Some human behaviors have serious societal consequences, but these consequences tend to be neglected in online research on societally relevant behaviors. For example, human activities contribute to climate change and biodiversity loss, but pro-environmental behavior is often studied using inconsequential self-reports and hypothetical scenarios. Such measures can easily be administered online, but suffer from severe validity problems. To address these problems, we developed a multi-trial web-based procedure for the study of consequential pro-environmental behavior. On the Work for Environmental Protection Task (WEPT), participants can choose to exert voluntary extra efforts screening numerical stimuli in exchange for donations to an environmental organization. They thus have the opportunity to produce actual environmental benefits at actual behavioral costs (i.e., to show actual pro-environmental behavior). In a preregistered validation study (N = 209), we found WEPT performance to systematically vary with these consequences, that is, the implemented costs and benefits were large enough for participants to effectively take them into account. In addition, aggregated WEPT performance was found to be highly reliable and to be correlated to self-reports and objective observations of other pro-environmental behaviors and conceptually related measures. These findings support the validity of the WEPT as an online procedure for the study of actual pro-environmental behavior. We discuss how the WEPT can advance the experimental analysis of pro-environmental behavior, help to address problems of common-method variance in individual difference research, and be adapted for the consequential study of other societally relevant behaviors.

一些人类行为具有严重的社会后果，但在对社会相关行为的在线研究中，这些后果往往被忽视。例如，人类活动导致气候变化和生物多样性丧失，但通常使用无关紧要的自我报告和假设情景来研究环保行为。这些措施可以很容易地在线管理，但存在严重的有效性问题。为了解决这些问题，我们开发了一个基于网络的多试验程序，用于研究相应的亲环境行为。在环保工作任务 (WEPT) 中，参与者可以选择自愿付出额外的努力来筛选数字刺激，以换取对环保组织的捐赠。因此，他们有机会以实际的行为成本产生实际的环境效益（即展示实际的环保行为）。在一项预先注册的验证研究中（N = 209），我们发现 WEPT 的性能会随着这些后果而系统地变化，也就是说，实施的成本和收益足够大，参与者可以有效地考虑到它们。此外，发现综合 WEPT 性能高度可靠，并且与自我报告和其他亲环境行为的客观观察和概念相关措施相关。这些发现支持 WEPT 作为研究实际亲环境行为的在线程序的有效性。我们讨论了 WEPT 如何推进亲环境行为的实验分析，帮助解决个体差异研究中的共同方法差异问题，并适用于其他社会相关行为的相应研究。