Announcement

Collapse
No announcement yet.

Coronavirus update - NO MORE RESTRICTIONS

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Re: Coronavirus update

    Originally posted by CCFCC3PO View Post
    Yes, but my question shows it isn't as clear cut as a weather person getting the weather wrong 70% of the time. It's also lacking an article from the science community, as far as I can tell.

    Basically saying a test result is x% accurate is meaningless in my opinion.

    You need to know if it is specific or sensitive. The other thing that is missing here is the base rate - which is what we don't know with covid. The base rate will be different dependent on if you are testing only symptomatic or sick people, or if you are testing the whole population. If you are only testing people who you believe have it, then the false negative rate will be higher (in other words, the specificity of the test becomes skewed towards the 70% figure you are quoting).

    If you are testing people who most likely don't have it, then you have more potential of false positive tests. However, there is no way of knowing that a test subject definitely doesn't have the virus.

    I put very little weight on the 70% accuracy figures based on the above, and also on the lack of citation from the science community.
    Have you seen the WSJ article ?

    Comment


    • Re: Coronavirus update

      Originally posted by trampie09 View Post
      Have you seen the WSJ article ?
      No, I was waiting for a link.

      Comment


      • Re: Coronavirus update

        Originally posted by CCFCC3PO View Post
        No, I was waiting for a link.
        I've told you how to find it and I can't get in now, I only wanted to find references to the 70% thing for you which I did, I only glanced at the Wall Street thing and it was all algebra, logarithms, equations etc, I don't know but I would be surprised if it did not meet your criteria.
        I seen an expert on TV saying something along the lines of if somebody has the full blown symptoms he would be 90% certain that person has it yet the test we were using was only 70% accurate, and he apparently is an expert and he was talking about the 70% success rate or should that be ,30% failure rate.

        Comment


        • Re: Coronavirus update

          Originally posted by trampie09 View Post
          I've told you how to find it and I can't get in now, I only wanted to find references to the 70% thing for you which I did, I only glanced at the Wall Street thing and it was all algebra, logarithms, equations etc, I don't know but I would be surprised if it did not meet your criteria.
          I seen an expert on TV saying something along the lines of if somebody has the full blown symptoms he would be 90% certain that person has it yet the test we were using was only 70% accurate, and he apparently is an expert and he was talking about the 70% success rate or should that be ,30% failure rate.
          I have found it. The table is based on 70% sensitivity and 95% specifity. It doesn't take into account the unknown base rates, and it is based on a report from China that tests could have a 70% sensitivity rate.

          The article is also a month old, but that is not my issue with the 70% figure. My issue is that, without knowing the base rate of the sub population being tested, and without considering there will be more false negatives if tests are only carried out on a population based on the assumption that the subject is likely to have covid at the time of testing.

          Comment


          • Re: Coronavirus update

            Originally posted by CCFCC3PO View Post
            I have found it. The table is based on 70% sensitivity and 95% specifity. It doesn't take into account the unknown base rates, and it is based on a report from China that tests could have a 70% sensitivity rate.

            The article is also a month old, but that is not my issue with the 70% figure. My issue is that, without knowing the base rate of the sub population being tested, and without considering there will be more false negatives if tests are only carried out on a population based on the assumption that the subject is likely to have covid at the time of testing.
            You disbelieve the 70% figure ?

            What base rate are you talking about ?

            Do you know if they did or didn't take the unknown base rate into consideration ?

            Comment


            • Re: Coronavirus update

              Originally posted by trampie09 View Post
              You disbelieve the 70% figure ?

              What base rate are you talking about ?

              Do you know if they did or didn't take the unknown base rate into consideration ?
              As I said earlier, the 70% accuracy is not relevant unless we know what the base rate is. And, with this virus, we won't know the base rate. Ergo, I am not putting any value on the 70% accuracy. Of course, if you can provide links that show the specificity and sensitivity of the various tests, based on different populations and with a known base rate, I am open minded enough to have my mind changed.

              I have shown why I am skeptical about the 70% figure, can you explain why you are certain the figure is correct?

              Comment


              • Re: Coronavirus update

                Originally posted by trampie09 View Post
                You disbelieve the 70% figure ?

                What base rate are you talking about ?

                Do you know if they did or didn't take the unknown base rate into consideration ?
                Also what is sensitivity and specificty in the context you are using it ?

                Comment


                • Re: Coronavirus update

                  I've been looking but I still cannot find any information on how one volunteers to be one of the 'army' of people to monitor this tracing app. No info on it what so ever!
                  Can anyone steer me in the right direction?

                  Is there a Welsh government hotline or something?

                  Comment


                  • Re: Coronavirus update

                    Originally posted by CCFCC3PO View Post
                    As I said earlier, the 70% accuracy is not relevant unless we know what the base rate is. And, with this virus, we won't know the base rate. Ergo, I am not putting any value on the 70% accuracy. Of course, if you can provide links that show the specificity and sensitivity of the various tests, based on different populations and with a known base rate, I am open minded enough to have my mind changed.

                    I have shown why I am skeptical about the 70% figure, can you explain why you are certain the figure is correct?
                    No no butt, you have showed Porthcawl.
                    What is this base rate you refer too ?
                    What is specificty and sensitivity you refer too ?

                    If you are not putting a value on anything then your posts are meaningless, if you have no idea if the accuracy of the tests are very close to the 70% figure often routinely quoted, or if they are higher or lower, then you going on about the testing accuracy is pointless.

                    I'm not saying the figure is or isn't correct but it seems to be widely accepted, you dispute it yet you don't say if in your view the figure should be higher or lower.

                    Comment


                    • Re: Coronavirus update

                      Originally posted by xsnaggle View Post
                      I've been looking but I still cannot find any information on how one volunteers to be one of the 'army' of people to monitor this tracing app. No info on it what so ever!
                      Can anyone steer me in the right direction?

                      Is there a Welsh government hotline or something?
                      My sister is doing it, they’ve moved people who were doing the phones onto the app.

                      Comment


                      • Re: Coronavirus update

                        Originally posted by trampie09 View Post
                        Also what is sensitivity and specificty in the context you are using it ?

                        A sensitive test will correctly identify people with covid. Sensitivity measures correct positive results.

                        If a test is 90 percent sensitive, it will correctly identify 90 percent of people who are infected. However, 10 percent of people who are infected and tested would get a false negative result—they have the virus, but the test said they don’t.

                        A specific test will accurately identify people without covid-19. Specificity measures correct negatives.

                        If a test is 90 percent specific, it will correctly identify 90 percent of people who are not infected. However, 10 percent of people who are not infected will test positive for the virus and receive a false positive.

                        Comment


                        • Re: Coronavirus update

                          Originally posted by xsnaggle View Post
                          I think that some people don't want pubs to re-open, whether they are the biggest source of infection or not. As I said I believe that there are people who just are happy to see them shut and if they had their way never reopen, and I believe that such people hold a disproportionate influence in wales.
                          'im not arguing one way or the other, it is just my opinion and I'm not going to change it.
                          I just don’t understand what’s made you think this? The majority of the country loves a pub whatever class they are and they bring in a lot of money.

                          The only reason they aren’t open is because of the virus.

                          Seriously what people in positions of power don’t want them open?

                          Comment


                          • Re: Coronavirus update

                            Originally posted by trampie09 View Post
                            No no butt, you have showed Porthcawl.
                            What is this base rate you refer too ?
                            What is specificty and sensitivity you refer too ?

                            If you are not putting a value on anything then your posts are meaningless, if you have no idea if the accuracy of the tests are very close to the 70% figure often routinely quoted, or if they are higher or lower, then you going on about the testing accuracy is pointless.

                            I'm not saying the figure is or isn't correct but it seems to be widely accepted, you dispute it yet you don't say if in your view the figure should be higher or lower.
                            I have already explained what the base rate is, and how not knowing the base rate (which we don't) will effectively mean the "70% accuracy" figure is meaningless.

                            I have also explained how, in testing only people believed to have the infection, we will get a higher number of false negatives by the very nature of the testing parameters.

                            I have explained, in a few posts, why I am not giving too much credence to the 70% accuracy figure. You haven't explained (other than "a lot of people say so, but I can't tell you who") why you are putting a lot of credence into the 70% accuracy figure.

                            Comment


                            • Re: Coronavirus update

                              Originally posted by CCFCC3PO View Post
                              I have already explained what the base rate is, and how not knowing the base rate (which we don't) will effectively mean the "70% accuracy" figure is meaningless.

                              I have also explained how, in testing only people believed to have the infection, we will get a higher number of false negatives by the very nature of the testing parameters.

                              I have explained, in a few posts, why I am not giving too much credence to the 70% accuracy figure. You haven't explained (other than "a lot of people say so, but I can't tell you who") why you are putting a lot of credence into the 70% accuracy figure.
                              What is the base rate you refer too ???

                              Comment


                              • Re: Coronavirus update

                                Originally posted by CCFCC3PO View Post
                                Much of the public appear to be following the rules that best suit them - i.e. Johnson's.

                                Today I have seen a house with 8 people in the front yard. The house is lived in by an elderly couple, so they have 6 people visiting.

                                The Barry Island live webcam sees a dozen people outside O'Shea's fish and chip shop on the promenade. The fish shop has changed its message from "stay home" to "Welcome to O'Sheas". Barry Island has been dead of late, it is definitely busier than this time last week (with similar weather).

                                The Stay at Home message that the Welsh Government are churning out is rendered invalid when trips to garden centres are allowable. It's also been rendered invalid by Johnson's completely incomprehensible message, a message that required a 50 page document to clear up!

                                This wave isn't even finished, hopefully we will avoid a second wave but today it looks like an increasing number of people are becoming more blase about social distancing.

                                Drakeford makes no sense, he said a few weeks ago he was waiting for 5 key points to occur, which they already are, yet instead of easing the lock down he hasn’t. He wants to be like New Zealand now, despite the fact they were in lockdown for less time than us, I didn’t realise Wales was an island 500 miles off the coast of England either.....the blundering fool is going to see thousands on the dole I tell u and then we will get two fingers back from boris......

                                Comment

                                Working...
                                X