The news, considered marvelous to some because it seems to confirm how good MRIs are at uncovering secrets of the brain, is that MRIs can predict “sexual orientation” with 92% accuracy. The peer-reviewed paper is “Accurate machine learning prediction of sexual orientation based on brain morphology and intrinsic functional connectivity” in
I'm picking the Schmengee Mediscan 3000™ just for the remote scanning and associated billing opportunities. Dial me up. We can make a mint on the non-medical research front.
That, and no quenching; no gurneys, IV stands, or oxygen tanks killing patients; and no having to refresh the liquid helium. And no size restrictions on the magnet bore, which is a real consideration these days. I have stories.
Agreed, 93% is the high end of this model. The 95% for the normals is the high end, which brings us back 92% + or - 9.8%. After reading your excellent article, my barely educated guess would be that this MRI’s accuracy would be 85% + or - 3% if tested on a sample of 10,000 people who exactly fit the national statistical profile of 7.1% other orientations in the real world. This test could be repeated 5 or 10 times to get a better sense of accuracy. Of course, the chances of them testing this machine in such sample is much less than 1%. This a great lesson about statistical models, funding for science, and the current state of science. Thank you!
Too many "soft studies" (i.e. not-Physics, not-Chemistry) think functional MR and PET-CT (nuclear medicine) are magic 8-balls, and unfortunately, no small number are health care providers. Researchers and journalists are the people Mr Barnum wished he had for customers.
Can I invest? I suggest we incorporate in the Grand Caymans. If that’s too close to home, there’s always the Cook Islands (near French Polynesia). They generally disregard foreign court orders, and thanks to their discretionary laws it’s illegal to disclose any names of information related to Cook trusts.
" Why would you want to use an MRI when the Schmengee is better?"
Because this is how I pay tribute to THE CULT OF DECEPTION which I have joined by preferring bribes over truth. As I have been scared to death by all the other CULT MEMBERS which have convinced me about their own conviction of the existence of death. Which hasn't left me with any other choice but obfuscating the truth for FIAT CURRENCY.
Well done example of modeling nonsense William especially the part about the Schmengee Mediscan 3000! Reminds one of Calvin and Hobbs Transmogrifier. Any well-crafted model might indeed boast near 100% accuracy on in-sample data. Yet, don’t be fooled by this superficial success. Any mathematician worth their salt understands that its smoke and mirrors at work here. It's all too easy to manipulate a model to achieve impeccable in-sample fit, but applied to real-world data is another story.
Well, this is compelling data indeed Mr. Briggs. I'm to meet with our Head of Procurement and Chief Equivocation Officer first thing on Monday. We will need one more bit of data, namely the name of the patent office you filed with for the Mediscan 3000. Not sure how to put this ... rather sensitive matter for a public forum ... but given the rise of offshore patents in recent years, I must inform you from the start that we are quite leery of equipment whose patents are registered with the Patents Ministry of Lower Slobbovia. I am confident, however, that this concern does not apply to the Schmengee.
I'm picking the Schmengee Mediscan 3000™ just for the remote scanning and associated billing opportunities. Dial me up. We can make a mint on the non-medical research front.
That, and no quenching; no gurneys, IV stands, or oxygen tanks killing patients; and no having to refresh the liquid helium. And no size restrictions on the magnet bore, which is a real consideration these days. I have stories.
I want to know about the 101.89%
I had no idea until now that the Schmengee Brothers were scientists/inventors in addition to being first-rate accordionists. Renaissance men.
Watching men throw a baseball is also 93% accurate in predicting orientation.
Agreed, 93% is the high end of this model. The 95% for the normals is the high end, which brings us back 92% + or - 9.8%. After reading your excellent article, my barely educated guess would be that this MRI’s accuracy would be 85% + or - 3% if tested on a sample of 10,000 people who exactly fit the national statistical profile of 7.1% other orientations in the real world. This test could be repeated 5 or 10 times to get a better sense of accuracy. Of course, the chances of them testing this machine in such sample is much less than 1%. This a great lesson about statistical models, funding for science, and the current state of science. Thank you!
Too many "soft studies" (i.e. not-Physics, not-Chemistry) think functional MR and PET-CT (nuclear medicine) are magic 8-balls, and unfortunately, no small number are health care providers. Researchers and journalists are the people Mr Barnum wished he had for customers.
Whatever happened to waiting for the birth of the baby to find out what the gender is going to be? Plus, it drives the liberals nuts!
This is one of my all-time favorite examples!
https://spiral.imperial.ac.uk/handle/10044/1/78872
Can I invest? I suggest we incorporate in the Grand Caymans. If that’s too close to home, there’s always the Cook Islands (near French Polynesia). They generally disregard foreign court orders, and thanks to their discretionary laws it’s illegal to disclose any names of information related to Cook trusts.
The others advantage of the ‘Schmengee Mediscan 3000’ is that it can be used on persons with pacemakers and metallic implants..
" Why would you want to use an MRI when the Schmengee is better?"
Because this is how I pay tribute to THE CULT OF DECEPTION which I have joined by preferring bribes over truth. As I have been scared to death by all the other CULT MEMBERS which have convinced me about their own conviction of the existence of death. Which hasn't left me with any other choice but obfuscating the truth for FIAT CURRENCY.
Mr. Briggs, does your Schmengee Mediscan 3000 have a recipromatic bucky? This is important for me to know.
My god, this is hilarious!
SCTV for the win! LOL Thanks!
Well done example of modeling nonsense William especially the part about the Schmengee Mediscan 3000! Reminds one of Calvin and Hobbs Transmogrifier. Any well-crafted model might indeed boast near 100% accuracy on in-sample data. Yet, don’t be fooled by this superficial success. Any mathematician worth their salt understands that its smoke and mirrors at work here. It's all too easy to manipulate a model to achieve impeccable in-sample fit, but applied to real-world data is another story.
So William, is that Schmengee Mediscan along the lines of the infamous "Bassomatic 76"?
Check it out here: https://www.facebook.com/watch/?v=276052867040301
Well, this is compelling data indeed Mr. Briggs. I'm to meet with our Head of Procurement and Chief Equivocation Officer first thing on Monday. We will need one more bit of data, namely the name of the patent office you filed with for the Mediscan 3000. Not sure how to put this ... rather sensitive matter for a public forum ... but given the rise of offshore patents in recent years, I must inform you from the start that we are quite leery of equipment whose patents are registered with the Patents Ministry of Lower Slobbovia. I am confident, however, that this concern does not apply to the Schmengee.