1960) operationalized the construct as a multifaceted matrix of evaluative polarities situated along 45 distinct relational continua.

Each continuum represents an independent psychological domain encompassing five graded levels of affective valence, spanning from maximum approval through moderate approval, neutrality/ambivalence, and moderate disapproval, to maximum disapproval. The scale assesses several core relational and institutional dimensions:

1. Primary Parental Alignments (Paternal and Maternal Objects)

Rooted in psychodynamic and developmental theory, this dimension measures the internal mental representations and affective valences associated with parental figures. Paternal continua capture evaluations of paternal authority, affection, fairness, and emotional proximity (e.g., spanning from idealized statements such as “I loved my father best of all” to intense repudiation such as “It should be different, but I can’t help hating my father”). Maternal continua assess perceptions of maternal warmth, maternal protectiveness, intrusiveness, and rejection. The balance between maternal and paternal scores yields insights into early socialization experiences and identification patterns.

2. Authority Figures and Institutional Systems

Given the institutionalized population in which the scale was developed, a major dimension captures attitudes toward societal rules, legal frameworks, custodial officers, administrative leaders, and law enforcement. Continua within this cluster evaluate whether authority is perceived as benevolent, protective, and legitimate, or conversely as arbitrary, punitive, persecutory, and exploitative. This construct directly informs theories of institutional defiance, psychopathy, and antisocial conduct.

3. Peer Relations and Horizontal Social Affiliation

This dimension examines how individuals perceive their contemporaries, companions, and broader peer networks. Continua range from high trust, solidarity, and collaborative identification to intense suspicion, cynicism, competitive hostility, and perceived alienation. High levels of peer disapproval frequently correlate with interpersonal estrangement or paranoid social ideation.

4. Evaluative Polarization and Attitudinal Ambivalence

A unique construct measured by the scale’s structural architecture is attitudinal ambivalence. Unlike conventional Likert-type scales that force respondents onto a single midpoint score when feelings are mixed, this scale’s five-item continua allow an individual to endorse both strongly positive and strongly negative items within the same thematic cluster. Psychologically, this reflects emotional splitting, compartmentalization, or deep structural conflict regarding the target object, distinguishing true indifference from active, conflicted ambivalence.

Theoretical Framework

The construction and validation of the Attitude Scale are grounded in the convergence of two major twentieth-century psychological traditions: L. L. Thurstone’s psychophysical approach to attitude measurement and the psychoanalytically informed Frustration-Aggression Hypothesis.

Thurstone’s Equal-Appearing Intervals and Psychophysical Scaling

In the late 1920s and 1930s, Louis Leon Thurstone revolutionized social psychology by demonstrating that subjective psychological constructs, such as attitudes and values, could be measured with the same quantitative precision as physical stimuli. Thurstone postulated that an attitude represents the degree of positive or negative affect associated with some psychological object. By employing judges to sort statements along an objective psychological continuum, Thurstone established the method of equal-appearing intervals.

Kaswan, Wasman, and Freedman (1960) adopted this foundational psychometric logic by organizing their 205 items into 45 distinct five-item continua. Rather than assigning arbitrary weights, each item in a five-point continuum was scaled to represent an equidistant graduation of affect—from maximum positive endorsement (Position 1), through mild positive affect (Position 2), neutral or conditional acceptance (Position 3), moderate negative affect (Position 4), to maximum negative rejection (Position 5). This structure allowed the researchers to treat the responses not merely as isolated item endorsements, but as systematic points along an underlying psychological continuum.

The Frustration-Aggression Paradigm and Rosenzweig’s Ego Defense Model

The conceptual framework of the scale was directly tied to the theoretical debates surrounding the Frustration-Aggression Hypothesis articulated by John Dollard, Neal Miller, and colleagues in 1939. Dollard et al. posited that frustration invariably produces a drive toward aggression. Building upon this, Saul Rosenzweig (1945) proposed a sophisticated tripartite model of frustration reactions categorized by direction of aggression:

• Extrapunitive: Aggression is turned outward toward the environment, blaming external persons, institutions, or objects.
• Intropunitive: Aggression is turned inward, resulting in self-blame, guilt, remorse, and internal suffering.
• Impunitive: Aggression is evaded or glossed over, viewing the frustrating event as unavoidable or devoid of malicious intent.

Kaswan et al. (1960) sought to ascertain whether individuals who display high extrapunitiveness on projective measures like the Rosenzweig Picture-Frustration Study genuinely harbor conscious, generalized negative attitudes toward social objects, or whether projective extrapunitiveness represents an unconscious defense mechanism detached from direct self-report. By providing an objective mapping of conscious attitudes across 45 interpersonal domains, the Attitude Scale served as the theoretical counterweight to projective aggression assessments.

Validity

Validity evaluation for the Attitude Scale primarily centered on construct validity, external criterion-related validity, and convergent-divergent comparisons with established projective and behavioral markers of aggression.

Construct and Convergent Validity: The Rosenzweig P-F Comparison

The scale was administered alongside the Rosenzweig Picture-Frustration Study to assess whether conscious, verbalized attitudes correspond with projective extrapunitiveness. The empirical findings demonstrated a complex, non-linear relationship between conscious attitudes and projective aggression scores:

• Dissociation Between Projective and Direct Hostility: Inmates who exhibited pronounced extrapunitive tendencies on the Rosenzweig P-F Study did not consistently display across-the-board negative attitudes on the 205-item Attitude Scale. Instead, their negative attitudes were selective, targeting specific authority structures rather than peers or maternal figures.
• Ego-Defensive Masking: A significant subset of participants who scored exceptionally high in projective extrapunitiveness produced paradoxically elevated “maximum approval” scores regarding parental figures on the Attitude Scale. This divergence provided empirical evidence for defensive idealization and reaction formation, wherein overt conscious endorsement of love masks unconscious hostility or frustration toward primary socialization agents.

Criterion and Discriminant Validity

To evaluate criterion-related validity, Kaswan et al. (1960) examined how scores on the Attitude Scale corresponded with institutional behavioral records, particularly disciplinary infractions involving verbal and physical aggression within the prison facility:

• Behavioral Infractions: Inmates with high frequencies of institutional conduct violations displayed significantly higher concentrations of negative endorsements across the authority-specific continua (Positions 4 and 5) compared to non-infraction inmates.
• Discriminant Power of Continua: The multi-continua approach successfully discriminated between generalized misanthropy and specific anti-authority hostility. Individuals who exhibited behavioral conflict with custodial staff showed preserved positive attitudes toward peers and specific family members, confirming that the scale does not measure a single undifferentiated complaint trait, but rather specific relational orientations.

Methodological Considerations and Historical Limitations

While the scale demonstrated meaningful construct and criterion correlations, validity metrics in the 1960 study were primarily analyzed using group contrast methodologies, chi-square analyses, and non-parametric rank comparisons. Modern validity standards—such as multitrait-multimethod matrices (MTMM) or structural equation modeling (SEM)—were not yet part of standard psychometric practice, leaving some aspects of discriminant validity open to contemporary re-examination.

Reliability

The reliability of the Attitude Scale was evaluated using an empirical scaling verification methodology rooted in Thurstone’s original classification principles, rather than contemporary internal consistency metrics like Cronbach’s coefficient alpha (which was introduced by Lee Cronbach in 1951 but was not yet universally adopted for large-scale dichotomous sorting tasks in 1960).

Item-Sorting Agreement and Continuum Stability

To establish that the 205 items accurately and reliably occupied their theoretical positions along the 45 five-point continua, an independent psychometric reliability check was executed:

• Independent Naïve Rater Evaluation: A naïve judge, completely blind to the original continuum placements and theoretical hypotheses of the study, was provided with the randomized pool of 205 statements. The rater was instructed to sort the items into their corresponding thematic continua and arrange them from maximum approval to maximum disapproval.
• Discrepancy Rate: Across the entire instrument of 205 items, only 3.6% of the naïve classifications diverged from the original psychometric scaling established by Kaswan, Wasman, and Freedman.
• Implication for Scale Stability: This exceptionally low error rate (3.6%) demonstrated robust inter-judge agreement and confirmed that the semantic and affective gradients of the items possessed high stimulus clarity, providing strong evidence for the structural stability and objectivity of the equal-appearing continua.

Measurement Consistency Across the Continua

Because each of the 45 continua is anchored by five monotonically graded statements, internal reliability within a continuum is reflected in the consistency of respondent endorsement patterns. A respondent whose endorsements follow the expected Guttman-like or Thurstone-like single-peaked distribution (endorsing adjacent items and rejecting opposing extremes) exhibits high individual response reliability. The authors observed that across the prison sample, random or incoherent endorsement patterns (such as simultaneously endorsing maximum approval and maximum disapproval while rejecting the middle statements) occurred in fewer than 5% of all individual continua administrations, further substantiating the operational reliability of the measure.

Factor Analysis

At the time of the scale’s publication in 1960, computational factor analysis was severely constrained by the lack of accessible electronic computing infrastructure. Consequently, no formal exploratory factor analysis (EFA) or confirmatory factor analysis (CFA) was conducted in the original Kaswan, Wasman, and Freedman investigation. The scale was instead constructed via rational-theoretical item generation followed by empirical interval sorting.

Implied Latent Dimensionality

Although an empirical factor matrix was not extracted, the theoretical architecture of the 45 five-item continua points to a distinct multidimensional factor structure. Based on modern psychometric principles, the 205 items would likely decompose into a hierarchical model comprising several first-order factors and higher-order latent constructs:

• Factor I: Paternal Authority and Affect — Encompassing continua measuring reverence, fear, resentment, and dependency regarding the father figure.
• Factor II: Maternal Warmth and Autonomy — Capturing continua addressing maternal nurturance, maternal overcontrol, and perceived emotional abandonment.
• Factor III: Institutional and Legal Cynicism — Comprising continua targeting legal systems, police, judicial integrity, and correctional authorities, reflecting an extrapunitive socio-legal orientation.
• Factor IV: Horizontal Peer Solidarity vs. Alienation — Grouping continua focused on interpersonal trust, betrayal sensitivity, camaraderie, and social withdrawal among equals.
• Factor V: Generalized Extrapunitive Resentment — A potential second-order factor accounting for common variance across items reflecting generalized grievance, externalization of blame, and relational hostility.

Modern Factor-Analytic Perspectives

If evaluated under contemporary psychometric frameworks, such as Item Response Theory (IRT) or multidimensional CFA, the 205 dichotomous items would require specialized estimation techniques (e.g., diagonally weighted least squares [WLSMV] or full-information maximum likelihood). Such analyses would be necessary to evaluate local item independence and determine whether the 45 continua operate as 45 separate micro-factors or coalesce into a more parsimonious set of broad relational dimensions.

Instrument / Measurement Tool

The structure, administration, and scoring parameters of the Attitude Scale are summarized below:

• Name of Instrument: Attitude Scale
• Alternative / Historical Names: Kaswan-Wasman-Freedman Attitude Inventory; 205-Item Thurstone Attitude Continua
• Authors: Jaques Kaswan, Morris Wasman, and Lawrence Zelic Freedman
• Publication Date: 1960
• Construct Assessed: Conscious evaluative attitudes toward significant persons (father, mother, peers) and social objects (authority, societal rules)
• Assessment Type: Self-report rating scale based on Thurstone scaling methodology
• Number of Items: 205 items
• Structural Organization: Organized conceptually across 45 discrete continua, with each continuum comprising exactly 5 graded items representing a spectrum of affect:
• Position 1: Maximum Approval / Idealization
• Position 2: Moderate Approval / Positive Regard
• Position 3: Equivocal / Conditional / Neutral Acceptance
• Position 4: Moderate Disapproval / Grievance
• Position 5: Maximum Disapproval / Rejection / Hostility
• Presentation Format: Items are presented in a randomized sequence throughout the test booklet to prevent patterned or defensive responding.
• Response Options: Forced-choice dichotomous format: “Yes” or “No” for each statement.
• Administration Method: Paper-and-pencil administration (individual or group setting).
• Administration Time: Approximately 45 to 60 minutes for complete completion.
• Target Population: Adults (18 years and older); originally validated on male state prison inmates in the United States.
• Scoring and Classification Protocol:
• Positive Attitude: Scored when an examinee’s agreements cluster predominantly at Position 1 and Position 2 (items closest to maximum approval).
• Negative Attitude: Scored when an examinee’s agreements cluster predominantly at Position 4 and Position 5 (items closest to maximum disapproval).
• Ambivalent / Mixed Attitude: Scored when an examinee simultaneously endorses opposing ends of a continuum (e.g., agreeing with both Position 1 and Position 5) or selectively endorses middle-range/neutral statements (Position 3).

Permissions & Fee and Test Year

The Attitude Scale was published in 1960 in the Journal of Consulting Psychology by the American Psychological Association (APA). The underlying research was conducted under the auspices of research grants supporting psychiatric and criminological investigations at Yale University and the University of California.

Licensing and Academic Reproduction: The original scale and its associated experimental data reside within the academic literature published by the APA. Because the instrument was developed for empirical research rather than commercial test distribution, it has never been marketed as a proprietary psychological test kit. Researchers seeking to reproduce, adapt, or deploy the complete 205-item inventory for non-commercial scholarly purposes must consult the original publication (Kaswan et al., 1960) and adhere to standard academic fair-use guidelines and APA permissions policies regarding archival journal material.

References

Below is a curated list of foundational literature, theoretical references, and empirical studies associated with the Attitude Scale and its conceptual development:

• Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. Yale University Press. https://doi.org/10.1037/10022-000
• Freedman, L. Z. (1958). Mental disease, unnatural deaths, and violence: A study of 1,000 inmates. American Journal of Psychiatry, 115(5), 422–425. https://doi.org/10.1176/ajp.115.5.422
• Kaswan, J., Wasman, M., & Freedman, L. Z. (1960). Aggression and the Picture-Frustration Study. Journal of Consulting Psychology, 24(5), 446–452. https://doi.org/10.1037/h0046233
• Rosenzweig, S. (1945). The picture-association method and its application in a study of reactions to frustration. Journal of Personality, 14(1), 3–23. https://doi.org/10.1111/j.1467-6494.1945.tb01036.x
• Rosenzweig, S. (1978). The Rosenzweig Picture-Frustration (P-F) Study: Basic manual. Psychological Diagnostic Service.
• Thurstone, L. L. (1928). Attitudes can be measured. American Journal of Sociology, 33(4), 529–554. https://doi.org/10.1086/214483
• Thurstone, L. L., & Chave, E. J. (1929). The measurement of attitude: A psychophysical method and some experiments with a scale for measuring attitude toward the church. University of Chicago Press.

Items of the Scale