The research, conducted by Fangwen Zhang and Yang He, who are co-first authors, together with colleagues including corresponding authors Tao Xu and Shengjun Wu, builds on a well-established finding: causal attributions about a recipient’s difficulties can influence helping behavior and donation decisions. But the team noticed that the literature had focused largely on the attributions people reach, rather than on the raw evidence those attributions are built from. Drawing on the perspective of causal reasoning, they distinguished two fundamental types of evidence that people use to infer cause and effect. Covariation evidence concerns how frequently a cause and an effect occur together across cases, the statistical regularity that Hume famously identified as the basis of causal inference. Mechanism evidence, by contrast, concerns the process connecting cause to effect, the how rather than the how often, and is generally regarded in the reasoning literature as the more diagnostic of the two.

To test whether this distinction matters for charitable giving, the researchers ran two experiments. Study 1 recruited 243 participants and employed a two-by-two between-subjects design. The first factor was evidence type, contrasting mechanism evidence with covariation evidence. The second factor was evidence presentation, contrasting a condition in which no causal evidence was provided with a condition in which evidence was presented. Participants were randomly assigned to one of the resulting four groups, and the outcome measure was donation intention, the self-reported willingness to give to the cause described in the experimental materials.

The results of Study 1 were clear on both factors. Donation intention was significantly higher in the mechanism evidence group than in the covariation evidence group, confirming that the two types of causal evidence are not psychologically equivalent when people decide whether to give. Donation intention was also significantly higher in the evidence-present group than in the no-evidence group, indicating that supplying any causal evidence at all outperforms leaving the causal question unanswered. Notably, however, the interaction between evidence type and evidence presentation was not significant. In other words, the advantage of mechanism evidence over covariation evidence did not depend on whether evidence was present or absent, and the benefit of presenting evidence did not differ between the two types. The two factors operated independently, each adding its own contribution to donation intention.

That pattern of independent main effects is theoretically meaningful. If mechanism evidence had mattered only when it was highlighted, or if presenting evidence had helped only for one type of causal information, the interaction would have captured it. Instead, the data suggest that the mere presence of causal evidence lifts donation intention, and that the kind of evidence presented shifts intention further, with mechanism evidence holding the edge. For a charity crafting an appeal, this implies two separable levers: first, explain why the problem exists at all; second, prefer an explanation grounded in a causal process over one grounded in statistical co-occurrence.

Study 2 pushed the question further by treating evidence strength as an experimental variable in its own right. The researchers recruited 276 participants and employed a three-by-three between-subjects design, crossing mechanism evidence strength, at low, medium, and high levels, with covariation evidence strength, also at low, medium, and high levels. Participants were randomly assigned to one of nine experimental groups, allowing the team to examine not only whether each type of evidence matters but how graded increases in evidential force translate into graded changes in willingness to donate, and whether the two types of evidence interact when both are manipulated simultaneously.

The findings again revealed significant main effects of both covariation evidence strength and mechanism evidence strength on donation intention. As the strength of either type of causal evidence increased, donation intention showed a significant positive linear trend. Stronger statistical evidence that a cause and an outcome travel together produced proportionally greater willingness to give, and stronger evidence of an underlying causal mechanism did the same. The dose-response character of these trends is important: it suggests that the effect of causal evidence on donation intention is not a threshold phenomenon that switches on once some minimal credibility is reached, but a graded one in which each increment of evidential strength buys an increment of persuasive power.

Yet the interaction between covariation evidence strength and mechanism evidence strength was again not significant. Even when both types of evidence were manipulated across three levels each, they did not amplify or attenuate one another. The effects appear additive rather than synergistic. Strengthening mechanism evidence raises donation intention by roughly the same amount whether covariation evidence is weak or strong, and vice versa. For theories of causal reasoning applied to prosocial behavior, this is a telling result: the two evidence types seem to feed donation decisions through parallel channels rather than converging into a single integrated judgment whose effect depends on the combination. For practitioners, it means there is no obvious shortcut of pairing a strong mechanism with a strong statistic and expecting a multiplicative boost; each strand of evidence earns its own, independent return.

The authors conclude that, in the experimental contexts examined, donation intention varied with the type and strength of causal evidence. By focusing on the evidence that underlies causal attribution rather than on the attributions themselves, the study extends the literature on charitable giving and offers practical implications for charitable organizations seeking to develop evidence-based donation appeals. The upshot for fundraisers is concrete: appeals that explain the mechanism behind a recipient’s hardship should outperform appeals that merely report how often hardship and a purported cause coincide, and investing in more credible, more forceful evidence of either kind should yield measurable gains in intended giving. The researchers are careful to scope these conclusions to the experimental settings they tested, and donation intention, a stated willingness rather than an observed transfer of money, remains the measured outcome, a standard but real gap between what people say they will do and what they ultimately do at the checkout.

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Review Committee of Xijing Hospital, with all participants providing informed written consent, and the authors declare no competing interests. The work was supported by the Gansu Provincial Basic Research Program-Soft Science Project and the Gansu Provincial University Young Doctor Support Program. Accepted on October 1, 2026 and published on October 19, 2026 in BMC Psychology as an open-access article under a Creative Commons Attribution 4.0 license, the paper is being shared early as a citable, peer-reviewed accepted manuscript. For a field that has long known that causal stories move people to help, the contribution here is to specify the ingredients of those stories, showing that the recipe matters, that the potency of each ingredient matters, and that the ingredients work side by side rather than hand in hand.

Subject of Research: Experimental study of how causal evidence type and strength affect charitable donation intention

Article Title: Causal reasoning and donation intention: an experimental study on evidence type, evidence strength, and evidence combination

Article References: Zhang, F., He, Y., Xu, T., Guo, Q., Zhu, T., Cheng, H., Tao, Y., Zheng, W., & Wu, S. (2026). Causal reasoning and donation intention: an experimental study on evidence type, evidence strength, and evidence combination. BMC Psychology. https://doi.org/10.1186/s40359-026-05754-1

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05754-1

Keywords: donation intention, causal reasoning, mechanism evidence, covariation evidence, evidence strength, charitable giving, causal attribution, experimental psychology, helping behavior, BMC Psychology, fundraising appeals, prosocial behavior