Thomas-Kilmann Conflict Mode Instrument (TKI)
Kenneth Thomas and Ralph Kilmann (1974) developed the TKI based on the premise that conflict resolution reflects two behavioral dimensions: assertiveness (degree to which person pursues own concerns) and cooperativeness (degree to which person pursues others' concerns). This 2×2 matrix yields five modes: (1) Competing (high assertiveness, low cooperativeness)—forcing one's preferred outcome, disregarding others' concerns, appropriate for urgent decisions or when one party has clearly superior information.
(2) Accommodating (low assertiveness, high cooperativeness)—yielding to others' preferences, supporting group cohesion, used when issue is less important to oneself. (3) Avoiding (low on both)—postponing confrontation, appropriate when emotions are high or issue is trivial.
(4) Compromising (moderate on both)—mutual concessions, splitting differences, used when parties have equal power and time pressure exists. (5) Collaborating (high on both)—integrating concerns through joint problem-solving, seeking win-win solutions, most resource-intensive but produces highest satisfaction when feasible.
The TKI is a 30-item questionnaire (15 paired scenarios) administered in ~10 minutes, with demonstrated reliability (Cronbach's α=0 73-0 82 per dimension) and test-retest stability (r=0
80 over 4 weeks). The instrument has been translated into 20+ languages and is validated across cultures, though cultural differences exist in default styles (Thomas & Kilmann 2008).
Effectiveness Predictions by Context
No single conflict style is universally best; effectiveness depends on context. Meta-analysis (Cai & Fink 2002, N=60+ studies) shows: Competing produces fastest decisions (appropriate for emergencies, hierarchical organizations) but high resentment (employee satisfaction r=-0
38). Collaborating produces highest satisfaction (r=0 52) and best long-term outcomes but requires time and mutual goodwill. Compromising shows moderate effectiveness (r=0 15-0 25) but higher than avoiding.
Avoiding increases unresolved tensions and rumination. Accommodating works well for relationship maintenance (r=0 38 for liking) but can enable poor decisions. Longitudinal study by Gross & John (2003) found people using collaborative conflict resolution show lower depression and anxiety, higher relationship satisfaction at 1-year follow-up.
However, habitual competing style predicts cardiovascular disease (Dembroski et al. 1989), consistent with sustained stress activation.
Blake-Mouton Managerial Grid
Robert Blake and Jane Mouton (1964) developed the Managerial Grid, applying the assertiveness-cooperativeness framework to leadership. The grid maps 'concern for production' (task focus, x-axis) and 'concern for people' (relationship focus, y-axis), yielding five leadership styles at 9×9 positions: (1) Impoverished (1,1)—minimal effort on task or people, delegative to the point of absenteeism.
(2) Produce-or-perish (9,1)—high task focus, low relationship focus, authoritarian, efficiency-driven but demoralizing. (3) Middle-of-the-road (5,5)—moderate balance, conventional management, adequate but uninspiring.
(4) Country club (1,9)—high relationship focus, low task focus, leader attempts to be liked, productivity suffers. (5) Team management (9,9)—high on both dimensions, integrating task and relationship concerns, leader simultaneously pursues excellence and team commitment.
Blake & Mouton proposed the 9,9 style as ideal, leading to higher performance and satisfaction. Research partially supports this: 9,9 leaders show higher team performance (d=0 65) and employee satisfaction (r=0
48), but 9,1 leaders can achieve comparable performance in short-term, high-pressure contexts with lower satisfaction costs (Northouse 2016). The Managerial Grid saw broad adoption in organizational development training, though later research showed cultural context matters: 9,1 style is more culturally syntonic in high power-distance cultures, whereas 9,9 is preferred in low power-distance cultures (Hofstede & McCrae 2004).
Rahim's Organizational Conflict Inventory (ROCI-II)
Ali Rahim (1983) extended the conflict styles model with the ROCI-II, measuring concern for self and concern for others, identical to Thomas-Kilmann but with expanded organizational validation. Rahim's key contribution is demonstrating that integrating (collaborative) conflict resolution predicts organizational effectiveness (measured by productivity, innovation, employee retention) across hierarchical levels and industry types.
Meta-analysis of ROCI-II across 30 organizational studies (Rahim & Magner 1995) showed: Integrating style correlates with subordinate-supervisor satisfaction (r=0 42), trust (r=0 48), and job performance (r=0
35). Avoiding style predicts job dissatisfaction (r=-0 36) and turnover intention (r=0 38). Compromising is context-dependent: effective when resources are limited but reduces innovation (r=-0
25 with new ideas). Dominating (competing) style, when paired with expert power (leader has relevant expertise), shows positive outcomes (r=0 38), but with coercive power (punishment threats), leads to resentment (r=-0
42). Accommodating by leaders predicts employee liking (r=0 35) but not performance. Rahim's organizational research influenced HR practices; companies using integrative conflict resolution training show 25% reduction in litigation, 18% improvement in retention (Constantino & Merchant 1996).
Cultural and Gender Differences
Cross-cultural research (Cai & Fink 2002) shows significant variation in default conflict styles. High power-distance cultures (Malaysia, Mexico, India) show higher competing/accommodating; low power-distance cultures (Denmark, Australia, Canada) show higher collaborating.
Collectivist cultures (China, Japan) show higher avoiding initially (to preserve group harmony) before potentially escalating; individualist cultures show earlier direct engagement. Gender differences are modest but consistent: women on average show slightly higher accommodating (+0
25 SD on scale) and lower competing (-0 15 SD), attributed to socialization rather than intrinsic difference (Brewer et al. 2003). However, context matters: women in leadership positions show competing frequencies approaching male levels (Jarboe & Witteman 1996).
Longitudinal interventions teaching collaborative conflict styles to adults show sustained behavior change at 3-month follow-up (d=0 48), with greatest gains for those starting with avoiding/accommodating styles (Merrill & Merrill 2001).