Many seem to believe that researcher-made tests are unnecessary, if not inappropriate, for evaluating reading comprehension interventions. We suggest that this view reflects a zeitgeist in which researcher-made (proximal) tests that align with the researchers’ interventions are closely scrutinized and often devalued, whereas commercially developed (distal) tests, typically unaligned with the researchers’ interventions, escape such examination and judgment. We take issue with the zeitgeist. We object to what appears as an unthoughtful rejection of proximal tests and acceptance of the distal ones. We do so first by discussing the multidimensionality of reading comprehension; then, we present evidence that commercial tests explore few of the construct’s many dimensions, differ among themselves in the dimensions they address, and often have little to do with the aims and substance of researchers’ comprehension interventions. We argue that these facts are reason enough to reconsider commercial tests as the gold standard in program evaluation. Finally, we make a case for the supplemental use of researcher-made tests and offer a framework to help researchers develop tests of reading comprehension that align more and less with their intervention programs.

中文翻译：

---

商业开发的阅读理解测试：黄金标准还是傻瓜黄金？

许多人似乎认为研究人员进行的测试对于评估阅读理解干预是不必要的，如果不是不合适的话。我们认为这种观点反映了一种时代精神，在这种时代精神中，与研究人员的干预措施相一致的研究人员制造的（近端）测试受到了严格审查并且经常贬值，而商业开发的（远端）测试通常与研究人员的干预措施不一致，却逃脱了这种审查和判断。我们对时代精神提出质疑。我们反对看似轻率地拒绝近端测试和接受远端测试的做法。我们首先讨论阅读理解的多维性；然后，我们提供的证据表明，商业测试探索了构造的许多维度中的少数几个，它们所涉及的维度彼此不同，并且通常与研究人员理解干预的目的和内容无关。我们认为，这些事实足以重新考虑将商业测试作为项目评估的黄金标准。最后，我们为补充使用研究人员制作的测试提供了一个案例，并提供了一个框架来帮助研究人员开发与他们的干预计划越来越一致的阅读理解测试。