Designing messages which appeal to the generosity of prospective donors is critically important to those in harm's way, and secondly to those who advocate for their care, such as nonprofit organizations (NPOs). Contributions needed to combat disease and malnutrition are far in excess of the relief agencies' financial resources.

Donors typically respond first to the appeals of family and friends, before those of strangers. This suggests that people categorize others as socially close or distant. Donors also typically respond first to urgent appeals for humanitarian aid in response to large-scale natural catastrophes, before they will donate to less urgent appeals such as economic development or a capital campaign. This suggests that people categorize situations on a temporal basis. Two psychological distances (social and temporal) are theorized by construal level theory (CLT) to create different mental representations of an object. CLT's explanation of this phenomenon is based on a representational theory of the mind, rooted in cognitive science, which takes on the form of semantics to describe mental states such as the accuracy and truthfulness of thoughts, beliefs, desires, perceptions, and images about an object . CLT theorizes that mental representations are formed in concrete language (low-level construals), or abstract language (high-level construals). The main hypothesis is that donors evaluate charitable messages more favorably when the diction matches their construal.

A 2 x 2 x 2 factorial design successfully manipulated three independent variables into contrasting levels: social distance (near-distant) as the between-subject variable; with temporal distance (urgent/less-urgent) and diction (abstract-concrete) as the two within-subject variables. The dependent variables are: (a) financial donations toward the charitable appeal, and (b) self-report of motivation based upon the value-expressive attitude function versus the utilitarian attitude function. Data was solicited among 2,076 employees of a major international firm headquartered in the Mid-Atlantic region and randomly assigned to one study: a manipulation check (N=124), a revised check (N=26), a pretest (N=26), and the main experiment (N=221). A random drawing for $1,000 allowed respondents to realistically consider their allocation to a charity, as if the money were their own. Evidence to support the presence of mental representation as defined by CLT would have appeared as higher donations in the treatments which matched the diction (level of abstraction) to the construal. A three-way ANOVA of mean donor contributions determined that social distance and temporal distance interacted to significantly influence donor behavior independently of the use of abstract or concrete language in the stimuli materials, which appear in the form of magazine advertisements. Evidence does not support the contention that people produce low- or high-level construals. Significant but low correlations were observed between larger contributions and (a) higher household incomes, (b) extrinsic religious orientation, (c) interdependent self construal, and (d) personal involvement with charitable issues; but none for (e) value-expressive attitudes.