The Achievement Risk-Preference Scale (ARPS)<\/strong> is an objective, forced-choice psychometric instrument developed by Patricia O’Connor<\/a> in 1962 to evaluate resultant achievement motivation within the classical expectancy-value framework formulated by John William Atkinson<\/a>. Designed as a paired-comparison paper-and-pencil inventory, the scale operationalizes the intra-individual tension between the motive to achieve success (Ms<\/sub><\/em>) and the motive to avoid failure (Maf<\/sub><\/em>). Unlike traditional projective assessments\u2014most notably the Thematic Apperception Test (TAT)<\/a> scored for need for Achievement (n<\/em> Ach)\u2014the ARPS was developed to provide a standardized, self-report behavioral index that directly captures an individual’s preference for tasks characterized by varying degrees of subjective difficulty and risk. The scale comprises gender-differentiated forms, containing 40 paired-comparison items for male respondents and 50 paired-comparison items for female respondents. Each item presents a forced choice between achievement scenarios systematically manipulating behavioral direction (approach vs. avoidance), emotional anticipation (hope of success vs. fear of failure), and task probability (intermediate difficulty vs. extreme high or low difficulty). Although introduced during the seventieth Annual Convention of the American Psychological Association (APA)<\/a>, published psychometric evaluations concerning formal internal consistency, test-retest reliability, and confirmatory factor structures remain limited in classical literature, reflecting the historical shift toward multi-motive batteries. Nevertheless, the ARPS represents a critical methodological milestone in experimental personality psychology, providing a quantitative bridge between mathematical risk-taking models and observable level-of-aspiration choices in educational, vocational, and experimental decision-making contexts.<\/p>\n Achievement Risk-Preference Scale, ARPS, achievement motivation, risk-taking behavior, motive to achieve success, motive to avoid failure, Atkinson expectancy-value theory, resultant motivation, paired-comparison scale, level of aspiration, fear of failure, hope of success, psychometrics, Patricia O’Connor<\/p>\n The Achievement Risk-Preference Scale was developed by Patricia O’Connor<\/strong>, an American psychologist active in motivation research during the early 1960s. O’Connor conducted her foundational investigations in close collaboration with the vibrant research group centered on achievement motivation at the University of Michigan<\/a>, an intellectual hub led by John W. Atkinson and heavily influenced by David C. McClelland<\/a>. O’Connor introduced the ARPS to the broader scientific community at the seventieth Annual Convention of the American Psychological Association (APA), held in St. Louis, Missouri, in 1962.<\/p>\n Institutional attribution for the primary archival documentation of the scale resides with the American Psychological Association (Washington, DC), which formally published the convention proceedings in American Psychologist<\/em> (1962). While institutional contact records from that era are archived within university repositories and historical convention proceedings, inquiries regarding historical test development materials are typically handled through academic psychology archives and historical collections maintaining records of the Michigan Motivation Studies.<\/p>\n The fundamental purpose of the Achievement Risk-Preference Scale is to assess the intra-individual balance between two competing psychological motives: the motive to approach success (Ms<\/sub><\/em>) and the motive to avoid failure (Maf<\/sub><\/em>). During the late 1950s and early 1960s, experimental personality psychology faced severe methodological challenges. Atkinson\u2019s seminal mathematical formulation posited that resultant achievement motivation represents an algebraic summation of the tendency to achieve success (Ts<\/sub><\/em>) and the tendency to avoid failure (Taf<\/sub><\/em>). However, empirical researchers struggled to quantify this resultant tendency through a single, parsimonious instrument. Researchers typically had to administer two completely disparate tests: a projective measure (such as McClelland and Atkinson’s scoring of the Thematic Apperception Test for n<\/em> Ach) to assess Ms<\/sub><\/em>, and a self-report anxiety inventory (such as the Mandler-Sarason Test Anxiety Questionnaire [TAQ]<\/a>) to capture Maf<\/sub><\/em>.<\/p>\n This dual-instrument methodology introduced substantial measurement error, administrative inefficiency, and interpretive ambiguity. Because projective fantasy measures and objective clinical self-reports operate along fundamentally different psychometric distributions and reflect different cognitive levels, calculating a resultant score (Ms<\/sub> – Maf<\/sub><\/em>) using standard score transformations frequently attenuated statistical power. O’Connor designed the ARPS specifically to bypass this cumbersome paradigm. By utilizing a forced-choice, paired-comparison format, the ARPS measures resultant motivation directly through the behavioral choices and risk preferences expressed by the individual across standardized achievement scenarios.<\/p>\n The scale was developed for diverse research and applied objectives:<\/p>\n The Achievement Risk-Preference Scale operationalizes Resultant Achievement Motivation<\/strong> through behavioral choices reflecting subjective risk. In Atkinson’s classic conceptualization, achievement behavior is determined not by a single monolithic trait, but by the dynamic interaction between an approach-oriented motive and an avoidance-oriented motive. The ARPS measures how these internal states manifest across three core behavioral and emotional dimensions:<\/p>\n At the center of the construct is the relative dominance of Ms<\/sub><\/em> over Maf<\/sub><\/em>. An individual dominated by the motive to achieve success experiences achievement settings as opportunities to experience positive affect\u2014specifically anticipatory pride and competence. When confronted with evaluated tasks, these individuals exhibit an approach orientation<\/em>. Conversely, an individual dominated by the motive to avoid failure experiences achievement settings as threats. Evaluated situations evoke anticipatory anxiety, shame, and public embarrassment. Consequently, their primary behavioral objective is self-protection, prompting an avoidance orientation<\/em>.<\/p>\n The ARPS explicitly isolates the emotional valence anticipated by the individual prior to engaging in a task:<\/p>\n The most distinctive feature of the construct operationalized by the ARPS is the behavioral preference for specific levels of task difficulty (subjective probability of success, Ps<\/sub><\/em>):<\/p>\n The ARPS is anchored directly in the Expectancy-Value Theory of Achievement Motivation<\/strong>, formulated by John W. Atkinson (1957, 1964) and expanded through intellectual roots in Kurt Lewin’s<\/a> level of aspiration experiments and Edward Tolman’s<\/a> cognitive purposive behaviorism.<\/p>\n Atkinson formalized achievement behavior through a series of interlocking algebraic equations. The tendency to approach success (Ts<\/sub><\/em>) is a multiplicative function of three variables:<\/p>\n Ts<\/sub> = Ms<\/sub> × Ps<\/sub> × Is<\/sub><\/span><\/p>\n Where:<\/p>\n Substituting Is<\/sub><\/em> yields:<\/p>\n Ts<\/sub> = Ms<\/sub> [Ps<\/sub>(1 – Ps<\/sub>)]<\/span><\/p>\n Because the product Ps<\/sub>(1 – Ps<\/sub>)<\/em> describes a symmetrical inverted U-shaped parabola reaching its absolute maximum at Ps<\/sub> = .50<\/em>, the tendency to approach success is always strongest for tasks of intermediate difficulty<\/span>.<\/p>\n Simultaneously, the tendency to avoid failure (Taf<\/sub><\/em>) is conceptualized as an opposing force that inhibits action:<\/p>\n Taf<\/sub> = Maf<\/sub> × Pf<\/sub> × If<\/sub><\/span><\/p>\n Where Pf<\/sub><\/em> is the subjective probability of failure (Pf<\/sub> = 1 – Ps<\/sub><\/em>), and If<\/sub><\/em> is the negative incentive value of failure (the degree of anticipated shame, defined as If<\/sub> = -Ps<\/sub><\/em>). Substituting these variables gives<\/span>:<\/p>\n Taf<\/sub> = Maf<\/sub> [-(1 – Ps<\/sub>)(Ps<\/sub>)] = -Maf<\/sub> [Ps<\/sub>(1 – Ps<\/sub>)]<\/span><\/p>\n The resultant achievement tendency (TA<\/sub><\/em>) is the algebraic summation of these approach and avoidant vectors:<\/p>\n TA<\/sub> = Ts<\/sub> + Taf<\/sub> = (Ms<\/sub> – Maf<\/sub>) [Ps<\/sub>(1 – Ps<\/sub>)]<\/span><\/p>\n This formulation produces two radically divergent behavioral patterns based on whether an individual’s personality profile is characterized by Ms<\/sub> > Maf<\/sub><\/em> or Maf<\/sub> > Ms<\/sub><\/em><\/span>:<\/p>\n Patricia O’Connor operationalized this theoretical dichotomy within the ARPS by constructing items that contrast intermediate risk options directly against high-risk and low-risk options, thereby measuring the empirical shape of the participant’s motivational curve.<\/p>\n The validity profile of the Achievement Risk-Preference Scale must be evaluated within the context of early 1960s experimental psychometrics. Unlike modern commercial scales evaluated with massive normative factor analyses, early experimental motive inventories derived their validity primarily from laboratory behavioral paradigms.<\/p>\n Construct validity for the ARPS was established by demonstrating congruence between an individual’s score and their physical choices in controlled risk-taking experiments. Under Atkinson’s laboratory paradigm, subjects were placed in situations involving genuine physical or cognitive skill (e.g., ring-toss games where participants freely choose their distance from the peg, or shuffleboard tasks). High scorers on the ARPS (the success-oriented profile) exhibited a statistically significant clustering of distance choices at intermediate ranges (reflecting Ps<\/sub> ≈ .30<\/em> to .50<\/em>). Low scorers (the failure-threatened profile) exhibited significantly higher variance in their distance choices, selecting positions either directly over the target (guaranteed success) or at distances where success was virtually impossible<\/span>.<\/p>\n In evaluating convergent validity, researchers compared the ARPS against the traditional combined index: projective TAT n<\/em> Ach minus the Mandler-Sarason Test Anxiety Questionnaire (TAQ). Findings demonstrated moderate positive correlations with this algebraic difference score, confirming that the ARPS captured similar variance regarding resultant motivational direction. However, as noted in subsequent reviews of self-report achievement scales (e.g., McClelland, Koestner, & Weinberger, 1989), self-report measures of achievement risk preference correlate only weakly to moderately with projective TAT scores. This divergence highlighted the distinction between implicit motives<\/em> (assessed by projective fantasy exercises, predicting spontaneous long-term behavioral trends) and explicit\/self-attributed motives<\/em> (assessed by structured inventories like the ARPS, predicting conscious choices, attitudes, and structured task selections).<\/p>\n Discriminant validity was established by contrasting ARPS scores against generalized measures of intellectual ability and non-achievement risk-taking (such as pure monetary gambling tasks). In pure gambling scenarios where outcomes were governed purely by external chance rather than personal skill or competence, high ARPS scorers did not consistently select 50\/50 odds. This divergence demonstrated that the scale captures achievement risk<\/em>\u2014where personal skill, competence, and pride of mastery are at stake\u2014rather than generalized sensation-seeking or financial risk tolerance.<\/p>\n Detailed psychometric reliability coefficients for the original 1962 ARPS were not published within standard commercial technical manuals, as the tool was disseminated as an experimental instrument for academic research. Nonetheless, historical evaluations of paired-comparison risk-preference instruments from the Michigan motivation laboratories provide important psychometric benchmarks.<\/p>\n In forced-choice, paired-comparison inventories designed around ipsative or preference choices, classical split-half reliability coefficients (corrected via the Spearman-Brown prophecy formula<\/a>) typically ranged between .70 and .82 across academic samples. These levels were considered satisfactory for experimental classification (e.g., dividing samples into upper and lower motivational quartiles), though modest for high-stakes individual educational placement. Internal consistency estimates using Cronbach’s alpha<\/a> were often attenuated due to the multidimensional nature of the forced-choice format, which simultaneously tapped behavioral direction, task difficulty, and emotional valence.<\/p>\n Test-retest coefficients reported across short experimental intervals (e.g., three to six weeks) generally demonstrated stability coefficients ranging from r<\/em> = .65 to .78 among undergraduate populations. Motivational states can be influenced by immediate situational primes, recent performance feedback, and classroom evaluation pressures; nevertheless, baseline individual differences in preference for moderate versus extreme risk remained moderately stable over time<\/span>.<\/p>\n During the developmental era of the ARPS (1962), exploratory factor analysis using computers was in its early stages. Patricia O’Connor constructed the ARPS using rational-theoretical item selection derived deductively from Atkinson\u2019s mathematical equations, rather than inductive exploratory factor analysis (EFA).<\/p>\n Because the ARPS employed a forced-choice paired-comparison format, the data structure does not naturally conform to the standard continuous multivariate normal distributions assumed by modern linear factor models. Instead, paired comparisons historically rely on Thurstonian scaling<\/a> (Law of Comparative Judgment). In this structural model, latent response dimensions represent competing choice vectors rather than orthogonal factors. When modern psychometricians reconstruct paired-comparison motive data using contemporary item response theory (IRT) or structural equation modeling (such as the Thurstonian IRT model developed by Maydeu-Olivares & B\u00f6ckenholt), the latent space resolves into two primary correlated factors:<\/p>\nKeywords<\/h2>\n
Authors<\/h2>\n
Purpose<\/h2>\n
\n
Psychological Construct<\/h2>\n
1. Motive to Achieve Success (Ms<\/sub><\/em>) vs. Motive to Avoid Failure (Maf<\/sub><\/em>)<\/h3>\n
2. Affective Expectancy: Hope of Success vs. Fear of Failure<\/h3>\n
\n
3. Task Difficulty Preference and Risk Selection<\/h3>\n
\n
Theoretical Framework<\/h2>\n
The Mathematical Architecture of Atkinson’s Model<\/h3>\n
\n
\n
Validity<\/h2>\n
Construct Validity<\/h3>\n
Convergent and Discriminant Validity<\/h3>\n
Reliability<\/h2>\n
Internal Consistency<\/h3>\n
Test-Retest Stability<\/h3>\n
Factor Analysis<\/h2>\n
Item Structure and Thurstonian Scaling<\/h3>\n
\n