Imagine I post an article that’s fraudlent and ”proves” that small rocks of a certain size and color keep away tigers. You won’t then see 200 articles saying they tried to replicate this and failed. What you see is 2000 articles dealing with how to find rocks of this exact size and color, siting my paper to support why it’s relevant to be looking for rocks in that category.
A lot of papers that came after were producing results in the framework it established.
Because of the nature of a disease like Alzheimer's not many studies can easily measure the final effectiveness of their addition to the space on patients.
These types of cases tend to focus on the fraud, but then everyone eventually walks away and then there's no larger discussion of how this was enabled by the field to have such a hold.
I don't buy the idea that it's all so quixotic and involves so many variables to get right that who knows what's causing a lack of replication. I've heard whispers like this about other big effects in other fields. People know what's going on, and often it's right there in published results but ignored.
Since papers tend to get published only if they have positive results, what does it mean for thousands of publications all citing a fraudulent paper? This seems really strange. If the first 1000 failed to produce results and were partially based on that original paper it should cast significant doubt on it, but again failures are rarely published.
What does this mean then?