Achievement Risk-Preference Scale (ARPS)

Abstract

The Achievement Risk-Preference Scale (ARPS) is an objective, forced-choice psychometric instrument developed by Patricia O’Connor in 1962 to evaluate resultant achievement motivation within the classical expectancy-value framework formulated by John William Atkinson. Designed as a paired-comparison paper-and-pencil inventory, the scale operationalizes the intra-individual tension between the motive to achieve success (M_s) and the motive to avoid failure (M_af). Unlike traditional projective assessments—most notably the Thematic Apperception Test (TAT) scored for need for Achievement (n Ach)—the 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), 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.

Keywords

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

Authors

The Achievement Risk-Preference Scale was developed by Patricia O’Connor, 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, an intellectual hub led by John W. Atkinson and heavily influenced by David C. McClelland. 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.

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 (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.

Purpose

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 (M_s) and the motive to avoid failure (M_af). During the late 1950s and early 1960s, experimental personality psychology faced severe methodological challenges. Atkinson’s seminal mathematical formulation posited that resultant achievement motivation represents an algebraic summation of the tendency to achieve success (T_s) and the tendency to avoid failure (T_af). 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 Ach) to assess M_s, and a self-report anxiety inventory (such as the Mandler-Sarason Test Anxiety Questionnaire [TAQ]) to capture M_af.

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 (M_s – M_af) 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.

The scale was developed for diverse research and applied objectives:

• Experimental Research: Evaluating how individuals choose between varying probabilities of success (P_s) in laboratory tasks such as ring-toss games, anagram resolution, and competitive problem-solving.
• Educational Assessment: Identifying students who demonstrate a realistic, intermediate-difficulty approach toward academic challenges versus those who exhibit defensive avoidance patterns by selecting unrealistically easy or impossibly difficult academic tasks.
• Vocational Counseling: Investigating the relationship between achievement risk preference and career aspiration levels, occupational mobility, and persistence in demanding professional trajectories.
• Theoretical Validation: Providing an objective behavioral instrument to test Atkinson’s mathematical derivations regarding subjective probability, incentive value, and resultant behavioral inertia without relying on subjective coding of projective fantasy stories.

Psychological Construct

The Achievement Risk-Preference Scale operationalizes Resultant Achievement Motivation 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:

1. Motive to Achieve Success (M_s) vs. Motive to Avoid Failure (M_af)

At the center of the construct is the relative dominance of M_s over M_af. An individual dominated by the motive to achieve success experiences achievement settings as opportunities to experience positive affect—specifically anticipatory pride and competence. When confronted with evaluated tasks, these individuals exhibit an approach orientation. 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.

2. Affective Expectancy: Hope of Success vs. Fear of Failure

The ARPS explicitly isolates the emotional valence anticipated by the individual prior to engaging in a task:

• Hope of Success: Characterized by optimism, constructive task-engagement, cognitive absorption, and the affective anticipation of mastery. In paired-comparison items, this is reflected in preferences for situations where personal effort determines the outcome and where success yields genuine pride.
• Fear of Failure: Characterized by cognitive worry, self-doubt, rumination over negative evaluations, and the anticipation of humiliation. Items measuring this facet capture choices aimed at minimizing the emotional cost of losing, even if it requires forfeiting potential success.

3. Task Difficulty Preference and Risk Selection

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, P_s):

• Intermediate Difficulty Preference (P_s ≈ .50): Atkinson’s theory mathematically dictates that the tendency to approach success is maximal when task uncertainty is highest—that is, when the probability of success is intermediate (moderate risk). At this point, the incentive value of success (I_s = 1 – P_s = .50) combined with probability produces the highest motivational pull (T_s = M_s × .50 × .50 = .25 M_s). Individuals high in resultant motivation prefer tasks where competence is genuinely tested and personal agency determines success.
• Extreme Difficulty Preference (P_s ≈ .10 or P_s ≈ .90): Conversely, when the motive to avoid failure predominates (M_af > M_s), the negative resultant tendency is most acute at intermediate difficulty, producing maximum conflict and anxiety. To defend the self against the psychological cost of failure, failure-threatened individuals gravitate toward the extremes. They choose either exceptionally easy tasks (where P_s ≈ .90, virtually guaranteeing success and eliminating failure threat) or exceptionally difficult tasks (where P_s ≈ .10, ensuring that failure is inevitable but socially blameless, preserving self-esteem because "no one could be expected to succeed").

Theoretical Framework

The ARPS is anchored directly in the Expectancy-Value Theory of Achievement Motivation, formulated by John W. Atkinson (1957, 1964) and expanded through intellectual roots in Kurt Lewin’s level of aspiration experiments and Edward Tolman’s cognitive purposive behaviorism.

The Mathematical Architecture of Atkinson’s Model

Atkinson formalized achievement behavior through a series of interlocking algebraic equations. The tendency to approach success (T_s) is a multiplicative function of three variables:

T_s = M_s × P_s × I_s

Where:

• M_s is the general, stable personality disposition to strive for success.
• P_s is the subjective cognitive probability or expectancy that performance will be successful (ranging from 0.0 to 1.0).
• I_s is the incentive value of success, defined axiomatically as an inverse linear function of difficulty: I_s = 1 – P_s. Thus, succeeding at a very difficult task (low P_s) provides high subjective pride (high I_s), whereas succeeding at an effortless task provides minimal pride.

Substituting I_s yields:

T_s = M_s [P_s(1 – P_s)]

Because the product P_s(1 – P_s) describes a symmetrical inverted U-shaped parabola reaching its absolute maximum at P_s = .50, the tendency to approach success is always strongest for tasks of intermediate difficulty.

Simultaneously, the tendency to avoid failure (T_af) is conceptualized as an opposing force that inhibits action:

T_af = M_af × P_f × I_f

Where P_f is the subjective probability of failure (P_f = 1 – P_s), and I_f is the negative incentive value of failure (the degree of anticipated shame, defined as I_f = -P_s). Substituting these variables gives:

T_af = M_af [-(1 – P_s)(P_s)] = -M_af [P_s(1 – P_s)]

The resultant achievement tendency (T_A) is the algebraic summation of these approach and avoidant vectors:

T_A = T_s + T_af = (M_s – M_af) [P_s(1 – P_s)]

This formulation produces two radically divergent behavioral patterns based on whether an individual’s personality profile is characterized by M_s > M_af or M_af > M_s:

• When M_s > M_af: Resultant motivation is positive. The function is an inverted U-curve peaking at P_s = .50. The individual exhibits high persistence, selects intermediate risks, and actively engages with performance standards.
• When M_af > M_s: Resultant motivation is negative. The function is an upright U-curve with its deepest negative trough at P_s = .50, where failure anxiety is most intense. When forced to choose, the individual experiences maximum repulsion away from moderate difficulty, retreating toward either minimal risk (P_s ≈ .90) or maximal risk (P_s ≈ .10).

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.

Validity

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.

Construct Validity

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 P_s ≈ .30 to .50). 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.

Convergent and Discriminant Validity

In evaluating convergent validity, researchers compared the ARPS against the traditional combined index: projective TAT n 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 (assessed by projective fantasy exercises, predicting spontaneous long-term behavioral trends) and explicit/self-attributed motives (assessed by structured inventories like the ARPS, predicting conscious choices, attitudes, and structured task selections).

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—where personal skill, competence, and pride of mastery are at stake—rather than generalized sensation-seeking or financial risk tolerance.

Reliability

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.

Internal Consistency

In forced-choice, paired-comparison inventories designed around ipsative or preference choices, classical split-half reliability coefficients (corrected via the Spearman-Brown prophecy formula) 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 were often attenuated due to the multidimensional nature of the forced-choice format, which simultaneously tapped behavioral direction, task difficulty, and emotional valence.

Test-Retest Stability

Test-retest coefficients reported across short experimental intervals (e.g., three to six weeks) generally demonstrated stability coefficients ranging from r = .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.

Factor Analysis

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’s mathematical equations, rather than inductive exploratory factor analysis (EFA).

Item Structure and Thurstonian Scaling

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 (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 & Böckenholt), the latent space resolves into two primary correlated factors:

• Factor 1: Approach / Pride Valence (Success Seeking): Strong positive loadings from item pairs that contrast personal skill challenges with neutral or passive scenarios, heavily weighting intermediate probability of success.
• Factor 2: Avoidant / Defense Valence (Failure Prevention): Strong positive loadings from item pairs that favor the total elimination of failure risk (either through effortless tasks or tasks where failure carries zero social stigma).

Later psychometric revisions of achievement motivation (such as those by Andrew J. Elliot and colleagues in their 2×2 and 3×2 achievement goal frameworks) confirmed that approach and avoidance temperaments constitute distinct, correlated latent constructs rather than a perfectly polar unidimensional continuum. This finding aligns with the structural tensions inherent in the ARPS paired items.

Instrument / Measurement Tool

The operational specifications of the Achievement Risk-Preference Scale reflect standard experimental measurement protocols from 1962:

• Instrument Name: Achievement Risk-Preference Scale (ARPS)
• Primary Author: Patricia O’Connor
• Year of Development: 1962
• Administration Format: Standardized self-administered paper-and-pencil inventory
• Target Population: Adults and university undergraduate populations (≥ 18 years of age)
• Item Configuration: Gender-differentiated versions developed to account for sex-role differences in achievement situations during the 1960s:
  • Male Version: 40 paired-comparison items
  • Female Version: 50 paired-comparison items
• Item Presentation Format: Forced-choice paired comparisons. Each item presents the respondent with two distinct behavioral options, achievement scenarios, or psychological reactions (Option A vs. Option B), requiring the respondent to select the option that most closely reflects their personal preference.
• Core Item Contrast Categories:
  • Intermediate difficulty / moderate risk (P_s ≈ .50) vs. extreme low difficulty / safe outcome (P_s ≈ .90)
  • Intermediate difficulty / moderate risk (P_s ≈ .50) vs. extreme high difficulty / highly improbable outcome (P_s ≈ .10)
  • Approach / hope of success orientation vs. avoidant / fear of failure self-protection
  • Skill-dependent performance scenarios vs. non-skill or chance-determined outcomes
• Scoring Methodology:
  • Responses are coded dichotomously (1 or 0) for each pair.
  • A point is awarded whenever the respondent chooses the intermediate risk / approach / success-oriented alternative over the avoidant or extreme-risk alternative.
  • Total scores represent the cumulative frequency of intermediate risk choices. Higher scores indicate a predominance of resultant achievement motivation (M_s > M_af), whereas lower scores indicate a predominance of the motive to avoid failure (M_af > M_s).

Permissions & Fee and Test Year

The Achievement Risk-Preference Scale was developed in 1962 and documented within the official scientific convention program of the American Psychological Association (APA). The foundational reference appeared in the American Psychologist, Vol. 17, No. 6, during the seventieth Annual Convention held in St. Louis, Missouri.

The scale was developed as an experimental academic instrument and was never issued as a commercial, fee-bearing testing kit through a commercial test publisher. Consequently, there are no commercial royalties or purchase fees associated with its archival review. However, intellectual property and archival rights surrounding convention papers and proceedings from that period remain under the copyright stewardship of the American Psychological Association or the author’s academic estate. Researchers wishing to inspect, adapt, or utilize original historical materials must consult the historical archives of the APA or review archival research publications from the University of Michigan’s Department of Psychology. Any modern reproduction of test items in publications must comply with standard fair use and APA copyright policies.

References

• American Psychological Association. (1962). Program of the seventieth Annual Convention of the American Psychological Association: August 30–September 5, 1962, St. Louis, Missouri. American Psychologist, 17(6), 295–408. https://doi.org/10.1037/h0044593
• Atkinson, J. W. (1957). Motivational determinants of risk-taking behavior. Psychological Review, 64(6, Pt.1), 359–372. https://doi.org/10.1037/h0043445
• Atkinson, J. W. (1964). An introduction to motivation. D. Van Nostrand Company.
• Elliot, A. J., & Church, M. A. (1997). A hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology, 72(1), 218–232. https://doi.org/10.1037/0022-3514.72.1.218
• Elliot, A. J., & McGregor, H. A. (2001). A 2×2 achievement goal framework. Journal of Personality and Social Psychology, 80(3), 501–519. https://doi.org/10.1037/0022-3514.80.3.501
• Feather, N. T. (1961). The relationship of persistence at a task to the expectation of success and achievement related motives. The Journal of Abnormal and Social Psychology, 63(3), 552–561. https://doi.org/10.1037/h0045585
• Lewin, K., Dembo, T., Festinger, L., & Sears, P. S. (1944). Level of aspiration. In J. McV. Hunt (Ed.), Personality and the behavior disorders (Vol. 1, pp. 333–378). Ronald Press.
• Mandler, G., & Sarason, S. B. (1952). A study of anxiety and learning. The Journal of Abnormal and Social Psychology, 47(2), 166–173. https://doi.org/10.1037/h0062855
• Maydeu-Olivares, A., & Böckenholt, U. (2005). Structural equation modeling of paired-comparison and ranking data. Psychological Methods, 10(3), 285–304. https://doi.org/10.1037/1082-989X.10.3.285
• McClelland, D. C., Atkinson, J. W., Clark, R. A., & Lowell, E. L. (1953). The achievement motive. Appleton-Century-Crofts. https://doi.org/10.1037/11144-000
• McClelland, D. C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit motives differ? Psychological Review, 96(4), 690–702. https://doi.org/10.1037/0033-295X.96.4.690
• Ray, J. J. (1979). A quick measure of achievement motivation. Australian Psychologist, 14(3), 337–344. https://doi.org/10.1080/00050067908254366

Items of the Scale