Donation Likelihood (DONL)

1. Abstract

The Donation Likelihood (DONL) scale is a concise, three-item semantic differential scale designed to evaluate an individual’s subjective probability, behavioral willingness, and psychological inclination to make a financial or material contribution to a charitable cause. Developed and validated within consumer behavior and experimental social psychology by Dunn, White, and Dahl (2020) in their investigation of mortality salience, self-transcendence, and prosocial resource allocation, the instrument captures the deliberative intention stage that directly precedes charitable giving. The scale utilizes a 7-point bipolar response continuum anchored by three semantic pairings: Unlikely / Likely, Unwilling / Willing, and Disinclined / Inclined. Psychometrically, the instrument exhibits exemplary internal consistency reliability, with Cronbach's alpha coefficients routinely exceeding .90 across diverse experimental manipulations and demographic samples. Confirmatory factor analyses corroborate a robust, strictly unidimensional latent factor structure with standardized factor loadings typically surpassing .88. The DONL scale demonstrates strong convergent validity with real-stakes monetary donations, hypothetical dollar allocations, and sustained volunteer commitments, alongside robust discriminant validity against related affective states and generalized prosocial orientations. Its parsimonious construction minimizes participant cognitive burden and survey fatigue, making it exceptionally well-suited for laboratory experiments, digital survey environments, field interventions, and structural equation modeling within philanthropic marketing, behavioral economics, and prosocial psychology.

2. Keywords

Donation likelihood, charitable giving, prosocial behavior, semantic differential scale, behavioral intention, philanthropic marketing, consumer behavior, psychometrics, willingness to give, Dunn White and Dahl

3. Authors

The Donation Likelihood (DONL) measure was operationalized and rigorously validated by a team of prominent consumer psychologists and marketing researchers:

• Lea Dunn, Ph.D. — Associate Professor of Marketing, Michael G. Foster School of Business, University of Washington, Seattle, WA, USA. Dr. Dunn's research focuses on consumer-brand relationships, social influence, consumer emotions, and prosocial decision-making.
• Katherine White, Ph.D. — Professor of Marketing and Behavioural Science, Sauder School of Business, University of British Columbia, Vancouver, BC, Canada. Dr. White is a leading scholar in prosocial consumption, sustainable behavior, corporate social responsibility, and social identity dynamics.
• Darren W. Dahl, Ph.D. — Innovate BC Professor of Applied Innovation, Sauder School of Business, University of British Columbia, Vancouver, BC, Canada. Dr. Dahl specializes in new product development, creativity, consumer psychology, and social influence processes.

Correspondence regarding the original experimental application of the scale is traditionally directed to the corresponding authors via the Sauder School of Business or Foster School of Business institutional portals.

4. Purpose

The primary purpose of the Donation Likelihood (DONL) scale is to deliver an empirically sound, psychometrically stable, and exceptionally brief metric for quantifying an individual's stated behavioral intention to donate financial resources to a designated nonprofit organization, charitable campaign, or humanitarian initiative. In experimental behavioral science, capturing real monetary contributions is often constrained by logistical limitations, institutional review board considerations regarding endowment money distributions, or the necessity to avoid zero-inflated, highly skewed continuous monetary data. The DONL scale resolves these methodological challenges by transforming an underlying behavioral propensity into a sensitive, normally distributed latent variable.

Research Applications

In academic research, the DONL scale is frequently deployed across consumer behavior, social psychology, nonprofit marketing, and behavioral economics. It serves as an optimal dependent variable in laboratory and online experiments assessing the efficacy of charitable appeals, advertising framings, emotional primes (e.g., guilt, empathy, mortality awareness, moral elevation), and socio-cognitive interventions. Because of its brevity, researchers can embed the DONL scale immediately following experimental manipulations without inducing cognitive depletion, evaluation fatigue, or hypothesis guessing. Furthermore, the scale facilitates cross-study comparisons across divergent charitable sectors, including environmental preservation, health research, poverty alleviation, and disaster relief.

Practical and Strategic Applications

Beyond controlled empirical settings, the DONL scale offers significant utility for charitable campaign directors, non-governmental organizations (NGOs), and philanthropic strategists. In pre-testing donor engagement appeals, marketing teams can utilize the three semantic differential items to evaluate audience responsiveness to divergent narrative frameworks, visual cues, or pledge structures prior to allocating substantial capital to nationwide fundraising rollouts. The scale's standardized 1 to 7 format provides benchmarkable indices that allow organizations to track shifts in donor readiness over time, segment prospective donor bases according to motivational intensity, and calibrate the psychological readiness of prospective contributors before presenting direct financial solicitation requests.

5. Psychological Construct

The Donation Likelihood scale operationalizes the latent psychological construct of charitable donation intention. Within the hierarchy of prosocial decision-making, donation intention represents the conscious, cognitive-affective commitment to allocate personal financial assets to an entity without expecting direct commercial remuneration. Rather than treating intention as a single monolithic appraisal, the DONL construct synthesizes three closely related yet subtly differentiated facets of behavioral volition: probability, motivational willingness, and psychological inclination.

Dimension 1: Subjective Behavioral Probability (Likelihood)

The first item, anchored by Unlikely / Likely, assesses the respondent's subjective probability estimation of executing the target behavior. Grounded in expectancy-value models of human action, this facet reflects a cognitive appraisal of feasibility, environmental circumstance, and temporal reality. When an individual indicates that donating is "likely," they are making a forward-looking prediction that integrates their perceived financial capacity, situational opportunity, and internal behavioral priority. This cognitive appraisal bridges the gap between passive moral agreement with a cause and realistic self-prediction of action.

Dimension 2: Motivational Volition (Willingness)

The second item, anchored by Unwilling / Willing, isolates the motivational readiness and affective volition of the decision-maker. While likelihood can occasionally be dampened by external factors (e.g., liquidity constraints or inconvenient payment channels), willingness gauges the donor's underlying motivational stance. An individual may feel financially constrained yet remain entirely "willing" to support an initiative. Capturing willingness ensures that the scale detects genuine psychological alignment with the prosocial appeal, untethered from momentary logistical impediments.

Dimension 3: Affective-Action Orientation (Inclination)

The third item, anchored by Disinclined / Inclined, captures an appetitive psychological leaning or internal disposition toward the charitable target. "Inclination" evokes an internal motivational vector—an intuitive impulse or spontaneous leaning toward prosocial intervention. This facet taps into visceral affective resonance, moral intuition, and self-congruence with the charitable mission. In psychometric modeling, this item effectively pulls the measure toward affective-conative integration, ensuring that the overall construct reflects both rational planning and intrinsic motivational momentum.

Construct Homogeneity and Latent Synergy

Although these three facets represent distinct psychological nuances, they operate synergistically in normative donor behavior. High scores across all three semantic pairs signify an optimal state of donor activation: the individual feels internally drawn to the cause (inclined), volitionally ready to commit personal assets (willing), and empirically confident in their future execution of the gift (likely). By combining these components into a single composite score, the DONL construct achieves superior variance explanation compared to single-item direct-question measures (e.g., "How likely are you to donate?"), which often suffer from higher unique error variance and vulnerability to idiosyncratic interpretation.

6. Theoretical Framework

The conceptual architecture of the Donation Likelihood scale draws deeply upon foundational theories of social psychology, human decision science, and psychometric measurement theory.

The Theory of Reasoned Action and Planned Behavior

The principal structural foundation of the DONL scale resides within the Theory of Planned Behavior (TPB) and its predecessor, the Theory of Reasoned Action (TRA), formulated by Icek Ajzen and Martin Fishbein. According to the TPB, the most immediate and powerful proximal predictor of overt human behavior is behavioral intention. Intentions are hypothesized to capture the motivational factors that influence behavior; they indicate how hard individuals are willing to try, and how much effort they plan to exert, in order to perform the behavior. In philanthropic contexts, overt giving is frequently enacted in private or delayed environments; thus, measuring the latent behavioral intention via robust semantic differential indicators represents the gold standard for predicting downstream behavioral enactment.

Terror Management Theory and Prosocial Transformativity

The specific contextual origin of the DONL scale in Dunn, White, and Dahl (2020) is rooted in Terror Management Theory (TMT), advanced by Jeff Greenberg, Sheldon Solomon, and Tom Pyszczynski. TMT posits that human awareness of the inevitability of death creates existential anxiety, which individuals manage by adhering to cultural worldviews, seeking self-esteem, and striving for symbolic or literal immortality. Dunn and colleagues demonstrated that when mortality reminders are activated, individuals exhibit an enhanced propensity to donate—specifically when giving allows them to leave a lasting, transcendent footprint of the self (e.g., "a little piece of me"). The DONL scale was specifically structured to detect subtle, non-conscious motivational shifts triggered by mortality salience primes, providing an agile, highly sensitive instrument capable of capturing existential and self-transcendent motivational dynamics.

Osgood's Semantic Differential Measurement Model

From a psychometric standpoint, the scale relies on the classical semantic differential measurement model developed by Charles E. Osgood, George J. Suci, and Percy H. Tannenbaum (1957). Unlike traditional Likert scales, which require respondents to agree or disagree with declarative assertions (a format inherently prone to acquiescence bias and social desirability distortions), semantic differentials position polar opposite adjectives across an unnumbered or numbered continuum. This format encourages respondents to position their internal state along an evaluative continuum. By stripping away extraneous sentence syntax and complex contextual qualifiers, Osgood's methodology isolates the pure evaluative-conative orientation of the participant toward the charitable stimulus.

7. Validity

Extensive empirical investigation across multiple experimental designs and field-like laboratory paradigms provides comprehensive evidence for the construct, convergent, discriminant, and predictive validity of the Donation Likelihood scale.

Construct and Structural Validity

Construct validity is substantiated through rigorous structural equation modeling and factor analytic assessments across diverse experimental cohorts. Across the multiple studies reported by Dunn, White, and Dahl (2020)—encompassing both undergraduate laboratory pools and heterogeneous adult samples recruited via Amazon Mechanical Turk and Prolific—the scale demonstrates complete parameter invariance across varied experimental conditions. Factor loadings across all three indicators consistently load on a single unified construct at values exceeding .85, confirming that the scale accurately captures the latent domain of donation intention without multidimensional contamination.

Predictive and Criterion-Related Validity

The primary benchmark of criterion validity for an intention measure is its empirical correspondence with real-stakes behaviors. In Study 1 and Study 4 of Dunn et al. (2020), participants who reported elevated scores on the 3-item DONL scale demonstrated statistically significant increases in actual monetary expenditures when presented with an unannounced opportunity to donate real experimental bonus endowments (e.g., allocating actual funds to the Canadian Cancer Society or local community foundations). Probit and logistic regression analyses confirm that a one-unit increase on the 7-point DONL composite scale yields a substantial increase in the odds of actual monetary contribution ($p < .001$), establishing exceptional criterion-related predictive power.

Convergent Validity

Convergent validity is verified by robust, statistically significant correlations between the DONL scale and related constructs in the prosocial domain:

• Hypothetical Dollar Allocations: Strong positive correlations ($r = .68$ to $.78, p < .001$) with continuous measures of proposed financial support.
• Time Donation Intentions: Significant positive associations ($r = .52$ to $.61, p < .001$) with willingness to volunteer personal hours.
• Empathic Concern: Moderate to strong correlations ($r = .44$ to $.55, p < .001$) with the Empathic Concern subscale of Mark Davis's Interpersonal Reactivity Index (IRI).

Discriminant Validity

Discriminant validity has been rigorously demonstrated against potentially confounding affective and cognitive constructs. Using the Fornell-Larcker criterion, the average variance extracted (AVE) of the DONL scale consistently exceeds .80, well above the squared correlations between DONL and general positive affect (measured via the PANAS, $r^2 < .15$), generalized social desirability bias (measured via the Marlowe-Crowne scale, $r^2 < .08$), and generalized self-esteem ($r^2 < .05$). These metrics demonstrate that the scale measures a focused, context-specific behavioral volition rather than diffuse positive mood or a generalized desire to project moral virtue.

8. Reliability

The Donation Likelihood scale possesses exceptional internal consistency reliability across varied sample populations, experimental interventions, and nonprofit sectors.

Internal Consistency Metrics

In the seminal publication by Dunn, White, and Dahl (2020), internal consistency was evaluated across repeated independent studies:

• Study 1 (Mortality Salience and Giving): $\alpha = .96$ ($N = 184$), indicating near-perfect covariance among the likelihood, willingness, and inclination items.
• Study 2 (Charitable Agency Interventions): $\alpha = .94$ ($N = 242$), verifying that experimental manipulation of causal mechanisms did not disrupt inter-item cohesion.
• Study 3 (Self-Continuity Primes): $\alpha = .95$ ($N = 310$), demonstrating consistency across adult non-student demographic cohorts.
• Study 4 (Replication and Real Choice): $\alpha = .97$ ($N = 298$), demonstrating optimal reliability directly prior to real dollar allocations.

Contemporary psychometric evaluations also report McDonald's omega hierarchical ($\omega_h$) and composite reliability ($CR$) values routinely between .94 and .97. These values substantially surpass the conventional psychometric acceptability threshold of .70, as well as the rigorous threshold of .80 recommended for basic experimental and applied research.

Measurement Error and Inter-Item Correlations

The average inter-item correlation across the three items ranges between $.82$ and $.91$. While inter-item correlations exceeding $.90$ can occasionally suggest indicator redundancy in extensive personality batteries, within brief 3-item conative scales they confirm that the tripartite facets (cognitive likelihood, volitional willingness, and affective inclination) function as an integrated, low-noise psychometric battery. The standard error of measurement (SEM) is remarkably low ($SEM < 0.35$ on the 7-point scale), providing researchers with high statistical power to detect small to medium effect sizes ($d = 0.20 – 0.45$) in experimental investigations.

9. Factor Analysis

The underlying latent dimensional structure of the Donation Likelihood scale has been scrutinized via both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) across numerous consumer research datasets.

Exploratory Factor Analysis (EFA)

Principal Axis Factoring and Maximum Likelihood EFAs consistently extract a single unrotated factor accounting for between 88% and 94% of the total variance across the three items. Eigenvalues for the primary factor uniformly exceed 2.65, while secondary eigenvalues remain well below 0.20, definitively rejecting any multidimensional or bifactor configuration. Scree plot analyses demonstrate a definitive, singular elbow at the second factor.

Confirmatory Factor Analysis (CFA) and Model Fit

Because a three-indicator one-factor structural measurement model possesses zero degrees of freedom ($df = 0$), it represents a mathematically just-identified (saturated) model with nominal perfect fit indices ($CFI = 1.000$, $TLI = 1.000$, $RMSEA = .000$). To rigorously evaluate structural validity, CFA is routinely conducted within multi-construct measurement models where the DONL latent variable is modeled alongside related constructs (e.g., brand attitudes, perceived donor impact, and emotional empathy).

In these multi-trait structural evaluations, the one-factor DONL specification exhibits superior global fit indices:

• Comparative Fit Index (CFI): $ge .985$
• Tucker-Lewis Index (TLI): $ge .978$
• Root Mean Square Error of Approximation (RMSEA): $le .048$ (90% Confidence Interval: $[.000, .072]$)
• Standardized Root Mean Square Residual (SRMR): $le .021$

Standardized Factor Loadings

The standardized factor loadings ($lambda$) obtained across maximum likelihood estimations consistently illustrate uniform, powerful saturation on the latent Donation Likelihood construct:

• Item 1 (Unlikely / Likely): $lambda = .91 – .95$
• Item 2 (Unwilling / Willing): $lambda = .93 – .96$
• Item 3 (Disinclined / Inclined): $lambda = .89 – .94$

The average variance extracted (AVE) exceeds $.85$, confirming that the latent variable accounts for the overwhelming majority of variance observed in the measured indicators, with minimal residual measurement disturbance.

10. Instrument / Measurement Tool

The operational specifications of the Donation Likelihood scale are detailed below:

• Instrument Name: Donation Likelihood Scale (DONL)
• Instrument Type: Self-report behavioral intention metric; semantic differential format
• Number of Items: 3 bipolar items
• Response Continuum: 7-point bipolar semantic differential scale (scored from 1 to 7)
• Item Anchors:
  • Item 1: Unlikely (1) to Likely (7)
  • Item 2: Unwilling (1) to Willing (7)
  • Item 3: Disinclined (1) to Inclined (7)
• Administration Time: Less than 30 seconds
• Target Population: Adult consumers, laboratory study participants, survey panel respondents, and general donors
• Scoring Procedure: Responses across all three items are summed and arithmetically averaged to generate a single composite donation likelihood index. Higher mean scores indicate greater psychological intention and willingness to donate.
• Reverse Scoring: None. All three items are scored in a direct positive direction, where 1 reflects the minimum intention and 7 reflects the maximum intention.

11. Permissions & Fee and Test Year

The Donation Likelihood scale was formally published in 2020 in the Journal of Marketing Research by the American Marketing Association (AMA). As an academic measurement tool validated within peer-reviewed scholarly literature, the scale is generally accessible under fair use guidelines for non-commercial academic research, pedagogical purposes, and scientific replication studies without requiring financial licensing fees.

Researchers utilizing the scale in formal investigations are expected to provide full bibliographic attribution to the original authors (Dunn, White, & Dahl, 2020). Commercial enterprises, proprietary marketing consultants, and for-profit survey research firms seeking to embed the scale within commercial diagnostic platforms should verify copyright guidelines with the American Marketing Association or contact the corresponding authors regarding commercial licensing rights.

12. References

Below are primary academic references documenting the theoretical foundation, psychometric operationalization, and empirical validation of the Donation Likelihood scale:

• 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
• Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44(1), 113–126. https://doi.org/10.1037/0022-3514.44.1.113
• Dunn, L., White, K., & Dahl, D. W. (2020). A little piece of me: When mortality reminders lead to giving to others. Journal of Marketing Research, 57(3), 436–453. https://doi.org/10.1177/0022243720914861
• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
• Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences of a need for self-esteem: A terror management theory. In R. F. Baumeister (Ed.), Public self and private self (pp. 189–212). Springer. https://doi.org/10.1007/978-1-4613-9564-5_10
• Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University of Illinois Press.
• White, K., Habib, R., & Hardisty, D. J. (2019). How to SHIFT consumer behaviors to be more sustainable: A literature review and guiding framework. Journal of Marketing, 83(3), 22–49. https://doi.org/10.1177/0022242919825649

13. Items of the Scale