The study they were citing was a typically underpowered mouse study with p-hacked groups. Why did they choose to run 2 groups, each with their own controls, on the same tests? The only significant behavioral difference between the groups was that controls performed better on the memory test.
But 1.5 mg/kg and 8.3 mg/kg are normal human-level AL intakes! And look at those error bars! Also, they ran their analysis completely wrong. This null hypothesis you’re supposed to be testing against is “there is no difference in exploration time (c-a) between each groups”. Instead, they were testing against “there is no difference in exploration time between a and c”. That is, their null hypothesis was that their mice had no memory, they found that their control group definitely did have memory, and concluded that AL had an effect… But then they used the wrong control group? Everything in this study is wrong and the authors and reviewers should feel bad.
Let me reiterate that. THEIR ONLY STATISTICALLY SIGNIFICANT RESULT WAS THAT CONTROL MICE HAD FUNCTIONAL MEMORY. THE ERROR BARS ON TEST GROUP MICE WERE SO HIGH YOU CAN’T TELL EITHER WAY. THEY DID THEIR ANALYSIS WRONG AND THEIR CONCLUSIONS SHOULD BE DISCARDED. So yeah, it’s completely bunk.
https://sci-hub.ru/10.1007/s12640-016-9656-y
The study they were citing was a typically underpowered mouse study with p-hacked groups. Why did they choose to run 2 groups, each with their own controls, on the same tests? The only significant behavioral difference between the groups was that controls performed better on the memory test.
But 1.5 mg/kg and 8.3 mg/kg are normal human-level AL intakes! And look at those error bars! Also, they ran their analysis completely wrong. This null hypothesis you’re supposed to be testing against is “there is no difference in exploration time (c-a) between each groups”. Instead, they were testing against “there is no difference in exploration time between a and c”. That is, their null hypothesis was that their mice had no memory, they found that their control group definitely did have memory, and concluded that AL had an effect… But then they used the wrong control group? Everything in this study is wrong and the authors and reviewers should feel bad.
Let me reiterate that. THEIR ONLY STATISTICALLY SIGNIFICANT RESULT WAS THAT CONTROL MICE HAD FUNCTIONAL MEMORY. THE ERROR BARS ON TEST GROUP MICE WERE SO HIGH YOU CAN’T TELL EITHER WAY. THEY DID THEIR ANALYSIS WRONG AND THEIR CONCLUSIONS SHOULD BE DISCARDED. So yeah, it’s completely bunk.