CPI Good Impression (Gi) Scale

1. Abstract

The California Psychological Inventory – Good Impression (Gi) Scale is one of the foundational validity indices developed by Harrison G. Gough to assess positive response distortion, deliberate impression management, and conventional social desirability within self-report personality assessment. Originally introduced in the 1957 edition of the California Psychological Inventory (CPI) and refined in subsequent revisions (Gough, 1987), the Gi scale comprises 40 dichotomous (True/False) items constructed through an empirical contrast-group criterion method derived from Floyd L. Ruch’s (1942) instructional faking paradigms. Items were selected based on their capacity to differentiate respondents instructed to complete the inventory under standard testing conditions from those instructed to simulate an exceptionally favorable impression, such as when competing for a high-stakes professional position. Psychometrically, the scale demonstrates robust internal consistency (Cronbach’s α = .77 across collegiate cohorts) and acceptable test-retest reliability over a one-year interval (r = .68 in adolescent cohorts).

Crucially, psychometric and behavioral research demonstrates that the Gi scale exhibits an interpretative duality. At moderate-to-elevated raw scores (ranging from approximately 18 to 29), the scale measures substantive personality variance, capturing individuals who are genuinely conscientious, tactful, socially cooperative, industrious, and observant of normative social conventions. Independent observer ratings, including spouse Q-sorts (N = 400) and trained assessor evaluations (N = 793), consistently substantiate that moderate Gi elevations correspond to verifiable interpersonal dependability and impulse control. However, at extreme elevations—specifically at or above the established clinical cutoff score of 30—the scale functions as a robust dissimulation detector, identifying protocols compromised by intentional “faking good,” severe defensiveness, or an implausible denial of common human frailties. This comprehensive article examines the historical derivation, theoretical architecture, latent factor structure, construct validity, clinical and industrial utility, and psychometric operationalization of the CPI Gi scale.

2. Keywords

California Psychological Inventory, Good Impression scale, CPI Gi, Harrison G. Gough, impression management, response distortion, faking good, social desirability, validity scales, personality assessment, psychometrics, folk concepts.

3. Authors

The Good Impression (Gi) Scale was developed by Harrison G. Gough, Ph.D. (1921–2014), Professor Emeritus of Psychology at the University of California, Berkeley, and former Director of the Institute of Personality Assessment and Research (IPAR, later known as the Institute of Personality and Social Research [IPSR]). Dr. Gough was one of the preeminent psychometricians of the twentieth century, widely celebrated for pioneering the empirical criterion-keying methodology within non-clinical populations and conceptualizing personality measurement through universal “folk concepts.”

Throughout his tenure at UC Berkeley, Dr. Gough authored and refined the California Psychological Inventory across multiple editions (1957, 1987, and the 1996 CPI-434 / CPI-260 revisions). His research focused on personality dynamics, social competence, creativity, academic achievement, executive leadership, and the psychometric detection of test-taking attitudes, including dissimulation, defensiveness, and malingering. Historical correspondence and archival inquiries regarding the CPI development are maintained within the Department of Psychology at UC Berkeley and Consulting Psychologists Press (now The Myers-Briggs Company).

4. Purpose

The primary purpose of the CPI Good Impression (Gi) Scale is twofold: first, to serve as an objective psychometric screen for positive dissimulation (“faking good”) in self-administered personality protocols; and second, to measure substantive interpersonal dispositions toward social conformity, emotional overcontrol, and the maintenance of positive interpersonal relations under standard conditions. In applied psychological testing—spanning industrial and organizational psychology, personnel selection, executive coaching, high-security clearance screening, and forensic evaluations—examinees frequently encounter incentives to present themselves in an excessively favorable light. The Gi scale was specifically engineered to identify individuals who consciously or semi-consciously distort their self-descriptions to project moral infallibility, emotional stability, and professional perfection.

Beyond its function as a response bias detector, Gough designed the CPI not as a psychiatric diagnostic tool like the Minnesota Multiphasic Personality Inventory (MMPI), but as an inventory of “folk concepts”—personality characteristics that emerge naturally across cultures and everyday social interaction. Consequently, the Gi scale was intentionally constructed to capture how individuals manage their social reputation. While excessive scores compromise test validity, normative elevations reflect adaptive interpersonal tendencies: high scorers are often seen by peers and organizational supervisors as responsible, polite, peace-seeking, and industrious employees who actively adhere to organizational norms and avoid interpersonal friction.

In clinical and counseling settings, the Gi scale provides vital contextualization for interpreting the broader CPI profile. An elevated Gi score warns the clinician that other clinical or folk scales (such as Well-Being [Wb], Tolerance [To], or Self-Control [Sc]) may be artificially inflated, whereas scales measuring pathology, neuroticism, or interpersonal conflict may be suppressed. Conversely, an unusually depressed Gi score (e.g., raw scores below 10 or standard scores below 35T) alerts the practitioner that the individual may be engaging in “faking bad,” self-deprecating response distortion, severe acute distress, or an overt rejection of conventional societal expectations.

5. Psychological Construct

The psychological construct evaluated by the Gi scale lies at the intersection of impression management, social desirability, and personality overcontrol. Within contemporary personality science, socially desirable responding is conventionally bifurcated into two distinct psychological components, as articulated by Paulhus (1984): Self-Deceptive Enhancement (SDE), an unconscious, honest bias toward presenting oneself favorably; and Impression Management (IM), a deliberate, conscious manipulation of self-presentation tailored toward an external audience. The Gi scale primarily captures Impression Management, specifically conditioned by organizational and conventional social expectations.

To fully understand the construct of the Gi scale, one must examine its distinct operational components:

• The Conventional Persona: High Gi scorers systematically endorse items that reflect unwavering adherence to conventional social mores, moral propriety, and task persistence. They subscribe to classical work-ethic maxims (e.g., placing business strictly before pleasure) and express intellectual, cultural, and civic interests that are socially revered.
• Denial of Common Negative Affect and Minor Transgressions: The scale incorporates statements reflecting ubiquitous, normative human frailties (e.g., minor irritations, mild interpersonal resentments, procrastination, or occasional white lies). Individuals high on the Gi construct systematically reject these normative shortcomings, claiming an implausible degree of emotional equanimity, patience, and altruism.
• Behavioral and Characterological Overcontrol: As substantiated by observational research (Gough, 1987; Lanning, 1989), the psychological core of the non-faking high scorer is characterized by strict self-regulation, inhibition of aggressive or erotic impulses, fastidiousness, and excessive binding of inner tension. Such individuals delay gratification to an extraordinary degree and prioritize social harmony over individual self-expression.
• Social Convergence and Cooperativeness: The construct entails an acute sensitivity to social expectations. Individuals elevated on this dimension avoid controversy, defer to authority, exhibit formal etiquette, and display an intense desire to “fit in” smoothly across diverse social and institutional matrices.

The defining psychometric hallmark of the Gi construct is its non-linear interpretation. As Gough observed, the scale does not operate on a simple unipolar spectrum where “more is always better.” Rather, when raw scores surpass approximately 30 (out of 40), the probability of genuine behavioral conscientiousness declines precipitously, replaced almost entirely by conscious dissimulation and psychological defensiveness.

6. Theoretical Framework

The theoretical framework underpinning the CPI Gi scale is grounded in Harrison Gough’s Folk Concept Theory of Personality, integrated with the experimental role-playing paradigms pioneered by Floyd L. Ruch (1942). Gough rejected the premise that personality scales should strictly measure formal, abstract clinical taxonomies. Instead, he argued that psychometric instruments should measure psychological processes that everyday social observers naturally recognize and utilize to describe, predict, and evaluate human behavior in daily life.

In Gough’s structural framework, folk concepts are interpersonal currencies. When individuals interact, they continually evaluate one another along axes of dominance, sociability, responsibility, tolerance, and the integrity of their self-presentation. The management of one’s public persona—what Erving Goffman termed the “presentation of self” in sociological dramaturgy—is a fundamental human social function. Gough recognized that any comprehensive assessment system must include internal calibrators capable of gauging the degree to which an individual is manipulating their social presentation to achieve an instrumental social or organizational objective.

Methodologically, the construction of the Gi scale utilized Ruch’s (1942) contrast-group methodology. Rather than relying on theoretical derivation or subjective item writing, Gough administered a large experimental pool of personality items to research participants under two distinct instructional conditions:

1. Standard Administration: Participants responded to the items honestly and naturally under standardized psychological testing conditions.
2. Good Impression Simulation: The same participants (or matched contrast cohorts) were instructed to complete the booklet with explicit role-playing instructions: “Try to give just as favorable an impression of yourself as you would if you were actually applying for an important position, or were trying to create a very favorable impression…”

By contrasting the response percentages between these two conditions, Gough identified items that demonstrated dramatic statistical shifts. Items that showed the greatest divergence—those endorsed with significantly higher frequency under the simulation instructions than under baseline conditions—were selected for the preliminary scale. Because the simulation prompt explicitly invoked the scenario of a high-stakes job interview, the resulting scale intrinsically selected for items reflecting the “ideal corporate or institutional citizen”: dependability, emotional stability, deference, conscientiousness, and social cooperativeness.

7. Validity

The construct, convergent, and discriminant validity of the CPI Good Impression Scale has been extensively documented through decades of empirical testing across clinical, educational, occupational, and correctional populations.

Criterion and Experimental Validity

The initial criterion validity was established directly through the contrast-group item selection method. Studies consistently show that when normal individuals are instructed to fake good or present an idealized self-image, their mean scores on the Gi scale increase dramatically—typically shifting from a normative mean of ~18.5 to scores well above 30, often reaching raw scores between 32 and 37 (Gough, 1987). This demonstrated sensitivity makes Gi one of the most effective non-clinical scales for detecting instructional dissimulation.

Convergent Validity: Observer and Spouse Q-Sorts

To investigate what the Gi scale measures when respondents are not under experimental faking instructions, Gough conducted extensive criterion-matching studies using the California Q-Set (CAQ). In a landmark study involving 400 married couples, respondents completed the CPI, while their spouses independently rated them using the CAQ without access to the test results. The four Q-sort items displaying the highest positive correlations with respondents’ Gi scores were:

1. “A conscientious and serious-minded person.”
2. “Well-organized, capable, patient, and industrious; values achievement.”
3. “Gentle, considerate, and tactful in dealing with others.”
4. “Gets along well with others; able to ‘fit in’ easily in most situations.”

In an independent investigation involving 793 adult participants evaluated at the Institute of Personality Assessment and Research (IPAR) by teams of trained psychological assessors (who observed participants across intensive weekend living-in assessment scenarios), respondents’ Gi scores correlated most positively with the following trained assessor Q-sort items:

1. “Is fastidious.”
2. “Favors conservative values in a variety of areas.”
3. “Is a genuinely dependable and responsible person.”
4. “Tends toward overcontrol of needs and impulses; binds tensions excessively; delays gratification unnecessarily.”

These findings provide compelling convergent evidence for the dualistic nature of the Gi scale: under non-faked conditions, moderate elevations correspond directly to real-world, externally observable conscientiousness, impulse control, and social adaptability.

Discriminant and Profile Validity

The Gi scale correlates moderately to strongly with other CPI scales within Class II (measures of socialization, maturity, and responsibility), particularly Self-Control (Sc) (r ≈ .50 to .65) and Good Well-Being (Wb). However, it diverges significantly from Class I scales measuring assertive, outgoing leadership (e.g., Dominance [Do] and Sociability [Sy]), where correlations are modest, reflecting that Gi prioritizes compliance, self-restraint, and harmony over forceful dominance or extraverted risk-taking. Furthermore, Gi demonstrates strong positive correlations with the Lie (L) and Defensiveness (K) scales of the MMPI (ranging from .55 to .70), confirming its shared sensitivity to deliberate positive impression management and psychological defensiveness.

8. Reliability

The psychometric stability and internal consistency of the CPI Gi Scale have been validated across diverse demographic, educational, and professional cohorts.

Internal Consistency

In normative samples presented in the California Psychological Inventory Administrator’s Guide (Gough, 1987), the scale demonstrates robust internal consistency. Across a primary validation cohort of 4,126 college students (including both men and women), the Gi scale yielded a Cronbach’s alpha (α) coefficient of .77 for males and .77 for females. Given that the scale samples a broad array of behavioral content—spanning emotional control, work habits, aesthetic interests, and interpersonal manners—an internal consistency coefficient approaching .80 reflects high psychometric coherence without excessive, redundant item phrasing.

Test-Retest Stability

Temporal stability evaluations demonstrate that when individuals complete the scale under standard testing instructions, the underlying construct remains remarkably stable over time. In a longitudinal cohort of male and female high school students re-assessed after a one-year interval, Gough (1987) reported a test-retest correlation coefficient of r = .68. Shorter test-retest intervals (e.g., one to four weeks among undergraduate cohorts) typically yield stability coefficients ranging from .80 to .86, confirming that absent deliberate faking or substantial structural environmental interventions, baseline impression management and conventional self-control reflect stable trait-like dispositions.

Standard Error of Measurement

With standard deviations consistently hovering between 5.5 and 7.0 across diverse normative samples (e.g., college students: SD = 5.9; nurses: SD = 5.5; prison inmates: SD = 7.0), the standard error of measurement (SEM) for the raw Gi scale is approximately 2.8 raw score points. This permits clinicians to establish reliable 95% confidence intervals around observed scores and provides clear boundaries for detecting faked profiles.

9. Factor Analysis

Extensive exploratory factor analytic (EFA) and confirmatory factor analytic (CFA) investigations into the CPI have clarified the structural positioning of the Gi scale within the broader inventory and delineated its internal dimensionality.

Structural Position Within the CPI Mega-Dimensions

In Gough’s three-vector structural model of the CPI (which organizes the 20 folk scales into three broad structural axes: Vector 1 – Internality/Externality; Vector 2 – Norm-Favoring/Norm-Questioning; and Vector 3 – Realization/Ego Integration), the Gi scale loads heavily on Vector 2 (Norm-Favoring vs. Norm-Questioning). Factor analytic studies conducted by Megargee (1972) and Lanning (1989) reveal that Gi exhibits its primary factor loading (frequently exceeding .70) on the overarching “Socialization / Rule Compliance” factor, alongside the Socialization (So), Self-Control (Sc), and Tolerance (To) scales. Gi also demonstrates a secondary, moderate loading on Vector 3, reflecting psychological adjustment and personal integration.

Latent Factor Structure of the Gi Scale

When the 40 items of the Gi scale are subjected to independent factor analysis, the scale consistently reveals a multi-faceted internal architecture reflecting three distinct latent components:

• Factor I: Denial of Common Flaws and Hostile Impulses (Impression Management Proper): This primary factor accounts for the largest proportion of common variance. Items loading on this factor reflect the absolute repudiation of anger, petty resentments, procrastination, and untruthfulness. In faking-good simulation conditions, this factor accounts for almost the entirety of the variance shift.
• Factor II: Conscientious Work Ethic and Conventional Industriousness: The second factor comprises items emphasizing orderliness, business before pleasure, devotion to duty, and adherence to social obligations. This factor accounts for the high ratings of dependability and competence rendered by spouses and external assessors.
• Factor III: Cultural Refinement and Intellectual Openness: A smaller, distinct factor captures claims of sophisticated, socially esteemed interests, such as enjoying lectures on international affairs, reading educational literature, and displaying refined aesthetic taste.

CFA studies evaluating unifactorial versus bifactorial models of response distortion (e.g., Paulhus, 1984; Lanning, 1989) confirm that while a single general factor (“Good Impression”) adequately summarizes the total score for screening purposes, the scale functions simultaneously as a substantive measure of emotional overcontrol and normative social conformity.

10. Instrument / Measurement Tool

The operational specifications of the CPI Good Impression (Gi) Scale are summarized below:

• Instrument Name: California Psychological Inventory – Good Impression Scale (CPI-Gi).
• Original Author: Harrison G. Gough, Ph.D.
• Test Type: Standardized self-report validity scale / objective personality assessment scale.
• Format: Dichotomous True/False self-administered paper-and-pencil or computerized questionnaire.
• Item Count: 40 items (35 retained intact from the 1957 edition, 5 slightly modified or updated for the 1987 revision).
• Response Options: Dichotomous: TRUE or FALSE.
• Keying Direction:
  • 9 items keyed True: Endorsement of conventional conscientiousness and cultural interests (CPI booklet item numbers: 14, 103, 127, 133, 140, 165, 195, 222, 254).
  • 31 items keyed False: Rejection of normative social infractions, minor irritations, and anti-conventional impulses (CPI booklet item numbers: 10, 30, 34, 38, 42, 44, 48, 56, 66, 70, 78, 81, 91, 101, 102, 109, 120, 150, 153, 159, 170, 178, 203, 207, 231, 238, 248, 262, 268, 273, 289, 293).
• Scoring and Score Range: Raw scores range from 0 to 40 points (1 point awarded for each item matched to the scoring key). Raw scores are converted to normalized standard T-scores (Mean = 50, SD = 10) based on appropriate normative reference groups.
• Normative Reference Means (Gough, 1987):
  • College Students (N = 4,126): Mean = 18.5, SD = 5.9
  • Registered Nurses (N = 100): Mean = 18.6, SD = 5.5
  • Prison Inmates (N = 345): Mean = 17.9, SD = 7.0
• Interpretative Cutoff Scores:
  • Raw Score ≤ 10 (T < 35): Marked negative dissimulation (“faking bad”), acute distress, alienation, or severe defiance of conventional norms.
  • Raw Score 11–26 (T 36–62): Normal, valid range. Reflects balanced self-appraisal with typical social adaptation.
  • Raw Score 27–30 (T 63–69): Heightened conventionality, high conscientiousness, strong need to please, or moderate defensiveness.
  • Raw Score > 30 (T ≥ 70): Strongly suggestive of deliberate positive distortion (“faking good”) or severe lack of self-awareness. At this level, the substantive validity of the entire CPI profile is severely compromised.

11. Permissions & Fee and Test Year

The Good Impression Scale was originally published in 1957 as an integral scale of the California Psychological Inventory by Consulting Psychologists Press, Inc. (Palo Alto, CA). Revisions were released in 1987 and 1996 (incorporating the CPI-434 and the shortened CPI-260). The inventory and its constituent scales, including Gi, are proprietary intellectual property owned and copyrighted by The Myers-Briggs Company (formerly CPP, Inc., Sunnyvale, CA).

The scale is not in the public domain. It is a commercial, fee-based psychometric instrument. Administration booklets, scoring keys, computer scoring services, and manual guides must be purchased directly through The Myers-Briggs Company or its licensed international distributors. In accordance with standard professional test publishing standards, access is restricted to qualified users meeting Qualification Level B or C (e.g., licensed psychologists, certified psychometricians, or individuals who have completed formal graduate-level training in psychometrics and personality assessment). Researchers seeking to use individual CPI scales for non-commercial academic research must apply for written research permissions or a formal academic licensing agreement through The Myers-Briggs Company permissions portal.

12. References

Goffman, E. (1959). The presentation of self in everyday life. Anchor Books.

Gough, H. G. (1957). California Psychological Inventory manual. Consulting Psychologists Press.

Gough, H. G. (1987). California Psychological Inventory administrator’s guide. Consulting Psychologists Press.

Gough, H. G., & Bradley, P. (1996). CPI manual (3rd ed.). Consulting Psychologists Press.

Lanning, K. (1989). Detection of invalid response patterns on the California Psychological Inventory. Applied Psychological Measurement, 13(1), 45–56. https://doi.org/10.1177/014662168901300105

Megargee, E. I. (1972). The California Psychological Inventory handbook. Jossey-Bass. https://doi.org/10.1037/h0033878

Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46(3), 598–609. https://doi.org/10.1037/0022-3514.46.3.598

Ruch, F. L. (1942). A technique for detecting attempts to fake performance on a self-inventory type of personality test. In Studies in Personality (pp. 229–234). McGraw-Hill.

13. Items of the Scale