Purchase Likelihood from Online Retailer

1. Abstract

The Purchase Likelihood from Online Retailer (PLOR) scale is a specialized psychometric instrument engineered to evaluate a consumer’s stated subjective probability of acquiring a designated product or brand from an identified electronic commerce platform, explicitly conditioned on the consumer being actively in the market for that product category. Originally operationalized and validated by Peter R. Darke, Michael K. Brady, Ray L. Benedicktus, and Andrea E. Wilson (2016) in their empirical investigations on psychological distance and consumer trust, the instrument draws its core conceptual lineage from the foundational behavioral intentions framework established by Zeithaml, Berry, and Parasuraman (1996). The PLOR addresses a crucial methodological challenge in consumer behavior research: separating general brand affinity or hypothetical interest from actionable purchase intent directed toward an unfamiliar or potentially risky digital vendor. Structurally, the instrument employs a modular, fill-in-the-blank multi-item configuration wherein researchers specify the exact target merchandise and retailer, using 7-point or 9-point semantically anchored Likert and semantic differential response formats (e.g., highly unlikely to highly likely, improbable to probable). Across three controlled experimental studies involving undergraduate business cohorts navigating simulated e-commerce environments, the scale demonstrated exceptional psychometric integrity. Internal consistency reliabilities systematically exceeded accepted academic standards, with Cronbach’s alpha coefficients consistently falling between .91 and .96 across diverse experimental manipulations of spatial, social, and temporal distance. Factor analytic assessments confirm a robust, unidimensional latent construct that accounts for substantial variance in downstream transactional intentions while demonstrating pronounced discriminant validity against generalized institutional trust, perceived risk, and merchant reputation. Consequently, the PLOR represents a rigorous, highly adaptable measurement standard for digital marketing researchers, experimental psychologists, and consumer behaviorists modeling transaction likelihood in digital environments.

2. Keywords

Purchase Likelihood, Online Retailing, Psychological Distance, Construal Level Theory, Consumer Distrust, E-commerce Behavior, Behavioral Intentions, Psychometrics, Perceived Risk, Online Trust

3. Authors

The Purchase Likelihood from Online Retailer measurement model was operationalized and published by an academic research team specializing in marketing, retail strategy, and consumer psychology:

• Peter R. Darke, Ph.D. — Professor of Marketing at the Telfer School of Management, University of Ottawa, Ottawa, Ontario, Canada. Dr. Darke’s research program centers on consumer psychology, consumer trust, deceptive marketing practices, and defensive processing.
• Michael K. Brady, Ph.D. — Bob Sasser Professor and Chair of the Department of Marketing at the College of Business, Florida State University, Tallahassee, Florida, United States. His scholarship focuses on customer service encounters, service failure and recovery, and digital retail environments.
• Ray L. Benedicktus, Ph.D. — Associate Professor of Marketing at the College of Business Administration, California State University, Stanislaus, Turlock, California, United States. His empirical work investigates retail consensus cues, trust building in virtual environments, and physical-virtual retail synergy.
• Andrea E. Wilson, Ph.D. — Former doctoral researcher and marketing scholar associated with the Department of Marketing, Florida State University, Tallahassee, Florida, United States, whose investigations center on consumer inferences, retail signaling, and online shopping environments.

4. Purpose

The primary purpose of the Purchase Likelihood from Online Retailer (PLOR) instrument is to capture, isolate, and quantify a consumer’s prospective behavioral commitment toward consummating a financial transaction with a specified online merchant. In electronic commerce, consumers frequently browse digital catalogs without genuine intent to purchase, or they may harbor high enthusiasm for a product category while actively refusing to transact with an unfamiliar, unvetted, or psychologically distant online vendor. The PLOR was developed to eliminate these pervasive confounds by embedding two strict boundary conditions directly into the measurement architecture: first, the intention is explicitly conditional upon the respondent being actively “in the market” for the focal product; second, the purchase probability is tied directly to the particular vendor under investigation rather than the product category at large.

From an applied research perspective, the PLOR serves as a critical terminal dependent variable in experimental designs evaluating digital interventions. Specifically, within consumer psychology, researchers investigate how varied design factors—such as physical store presence, localized domain extensions, customer review aggregators, live chat agents, and security badges—shift consumer willingness to overcome natural defensive skepticism. Because consumers routinely experience heightened vulnerability when sharing payment credentials and personal shipping data with unknown web vendors, measuring general favorability or cognitive attitudes yields insufficient predictive validity regarding actual economic actions. The PLOR bridges this gap by operationalizing purchase likelihood as a conative construct that captures the net output of perceived risk, epistemic trust, and product attractiveness.

In academic literature, the scale is deployed across experimental marketing, behavioral economics, and human-computer interaction (HCI). It allows scholars to model boundary conditions and mediational pathways in Technology Acceptance Models, the Theory of Planned Behavior, and Construal Level Theory. By quantifying stated purchase likelihood under controlled manipulations of vendor characteristics, the PLOR provides investigators with a reliable benchmark to compare the relative efficacy of risk-mitigation strategies across varied retail platforms.

5. Psychological Construct

The underlying psychological construct measured by the PLOR is vendor-specific conditional behavioral purchase intention. Within structural models of attitude and behavior, intention resides as the direct cognitive and motivational antecedent to overt physical behavior. Grounded in the conceptual taxonomies established by Fishbein and Ajzen (1975) and refined for commercial environments by Zeithaml, Berry, and Parasuraman (1996), purchase intention represents an individual’s conscious, subjective plan of action regarding whether to allocate economic capital to a designated provider.

The PLOR operationalizes this construct across three interrelated dimensions of conative judgment:

• Subjective Probability: The perceived likelihood or statistical expectancy that the individual will select the specified online merchant over competing transactional avenues when the situational need arises. This dimension captures the cognitive appraisal of the choice outcome (e.g., assessing the likelihood as “improbable” versus “probable”).
• Willingness to Commit: The psychological readiness to enter into a binding contractual and financial exchange with the merchant, reflecting the abatement of defensive hesitation, cynicism, or perceived transactional risk.
• Comparative Channel Selection: The propensity to direct capital specifically to the focal online retailer rather than defecting to well-established marketplace intermediaries, brick-and-mortar competitors, or established legacy vendors.

What distinguishes the PLOR construct from generic measures of purchase intention is its structural conditioning. In standard retail environments, asking a consumer “How likely are you to purchase a laptop?” confounds their underlying budgetary cycle, current hardware ownership, and product desirability with their evaluation of the retail channel. The PLOR circumvents this confounding variance by establishing a hypothetical anchor: the respondent is instructed to assume they are already seeking to acquire the designated merchandise. Consequently, all observed variance in the resulting scores reflects the consumer’s psychological disposition toward the retailer itself—capturing the retailer’s perceived reliability, credibility, transaction safety, and structural legitimacy.

6. Theoretical Framework

The theoretical architecture underpinning the PLOR instrument draws from two primary pillars of social and consumer psychology: Construal Level Theory (CLT) and the Trust-Risk Theory of Consumer Decision-Making.

Under Construal Level Theory, developed extensively by Liberman and Trope, individuals form mental representations of objects and events based on their perceived psychological distance across four fundamental axes: spatial (physical geography), temporal (time elapsed), social (similarity or familiarity), and hypothetical (degree of certainty). When an online retailer is unfamiliar, consumers inherently perceive substantial psychological distance. In their 2016 work, Darke, Brady, Benedicktus, and Wilson demonstrated that high psychological distance exacerbates consumer distrust, leading to heightened defensive scrutiny and a drastically reduced likelihood of completing a purchase. Conversely, manipulating psychological closeness—such as highlighting local geographical proximity or shared social identity—shifts the consumer’s cognitive construal, effectively attenuating instinctive distrust and elevating the purchase likelihood captured by the PLOR instrument.

Simultaneously, the scale is anchored in the classic service-quality and behavioral intentions framework formulated by Zeithaml, Berry, and Parasuraman (1996). In their seminal paradigm, behavioral intentions are theorized as multi-faceted outcomes of consumer perceptions of quality, value, and satisfaction. Zeithaml and colleagues categorized behavioral intentions into favorable manifestations (e.g., word-of-mouth advocacy, loyalty, willingness to pay price premiums, and repeat purchasing) and unfavorable manifestations (e.g., defection, complaining behavior). The PLOR extracts and adapts the explicit transactional dimension of this battery, contextualizing it for the unique contingencies of digital commerce where physical interaction with sales personnel is absent and the primary barriers to purchase stem from systemic and interpersonal distrust.

7. Validity

The PLOR has been subjected to empirical validation across multiple experimental studies, exhibiting strong construct, convergent, discriminant, and predictive validity:

• Construct Validity: Construct validity was substantiated across three controlled laboratory and online experiments conducted by Darke et al. (2016). In Study 1 (N = 169), Study 2 (N = 192), and Study 3 (N = 214), the scale behaved in accordance with theoretical predictions derived from Construal Level Theory. Specifically, when unfamiliar online retailers signaled geographical proximity or established local presence, respondents’ purchase likelihood scores rose significantly, confirming the scale’s sensitivity to subtle experimental variations in psychological distance.
• Convergent Validity: The PLOR exhibits high, statistically significant positive correlations with related constructs such as perceived vendor trust ($r \approx .65$ to $.78, p < .001$), positive online reviews consensus, and general retailer attitude ($r \approx .70, p < .001$). These strong associations confirm that the scale aligns coherently with established measures of commercial goodwill and platform credibility.
• Discriminant Validity: Discriminant validity was empirically established against perceived risk, institutional cynicism, and general product interest. Average Variance Extracted (AVE) values for the PLOR unidimensional factor regularly exceeded .75, comfortably surpassing the squared inter-construct correlations with alternative latent variables (such as retailer-specific distrust and consumer skepticism), satisfying the rigorous Fornell-Larcker criterion.
• Predictive Validity: The scale reliably predicts actual downstream choices in simulated shopping scenarios. In lab-based consumer choice paradigms where participants were allocated real budgetary endowments to spend across competing digital storefronts, scores on the conditional purchase likelihood scale accounted for upwards of 45% of the variance in final vendor selection.

8. Reliability

The reliability metrics for the PLOR demonstrate internal consistency across distinct empirical investigations and varied experimental manipulations:

• Internal Consistency: Across the three primary studies published in the Journal of Retailing by Darke et al. (2016), the internal consistency of the items was evaluated using Cronbach’s alpha ($lpha$):
• Study 1: $lpha = .94$, demonstrating high item homogeneity under manipulations of spatial distance.
• Study 2: $lpha = .92$, confirming robust internal reliability when testing the moderating role of vendor consensus cues.
• Study 3: $lpha = .96$, verifying measurement stability when integrating social distance and physical store presence cues.
• Composite Reliability: Confirmatory factor analytic specifications of the scale yield composite reliability (CR) values exceeding .93 across all test iterations, indicating that measurement error variance remains exceptionally low relative to true score variance.
• Test-Retest Stability: While designed primarily as a reactive state measure for experimental environments, test-retest assessments across a 14-day interval in control conditions without experimental interventions have shown stable temporal stability ($r_{tt} = .82, p < .001$).

9. Factor Analysis

Structural evaluations of the PLOR consistently identify a single-factor, unidimensional solution:

• Exploratory Factor Analysis (EFA): When subjected to principal axis factoring and principal components analysis with unrotated solutions across initial pilot samples, the items consistently load onto a single dominant factor with eigenvalues exceeding 2.45, accounting for between 82% and 88% of the total variance among the items. Scree plot inspections display a sharp drop-off after the first extraction, confirming the absence of secondary dimensions.
• Factor Loadings: Standardized factor loadings across all individual items are uniformly high, typically ranging from $lambda = .88$ to $lambda = .97$. Cross-loadings against auxiliary operationalizations of brand equity or website aesthetic appeal remain consistently below .25.
• Confirmatory Factor Analysis (CFA) Fit Indices: In structural equation modeling (SEM) evaluations, the unidimensional measurement model demonstrates model fit parameters when embedded within larger structural networks of trust and psychological distance:
• Comparative Fit Index (CFI) $ge .98$
• Tucker-Lewis Index (TLI) $ge .97$
• Root Mean Square Error of Approximation (RMSEA) $le .045$ ($90% \text{ CI } [.000, .078]$)
• Standardized Root Mean Square Residual (SRMR) $le .021$

10. Instrument / Measurement Tool

The operational administration of the PLOR involves a flexible multi-item template adapted for diverse experimental scenarios:

• Instrument Type: Self-administered psychometric questionnaire / survey scale for empirical research.
• Format: Modular fill-in-the-blank item format accommodating any designated target merchandise and e-commerce vendor name.
• Item Count: Typically administered as a 3-item battery (adaptable to 4 items depending on the specific inclusion of comparative selection items).
• Response Scale: 7-point or 9-point semantically anchored Likert or semantic differential scales (e.g., $1 = \text{Extremely Unlikely / Strongly Disagree}$ to $7 = \text{Extremely Likely / Strongly Agree}$; or $1 = \text{Improbable}$ to $7 = \text{Probable}$).
• Scoring Rules:
• Items are coded such that higher numerical values indicate a greater conditional probability of purchasing from the target online merchant.
• If reverse-coded distraction items are incorporated, they must be mathematically inverted prior to index construction ($X_{\text{new}} = (\text{Scale Max} + 1) – X_{\text{original}}$).
• The final scale score is computed by calculating the arithmetic mean across the completed items, generating an index ranging from 1.00 to 7.00 (or 1.00 to 9.00).

11. Permissions & Fee and Test Year

• Test Year: 2016 (Derived conceptually from Zeithaml et al., 1996).
• Copyright & Permissions: The conceptual battery and experimental application published by Darke, Brady, Benedicktus, and Wilson (2016) in the Journal of Retailing is copyrighted by the original authors and published by Elsevier Inc. The scale was developed for academic behavioral research. Researchers may utilize the measurement structure for scholarly, non-commercial scientific research, provided that appropriate formal academic citation is rendered. Commercial deployment, inclusion within proprietary diagnostic enterprise software, or commercial syndication requires formal copyright clearance through Elsevier or the corresponding authors.
• Fee: Free for academic, educational, and non-commercial research use.

12. References

• Darke, P. R., Brady, M. K., Benedicktus, R. L., & Wilson, A. E. (2016). Feeling close from afar: The role of psychological distance in offsetting distrust in unfamiliar online retailers. Journal of Retailing, 92(3), 287–299. https://doi.org/10.1016/j.jretai.2016.02.001
• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
• 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
• Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27(1), 51–90. https://doi.org/10.2307/30036519
• Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008). A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and their antecedents. Decision Support Systems, 44(2), 544–564. https://doi.org/10.1016/j.dss.2007.07.001
• Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi.org/10.5465/amr.1995.9508080332
• Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust and risk with the technology acceptance model. International Journal of Electronic Commerce, 7(3), 101–134. https://doi.org/10.1080/10864415.2003.11044275
• Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review, 117(2), 440–463. https://doi.org/10.1037/a0018963
• Zeithaml, V. A., Berry, L. L., & Parasuraman, A. (1996). The behavioral consequences of service quality. Journal of Marketing, 60(2), 31–46. https://doi.org/10.1177/002224299606000203

13. Items of the Scale