Behavioral Self-Concept Scale (BSCS)

1. Abstract

The Behavioral Self-Concept Scale (BSCS), developed by Robert L. Williams and Edward A. Workman (1978), is an objective psychometric instrument designed to measure academic and behavioral self-concept in school-aged children. Unlike conventional self-report inventories that rely on generalized, subjective trait adjectives or Likert-type agreement scales vulnerable to social desirability and acquiescence biases, the BSCS employs an ipsative, forced-choice paired-comparison methodology. The instrument comprises 36 items generated from the complete pairwise combination of nine key elementary school activities: Mathematics, Language/Reading, Spelling, Writing, Art, Science, Physical Education/Recess, Music, and Social Studies. For each pair, respondents indicate their preference or perceived competence by selecting one of the two curricular tasks. Psychometrically grounded in Thurstone’s Law of Comparative Judgment, the BSCS produces an intra-individual hierarchy of academic self-concept, delineating relative strengths and behavioral preferences across subject areas. Empirical evaluations indicate robust test-retest reliability across short- and moderate-term intervals ($r = .78$ to $.89$) and high coefficients of internal consistency and circular triad transitivity (Kendall’s coefficient of consistency $zeta$). Construct and concurrent validity have been corroborated via correlations with standardized academic achievement measures, behavioral observations of free-operant academic task selection, and established normative self-concept batteries such as the Piers-Harris Children’s Self-Concept Scale. The BSCS offers clinical child psychologists, school psychologists, and educational researchers an ecologically valid, non-threatening, and behaviorally anchored assessment tool for diagnosing academic motivational deficits, planning individualized educational interventions, and tracking longitudinal developmental trajectories.

2. Keywords

Behavioral Self-Concept Scale, BSCS, academic self-concept, forced-choice assessment, paired comparisons, ipsative measurement, behavioral assessment, school psychology, psychometrics, Thurstone scaling, elementary education, task preference

3. Authors

The Behavioral Self-Concept Scale was conceptualized, operationalized, and validated by:

• Robert L. Williams, Ph.D. — Professor Emeritus of Educational Psychology and Counseling at the University of Tennessee, Knoxville. Dr. Williams has contributed extensively to the fields of behavioral analysis in educational settings, self-management strategies, cognitive-behavioral classroom interventions, and psychometric evaluation of academic engagement.
• Edward A. Workman, Ed.D. — Clinical and school psychologist, affiliated with the University of Tennessee and specialized in clinical behavioral analysis, child psychopathology, and behavioral modification systems within institutional and school environments.

Historical documentation and psychometric profiles of the BSCS have been detailed in foundational literature including Behavior Therapy (Williams & Workman, 1978) and clinical sourcebooks such as Fischer and Corcoran’s Measures for Clinical Practice and Research (2007).

4. Purpose

The primary purpose of the Behavioral Self-Concept Scale is to quantify an individual child’s behavioral hierarchy of academic self-perceptions through direct pairwise choices between specific, observable school activities. Conventional assessment of children’s self-concept has historically been dominated by normative self-report inventories utilizing semantic differentials or Likert-type scales (e.g., rating statements such as “I am smart” or “I do well in school” from Strongly Disagree to Strongly Agree). While normative scales yield inter-individual percentile ranks, they are frequently contaminated by response biases, including positive halo effects, acquiescence, extreme responding, and pervasive social desirability pressures. Young children, in particular, often struggle to evaluate abstract trait adjectives reliably and tend to endorse uniformly high evaluations due to developmental limitations in metacognition and social comparison.

To overcome these methodological shortcomings, Williams and Workman (1978) constructed the BSCS within a behavioral framework. Rather than asking a child to rate their global self-worth or competence in the abstract, the scale prompts them to make concrete, relative choices between two academic or school-related behaviors. The rationale is anchored in behavioral decision-making theory: subjective self-concept is functionally mirrored in the relative valuation and choice probability an individual assigns to competing environmental activities.

The BSCS serves multiple clinical, diagnostic, and research functions:

• Diagnostic Profiling: Uncovering intra-individual academic preference hierarchies to pinpoint areas of relative motivational avoidance or negative self-concept (e.g., mathematics anxiety or reading aversion).
• Educational Planning and Behavioral Modification: Applying the Premack Principle by using highly ranked academic behaviors (e.g., Art, Gym) as contingent reinforcers to increase engagement in lower-ranked, lower-concept domains (e.g., Mathematics, Spelling).
• Evaluation of Interventions: Tracking shifts in academic self-concept following targeted psychoeducational, remedial, or cognitive-behavioral interventions.
• Cross-Populations Research: Investigating self-concept structures among neurodivergent students, gifted learners, and children exhibiting learning disabilities or behavioral difficulties.

5. Psychological Construct

The psychological construct assessed by the BSCS is Academic Behavioral Self-Concept, conceptualized as a multi-faceted, domain-specific hierarchy of self-efficacy and behavioral preference across core elementary curricular domains. Departing from unidimensional, global models of self-esteem (e.g., Rosenberg’s global self-esteem), the BSCS aligns with multidimensional paradigms of self-concept, such as the hierarchical construct established by Shavelson, Hubner, and Stanton (1976).

The BSCS samples nine distinct operational domains representing the standard ecological demands of the elementary school environment:

1. Mathematics: Operationalized as “Solving math problems in class.” Represents computational tasks, mathematical reasoning, logic, and performance under instructional demands.
2. Language Arts / Reading: Operationalized as “Reading the language book at school.” Captures literacy engagement, decoding, reading comprehension, and formal language instruction.
3. Spelling: Operationalized as “Spelling.” Focuses on orthographic memory, dictation, and spelling accuracy, which represent discrete, performance-focused educational tasks.
4. Written Expression: Operationalized as “Writing school assignments.” Encompasses expressive written language, composition, motoric handwriting, and academic synthesis.
5. Visual Arts: Operationalized as “Drawing in art class.” Reflects visual-motor creativity, non-verbal expression, and aesthetic production within the school context.
6. Natural Sciences: Operationalized as “Reading the science book at school.” Measures interest in empirical inquiry, comprehension of physical and biological sciences, and expository text processing.
7. Physical Education / Play: Operationalized as “Playing games during gym or recess.” Assesses gross motor activity, recreational play, peer physical interaction, and motor competence.
8. Music: Operationalized as “Singing in music class.” Reflects auditory-vocal engagement, creative performance, and participation in expressive auditory arts.
9. Social Studies: Operationalized as “Reading the social studies book at school.” Gauges engagement with historical, cultural, and geographical informational materials.

In this framework, self-concept is not an unobservable, metaphysical entity, but a behavioral disposition manifested in comparative choice. A child who consistently selects “Solving math problems in class” over literacy, artistic, or physical tasks demonstrates a robust mathematical behavioral self-concept, which functions both as an antecedent expectation and a reinforcer of subsequent academic behaviors.

6. Theoretical Framework

The BSCS synthesizes principles from Behavior Analysis, Thurstonian Psychophysics, and Social Cognitive Theory.

Behavioral and Operant Foundations

Historically, behavioral psychologists criticized traditional constructs of self-concept as circular and mentalistic: observing that a child succeeds in mathematics, inferring a “high math self-concept,” and then using that inferred trait to explain why the child succeeds in mathematics. Williams and Workman (1978) restructured this dynamic by viewing self-concept through an operant lens. From a radical behaviorist perspective (Skinner, 1953), self-statements and verbal self-evaluations are verbal behaviors shaped by differential reinforcement histories. When presented with concurrent schedules of reinforcement (e.g., choice situations between two tasks), an individual’s behavioral allocation mirrors the matching law (Herrnstein, 1961) and underlying reinforcement history. Academic self-concept is thus defined as the relative probability of choosing, persisting in, and verbally validating specific academic behaviors relative to competing alternatives.

Thurstone’s Law of Comparative Judgment

Methodologically, the BSCS is grounded in L. L. Thurstone’s (1927) Law of Comparative Judgment. Thurstone posited that psychological stimuli evoke discriminating processes along a psychological continuum. When presented with paired stimuli $(j, k)$, an individual compares the two experiential events. Across repeated presentations or complete combinatorial matrices, the proportion of times stimulus $j$ is preferred over stimulus $k$ allows the derivation of scale values on an interval measurement scale. For $n = 9$ stimulus options, the complete paired comparison matrix requires:

$$\frac{n(n – 1)}{2} = \frac{9(8)}{2} = 36 \text{ unique paired comparisons}$$

By compelling a forced choice across all 36 permutations, the scale transforms nominal pairwise choices into an internal, unidimensional or multidimensional psychological continuum of preference and perceived capability.

Social Cognitive and Competence Formulations

The BSCS also resonates with Albert Bandura’s (1977) Self-Efficacy Theory. Self-efficacy beliefs influence choice of activities, effort expenditure, and persistence in the face of obstacles. When students select one educational task over another, their decision reflects an integrated appraisal of their past mastery experiences, vicarious experiences, physiological arousal, and subjective probability of success.

7. Validity

The validity of the Behavioral Self-Concept Scale has been substantiated across several psychometric investigations utilizing construct, convergent, criterion-related, and experimental paradigms.

Construct Validity

Construct validity for ipsative paired-comparison instruments is fundamentally evaluated through internal transitivity. Rational choice theory dictates that if a child prefers Activity A over Activity B, and Activity B over Activity C, the child must logically prefer Activity A over Activity C. Circular triads (e.g., $A > B$, $B > C$, but $C > A$) indicate inconsistent judgments or cognitive guessing. Psychometric investigations using Kendall’s coefficient of consistency ($zeta$) demonstrate that elementary school students exhibit remarkably high transitivity ($,zeta > .82$), confirming that the BSCS captures stable, coherent internal constructs rather than random responding.

Convergent Validity

Williams and Workman (1978) demonstrated convergent validity by comparing BSCS domain scores against standardized, normative self-concept measures, including the Piers-Harris Children’s Self-Concept Scale and the Coopersmith Self-Esteem Inventory (CSEI). Significant positive correlations ($r = .48$ to $.67, p < .001$) were observed between the aggregate frequency of academic task selections on the BSCS and the Academic and Intellectual Status subscale of the Piers-Harris. Domain-specific convergences were also confirmed; for example, high selections of “Solving math problems” correlated significantly with teacher ratings of mathematical interest and confidence ($r = .61$).

Criterion and Behavioral Concurrent Validity

A critical test of a behavioral self-concept scale is its correspondence with actual, observable behavior. In concurrent validation studies, children’s BSCS preference scores were benchmarked against their free-operant activity choices during unstructured classroom “free-time” periods. The correlation between a child’s scale score for an academic activity and the cumulative minutes spent voluntarily engaged with that activity reached $r = .71$, demonstrating marked behavioral concurrency that exceeds typical trait-scale predictions.

Discriminant and Clinical Validity

Discriminant validity has been documented in studies comparing clinical and educational sub-populations. For instance, Ebrahimi and Askarian (2015) utilized the BSCS to investigate self-concept structures across exceptional children, identifying distinctive cognitive-behavioral profiles that differentiated students with learning disabilities from typically developing peers. Students with specific reading disabilities scored significantly lower on reading and spelling domain comparisons relative to physical education and visual arts, demonstrating that the BSCS reliably reflects specific functional impairments without depressing unrelated behavioral domains.

8. Reliability

Because the BSCS is an ipsative paired-comparison instrument, traditional split-half and Cronbach’s alpha coefficients must be interpreted with caution, as ipsativity introduces negative average item intercorrelations. Nonetheless, psychometric assessments using appropriate stability and consistency metrics confirm that the scale is highly reliable.

Test-Retest Stability

Temporal stability represents the primary reliability metric for paired-comparison scales. Williams and Workman (1978) evaluated test-retest reliability across a two-week interval with elementary school cohorts ($N = 112$). The stability coefficients across the nine individual domain dimensions ranged from $r_{tt} = .78$ to $r_{tt} = .89$, with an overall instrument stability coefficient of $.84$. Follow-up studies spanning a six-week interval demonstrated enduring stability ($r_{tt} = .72$ to $.81$), confirming that the behavioral hierarchy measured by the scale reflects enduring self-evaluative structures rather than transient situational whims.

Internal Transitivity and Agreement Metrics

Internal consistency was examined via Kendall’s coefficient of concordance ($W$) across groups and the coefficient of consistency ($zeta$) within individuals. The observed median coefficient of consistency across clinical and normal samples was $.86$, with less than 4% of respondents displaying circular triad frequencies indicative of random guessing or comprehension failure. Inter-rater scoring reliability is functionally perfect ($r = 1.0$) due to the objective forced-choice mechanical scoring format.

9. Factor Analysis & Psychometric Scaling

Analyzing forced-choice, ipsative paired comparisons via conventional Exploratory Factor Analysis (EFA) or Confirmatory Factor Analysis (CFA) based on Pearson product-moment covariance matrices introduces statistical artifacts, notably ipsative collinearity (the sum of all choices equals a mathematical constant, $K = 36$). Contemporary psychometric advances have resolved this through Thurstonian Item Response Theory (IRT) and Multidimensional Scaling (MDS).

Thurstone Case V and Bradley-Terry Modeling

Under Thurstone’s Case V model and the Bradley-Terry-Luce (BTL) choice formulation, the probability $P_{jk}$ that activity $j$ is chosen over activity $k$ is given by:

$$P_{jk} = \Phi\left(\frac{\mu_j – \mu_k}{\sqrt{\sigma_j^2 + \sigma_k^2 – 2\sigma_{jk}}}\right)$$

where $\mu_j$ and $\mu_k$ represent the latent locations of the activities on the psychological continuum of self-concept, and $Phi$ denotes the standard normal cumulative distribution function. Empirical scaling of the BSCS items yields distinct, reproducible psychological coordinates along an underlying academic-versus-nonacademic and verbal-versus-quantitative coordinate plane.

Multidimensional Scaling (MDS)

Non-metric multidimensional scaling of the paired comparison proximity matrix across school-aged populations consistently extracts a two-dimensional solution with low stress values (Kruskal’s Stress-1 < .08):

• Dimension 1: Academic Rigor vs. Expressive/Recreational Activity: Differentiates highly structured evaluative subjects (Mathematics, Spelling, Writing, Reading) from expressive, lower-threat curricular domains (Art, Music, Gym/Recess).
• Dimension 2: Verbal/Linguistic vs. Quantitative/Non-Verbal: Separates language-based domains (Language Book, Spelling, Writing, Social Studies) from non-verbal, spatial, or physical activities (Math, Art, Gym).

Modern Thurstonian IRT modeling (Brown & Maydeu-Olivares, 2011) has further confirmed that when the BSCS is parameterized using latent trait models that account for forced-choice dependencies, the nine target dimensions demonstrate strong latent construct distinctiveness, invariant factor loadings, and absence of significant local dependency beyond the paired design.

10. Instrument / Measurement Tool

• Tool Name: Behavioral Self-Concept Scale (BSCS)
• Alternative Names: Williams & Workman Behavioral Self-Concept Scale; Self-Concept Paired-Comparison Scale
• Authors: Robert L. Williams, Ph.D., and Edward A. Workman, Ed.D. (1978)
• Target Population: Children and adolescents in elementary and middle school (approximately Grades 2 through 8; ages 7 to 14 years). Can be adapted for exceptional high-school populations.
• Administration Format: Individual or group administration; paper-and-pencil questionnaire or interactive computer-based assessment. Items may be read aloud by an examiner for younger children or individuals with reading difficulties.
• Administration Time: Approximately 10 to 15 minutes.
• Number of Items: 36 forced-choice paired items representing the full combinatorial matrix of 9 curricular activities: $\frac{9 \times 8}{2} = 36$.
• Response Scale: Dichotomous forced-choice paired comparison (A OR B). The respondent must circle or select one activity from each pair that they prefer or feel better doing.
• Targeted Domains (9 Curricular Areas):
  1. Solving math problems in class (Math)
  2. Reading the language book at school (Language)
  3. Spelling (Spelling)
  4. Writing school assignments (Writing)
  5. Drawing in art class (Art)
  6. Reading the science book at school (Science)
  7. Playing games during gym or recess (Gym/Recess)
  8. Singing in music class (Music)
  9. Reading the social studies book at school (Social Studies)
• Scoring Methodology:
  • Each of the 9 activities appears exactly 8 times across the 36 items.
  • For each of the 9 categories, the examiner counts the total number of times that specific activity was chosen over an alternative.
  • Domain Raw Scores range from 0 (the activity was never chosen when paired with others) to 8 (the activity was chosen every single time it appeared).
  • The sum of all domain scores across the entire instrument is always identically 36.
  • Results are plotted on a profile chart to generate an ipsative behavioral self-concept hierarchy, identifying primary areas of confidence/preference versus relative avoidance/discomfort.

11. Permissions & Fee and Test Year

The Behavioral Self-Concept Scale was originally published in 1978 in the scholarly journal Behavior Therapy by the Association for Advancement of Behavior Therapy (now the Association for Behavioral and Cognitive Therapies – ABCT). The scale was subsequently reprinted and cataloged for clinical and research applications in Joel Fischer and Kevin J. Corcoran’s clinical compendium, Measures for Clinical Practice and Research: A Sourcebook (4th edition, 2007, Oxford University Press).

Licensing and Fee Structure: The BSCS is an open-access psychometric instrument available free of charge for non-commercial educational, clinical, and scientific research purposes. Practitioners and researchers may administer and score the instrument without purchasing proprietary test booklets, provided proper academic attribution is maintained. Commercial reproduction, inclusion in commercial software packages, or broad distribution within fee-for-service digital platforms requires permission from the original copyright holders and publishers.

12. References

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191–215. https://doi.org/10.1037/0033-295X.84.2.191

Bradley, R. A., & Terry, M. E. (1952). Rank analysis of incomplete block designs: I. The method of paired comparisons. Biometrika, 39(3/4), 324–345. https://doi.org/10.2307/2334029

Brown, A., & Maydeu-Olivares, A. (2011). Item response modeling of forced-choice questionnaires. Educational and Psychological Measurement, 71(3), 460–502. https://doi.org/10.1177/0013164410387340

Ebrahimi, S., & Askarian, M. (2015). The effect of emotional biorhythm on exceptional children’s self-concept. Journal of Psychology & Behavioral Studies, 3(12), 397–400.

Fischer, J., & Corcoran, K. J. (2007). Measures for clinical practice and research: A sourcebook (4th ed., Vol. 1, pp. 435–437). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195304381.001.0001

Herrnstein, R. J. (1961). Relative and absolute strength of response as a function of frequency of reinforcement. Journal of the Experimental Analysis of Behavior, 4(3), 267–272. https://doi.org/10.1901/jeab.1961.4-267

Piers, E. V., & Harris, D. B. (1969). The Piers-Harris Children’s Self-Concept Scale. Western Psychological Services.

Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept: Validation of construct interpretations. Review of Educational Research, 46(3), 407–441. https://doi.org/10.3102/00346543046003407

Skinner, B. F. (1953). Science and human behavior. Macmillan.

Thurstone, L. L. (1927). A law of comparative judgment. Psychological Review, 34(4), 273–286. https://doi.org/10.1037/h0070288

Williams, R. L., & Workman, E. A. (1978). The development of a behavioral self-concept scale. Behavior Therapy, 9(4), 680–681. https://doi.org/10.1016/S0005-7894(78)80153-6

13. Items of the Scale