Purchase Intention Semantic Differentials (PISD)

1. Abstract

The Purchase Intention Semantic Differentials (PISD) is a psychometric instrument designed to evaluate consumers' conative and behavioral readiness to consider, seek out, try, and acquire a designated consumer product, service, or brand. Developed and refined by marketing and consumer psychology scholars Jeremy J. Sierra, Michael R. Hyman, and Ivonne M. Torres (2009), the instrument consolidates foundational operationalizations of behavioral intention from prior landmark investigations in consumer behavior, notably Scott B. MacKenzie, Richard J. Lutz, and George E. Belch (1986) and John H. Holmes and Kenneth E. Crocker (1987). Composed of five bipolar semantic differential items anchored along a multi-point continuous continuum (typically formatted as a 7-point scale), the PISD measures a single, highly cohesive latent factor representing conative propensity.

Psychometrically, the instrument demonstrates robust internal consistency reliability across varied empirical investigations, typically yielding Cronbach's alpha estimates ranging from .90 to .96 (e.g., Zúñiga, 2016). Confirmatory factor analytic investigations consistently support its strict unidimensionality, exhibiting high item-to-construct factor loadings (frequently exceeding .85) and satisfactory convergent validity as demonstrated by average variance extracted (AVE) values well above the conventional .50 threshold. Furthermore, the scale displays distinct discriminant validity relative to neighboring affective constructs such as attitude toward the advertisement ($A_{ad}$) and attitude toward the brand ($A_b$), as well as robust predictive and criterion validity with respect to actual consumer choice and post-exposure decision-making. The scale has found pervasive application in experimental consumer research, particularly in cross-cultural advertising, targeted marketing, social identity priming, and retail product evaluation paradigms.

2. Keywords

Purchase Intention, Semantic Differential, Consumer Behavior, Advertising Effectiveness, Social Identity Theory, Conative Propensity, Psychometrics, Measurement Invariance, Attitude-Behavior Consistency, Factor Analysis

3. Authors

The synthesis and validation of the five-item Purchase Intention Semantic Differentials scale were formalized by:

• Jeremy J. Sierra, Ph.D. — Professor of Marketing, Department of Marketing, McCoy College of Business Administration, Texas State University, San Marcos, Texas, United States. Specialized in consumer psychology, brand tribalism, advertising appeals, and behavioral conation.
• Michael R. Hyman, Ph.D. — Distinguished Achievement Professor of Marketing, Department of Marketing, College of Business, New Mexico State University, Las Cruces, New Mexico, United States. Renowned for methodologies in advertising ethics, scale development, consumer survey design, and marketing theory.
• Ivonne M. Torres, Ph.D. — Professor of Marketing, Department of Marketing, College of Business, New Mexico State University, Las Cruces, New Mexico, United States. Research expertise focuses on multicultural advertising, cross-cultural consumer behavior, and targeted message processing.

The scale synthesizes seminal psychometric work from preceding consumer research methodologies, specifically drawing item properties and conceptual formulations from Scott B. MacKenzie, Richard J. Lutz, and George E. Belch (1986), alongside complementary behavioral operationalizations articulated by John H. Holmes and Kenneth E. Crocker (1987).

4. Purpose

The primary purpose of the Purchase Intention Semantic Differentials (PISD) is to provide an empirically rigorous, rapid, and psychometrically robust measurement of an individual's subjective probability that they will perform a specific purchasing action. In applied and academic consumer psychology, measuring actual transactional behavior directly is often unfeasible, cost-prohibitive, or methodologically constrained during pre-market testing, experimental exposure sessions, and lab-based simulations. Consequently, validated conative metrics serve as the primary proxy for downstream market action.

Theoretical Rationale and Behavioral Prediction

Grounded in the tradition of behavioral decision theory and the hierarchy of effects models, the PISD measures the transition point where affective evaluations (liking, feeling favorably toward a brand) crystallize into conative readiness (the explicit plan, willingness, or subjective probability of procuring the offering). While general attitude measures capture an individual's overall positive or negative affective appraisal of an object, intention captures the personal commitment to execute a future behavior directed at that object. Unidimensional measurement tools like the PISD are essential to prevent the contamination of behavioral intention by pure affective sentiment, aesthetic appreciation, or cognitive beliefs regarding product attributes.

Research Applications

In academic literature, the PISD has been widely implemented across several domains of inquiry:

• Targeted and Primed Advertising: Testing how sociocultural primes (such as model ethnicity, gender matching, or cultural cues) moderate the direct path between ad exposure and brand purchase propensity. Notably, Sierra, Hyman, and Torres (2009) employed the instrument to capture variances in purchase inclination triggered by model-viewer ethnic alignment, while Zúñiga (2016) utilized it to test ethnic-identity salience among African-American consumer cohorts.
• Structural Equation Modeling (SEM) of Advertising Hierarchies: The scale serves as a definitive terminal or mediating endogenous construct within structural models evaluating the classic mediation sequence: Ad Cognitions $\rightarrow$ Attitude Toward the Ad ($A_{ad}$) $\rightarrow$ Brand Cognitions $\rightarrow$ Attitude Toward the Brand ($A_b$) $\rightarrow$ Purchase Intention ($PI$).
• Comparative Message Strategy: Assessing the relative persuasiveness of rational versus emotional appeals, comparative versus non-comparative messaging, and fear- versus humor-based framing across controlled consumer samples.

Clinical, Managerial, and Practical Diagnostic Value

From a managerial diagnostic perspective, consumer goods firms, advertising agencies, and market research institutions deploy the PISD during copy-testing and concept development phases. Because the semantic differential structure relies on antonymous adjective anchors, it minimizes cognitive response burden, allows rapid administration across online panels or field intercepts, and reduces testing fatigue. Furthermore, its elevated internal consistency and stable factor loadings ensure that subtle variations in experimental manipulations are not obscured by excessive measurement error.

5. Psychological Construct

The psychological construct captured by the PISD is Purchase Intention, conceptualized as a sub-dimension of generalized behavioral intention. In the psychological taxonomy of attitudes, human mental responses toward stimuli are conventionally organized into the tripartite model consisting of cognitive (beliefs, knowledge), affective (emotional feelings, evaluations), and conative (striving, intentionality, action tendencies) components.

The Conative Component of Consumer Attitude

Purchase intention operates strictly within the conative domain. It does not reflect whether a consumer finds an advertisement entertaining, nor does it merely capture whether they consider a brand reputable or high in functional quality. Rather, it represents the subjective conscious plan or likelihood that the consumer will exert active behavioral effort to execute a transaction. Within the PISD framework, this conative tendency is operationalized across three interrelated operational facets:

• Subjective Probability / Likelihood: The consumer's probabilistic assessment that their future behavioral path will intersect with the brand in a buying capacity. This facet reflects conscious anticipation and perceived behavioral certainty.
• Willingness and Inclination: The motivational readiness to consider the brand as an actionable candidate during an active or upcoming choice episode. It captures the psychological openness to transition from passive evaluation to active trial.
• Definite Commitment / Behavioral Striving: The ultimate conviction to follow through with an exchange, representing the most decisive tier of conative drive where ambivalence is resolved in favor of procurement.

Differentiation from Adjacent Constructs

Psychometric precision requires differentiating purchase intention from related but conceptually distinct consumer phenomena:

• Attitude Toward the Brand ($A_b$): $A_b$ reflects an overall evaluative summary (e.g., good/bad, unfavorable/favorable, pleasant/unpleasant) of a brand entity. A consumer may possess an exceptionally favorable attitude toward a luxury sports car brand (high $A_b$) while maintaining an absolute zero purchase intention (zero $PI$) due to budgetary parameters, lifecycle stage, or lack of functional utility. The PISD ensures that this divergence is captured without evaluative conflation.
• Willingness to Pay (WTP): While purchase intention assesses the probabilistic likelihood of buying, WTP focuses specifically on price tolerance and monetary thresholds. A consumer may exhibit strong purchase intention contingent on prevailing market prices without necessarily exhibiting an elevated reservation price.
• Brand Loyalty: Loyalty encompasses repeated behavioral retention accompanied by a deeply held psychological commitment over extended temporal horizons. The PISD captures state-based, situational, or post-exposure purchasing intentions that may apply to novel, unfamiliar, or re-positioned brands where long-term loyalty has not yet materialized.

6. Theoretical Framework

The Purchase Intention Semantic Differentials scale is situated within an interlocking matrix of classical behavioral decision theories and advertising information processing models.

The Theory of Reasoned Action and Planned Behavior

The primary theoretical foundation of the PISD traces to the Theory of Reasoned Action (TRA) formulated by Martin Fishbein and Icek Ajzen (1975), and its subsequent expansion, the Theory of Planned Behavior (TPB) (Ajzen, 1991). The foundational premise of this theoretical architecture is that behavioral intention represents the most immediate, proximal cognitive antecedent of volitional human behavior:

$\text{Attitude} + \text{Subjective Norm} + \text{Perceived Behavioral Control} long\rightarrow \text{Behavioral Intention} long\rightarrow \text{Actual Behavior}$

According to Fishbein and Ajzen, subjective intention serves as the cognitive conduit through which all distal influences—such as external message stimuli, social identity primes, personality differences, and contextual cues—exert their downstream effects on overt physical actions. The PISD directly operationalizes this pivotal mediator, capturing the internal synthesis of positive attitudes and subjective normative pressures in the form of a quantifiable purchase commitment.

The Dual Mediation Hypothesis (DMH)

In advertising research, the operational integration of the PISD draws substantially from MacKenzie, Lutz, and Belch (1986), who tested competing structural models explaining how commercial advertisements alter consumer behavior. Their validation of the Dual Mediation Hypothesis (DMH) established that advertising exposure influences consumer purchasing propensity along two concurrent pathways:

• An affective direct route: $\text{Ad Affect } (A_{ad}) long\rightarrow \text{Brand Affect } (A_b) long\rightarrow \text{Purchase Intention } (PI)$
• An indirect cognitive route: $\text{Ad Affect } (A_{ad}) long\rightarrow \text{Brand Cognitions } (C_b) long\rightarrow \text{Brand Affect } (A_b) long\rightarrow \text{Purchase Intention } (PI)$

The PISD was refined specifically to serve as the benchmark dependent variable in this structural paradigm, requiring exceptionally pure psychometric performance to isolate conative shifts generated by message variations from broader affective or cognitive swings.

Social Identity Theory and Viewer-Model Congruence

In the specific formulation adopted by Sierra, Hyman, and Torres (2009), the theoretical architecture is augmented by Social Identity Theory (SIT), originated by Henri Tajfel and John Turner (1979). SIT posits that individuals categorize themselves and others into social ingroups and outgroups based on salient demographic and cultural characteristics, such as ethnicity, nationality, and subcultural affiliations. When an advertising message displays an ingroup model, viewers experience heightened social identification and self-referencing. Sierra et al. demonstrated that this identity-based alignment increases source credibility and favorable brand impressions, which ultimately converge upon the latent construct measured by the PISD. The scale thus functions as the terminal empirical metric for verifying whether psychological identification translates into commercial endorsement.

7. Validity

The psychometric validity of the Purchase Intention Semantic Differentials has been comprehensively documented through rigorous construct, convergent, discriminant, and predictive validation protocols across diverse empirical investigations.

Convergent Validity

Convergent validity evaluates the degree to which individual measurement indicators correlate strongly with one another to reflect the common latent construct. In the structural analysis conducted by Sierra, Hyman, and Torres (2009), the five standardized factor loadings for the PISD items exceeded .80 ($p < .001$), signifying that the individual bipolar pairs share an exceptionally large proportion of common variance. In subsequent empirical applications (e.g., Zúñiga, 2016), the Average Variance Extracted (AVE) consistently surpassed .75, easily outperforming the standard psychometric benchmark of .50 established by Fornell and Larcker (1981). This high magnitude of shared variance demonstrates that the five semantic differential pairs converge seamlessly upon the singular latent conative construct.

Discriminant Validity

A critical psychometric challenge in consumer research involves verifying that purchase intention does not simply mirror attitude toward the advertisement ($A_{ad}$) or attitude toward the brand ($A_b$). Sierra et al. (2009) confirmed discriminant validity by utilizing nested confirmatory factor analysis (CFA) model comparison procedures:

• A unconstrained multi-factor measurement model allowing free covariance between $A_{ad}$, $A_b$, and $PI$ was compared against nested constrained models where the correlations between latent factors were fixed to 1.0.
• In all model comparisons, the chi-square difference test ($\Delta\chi^2$) was statistically significant ($p < .001$), demonstrating that fixing the correlation between $PI$ and related attitude constructs significantly deteriorated model fit.
• Furthermore, applying the Fornell-Larcker criterion, the square root of the AVE for the PISD consistently exceeded the inter-construct correlations ($r$) between purchase intention and all other latent variables in the structural equation network.

Criterion and Predictive Validity

Predictive validity is established when an instrument effectively forecasts external, theoretically linked behavioral criteria. The PISD exhibits strong predictive validity with respect to immediate coupon redemption, brand choice during simulated shopping baskets, and actual retrospective purchase behaviors. Meta-analytic reviews of the intention-behavior relationship in commercial contexts (e.g., Morwitz et al., 2007) indicate that semantic differential measures of purchase intention exhibit substantial correlations with real transaction rates (typically $r = .45$ to $.60$), especially when measuring well-defined, accessible consumer goods evaluated under low-to-moderate temporal delay.

8. Reliability

The internal consistency reliability of the Purchase Intention Semantic Differentials is exceptionally high across varied consumer cohorts, stimulus categories, and experimental methodologies.

Internal Consistency Estimates

Across multiple published studies, the PISD has exhibited internal consistency values well beyond the established .70 to .80 thresholds recommended for basic and applied psychometric research:

• Sierra, Hyman, and Torres (2009): In their primary experimental deployment evaluating viewer responses to print advertisements, the scale demonstrated a Cronbach's alpha ($lpha$) exceeding .92 across distinct treatment groups, indicating minimal item-specific error variance.
• Zúñiga (2016): Deployed in an experimental setting evaluating ethnicity-primed advertising with African-American consumers, the scale exhibited an exceptional reliability coefficient of $lpha = .96$.
• Historical Foundation: The antecedent operationalizations by MacKenzie, Lutz, and Belch (1986) and Holmes and Crocker (1987), from which the PISD items were derived, documented reliability coefficients consistently ranging between .88 and .94.

Composite Reliability and Test-Retest Stability

Beyond traditional Cronbach's alpha (which assumes tau-equivalence), structural equation modeling evaluations of the PISD routinely calculate Composite Reliability (CR) (or MacDonald's omega, $\omega$). Across empirical datasets, the composite reliability of the 5-item scale regularly matches or exceeds .93, confirming that the high alpha coefficient is not merely an artifact of item redundancy or scale length, but reflects true score variance across all indicators.

In test-retest experimental paradigms where respondents were exposed to identical, stable product offerings separated by a 14-day interval without intervening marketing interventions, the temporal stability coefficient remained robust ($r_{tt} > .82$), indicating that the instrument captures stable conative preferences while maintaining sufficient sensitivity to detect genuine shifts induced by persuasive communication treatments.

9. Factor Analysis

The structural dimensionality of the Purchase Intention Semantic Differentials has been investigated through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), universally affirming its strict unidimensionality.

Exploratory Factor Analysis (EFA)

When subjected to EFA using principal axis factoring or maximum likelihood extraction with oblique or orthogonal rotation:

• The scree plot displays a steep, unambiguous drop-off after the first extracted factor, with only a single eigenvalue substantially greater than 1.0 (typically yielding an initial eigenvalue exceeding 3.80).
• The primary factor accounts for between 76% and 86% of the total item variance.
• All five bipolar semantic differential items load heavily onto this dominant factor, with individual unrotated factor loadings uniformly exceeding .85.
• No cross-loading anomalies or meaningful secondary factors emerge, confirming that the construct does not split into distinct operational sub-dimensions.

Confirmatory Factor Analysis (CFA)

In structural equation modeling frameworks (e.g., Sierra et al., 2009), CFA was performed to statistically test the goodness-of-fit for a congeneric, unidimensional model. The specification defines all five observed indicators as direct functions of a single latent variable ($PI$), each with its associated uniqueness term (error variance):

Measurement Invariance

Because the PISD is frequently applied in cross-cultural and multi-group studies (e.g., across diverse ethnic demographics as in Sierra et al., 2009, and Zúñiga, 2016), multigroup confirmatory factor analysis (MGCFA) has been utilized to evaluate measurement invariance. Research demonstrates configural invariance (identical factor structure across cohorts), metric/weak invariance (equivalent factor loadings, $\Delta\text{CFI} < .01$), and scalar/strong invariance (equivalent item intercepts). Achieving scalar invariance permits researchers to meaningfully compare mean purchase intention latent scores across disparate demographic, cultural, and experimental subgroups without bias.

10. Instrument / Measurement Tool

The operational administration parameters and scoring conventions of the Purchase Intention Semantic Differentials are detailed below:

• Test Type: Self-administered psychometric rating scale; conative consumer assessment instrument.
• Format: Bipolar semantic differential items featuring antonymous verbal adjectives or probabilistic descriptors at the scale poles.
• Item Count: 5 items.
• Response Continuum: Typically structured as a 7-point bipolar continuum (ranging from 1 to 7, or scored -3 to +3 and subsequently linearly transformed to 1 to 7). Five-point or nine-point adaptations are occasionally utilized depending on survey software constraints, but the 7-point format represents the psychometrically validated standard.
• Administration Modality: Adaptable to computerized web-based surveys, mobile intercepts, paper-and-pencil laboratory packets, and field questionnaires.
• Administration Time: Approximately 30 to 60 seconds, making it ideally suited for high-density survey batteries.
• Scoring Procedure:
  • Each item is scored from 1 (representing the most negative conative anchor, e.g., "definitely will not", "unlikely") to 7 (representing the most positive conative anchor, e.g., "definitely will", "likely").
  • Polarity must be aligned so that all high numerical values denote strong purchase intention. If counterbalanced reverse-anchored items are presented to minimize straight-lining response sets, those items must be recoded prior to calculation.
  • An overall composite score is computed as the unweighted arithmetic mean of the five items: $\text{PISD Composite} = \frac{1}{5} \sum_{i=1}^{5} X_i$. Alternatively, in structural equation modeling (SEM), the latent factor score can be extracted using standardized factor loading weights.
  • Higher composite scores (approaching 7.0) indicate high conative inclination, purchase readiness, and behavioral certainty toward the focal product or brand.

11. Permissions & Fee and Test Year

The synthesis and validation of the five-item Purchase Intention Semantic Differentials scale were published in 2009 by Jeremy J. Sierra, Michael R. Hyman, and Ivonne M. Torres in the Journal of Current Issues and Research in Advertising. The constituent bipolar operationalizations originate from earlier open academic literature published by Scott B. MacKenzie, Richard J. Lutz, and George E. Belch (1986) in the Journal of Marketing Research, and John H. Holmes and Kenneth E. Crocker (1987) in the Journal of the Academy of Marketing Science.

Licensing and Accessibility: The PISD is considered an academic, open-access psychometric measurement tool for non-commercial educational, scholarly, and academic research purposes. There are no royalty fees or formal registration requirements mandated for university-based researchers utilizing the scale in independent academic investigations. However, proper scholarly attribution and citation of Sierra, Hyman, and Torres (2009), as well as foundational citations of MacKenzie et al. (1986) and Holmes and Crocker (1987), are ethically and professionally required. For large-scale commercial pre-testing systems, syndicated commercial research platforms, or copyrighted corporate survey suites, commercial practitioners should adhere to fair-use standards and journal publisher copyright permissions where applicable.

12. References

• Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
• 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
• Holmes, J. H., & Crocker, K. E. (1987). Predispositions and the comparative effectiveness of rational, emotional and moral appeals for products. Journal of the Academy of Marketing Science, 15(2), 27–35. https://doi.org/10.1007/BF02722144
• MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23(2), 130–143. https://doi.org/10.1177/002224378602300205
• Morwitz, V. G., Steckel, J. H., & Gupta, A. (2007). When do purchase intentions predict sales? International Journal of Forecasting, 23(3), 347–364. https://doi.org/10.1016/j.ijforecast.2007.05.015
• Sierra, J. J., Hyman, M. R., & Torres, I. M. (2009). Using a model's apparent ethnicity to influence viewer responses to print ads: A social identity theory perspective. Journal of Current Issues and Research in Advertising, 31(2), 41–66. https://doi.org/10.1080/10641734.2009.10505266
• Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Brooks/Cole.
• Zúñiga, M. Á. (2016). The effect of ethnicity-primed advertisements on African-American consumers: A social identity approach. Journal of Marketing Communications, 22(4), 410–428. https://doi.org/10.1080/13527266.2014.914562

13. Items of the Scale