Interpersonal Power Inventory (IPI)

1. Abstract

The Interpersonal Power Inventory (IPI) is a multidimensional psychometric instrument developed by Bertram H. Raven, Joseph Schwarzwald, and Meni Koslowsky in 1998. Designed to operationalize Raven’s expanded Power/Interaction Model of Interpersonal Influence, the scale refines the classic five bases of social power originally conceptualized by John R. P. French and Bertram Raven (1959) into an empirically differentiated 11-factor taxonomy. The inventory measures how an influencing agent (such as a supervisor, counselor, or educator) motivates or induces compliance in an influence target (such as an employee, client, or student), or conversely, how targets perceive the power strategies deployed against them.

Comprising 44 items, the IPI bifurcates traditional power bases into distinct interpersonal and structural manifestations: Impersonal Reward, Personal Reward, Impersonal Coercion, Personal Coercion, Legitimate Position, Legitimate Reciprocity, Legitimate Equity, Legitimate Dependence, Expert, Referent, and Information power. Each item represents an explicit rationale or mechanism of compliance. Administered across dual comparative perspectives—the Superiors/Agent Version (measuring self-reported use or perceived efficacy of tactics) and the Subordinates/Target Version (measuring attributed bases of compliance)—the instrument utilizes a 7-point Likert-type response format assessing the likelihood of compliance. Extensive psychometric evaluations demonstrate sound construct validity, robust structural replicability through confirmatory factor analyses, and internal consistency coefficients ($lpha$) typically ranging from .68 to .89 across subscales. The IPI remains a gold standard in organizational behavior, industrial-organizational psychology, clinical supervision, and therapeutic alliance research.

2. Keywords

Interpersonal Power Inventory, social power bases, French and Raven, Power/Interaction Model, interpersonal influence, legitimate power, expert power, referent power, informational power, coercive power, reward power, organizational compliance, psychometrics.

3. Authors

The Interpersonal Power Inventory was conceptualized and validated by three prominent social and organizational psychologists:

• Bertram H. Raven, Ph.D.: Professor Emeritus of Psychology at the University of California, Los Angeles (UCLA). A seminal figure in group dynamics and social psychology, Raven co-authored the foundational 1959 social power taxonomy with John R. P. French Jr. and later authored the comprehensive Power/Interaction Model of Interpersonal Influence.
• Joseph Schwarzwald, Ph.D.: Professor of Social and Organizational Psychology at Bar-Ilan University, Ramat Gan, Israel. His research specialized in interpersonal relations, intergroup conflict, social influence strategies, and compliance behaviors in organizational and educational contexts.
• Meni Koslowsky, Ph.D.: Professor of Psychology at Bar-Ilan University and Ariel University, Israel. A renowned methodologist and organizational psychologist, Koslowsky has published extensively on workplace stress, employee turnover, statistical modeling, and the empirical measurement of organizational power dynamics.

4. Purpose

The fundamental purpose of the Interpersonal Power Inventory is to provide a granular, psychometrically validated methodology for assessing the multidimensional strategies that individuals utilize to exert influence or comply with authority within hierarchical and interpersonal relationships. For several decades following French and Raven’s (1959) original framework, empirical investigations relied heavily on simplistic, single-item measures or crude five-factor operationalizations (Reward, Coercive, Legitimate, Referent, Expert). These earlier instruments suffered from chronic methodological limitations, including low construct reliability, social desirability distortion, conflation between personal and institutional authority, and an inability to distinguish informational persuasion from perceived cognitive expertise.

Raven, Schwarzwald, and Koslowsky (1998) engineered the IPI to resolve these psychometric deficiencies. The instrument serves several distinct theoretical, organizational, and clinical objectives:

• Deconstruction of Coarse Power Categories: The inventory systematically dissociates traditional broad constructs into subtle psychological dimensions. Reward and Coercive powers are partitioned into formal/tangible (impersonal) and relational/affective (personal) mechanisms. Similarly, Legitimate power is divided into four distinct normative obligations: formal position, reciprocity, equity (compensating for past harm), and dependence (the social responsibility norm).
• Investigation of Relational Dyads: The IPI is specifically constructed with parallel phrasing for influencing agents (Superiors/Counselors/Supervisors) and targets of influence (Subordinates/Clients/Employees). This enables researchers to examine perceptual congruence or asymmetry within dyads, identifying how discrepancies in perceived power dynamics affect relationship satisfaction, trust, resistance, and psychological reactance.
• Clinical and Counseling Assessment: Beyond corporate structures, the IPI evaluates clinical and therapeutic dynamics. In supervisory and mental health settings, counselors deploy social influence to foster client adherence to therapeutic interventions, court-mandated treatment programs, and behavioral homework assignments. The instrument permits evaluation of whether therapists rely on coercive/positional authority versus collaborative informational and expert strategies.
• Organizational Diagnostic Application: In industrial-organizational contexts, human resource specialists and organizational development consultants use the IPI to diagnose leadership styles, evaluate management training programs, assess dysfunctional workplace climates, and study employee burnout, turnover intentions, and counterproductive work behaviors.

5. Psychological Construct

The Interpersonal Power Inventory measures eleven distinct sub-constructs derived from Raven’s expanded taxonomy of social power bases. Social power is formally defined as the potential ability of one social actor (the agent, $O$) to influence the beliefs, attitudes, or behaviors of another designated social actor (the target, $P$), utilizing resources controlled or perceived to be controlled by $O$. The eleven dimensions are operationalized as follows:

1. Impersonal Reward Power

This dimension reflects compliance motivated by the target’s desire to obtain tangible, institutional, or administrative benefits directly controlled by the agent. In workplace settings, these incentives include salary increments, promotions, preferable scheduling, or coveted job assignments. In therapeutic and correctional settings, this encompasses favorable evaluations that secure reduced fees or earlier discharge from mandated oversight (e.g., “A good evaluation from me could lead to an earlier release from court-ordered counseling or other benefits”).

2. Personal Reward Power

Unlike bureaucratic rewards, personal reward power operates via affective and relational approval. The target complies because the agent’s warmth, praise, social validation, and personal acceptance serve as reinforcing stimuli (e.g., “It makes the client feel better to know I like him/her”; “My supervisor made me feel more valued when I did as requested”). The motivating resource is interpersonal intimacy and emotional validation.

3. Impersonal Coercion Power

Impersonal coercion involves compliance compelled by fear of tangible, structural, or administrative penalties. Targets conform to prevent the agent from administering sanctions such as demotion, undesirable task assignments, negative formal performance reviews, salary withholding, or prolonged judicial oversight (e.g., “I can give the client more difficult assignments”; “My supervisor could make things unpleasant for me”).

4. Personal Coercion Power

Personal coercion relies on the threat of relational rejection, emotional withdrawal, or interpersonal disapproval. The agent induces compliance by signaling that noncompliance will result in coldness, ostracism, or personal dislike (e.g., “I may be cold and distant if he/she does not do as requested”; “He/she would be upset knowing that he/she was on the bad side of me”). Compliance is driven by anxiety surrounding interpersonal estrangement.

5. Legitimate Position Power

Rooted in structural hierarchies, societal norms, and mutual role expectations, legitimate position power reflects the target’s internal belief that the agent has the formal right to prescribe behavior by virtue of their status or office, and that the target has a corresponding obligation to comply (e.g., “After all, I am the counselor, and the client should feel some obligation to go along”; “It was his/her job to tell me how to do my work”).

6. Legitimate Reciprocity Power

Grounded in the universal social norm of reciprocity (Alvin Gouldner, 1960), this construct captures compliance elicited because the agent has previously performed favors, concessions, or positive deeds for the target. The target complies out of a perceived psychological indebtedness (e.g., “I have done some nice things for him/her in the past and so he/she does this in return”).

7. Legitimate Equity Power

Based on equity theory (J. Stacy Adams, 1965) and compensatory morality, legitimate equity represents an obligation to comply to compensate for past mistakes, inconvenience, or distress the target has caused the agent (e.g., “By doing so, he/she can make up for some difficulties he/she may have caused in the past”; “He/she has made some mistakes and therefore feels that he/she owes this to me”).

8. Legitimate Dependence Power

Reflecting the cultural “norm of social responsibility” (Leonard Berkowitz, 1972), this base compels compliance when the target recognizes that the agent is in a state of vulnerability or reliance. The target assists because the agent explicitly needs help and cannot achieve their objective without cooperation (e.g., “It is clear that I really depend on the client to do this for me”; “Unless I did so, his/her job would be more difficult”).

9. Expert Power

Expert power is driven by the target’s perception that the agent possesses superior knowledge, cognitive competence, credentials, or specialized insight regarding the task at hand. Compliance occurs through cognitive deference and trust in the agent’s professional judgment (e.g., “He/she feels I probably know more about this particular situation”; “My supervisor probably had more technical knowledge about this than I did”).

10. Referent Power

Referent power stems from the target’s identification with, admiration for, or desire to emulate the agent. The agent serves as a psychological model or reference group. The target adopts attitudes or executes behaviors to foster a sense of psychological oneness and shared identity (e.g., “He/she sees me as someone he/she can identify with”; “He/she looks up to me and generally models his/her behavior accordingly”).

11. Informational Power

Crucially distinguished from expert power, informational power is entirely cognitive and objective. Influence is achieved through persuasive communication, logical reasoning, and factual demonstration. The change in the target is independent of the agent’s persona; once the rationale or underlying logic is communicated, the target recognizes the inherent merit of the action itself (e.g., “Once I point it out, he/she can see why the change is necessary”; “I gave the client good reasons for changing how he/she handled the situation”).

6. Theoretical Framework

The conceptual foundation of the Interpersonal Power Inventory rests upon Bertram Raven’s Power/Interaction Model of Interpersonal Influence (Raven, 1992, 1993, 2008), which represents an extensive theoretical maturation of the classic French and Raven (1959) taxonomy. The original 1959 formulation postulated five discrete power bases: Reward, Coercive, Legitimate, Referent, and Expert. In 1965, Raven added Informational power to differentiate between influence derived from cognitive processing of content (persuasion) versus influence rooted in attribution of competence to the source (expertise).

During the 1990s, Raven synthesized social psychological research on social norms, social exchange theory, and cognitive dissonance into an overarching systemic model. The Power/Interaction Model does not treat influence as a static event; rather, it formalizes power as a cyclical, multi-stage interactional process consisting of the following sequential components:

1. Motivation to Influence: The agent experiences an internal drive or external task requirement to alter the target’s behavior, beliefs, or emotional states.
2. Assessment of Available Power Bases: The agent evaluates the repertoire of power bases potentially at their disposal. This assessment is conditioned by structural position, personal traits, organizational culture, and relational history.
3. Evaluation of Costs and Constraints: Prior to strategy selection, the agent calculates psychological, relational, and material costs. Coercive tactics require intensive surveillance and risk inducing hostility or psychological reactance; positional power requires mutual role validation; personal rewards demand emotional availability.
4. Implementation of Influence Strategy: The agent selects and operationalizes specific soft (e.g., informational, expert, personal reward) or harsh (e.g., impersonal coercion, legitimate position) power bases.
5. Target’s Perceptions and Cognitive Appraisals: The target evaluates the agent’s behavior, deciphering the motive, perceived legitimacy, and underlying relational contracts.
6. Behavioral Outcome and Feedback Loop: Compliance or noncompliance occurs, generating post-influence consequences. The outcome recalibrates future power dynamics, perceptions of the agent’s credibility, and the long-term relational climate.

A primary theoretical innovation of Raven’s 1998 model was the formal decomposition of the original broad categories into secondary and tertiary mechanisms. French and Raven had conceptualized Legitimate power as a unitary construct of authority. Raven observed that human obligation is deeply variegated: people comply not only because an individual holds a rank (Legitimate Position), but because universal moral norms compel fairness (Legitimate Equity), return of favors (Legitimate Reciprocity), and protection of the helpless (Legitimate Dependence). Furthermore, the model partitioned Reward and Coercion into impersonal (formal, organizational, resource-based) and personal (relational, love, rejection, esteem) dimensions, recognizing that interpersonal power operates across both institutional and socio-emotional currencies.

7. Validity

The construct, convergent, discriminant, and predictive validity of the Interpersonal Power Inventory has been confirmed across diverse organizational, clinical, educational, and cross-cultural investigations.

Construct and Factorial Validity

Initial validation by Raven, Schwarzwald, and Koslowsky (1998) utilized structural equation modeling (SEM) and confirmatory factor analysis (CFA) across diverse organizational cohorts. When contrasting the traditional five-factor model against the hypothesized eleven-factor structure, the eleven-factor formulation consistently yielded superior fit indices. The empirical separation between Informational power and Expert power was substantiated; while correlated ($r \approx .45$ to $.60$), they loaded onto orthogonal latent dimensions, demonstrating that logical cognitive persuasion is psychometrically distinguishable from reliance on an agent’s technical reputation.

Convergent and Discriminant Validity

Convergent validity has been established through strong correlations with parallel measures of managerial influence tactics, such as David Kipnis‘s Profile of Organizational Influence Strategies (POIS) and Gary Yukl’s Influence Behavior Questionnaire (IBQ). Research indicates:

• Informational Power strongly correlates with Yukl’s “Rational Persuasion” ($r = .68, p < .001$).
• Personal Coercion and Impersonal Coercion align robustly with “Pressure” and “Assertiveness” scales ($r = .58$ to $.64$).
• Legitimate Reciprocity exhibits high convergence with “Exchange” tactics ($r = .62$).
• Discriminant validity is evidenced by near-zero or weak negative correlations between Referent Power and Impersonal Coercion ($r = -.15$ to $-.28$), confirming that relational identification and punitive threats represent distinct psychological orientations.

Predictive and Criterion-Related Validity

In empirical outcome studies, the subscales of the IPI demonstrate differential predictive utility across critical organizational and clinical metrics:

• Subordinate Job Satisfaction and Organizational Commitment: In a seminal study of hospital staff and industrial managers by Koslowsky, Schwarzwald, and Ashuri (2001), personal power bases (Expert, Referent, Informational, and Personal Reward) were robust positive predictors of employee job satisfaction ($R^2 = .38$) and affective organizational commitment. Harsh bases (Impersonal Coercion, Legitimate Position) consistently predicted turnover intentions and organizational cynicism.
• Therapeutic Alliance in Counseling: In clinical settings evaluated by Anderson (2008), counselors’ self-reported reliance on Informational, Expert, and Personal Reward power correlated positively with client-rated Working Alliance Inventory (WAI) scores. Conversely, reliance on Impersonal Coercion and Legitimate Position in court-ordered rehabilitation counseling predicted elevated client resistance and premature therapeutic termination.

8. Reliability

The Interpersonal Power Inventory demonstrates sound internal consistency and temporal stability across international and occupational samples.

Internal Consistency

Cronbach’s alpha ($lpha$) coefficients documented in psychometric evaluations consistently meet or exceed acceptable standards for multidimensional personality and organizational scales:

• Impersonal Reward: $\alpha = .78 – .85$ (Employee version), $\alpha = .74 – .81$ (Counselor/Superior version)
• Personal Reward: $\alpha = .79 – .86$
• Impersonal Coercion: $\alpha = .77 – .84$
• Personal Coercion: $\alpha = .80 – .88$
• Legitimate Position: $\alpha = .72 – .82$
• Legitimate Reciprocity: $\alpha = .75 – .83$
• Legitimate Equity: $\alpha = .71 – .79$
• Legitimate Dependence: $\alpha = .68 – .77$
• Expert Power: $\alpha = .82 – .89$
• Referent Power: $\alpha = .81 – .88$
• Informational Power: $\alpha = .76 – .84$

While the Legitimate Dependence subscale occasionally hovers near the lower threshold of internal consistency ($lpha \approx .68$ to $.72$) due to the complexity of the social responsibility norm across varying corporate cultures, all other subscales consistently exhibit strong reliability ($lpha > .75$).

Test-Retest Stability

Stability over time has been evaluated across 4- to 8-week test-retest intervals within stable supervisor-subordinate dyads. Pearson correlation coefficients ($r_{tt}$) range between .71 and .84, indicating that perceived power usage patterns remain stable in established relationships, while maintaining sensitivity to changes resulting from targeted leadership intervention programs.

9. Factor Analysis

The latent structure of the 44-item Interpersonal Power Inventory has been thoroughly evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

During the initial development phase, Raven, Schwarzwald, and Koslowsky (1998) subjected the 44 items to principal axis factoring with oblique (Promax and Oblimin) rotations, accommodating theoretical intercorrelations among social influence tactics. Scree plot visual inspections, parallel analysis, and Kaiser’s criterion (eigenvalues $> 1.0$) converged on an eleven-factor solution, accounting for approximately 62% to 68% of the total variance across workplace and counseling samples. All 44 items loaded primarily onto their targeted conceptual factor, with primary factor loadings exceeding .50 and cross-loadings remaining low ($< .30$).

Confirmatory Factor Analysis (CFA)

Subsequent psychometric studies tested competing structural models to evaluate whether the 11-factor model provided superior empirical fit compared to simpler hierarchical models:

• Model 1 (Single-Factor Model): Positing that all power represents a global construct of interpersonal dominance. Fit was poor: $\chi^2/df = 5.82$, Comparative Fit Index (CFI) = .58, Root Mean Square Error of Approximation (RMSEA) = .114.
• Model 2 (Classic Five-Factor Model): Collapsing items into the original French and Raven (1959) categories (Reward, Coercion, Legitimate, Expert, Referent). Fit was modest but inadequate: $\chi^2/df = 3.41$, CFI = .76, RMSEA = .082.
• Model 3 (Hypothesized Eleven-Factor Model): Separating reward and coercion into personal/impersonal forms, legitimate power into four normative dimensions, and separating informational from expert power. This model yielded superior psychometric fit: $\chi^2/df = 1.68 – 1.94$, CFI = .93 – .95, Tucker-Lewis Index (TLI) = .92 – .94, and RMSEA = .039 – .048 (90% CI [.034, .046]), with Standardized Root Mean Square Residual (SRMR) = .044.

Higher-Order Structural Groupings

Second-order factor analyses consistently reveal two overarching meta-clusters of power tactics, frequently conceptualized in the literature as “Soft” versus “Harsh” Power Bases:

• Soft Power Cluster: Informational, Expert, Referent, Personal Reward, and Legitimate Dependence. These bases allow the target latitude in compliance, foster personal autonomy, and do not rely on institutional surveillance.
• Harsh Power Cluster: Impersonal Coercion, Personal Coercion, Impersonal Reward, and Legitimate Position. These bases constrain autonomy, emphasize status asymmetry, and require external surveillance to ensure sustained compliance.

10. Instrument / Measurement Tool

The structural characteristics, administration protocols, and scoring procedures of the Interpersonal Power Inventory are outlined below:

• Instrument Type: Standardized multi-dimensional self-report behavioral questionnaire (parallel forms available for influencing agents/superiors/counselors and targets/subordinates/clients).
• Administration Format: Paper-and-pencil questionnaire or computerized psychological assessment.
• Administration Time: Approximately 15 to 20 minutes.
• Item Count: 44 items, distributed symmetrically across 11 subscales (exactly 4 items per subscale).
• Response Scale: 7-point Likert-type scale assessing the target’s likelihood of compliance or the perceived efficacy of the rationale:
• 1 = Much less likely to comply
• 2 = Less likely to comply
• 3 = A bit less likely to comply
• 4 = Would not affect their tendency to comply
• 5 = A bit more likely to comply
• 6 = More likely to comply
• 7 = Much more likely to comply
• Subscale Item Allocation (Superiors / Counselor Version):
  • Impersonal Reward: Items 1, 22, 32, 41
  • Impersonal Coercion: Items 6, 13, 31, 39
  • Personal Reward: Items 8, 14, 29, 34
  • Personal Coercion: Items 10, 18, 23, 44
  • Legitimate Position: Items 2, 20, 28, 35
  • Legitimate Reciprocity: Items 7, 12, 33, 43
  • Legitimate Equity: Items 11, 21, 30, 37
  • Legitimate Dependence: Items 9, 16, 25, 40
  • Expert Power: Items 3, 19, 26, 38
  • Referent Power: Items 5, 15, 27, 36
  • Informational Power: Items 4, 17, 24, 42
• Scoring Instructions: Subscale scores are calculated by computing the mean or sum of the four items corresponding to each power base. Higher scores indicate a greater perceived likelihood of compliance or heavier reliance upon that specific social power base. Higher-order “Soft Power” and “Harsh Power” indices can be generated by averaging their respective constituent subscale means.

11. Permissions & Fee and Test Year

The Interpersonal Power Inventory was published by Bertram H. Raven, Joseph Schwarzwald, and Meni Koslowsky in 1998 in the Journal of Applied Social Psychology. The instrument was developed within academic research institutions (University of California, Los Angeles, and Bar-Ilan University).

The IPI is considered an open-access psychometric instrument for non-commercial academic, educational, and clinical research purposes. No licensing fees are required for academic investigators, master’s theses, or doctoral dissertations, provided that proper formal academic attribution is given to the authors and original publication. For commercial deployments, large-scale corporate consultations, or inclusion within proprietary psychological testing software, formal permission should be sought from the copyright holders and publishers (Wiley Periodicals / Journal of Applied Social Psychology).

12. References

• Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 2, pp. 267–299). Academic Press. https://doi.org/10.1016/S0065-2601(08)60108-2
• Anderson, R. S. (2008). Counselor gender self-confidence and social influence in counseling: Counselor perceptions of the therapeutic alliance (Doctoral dissertation, Ohio University). OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1212765389
• Berkowitz, L. (1972). Social norms, feelings, and other factors affecting helping and altruism. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 6, pp. 63–108). Academic Press. https://doi.org/10.1016/S0065-2601(08)60025-8
• French, J. R. P., & Raven, B. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in Social Power (pp. 150–167). University of Michigan Press.
• Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25(2), 161–178. https://doi.org/10.2307/2092623
• Koslowsky, M., Schwarzwald, J., & Ashuri, S. (2001). On the relationship between subordinates’ compliance to power sources and organizational attitudes. Journal of Applied Social Psychology, 31(3), 455–476. https://doi.org/10.1111/j.1559-1816.2001.tb02050.x
• Raven, B. H. (1965). Social influence and power. In I. D. Steiner & M. Fishbein (Eds.), Current Studies in Social Psychology (pp. 371–382). Holt, Rinehart, and Winston.
• Raven, B. H. (1992). A power/interaction model of interpersonal influence: French and Raven thirty years later. Journal of Social Behavior and Personality, 7(2), 217–244.
• Raven, B. H. (1993). The bases of power: Origins and recent developments. Journal of Social Issues, 49(4), 227–251. https://doi.org/10.1111/j.1540-4560.1993.tb01191.x
• Raven, B. H. (2008). The bases of power and the power/interaction model of interpersonal influence. Analyses of Social Issues and Public Policy, 8(1), 1–22. https://doi.org/10.1111/j.1530-2415.2008.00159.x
• Raven, B. H., Schwarzwald, J., & Koslowsky, M. (1998). Conceptualizing and measuring a power/interaction model of interpersonal influence. Journal of Applied Social Psychology, 28(4), 307–332. https://doi.org/10.1111/j.1559-1816.1998.tb01708.x
• Schwarzwald, J., Koslowsky, M., & Ochana-Levin, T. (2001). Usage of and compliance with power bases in formal and informal teacher-student interactions. The Journal of Social Psychology, 141(3), 297–314. https://doi.org/10.1080/00224540109600553

13. Items of the Scale (Questionnaire)