College this fall--parents' perspective

[Vkw10]

Sister-in-law texted photo of "TigerTown Bound Class of '24" yard sign that Clemson's Alumni Association mailed to nephew. Anyone know if such yard signs are a tradition? Or is this a new tactic to encourage enrollment?

I've never seen such, but I've been away from the area a long time.

[hmaria1609]

Sister-in-law texted photo of "TigerTown Bound Class of '24" yard sign that Clemson's Alumni Association mailed to nephew. Anyone know if such yard signs are a tradition? Or is this a new tactic to encourage enrollment?
I've never seen such, but I've been away from the area a long time.

A neighbor on another street had a sign like that in the front yard--it was for a college in KY.  Also there are signs for graduates of two local high schools.

[mythbuster]

We have yard signs in our neighborhood announcing who is graduating from high school. Not so much where they are headed. But I like the sentiment behind these.

[polly_mer]

On topic:  What would students do if they aren't in college?  Well, if colleges aren't really open for f2f, then likely not much of anything else is open for f2f and so a gap year of reading, family time, and just hanging out is what most of us will be doing anyway if we have the luxury of food, shelter, and basic needs being met.

Since I'm here and my professionalism has been questioned, I'm going to do science discussion for those who want to actually learn instead of immediately dismiss anything that contradicts their world view.

After leaving Super Dinky, I have been working in the area of VVUQ of scientific computational models.  The overly simplified explanation of VVUQ in this context is:

Verification: is the computer code right?

Validation: How good is the model compared to any sort of experimental or other relevant data out in the world?

Uncertainty Quantification: What kind of error bars do we have to put on the inputs to the models to have good estimates on uncertainty on the outputs?  How do we get the information to put either type of error bars on the relevant input/outputs, especially when we can't run the experiments with control on enough of the variables?

I don't just "work in the same building as people who study diseases".  I have had extended discussions for all the time since I left Super Dinky on what goes into models, what assumptions are made, and what do people do when they can't run standard experiments that control for all the variables.  Some of those conversations have exactly been on spread of disease and how one gets a handle on predicting how bad something will be based on information other than direct testing.  Waiting for direct testing often means waiting until the problem is really bad and impossible to ignore.  More useful all around are the constant monitoring of public big data to see certain patterns to have local health officials intervene before the problem is obvious to the casual observer.

A very common problem discussed regularly in scientific circles is the huge disconnect between what the peer-reviewed scientific article actually states with all the caveats and what is reported in the popular media.  The full-text of The Lancet article from The Guardian is publicly available so we can discuss the actual text, not the filtered media report.

The abstract itself leads to some immediate questions.

In response to the coronavirus disease 2019 (COVID-19) pandemic, 107 countries had implemented national school closures by March 18, 2020. It is unknown whether school measures are effective in coronavirus outbreaks (eg, due to severe acute respiratory syndrome [SARS], Middle East respiratory syndrome, or COVID-19). We undertook a systematic review by searching three electronic databases to identify what is known about the effectiveness of school closures and other school social distancing practices during coronavirus outbreaks. We included 16 of 616 identified articles. School closures were deployed rapidly across mainland China and Hong Kong for COVID-19. However, there are no data on the relative contribution of school closures to transmission control. Data from the SARS outbreak in mainland China, Hong Kong, and Singapore suggest that school closures did not contribute to the control of the epidemic. Modelling studies of SARS produced conflicting results. Recent modelling studies of COVID-19 predict that school closures alone would prevent only 2–4% of deaths, much less than other social distancing interventions. Policy makers need to be aware of the equivocal evidence when considering school closures for COVID-19, and that combinations of social distancing measures should be considered. Other less disruptive social distancing interventions in schools require further consideration if restrictive social distancing policies are implemented for long periods.

https://www.thelancet.com/journals/lanchi/article/PIIS2352-4642(20)30095-X/fulltext

This is a study of studies that only included a tiny fraction of the articles available (16 of 616).   Why so few articles included?  What happened to the others and why are they not relevant? 

No COVID-19 data related to school closures go into the modeling of the effect of closing schools on the spread of COVID-19.  From a simulation validation viewpoint, that's a big red flag on credibility for the model for the stated purpose.  Where they have data on SARS, the models don't align with the data, which again calls into question the validity of the models for this specific purpose.

In plain English, the statement that goes into The Guardian headline is purely based on models that have not been validated for this purpose, even using the SARS data (again, it's not good that the model results don't match the data we have).

Moving on to the meat of the article, the introduction relies heavily on influenza studies that show closing the schools matters if the closing happens early enough.  However, for an article in a medical journal, there's a lot of editorializing about economic effects in the introduction.

Regarding error bars on relevant inputs,

However, in the COVID-19 pandemic thus far, children appear to form a much lower proportion of cases than expected from their population, although evidence for this is mixed and some data suggest that children might be as likely to be infected as adults but largely remain asymptomatic or have a mild form of the disease [26]

From an uncertainty on the model inputs view, one might wonder if the lack of observed transmission from children to the general population is more a lack of knowledge of who is sick than not physically transmitting through children.  The higher-than-expected asymptomatic rates of infection will be a contributing factor to the uncertainty in transmission rates, particularly for children.

There's a whole paragraph of explanation on why we don't have data on COVID-19 school-related infection rates, including the fact that the schools closed very early in many places.  As one of my VVUQ colleagues is fond of saying, not having the information to put an error bar on an input should not then default to a guess on the value of the input with an error bar of zero.

In the methods section, the statement appears "We did not attempt to rate the quality of included studies in this Review."  Again, from a VVUQ standpoint, one generally tries to get the best data available, not just any old data.

In the results section, we find out that all the published articles regard the 2003 SARS outbreak, a preprint looked at other non-epidemic coronaviruses, and only 6 studies actually studied COVID-19, none of which have yet been published.  The studies that are actually relevant (school closures due to the current COVID-19 epidemic)

No data are available on the effectiveness of school closure as there was little variation in timing of closures (closures were reportedly applied in all Chinese cities uniformly and without delay) and school closures were part of a broad range of quarantine and social distancing measures. Both of these studies concluded that the overall package of quarantine and social distancing was effective in reducing the epidemic in mainland China,41,  42 although the relative contribution of school closures was not assessed.

Preprint studies of actions in Hong Kong related to COVID-19 noted that a 4-week school closure was initiated across the city on Feb 1, 2020, approximately 1 week after the first cases were identified in Hong Kong. School closures were implemented at the same time as a number of other stringent social distancing measures, with school closures extended initially to March, 2020, then to April, 2020.12,  44 Collectively, these measures were considered to have reduced the R below 1, controlling the spread of the outbreak.12 As in mainland China, no data were available from either paper on the effect of school closures separate from other measures. Cowling and colleagues12 noted that the social distancing measures implemented during the COVID-19 outbreak reduced community transmission by 44%, which was much greater than the estimated 10–15% reduction in influenza transmission conferred by school closures implemented alone during the 2009 pandemic in Hong Kong.44

The results of the studies that matter don't support the bare notion that school closures have minimal effect on transmission, although it is true that school closures were not studied separately from other aspects of simply closing the whole nation.


The paragraph that gets details ignored in The Guardian article is

Only one study examined the effect of school closures separately to other social distancing measures. In a non-peer-reviewed but widely cited report from an established group, Ferguson and colleagues31 modelled the estimated effects of a range of different social distancing measures and combinations of measures. They used UK population and schools data together with data on transmission dynamics reported from the COVID-19 outbreak in Wuhan. Using data from previous influenza outbreaks, they assumed that per-capita contacts within schools were double those in households, workplaces, or the community, and that, overall, approximately a third of transmission occurred in schools. They modelled a scenario in which all schools and 25% of universities were closed and where the effect on non-school social contacts was an increase of 50% in household contact rates for families with children and a 25% increase in community contacts during the closure. They concluded that school closure as an isolated measure was predicted to reduce total deaths by around 2–4% during a COVID-19 outbreak in the UK, whereas single measures such as case isolation would be more effective, and a combination of measures would be the most effective. The authors concluded that school closure is predicted to be insufficient to mitigate (never mind suppress) the COVID-19 pandemic in isolation, which is in contrast to seasonal influenza epidemics where children are the key drivers of transmission.31

From a layperson standpoint, this seems to be an argument against using school closures in isolation with no other actions taken. The VVUQ question is how good is mixing the Wuhan COVID-19 parameters with influenza parameters?  Is the model itself set up for UK behaviors with some unknown mixture of Chinese behaviors that influenced the output?  Yes, reported total death reduction is a relatively small range, but is that truly reflective of propagating the total uncertainty of the inputs or is that another case where when one doesn't know the uncertainty, one just puts in zero?

In other words, yeah, an article published in a reputable medical outlet was reported and discussed in a reputable news outlet, but the science involved is not really what the news would have the casual reader believe.  I see no scientific evidence supporting opening K-12 schools because they will be fine and not contribute greatly to overall deaths, but I see some evidence that a group motived by worry about economic decisions did something resembling science to support their predetermined view.

[marshwiggle]

Only one study examined the effect of school closures separately to other social distancing measures. In a non-peer-reviewed but widely cited report from an established group, Ferguson and colleagues31 modelled the estimated effects of a range of different social distancing measures and combinations of measures. They used UK population and schools data together with data on transmission dynamics reported from the COVID-19 outbreak in Wuhan. Using data from previous influenza outbreaks, they assumed that per-capita contacts within schools were double those in households, workplaces, or the community, and that, overall, approximately a third of transmission occurred in schools. They modelled a scenario in which all schools and 25% of universities were closed and where the effect on non-school social contacts was an increase of 50% in household contact rates for families with children and a 25% increase in community contacts during the closure. They concluded that school closure as an isolated measure was predicted to reduce total deaths by around 2–4% during a COVID-19 outbreak in the UK, whereas single measures such as case isolation would be more effective, and a combination of measures would be the most effective.

The 2-4% of deaths being single teachers who live alone and have no social contacts, so the only vector they have for getting infected is from students. Seriously, school closures as an isolated measure is ridiculous. And similarly, allowing schools to open as an isolated measure is also ridiculous unless kids are basically immune. Even if kids have an infection rate WAY below that of adults, becuase they have so much  interaction with so many other students at school, simple math dictates that they would represent a major disease transmission vector.

(If you know the "two people with the same birthday" game, you know that it only takes about 26 people for a greater than 50% chance that two will have the same birthday, even though in principle any particular person has only 1/365 chance of being born on a partciular date. So, even if there's only a 1% chance of any interaction passing on the virus, with the number of interactions that happen in a day a lot of transmission will occur.)

[Caracal]

Edited to add: And on the subject of children transmitting Covid-19, if they didn't, and there was evidence that they didn't, then nursing and retirement homes would be throwing open their doors to become day care facilities for pre-schoolers. Parents have first-hand experience with the disease-transmitting capabilities of small children who congregate in groups, as happens at schools.

I looked at this more and hadn't realized how different this was from the flu in terms of transmission as well as severity and the implications of that. On average, the flu has an R of 1.3. Someone, correct me if this is imprecise, but that is how many people on average an infected person is going to transmit the disease too. That isn't actually all that infectious and suggests that you aren't really aren't all that likely to get the flu from people who aren't putting their hands all over your face, blowing snot in your direction or hanging around your house all day with their germs. You can see why kids contribute so much to flu outbreaks. If you just closed the schools for three weeks whenever the flu got widespread, you'd  bring numbers way down without doing anything else.

COVID is, unfortunately, way more transmissible. The R is perhaps 2.3 or even higher.  That's bad, obviously, but it also really changes the approach towards kids and schools even if it turns out that kids are as effective as transmitting the virus as adults (which they may not be) Since the virus has so many other ways it can effectively be transmitted, school contact isn't as big a part of the overall picture. That's why that Lancet article is estimating that closing schools might reduce overall mortality by only 2-4 percent overall.

[pgher]

15 Fall Scenarios

I suppose what we need to do (as a family) is discuss some large fraction of these 15 scenarios to be prepared for whatever gets announced.

[spork]

[. . .]

As one of my VVUQ colleagues is fond of saying, not having the information to put an error bar on an input should not then default to a guess on the value of the input with an error bar of zero.

[. . .]

Or, sort of rephrased, the principle of "absence of evidence is not evidence of absence" -- which is how I got my PhD, because my dissertation committee was oblivious to it.

That study published by The Lancet is trash.

[Caracal]

From a layperson standpoint, this seems to be an argument against using school closures in isolation with no other actions taken.

In other words, yeah, an article published in a reputable medical outlet was reported and discussed in a reputable news outlet, but the science involved is not really what the news would have the casual reader believe.  I see no scientific evidence supporting opening K-12 schools because they will be fine and not contribute greatly to overall deaths, but I see some evidence that a group motived by worry about economic decisions did something resembling science to support their predetermined view.

Well sure, I think the original choice made almost everywhere to close schools made sense. You're trying to rapidly drive down transmission rates. The question after that becomes how do you balance trying to keep transmission rates down with other costs. The explanation on modeling is helpful. I think you're right that one of the problems with modeling is that it can look to non experts like a sort of magic, when it is obviously a bunch of decisions about what to factor in and what not to that can all be questioned. I'm not sure how fair it is to complain that they are relying on lots of studies of spread of other things besides Covid-19. How could they not be. There have been decades of studies on the transmission of flu, and this is a new disease that showed up five months ago, you aren't going to have enough data to model anything without taking those previous models and adjusting them based on the known characteristics of this new respiratory infection. My understanding is that most of the models being used to inform decisions are based on flu models. I know that the influential one by Neil Ferguson who is quoted in the Guardian article is.

For the actual questions, the point is that it actually contested and complicated. There are good reasons to think that schools aren't going to have the same outsized impact they do in the transmission of flu. From there it gets into questions of what those impacts are, which involves a lot of unknowns. I'd point out that you started this off questioning the very idea that there was any reason to doubt that schools and kids weren't a big source of transmissions and claiming that that I must be reading unreliable sources and now you are discussing the modeling assumptions of an article in the Lancet. Next time, maybe you could just write "I hadn't seen that, could you send me a link on the reporting of that?" I find your perspective on this stuff helpful, just not the tone.

[marshwiggle]

For the actual questions, the point is that it actually contested and complicated. There are good reasons to think that schools aren't going to have the same outsized impact they do in the transmission of flu.

But closing down schools an an isolated measure isn't even possible logistically. Who's going to care for those kids at home? Who are they going to interact with in that time that they're not at school?

The point is, it's not even really rational to try and consider the transmission of disease within schools unless you can include consideration of all of the social changes resulting from kids being out of school.

Trying to isolate variables in a model is highly misleading without understanding how interconnected all of the variables are in practice.

[Caracal]

For the actual questions, the point is that it actually contested and complicated. There are good reasons to think that schools aren't going to have the same outsized impact they do in the transmission of flu.

But closing down schools an an isolated measure isn't even possible logistically. Who's going to care for those kids at home? Who are they going to interact with in that time that they're not at school?

The point is, it's not even really rational to try and consider the transmission of disease within schools unless you can include consideration of all of the social changes resulting from kids being out of school.

Trying to isolate variables in a model is highly misleading without understanding how interconnected all of the variables are in practice.

Not sure I really get the point. Obviously closing down schools without doing anything else wouldn't have made much sense. And if I understand you right, I agree that closing them helped to shut down lots of other contacts, for example between and among parents and teachers. For that reason, I can't imagine you would see a scenario in which schools reopened but nothing else did. But, obviously it matters how much kids transmit at school and where that fits within the overall picture of transmission in terms of trying to figure out how to move to a new normal.

If kids are a big driver of overall transmission than schools could really only be reopened if you had very aggressive testing and contact tracing and even then you might need to make major changes to avoid big outbreaks showing up really quickly. If it turns out that kids actually don't transmit the virus as efficiently as adults, or if the larger dynamics mean that schools are low on the list of places where you get a lot of spreading, then you could reopen schools with more limited precautions.

[marshwiggle]

For the actual questions, the point is that it actually contested and complicated. There are good reasons to think that schools aren't going to have the same outsized impact they do in the transmission of flu.

But closing down schools an an isolated measure isn't even possible logistically. Who's going to care for those kids at home? Who are they going to interact with in that time that they're not at school?

The point is, it's not even really rational to try and consider the transmission of disease within schools unless you can include consideration of all of the social changes resulting from kids being out of school.

Trying to isolate variables in a model is highly misleading without understanding how interconnected all of the variables are in practice.

Not sure I really get the point. Obviously closing down schools without doing anything else wouldn't have made much sense. And if I understand you right, I agree that closing them helped to shut down lots of other contacts, for example between and among parents and teachers. For that reason, I can't imagine you would see a scenario in which schools reopened but nothing else did. But, obviously it matters how much kids transmit at school and where that fits within the overall picture of transmission in terms of trying to figure out how to move to a new normal.

I'll be a bit more blunt. Of all of the current disruptions, kids being out of school is a big one. Parents are on the hook for their education and supervision 24/7, even when they're having to work from home. So if if were possible to have kids back in school, that would make a lot of people happy. And it's tempting to consider since kids are under-represented in covid-19 cases needing serious medical intervention. So this is one of those cases where there is a lot of incentive to misuse statistics to tell people what they want to hear.

If kids are a big driver of overall transmission than schools could really only be reopened if you had very aggressive testing and contact tracing and even then you might need to make major changes to avoid big outbreaks showing up really quickly. If it turns out that kids actually don't transmit the virus as efficiently as adults, or if the larger dynamics mean that schools are low on the list of places where you get a lot of spreading, then you could reopen schools with more limited precautions.

But even if kids transmit the virus at a much lower rate than adults, because they have so many close interactions with other students at school, it still winds up being a big problem. (If a kid has a transmission rate that is 1/10 of the adult transmission rate, but has 10x as many interactions in a day, then it's no help.) And the kids are the bridge between adults (parens, grandparents, teachers). So again, even if a kid's transmission rate is 1/10 that of an adult, a teacher has 20 or 30 students in a class, so the probability of the teacher getting infected is high.

So talking about the rate of kids infecting each other is such a tiny part of the issue with kids being in or out of school that it's disingenuous to talk about it in isolation.

[Volhiker78]

15 Fall Scenarios

I suppose what we need to do (as a family) is discuss some large fraction of these 15 scenarios to be prepared for whatever gets announced.

Tough decisions for your family, especially your graduating high school senior.  I think I read that he(she) will be attending a large public university out of state.  My guess is those schools will be a mixture of virtual and in-person classes in the Fall.  I'd be concerned about the housing.  Large dorms, students coming and going, are a bad mixture.  My daughter is 2 years away from college but if a vaccine isn't available then, and this virus is still hanging around stubbornly,  I would definitely try and find an apartment off campus for her.

[Caracal]

But even if kids transmit the virus at a much lower rate than adults, because they have so many close interactions with other students at school, it still winds up being a big problem. (If a kid has a transmission rate that is 1/10 of the adult transmission rate, but has 10x as many interactions in a day, then it's no help.) And the kids are the bridge between adults (parens, grandparents, teachers). So again, even if a kid's transmission rate is 1/10 that of an adult, a teacher has 20 or 30 students in a class, so the probability of the teacher getting infected is high.

Right now at this precise moment, when we have lots of cases, and don't have any systems set up to figure out how many and where they are, you can't have schools reopen or really do much safely. But that's the point, none of this is sustainable. You have to get things down to a level where you can manage risks. But managing risks will mean having to deal with them too. That's why I find this so confusing "If a kid has a transmission rate that is 1/10 of the adult transmission rate, but has 10x as many interactions in a day, then it's no help." No help to what? These are imaginary numbers but that would mean that children going to school posed less risk than me having three people over to my house for dinner. That would be pretty important to know if you were trying to make decisions on what to allow in a month.

[marshwiggle]

Right now at this precise moment, when we have lots of cases, and don't have any systems set up to figure out how many and where they are, you can't have schools reopen or really do much safely. But that's the point, none of this is sustainable. You have to get things down to a level where you can manage risks. But managing risks will mean having to deal with them too. That's why I find this so confusing "If a kid has a transmission rate that is 1/10 of the adult transmission rate, but has 10x as many interactions in a day, then it's no help." No help to what? These are imaginary numbers but that would mean that children going to school posed less risk than me having three people over to my house for dinner. That would be pretty important to know if you were trying to make decisions on what to allow in a month.

Only if you're planning to have 3 people for dinner 5 days a week until a vaccine is found.