131–140 of 161 posts
This kind of makes sense in a strange way. Inexpensive red LED's have relatively narrow spectral line half width (~20nm) and you can set detection thresholds against measured values. This is much better than "I think I saw something in a dimly lit bathroom". This means that in the readout part (UI) it is hard to get wrong results ("Is it one or two?"). Then again it is a wasteful and overpriced product that can leave…
The issue here is that paper tests require the user to read and properly understand instructions. Teenagers in a panic can and will misread and misunderstand, as people often do when under duress. It may seem "pointless", but making the tests more user friendly and less prone to user error is super important.
Here is a paper showing how these tests have 99% sensitivity in the lab, but this drops to 91% when the test is read in the field by a volunteer, and down to 75% when the test is read by the subject themselves: https://archive.is/20081206110632/http://archfami.ama-assn.o... Moreover, if it's such that even a trace of a line indicates a high confidence of pregnancy, users may misinterpret a faint result as somehow mea…
The issue here is that paper tests require the user to read and properly understand instructions. Teenagers in a panic can and will misread and misunderstand, as people often do when under duress. It may seem "pointless", but making the tests more user friendly and less prone to user error is super important.
I wouldn't say super important. I'll probably get down voted for my opinion, but if we had proper sex education for everyone and valued women's health care than this wouldn't be an issue.
While the waste is lamentable, perhaps it's also worth considering this as a UI/UX problem. How many people mis-read the paper test vs how much extra error % does the circuitry introduce?
This is explained in the tweet here: https://twitter.com/RealSexyCyborg/status/130179943992159027...
Earlier quoted context omitted.
> I know that whoever made the test, if it's a reasonably competent enterprise, made it with human capacities in mind. Are you sure they weren't just constrained by technological limitations when they designed the paper test? Maybe the reason the digital test exists is that a competent enterprise wanted to adapt the paper technology to better fit human capacities?
Maybe, although I doubt it. It would mean that they put in extra research to, for example, plot an ROC curve and see what the optimal threshold for the test would be. Even if that was the case, you can still frame the question in terms of cost/benefits. The electronic test is, let's say, 10 times more expensive than a basic one. Is it more reliable than taking 10 basic tests (maybe at different times) and averaging t…
If you want, you can buy 10 cheaper strips, Try different lighting and times of day, then take the numeric average. I can buy a more expensive digital strip simply because I like the box's color.
Earlier quoted context omitted.
Would it be possible to make a paper one have words (such as pregnant/not pregnant) instead of one/two lines?
I don't think that would solve the problem. "The word "not" is faintly showing, what does that mean?" It it was trivial to show a single clear indicator word, they could just show a clear indicator line, then there'd be no ambiguity. It's not the second line that confuses people, it's figuring out if that line is really visible enough.
My 3rd world country has pregnancy tests[1] available in almost all vending machines, but unfortunately they're the paper kind. Obviously we managed to get them evenly distributed, but utterly failed to provide the future. [1] For the convenience of both those hoping for positive results and those hoping for negative. I'm guessing there are more of the latter than the former, because they're invariably located in the…
No reason to assume that being near condoms means people don't want to be pregnant.
The issue here is that paper tests require the user to read and properly understand instructions. Teenagers in a panic can and will misread and misunderstand, as people often do when under duress. It may seem "pointless", but making the tests more user friendly and less prone to user error is super important.
I wouldn't say super important. I'll probably get down voted for my opinion, but if we had proper sex education for everyone and valued women's health care than this wouldn't be an issue.
"Pregnancy tests are 99% accurate in the lab, 75% accurate in the wild due to misreads- mistakes which are highly dependent on education and socioeconomic status. No, it is not stupid or wasteful to use a hardware interface to help women with this." https://twitter.com/RealSexyCyborg/status/130178603213683507...
Had to turn it to racism, not surprised.
Going over the conversation here I'm a bit surprised by how many people seem to prefer an unnecessarily complicated and obfuscating solution to the straight-forward one. The Twitter thread clearly points out that the electronics serve strictly as an intermediary layer between the paper test and the user. They just binarize the paper reading. This in turn gives a false psychological sense of security ("it's the comput…
I actually saw this last night and initially had the same reaction but had some time to think on it. Remember, for a lot of people taking a pregnancy test is incredibly stressful. Reducing the complexity of reading by outputting a less ambiguous result has a lot of value to some people.