/sci/ - Science & Math

>>5097535
The problem can be read in the OP, and the first post after that.

>>5097509
Let me rephrase that. Do you think you can get rid of all pesticide by washing your corn? They wash them in the factory to begin with. There are always trace amounts.

>>5097494
You think you can sterilize everything by rinsing it with water?

>>5097465
Residues remain on the corn. You get the idea.

>>5097462
Yeah, no, it's not.
>>5097464
If you say so, chief.

>>5097460
I did. Before I even started the thread.

>>5097454
You people can't seem to read. A longitudinal study is not a 13 week study. You have to take the incubation period into account. 13 weeks is enough to test for immediate and medium incubation toxic effects. Not long term carcinogenic effects.

>>5097448
Because no one can seem to point me to a study. Not the posters on 4chan, the bloggers on the rest of the internet, or the news articles talking about how bad the study is. People keep talking about studies showing the opposite, that's fine, show them to me then.

>>5097444
True, I put <span class="math">some[/spoiler] trust in the results. Mainly because it's the only set of results that I have seen thus far. Show me something else to change my mind. I don't pretend these results are conclusive at all. Not in the best of cases. Still nothing else has come up to actually provide evidence for the alternative hypothesis though.

>>5097438
There is a reason to prove them wrong. A huge reason. You don't find it slightly odd that pesticides aren't tested for safe consumption? Especially if there's a risk that they interact with sex hormone metabolic pathways to cause cancer or liver problems. Why do you think we have drug treatment trials? We want to make sure that things are safe to stuff our faces with before we give it out to the masses.

>>5097437
Stop putting words in my mouth.

>>5097435
You propose that for each experimental group, there would be one control group? That's ridiculous. That doesn't solve the problem of statistical underpowerment at all. If anything it will increase the chances of type 2 errors, simply because of variance between control groups. The control group is not big enough, that's the problem. Not the fact that they had nine experimental groups. One group with the same number of animals as in the other groups combined would solve the problem.

>>5097430
How many times do I have to say it: Yes, the statistics are terrible. Seriously, this is like the fifth time. See :>>5097433

>>5097418
>You can't just believe whatever the fuck you want until someone "disproves" your bullshit.
Whoever said that I trust the results? They've given me reasons to doubt, that's all. Point me to a safety study on this crop, and I'll gladly leave with my tail between my legs.

>>5097424
Again, I admit the statistics are terrible (absent?). The fact that they used nine groups in itself is not part of the problem though.

>>5097417
What would be the precise reason for this being statistically terrible? Don't be afraid to go into detail, I know a thing or two about statistics. Multiple comparison problem? That can easily be corrected for. What else?

>>5097411
Right. That's incredibly common practice.

>>5097405
>Each of those articles is recounting what scientists have to say about the paper. It's seriously flawed at best, and deceitful at worst.
Right. But none of those articles actually refer to previous studies which have shown to disprove that this particular corp can cause cancer. If you would read back a bit, you'll see me agreeing that this study isn't particularly good. I'm not denying that the statistics are hugely flawed, and that there are considerable methodological problems. The fact remains, that thus far no one has been able to actually disprove the results. I say again, if you have concrete studies to point me to, I'd appreciate it.

>Are you a moron or a troll?
Talk about the contents. Do anything else one more time, and you'll be ignored permanently.

>>5097396
Unless you have a concrete study, not news articles to link to, or have something to say about the contents of the paper, I'll be ignoring your posts.

>>5097387
>But they already did a whole bunch of past research that disproves the idea. A bad study that contradicts a whole bunch of other studies does not "warrant future research".
What previous research? I asked a few times before, but so far no one has be able to point me to a longitudinal study on this crop, or on this pesticide. If you have a specific study in mind, by all means, post the link. I'd very much like to be corrected on this point.

>That's a terrible reason. Because people often contaminate their data you should too? or at least not take precautions against overfeeding?
You're looking at group differences here. Suppose you use rats or mice that are resistant against cancer, and so their lifetimes are too short for them to develop the necessary pathogenic mutations. Something like would increase the chances of statistical type II errors. It's called the floor effect. You need a population which is predisposed to develop cancer, so that you can interpret group differences as a matter of gradation. Besides, why isn't anyone attacking safety studies that use these types of mice, but when this study uses them all of a sudden, it's a huge problem?

>>5097379
I started this thread to talk about the study specifically. Not all about all the random bullshit that came along with it. If you have something to say about the contents, it would be my pleasure to engage you in delightfully pleasant discourse. Stick to the material published in the article. Otherwise, you shut the fuck up.

>>5097261
Yeah. But the fact that some cases of autism are severely debilitating justifies looking for a cure.

>>5097249
>Quality of life.
My point exactly. The quality of life of people with autism is severely worsened, hence we're looking for a cure.

>>5097237
Let's take that argument to the extreme, shall we? Why are we looking for ways to get people out of comas? Do you think it in itself isn't bad that these people are dependent upon support for the rest of their lives, and cannot live a life at all?