- “Light” Ph.D. — a less expensive research degree
- Umami Burger expands
- A diuretic reduces autism symptoms. Does water balance influence brain function in people without autism?
- This Amazon reviewer is almost always disappointed and his one-star reviews are fun to read. I suggest that ratings (book ratings, product ratings, etc.) compare the rating to other ratings given by the rater. A 5-star rating is more impressive if a rater’s average rating is 2 than if it is 5. I suggest percentiles. For example, rating = 5 (90%ile) is more impressive than rating = 5 (50%ile). I’d also like to know the average percentile across raters.
- Lack of variation in heart rate predicts infection in neonates. The writer (Mike Loukides) is too surprised (“astonishing connection”). Many studies have found associations between too-little variation in heart rate and serious health problems.
Thanks to Adam Clemens and Patrick Vlaskovits.
Okay devil’s advocate on the reviewing concept: I have personally observed that I am more likely to review products I have a problem with. Let’s say this is also true of the serial 1-star reviewer you linked to. Then there are many products that he *would have* rated more highly, had he written a review. If he had included these reviews would it change the meaning of any of his existing reviews? I say no. Therefore looking at his “spread” gives no extra information.
Seth: The interesting comparison is not Person X (fewer reviews) compared to Person X (more reviews). The interesting comparison is Person Y’s five-star review to Person X’s five-star review. Surely you agree that five stars means different things from different people. The additional information I suggest is one way to begin to take account of that — to try to adjust for individual differences.
My filter for amazon reviewers is whether they seem to care about the same things I care about. I’m not interested in how picky they are.
At Amazon, I usually start out looking at the 1-star ratings. Mostly they are worthless, but once in a while you can get a real sense of valid problems. It’s mostly a skimming operation. Then I look at the 2-stars. These seem to be rather better and more thoughtful. Then I look at the 4′s. I find that too many of the 5-stars are either fanboys or brand new users, and most of them sound pretty credulous.
This approach has worked pretty well for me as best I can tell. There are a lot of books I didn’t buy because of it, for one thing.
What I find astonishing about Mike Loukides article is how he can start with a fundamental misunderstanding of heart rate variability (HRV) and weave it into an elaborate story that is disconnected from reality. Heart Rate Variability occurs during respiration with heart rate being higher during inhalation and lower during exhalation. Lack of variation occurs when the body is stressed for a variety of reasons. Machine learning over vast amounts of real time data to detect subtle patterns is an over complication of the lessons learned at Sick Kids. Low HRV is a telltale “spike” that indicates the presence of a stressor. You only need a minute or so of heart rate data to calculate the HRV and the value could easily be reported by an inexpensive device like a finger Pulse Oximeter but they don’t currently. I think the lesson from the Sick Kids research is that HRV is simple and highly predictive compared to other data being gathered. HRV should be watched as closely as heart rate, blood pressure, blood oxygenation, and body temperature in patients. Perhaps IBM is partially at fault for trying to put a Real Time Big Data spin on results that show HRV is a powerful but essentialy low-tech tool.
Seth: Good point.
Shouldn’t a “light” PhD be called something else?
This is not related to the links, but cessation of masturbation and internet porn (abstaining for 4+ days) gave me greatly increased energy. I speculate that porn and masturbation are major dopamine sinks for the brain and when you stop, your dopamine responses get tied to something else.
I am surprised that so few people talk about this. Is it one of those obvious things? It’s an awkward topic and many people assume that you are a religious zealot when you start talking about the downfalls of masturbation.
A PhD is not necessary to get published. I did before getting mine and so have many. Affiliation tends to matter more for those that pervert the peer-review process.
I think RAD has identified why IBM is pushing their big data approach so hard — IBM doesn’t have an algorithm that can calculate HRV efficiently.
The algorithm does exist, though:
https://www.ncbi.nlm.nih.gov/pubmed?term=pd2i
I would disagree, however, with his assertion that HRV is an ‘essentially low-tech tool’ — the math that can extract the HRV needle from the haystack of autonomic noise is very complex.
Bookworm has put up a review of a book by an autistic teenager. It is absolutely stunning. Their minds are all there, but they are unable to express themselves through their bodies.
After the horrors of repetitive rote teaching of knowledge he had already absorbed, he finally was helped by a teacher who understood. He was finally able to communicate through an alphabet board.
https://pjmedia.com/lifestyle/2012/12/10/everything-you-think-you-know-about-autism-is-wrong/#comment-307587
His book:
https://www.amazon.com/exec/obidos/ASIN/0988324709/pjmedia-20
Barry Ritholz has a thoughtful post on Amazon reviews:
https://www.ritholtz.com/blog/2012/12/amazon-changes-its-review-policy/
I sample the 5 & 1 star reviews, tend to be swayed by the quality of the reviewer’s writing. Postmodern approach, which sometimes fails.