Satisfaction with the Problem’s Resolution

Abstract

The Satisfaction with the Problem’s Resolution scale is a specialized psychometric instrument designed to evaluate an individual’s evaluative assessment of how effectively a grievance, service failure, or technical issue was handled and rectified during an organizational or technological service interaction. Originating in empirical consumer behavior and service research by Crolic, Thomaz, Hadi, and Stephen (2022) in the Journal of Marketing, this three-item, unidimensional scale measures post-interaction sentiment, specifically capturing procedural and outcome-oriented satisfaction following interactions with human or automated conversational agents (e.g., artificial intelligence chatbots). Operating on a 7-point Likert scale ranging from 1 (“Strongly disagree”) to 7 (“Strongly agree”), the instrument yields a composite average score indicating the respondent’s overall evaluative contentment with the issue’s remedy. Psychometric validation across multiple randomized laboratory experiments and field investigations demonstrates outstanding internal consistency reliability (with Cronbach’s alpha and composite reliability coefficients routinely exceeding .90), clear unidimensionality via exploratory and confirmatory factor analyses, and robust convergent, discriminant, and predictive validity relative to downstream customer loyalty, positive word-of-mouth, and reduced brand defection. This article provides a comprehensive academic review of the scale’s theoretical foundation, psychometric properties, scoring procedures, and diverse research and applied applications in contemporary human-computer interaction (HCI) and customer relationship management (CRM).

Keywords

Satisfaction with the Problem’s Resolution, Customer Satisfaction, Service Recovery, Conversational AI, Anthropomorphism, Expectancy-Disconfirmation Theory, Justice Theory, Human-Computer Interaction, Psychometrics, Service Failure

Authors

The scale was formalized and validated by a team of leading consumer psychology and marketing science researchers at the Saïd Business School, University of Oxford:

• Cammy Crolic, Ph.D. — Associate Professor of Marketing, Saïd Business School, University of Oxford. Research focus: Sensory marketing, consumer well-being, customer experience, and human-technology interfaces.
• Felipe Thomaz, Ph.D. — Associate Professor of Marketing, Saïd Business School, University of Oxford. Research focus: Digital marketing networks, structural dynamics of digital ecosystems, and AI-mediated customer journeys.
• Rhonda Hadi, Ph.D. — Associate Professor of Marketing, Saïd Business School, University of Oxford. Research focus: Consumer psychology, affective technology, anthropomorphism, and embodied artificial intelligence.
• Andrew T. Stephen, Ph.D. — L’Oréal Professor of Marketing and Associate Dean of Research, Saïd Business School, University of Oxford. Research focus: The intersection of digital media, artificial intelligence, and marketing strategy.

Purpose

The primary purpose of the Satisfaction with the Problem’s Resolution scale is to provide a concise, psychometrically rigorous, and empirically validated measure of post-encounter outcome satisfaction when an individual seeks redress or assistance for a specific dilemma. In both commercial marketing and organizational psychology, service encounters often begin because of an antecedent failure—a defective product, a billing error, a software bug, or a missed flight. Measuring general customer satisfaction (cumulative satisfaction) fails to isolate whether the focal service recovery intervention succeeded in resolving the specific triggering problem.

The instrument was engineered to resolve critical empirical challenges in contemporary service research:

• Disentangling Process from Outcome: While interactional aspects of service (e.g., politeness, perceived empathy, speed) influence consumer evaluations, the structural remediation of the core issue is the primary driver of subsequent brand equity. The scale captures overall contentment focused strictly on the resolution outcome and procedural effectiveness.
• Contextualizing Human vs. Artificial Intelligence (AI) Interactions: In their foundational work, Crolic et al. (2022) evaluated how consumers respond to automated conversational agents compared to human customer service representatives. When users enter an interaction experiencing high negative affect (e.g., anger), anthropomorphic technological cues (giving a bot a human name, avatar, or empathetic phrasing) can backfire, reducing resolution evaluations because users perceive the agent as insincere or incapable of intentional restitution. The scale accurately tracks these subtle shifts in consumer evaluations across experimental manipulations.
• Minimizing Survey Fatigue in Longitudinal and Multi-Wave Designs: With only three items, the scale provides high measurement precision without placing excessive cognitive load on respondents. This makes it ideal for real-time customer feedback surveys, continuous customer journey monitoring, and multi-stage lab experiments where participant fatigue can compromise data quality.

Psychological Construct

The focal psychological construct measured by this instrument is transaction-specific resolution satisfaction. Unlike global satisfaction, which represents an accumulated, overall evaluation of a brand or ongoing relationship over time, transaction-specific satisfaction reflects an immediate, cognitive-affective post-purchase or post-interaction judgment regarding a single, discrete event.

Within the domain of organizational remediation and service recovery, this construct synthesizes several theoretical sub-dimensions into a unified evaluative judgment:

• Cognitive Appraisal of Issue Handling: Refers to the respondent’s rational evaluation of the operational competence and efficiency deployed to address the problem. This is reflected in item 1 (“I am satisfied with the handling of the issue“), where the user assesses whether the protocols, steps, and interventions executed by the agent met professional or logical standards of problem resolution.
• Perceived Outcome Equity and Goal Attainment: Evaluates whether the primary functional disruption that motivated the interaction was fully mitigated. Refers to the perceived outcome justice—whether the end result (e.g., refund, replacement, clarification, technical patch) successfully restored balance. This core cognitive assessment is captured by item 2 (“The problem was resolved to my satisfaction“).
• Affective Valence and Interaction Contentment: Captures the post-interaction emotional equilibrium. Resolution is rarely an entirely rational calculation; it produces an emotional residue. Item 3 (“I am pleased with the outcome of this interaction“) measures the respondent’s subjective sense of positive or relieved affect derived from the conclusion of the event.

Together, these facets reflect a unified, unidimensional construct: the degree to which an individual perceives that an anomalous or disruptive situation has been satisfactorily closed, returning the user to a baseline psychological and functional equilibrium.

Theoretical Framework

The scale is rooted in two foundational theories in social psychology and consumer behavior: Expectancy-Disconfirmation Theory and Justice Theory, supplemented by contemporary models of Anthropomorphism and Mind Perception.

1. Expectancy-Disconfirmation Theory (EDT)

Formulated by Richard L. Oliver (1980), EDT posits that customer satisfaction is a function of pre-encounter expectations compared against perceived actual performance. When actual performance surpasses expectations, positive disconfirmation occurs, leading to high satisfaction. Conversely, when performance falls short, negative disconfirmation results in dissatisfaction.

In the context of the Satisfaction with the Problem’s Resolution scale, consumers entering a service recovery touchpoint hold baseline expectations regarding the agent’s capability to fix the error. The scale measures the psychological output of the disconfirmation process specifically tied to the service outcome: did the resolution exceed, match, or fall short of the standard the consumer expected?

2. Justice Theory (Distributive, Procedural, and Interactional Justice)

In organizational psychology and service recovery frameworks (Tax, Brown, & Chandrashekaran, 1998), customers evaluate a dispute through three primary justice dimensions:

• Distributive Justice: Focuses on the fairness of the tangible outcome (e.g., compensation, repair, return).
• Procedural Justice: Focuses on the perceived fairness of the policies, timing, and structural processes used to arrive at the resolution.
• Interactional Justice: Focuses on the interpersonal treatment received during the communication (e.g., politeness, empathy, respect).

The Satisfaction with the Problem’s Resolution scale integrates these dimensions by measuring satisfaction with both the “handling” (procedural/interactional elements) and the ultimate “outcome” and “resolved” state (distributive elements), capturing the holistic perception of justice achieved during the encounter.

3. Mind Perception and Anthropomorphism in AI Recovery

Crolic et al. (2022) extended these frameworks by applying the theory of Mind Perception (Gray, Gray, & Wegner, 2007) and anthropomorphism models (Epley, Waytz, & Cacioppo, 2007). Mind perception divides human cognition into two dimensions: agency (the capacity for intentionality, planning, and goal execution) and experience (the capacity for feeling, emotion, and vulnerability).

When an automated agent is anthropomorphized, consumers intuitively expect both agency and emotional capacity. If an anthropomorphic chatbot fails to deliver adequate problem resolution, or if an angry consumer encounters humanlike cues that appear disingenuous, the consumer assigns blameworthiness to the agent, driving down resolution satisfaction scores. This psychometric scale serves as the primary dependent or mediating variable used to quantify this psychological dynamic.

Validity

The scale demonstrates robust psychometric validity across multiple independent experimental studies and field replications:

1. Construct and Factorial Validity

Construct validity has been established across diverse experimental samples (e.g., Prolific Academic panels, university laboratory subjects, and commercial online panels). Confirmatory Factor Analysis (CFA) shows that all three items load strongly on a single latent construct (standardized loadings $lambda > .88$). Goodness-of-fit statistics across iterations demonstrate excellent fit to the unidimensional model (e.g., $\chi^2/df < 2.5$, $CFI > .99$, $TLI > .98$, $RMSEA < .05$, $SRMR < .02$).

2. Convergent Validity

The scale achieves convergent validity as demonstrated by high Average Variance Extracted (AVE). In empirical testing, AVE consistently exceeds .80, well above the standard psychometric threshold of .50 (Fornell & Larcker, 1981). Furthermore, the scale correlates strongly with related constructs, including:

• Overall Service Evaluation ($r = .72$ to $.84, p < .001$)
• Net Promoter / Word-of-Mouth Intentions ($r = .65$ to $.78, p < .001$)
• Perceived Agent Competence ($r = .60$ to $.75, p < .001$)

3. Discriminant Validity

Using the Fornell-Larcker criterion, the square root of the AVE for the Satisfaction with the Problem’s Resolution scale (typically $\sqrt{AVE} > .90$) exceeds the inter-construct correlations between resolution satisfaction and other distinct variables, such as pre-interaction anger, perceived anthropomorphism, or general brand attitude. Heterotrait-Monotrait (HTMT) ratios of correlations systematically fall below the conservative threshold of .85, confirming that the scale is empirically distinct from general brand evaluation and affective venting.

4. Predictive and Criterion-Related Validity

The instrument exhibits strong predictive validity regarding behavioral intentions. In experimental designs examining chatbot interactions, resolution satisfaction directly predicted downstream repurchase intention ($b = .48, p < .001$) and significantly mediated the negative impact of consumer anger on overall brand loyalty. When users reported low satisfaction on this scale, their probability of switching to a competitor increased by over 40%.

Reliability

The scale exhibits high internal consistency reliability across varied empirical contexts, experimental conditions, and target demographics:

• Cronbach’s Alpha ($\alpha$): In the initial validation studies by Crolic et al. (2022), the scale consistently yielded Cronbach’s alpha coefficients between $.92$ and $.96$ across multiple experimental studies (e.g., Study 1, Study 2, and Study 3). These values indicate strong internal consistency without redundancy.
• Composite Reliability (CR): Structural equation modeling confirms composite reliability values exceeding $.93$, well above the accepted $.70$ benchmark.
• McDonald’s Omega ($\omega$): Omega hierarchical and total omega coefficients similarly fall in the $.92$ to $.95$ range, reinforcing that item inter-correlations reflect a single underlying latent variable.
• Test-Retest Stability: In short-interval repeated measurement settings (where resolution evaluations are measured immediately following the interaction and reassessed 24 hours later), the instrument demonstrates high temporal stability ($r_{tt} = .84, p < .001$), provided no new service interventions took place in the interim.

Factor Analysis

Both exploratory and confirmatory factor analyses confirm that the Satisfaction with the Problem’s Resolution scale operates as a strictly unidimensional instrument.

Exploratory Factor Analysis (EFA)

In exploratory factor analysis using Principal Axis Factoring (PAF) or Maximum Likelihood (ML) extraction with oblique rotation (Promax):

• Only one eigenvalue exceeds the Kaiser criterion ($Eigenvalue > 1.0$), with the primary factor yielding an eigenvalue typically between $2.55$ and $2.75$.
• This primary single factor accounts for between 82% and 91% of the total item variance.
• Scree plot visual inspection exhibits a sharp, unambiguous drop-off after the first factor, confirming the absence of secondary dimensions.

Confirmatory Factor Analysis (CFA)

Standardized factor loadings ($lambda$) from structural equation models across published laboratory data are uniformly high:

• Item 1 (“I am satisfied with the handling of the issue“): $lambda = .89 – .94$
• Item 2 (“The problem was resolved to my satisfaction“): $lambda = .91 – .96$
• Item 3 (“I am pleased with the outcome of this interaction“): $lambda = .88 – .93$

Because a three-item CFA model with one latent factor has exactly zero degrees of freedom ($df = 0$), it represents a just-identified (saturated) model. When embedded within larger structural models including antecedent (e.g., agent anthropomorphism, user emotional state) and consequence (e.g., brand repurchase, customer lifetime value) latent variables, the measurement model demonstrates exceptional fit (e.g., $\chi^2/df = 1.42$, $CFI = .994$, $TLI = .991$, $RMSEA = .036$, $SRMR = .018$).

Instrument / Measurement Tool

The structural details, administration parameters, and scoring conventions for the Satisfaction with the Problem’s Resolution scale are detailed below:

• Instrument Type: Self-report psychometric rating scale / Post-service survey instrument.
• Target Population: Consumers, end-users, or organizational stakeholders who have engaged in a problem-resolution touchpoint (via digital chatbot, automated IVR, live chat, email, or human representative).
• Total Number of Items: 3 items.
• Response Scale: 7-point Likert scale:
  • 1 = Strongly disagree
  • 2 = Disagree
  • 3 = Somewhat disagree
  • 4 = Neither agree nor disagree
  • 5 = Somewhat agree
  • 6 = Agree
  • 7 = Strongly agree
• Scoring and Indexing:
  • No reverse-scored items; all three items are framed positively.
  • To compute the final score, calculate the arithmetic mean of all three items: $$\text{Resolution Satisfaction} = \frac{\text{Item}_1 + \text{Item}_2 + \text{Item}_3}{3}$$
  • Higher scores (closer to 7.00) reflect complete satisfaction and positive closure. Lower scores (closer to 1.00) signify severe service recovery failure.
• Administration Time: Under 60 seconds.

Permissions & Fee and Test Year

The scale was developed and published in 2022 in the following peer-reviewed article:

Crolic, C., Thomaz, F., Hadi, R., & Stephen, A. T. (2022). Blame the Bot: Anthropomorphism and Anger in Customer–Chatbot Interactions. Journal of Marketing, 86(1), 132–148.

The article was published by the American Marketing Association (AMA) via SAGE Publications. The instrument items are accessible in the published academic literature for scholarly, non-commercial research, academic replication, and educational use under fair dealing and fair use conventions. Commercial deployments, integration into proprietary diagnostic software platforms, or large-scale enterprise customer experience tracking suites should consult AMA/SAGE copyright permissions policies or the authors for institutional licensing.

References

• Crolic, C., Thomaz, F., Hadi, R., & Stephen, A. T. (2022). Blame the bot: Anthropomorphism and anger in customer–chatbot interactions. Journal of Marketing, 86(1), 132–148. https://doi.org/10.1177/00222429211045687
• Epley, N., Waytz, A., & Cacioppo, J. T. (2007). On seeing human: A three-factor theory of anthropomorphism. Psychological Review, 114(4), 864–886. https://doi.org/10.1037/0033-295X.114.4.864
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of mind perception. Science, 315(5812), 619–619. https://doi.org/10.1126/science.1134475
• Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460–469. https://doi.org/10.1177/002224378001700405
• Tax, S. S., Brown, S. W., & Chandrashekaran, M. (1998). Customer evaluations of service complaint experiences: Implications for relationship marketing. Journal of Marketing, 62(2), 60–76. https://doi.org/10.1177/002224299806200205

Items of the Scale