Exercise 3: Why the 2015 Study Is Evidence FOR Science, Not Against It — Possible Solution ============================================================================================== The classmate's argument treats the reproducibility crisis as proof that science "can't be trusted." But this gets the real logic backward by looking at only half the picture. The Open Science Collaboration study wasn't an outside critic attacking science from beyond its own methods - it WAS science, applying science's own core tool (real, careful replication) to check its own prior work. No comparably rigorous, large-scale, systematic mechanism exists in astrology, homeopathy, or most other pseudoscientific fields to catch and publicly report a two-thirds failure rate in their own claims. The fact that psychology's own researchers deliberately went looking for this problem, measured it precisely, published the uncomfortable result in a top journal, and have since built real reform movements in direct response (Chapter 8's own preprints, open data, and pre-registration) is exactly the self-correcting behavior a genuinely trustworthy field of inquiry is supposed to exhibit. A field that could not be trusted at all would be one where flawed findings simply accumulated silently, with no mechanism ever exposing them - not one that produces a landmark, highly-cited paper doing exactly that exposing. ANSWER: The 2015 study is evidence FOR science's real trustworthiness because it demonstrates science's own self-correcting mechanism working exactly as intended - researchers used real replication to find and publicly report a serious problem within their own field, which is a strength, not a weakness; a field with no way to catch its own errors would be far more concerning than one that actively finds and publishes them. WHY THIS WORKS AS AN ANSWER ------------------------------ This directly rebuts the classmate's framing by identifying the real distinction between "science sometimes produces false findings" and "science has no way of catching false findings" - only the second would actually support the classmate's broader conclusion.