top of page

Texas A&M talk: Quantitative issues causing confusion for both skeptics and leaders in the pandemic

Last Friday gave a covid webinar at Texas A&M discussing various issues in the pandemic in which quantitative nuances have confused people, sometimes the pandemic skeptics, and sometimes the national leaders and media, including:


* Length of lockdowns

* Deniers vs. alarmists, vaccine enthusiasts vs. skeptics

* Vaccine efficacy vs. infection/transmission

* Vaccine safety

* Immunity after infection

* Age and Time confounding in the pandemic and Simpson's paradox

* Variants

* Censoring/fact-checking

* The role of statisticians and other biomedical data scientists in the pandemic

Here is the video:


1,073 views30 comments

Recent Posts

See All

Explanation of "Healthy Vaccinee Effect" and Confounding

Healthy vaccinee effect (HVE) has become frequent topic of discussion these days. I am glad to see this, as I have been talking about this phenomenon for the past 2.5 years, and for a long time no one

30 comentarios


su xeko
su xeko
12 jun

I would want to mention that this game is quite good and fun. Come to foodle free on this platform if you're bored in your free time and start playing this board game there without downloading anything or registering.

Me gusta

Knight Brett
Knight Brett
02 nov 2023

Is it feasible to conduct a cross-country comparison of vaccination rates and death rates? I was taken aback by the fact that some countries with low vaccination rates, like Haiti, have remarkably low death rates. I'm curious to know if this is a consistent trend or an exception. Additionally, considering the prevalence of multi-generational households in some cultures, I expected the virus to spread more easily in those areas. doodle baseball

Me gusta
Jeffrey Morris
Jeffrey Morris
02 nov 2023
Contestando a

There are places with such data, like ourworldindata.org


But you have to be careful of something called the ecological fallacy here — trying to draw conclusions about the effect of an individual exposure (vaccines) on an outcome (all cause deaths covid deaths) from looking at group (country level) associations of exposure rate (vaccination rate) and outcome rate (death rate). The ecological fallacy is that the group level correlation can completely differ from the individual level correlation (ie it could be vaccinated individuals have systematically lower death rate everywhere but highly vaccinated countries also have higher death rates for a period of time)


The reason for this is confounders — country level factors that impact both the exposure rate and the…


Me gusta

Hi Prof Morris,


would it be possible to do a vaccine rate/death rate comparison across all countries. I was surprised to see some highly unvaccinated countries (Haiti for example) have super low death rates and was wondering if that is consistent or an anomaly. The culture also has multi generational homes so I thought it would have spread easily there.

Me gusta
Thomas Clarke
Thomas Clarke
15 dic 2021
Contestando a

It is particularly difficult because, for example:

(1) accelerated vaccine campaigns (as now in the UK), and high population uptake of vaccines, are often caused by fears from increasing COVID rates!

(2) More broadly, countries less vaccinated can suffer very damaging COVID waves after which they do better because of high infection-based immunity than countries where only a few people have been infected. Thus low vaccination can lead indirectly to low COVID rate at some later time


These temporal relationships are so complex it adds to the difficulty of reading anything from a simple comparison.


There are strong other confounders - for example, poor countries tend to be less able to get vaccine, and also tend to be younger. As…

Me gusta

I enjoyed the presentation and don't think it can be emphasized enough that (bio)statisticians need a seat at the table.


During the Q&A you mentioned the importance of fleshing out the nuances in the data in regards to a question about outbreaks in highly vaccinated countries. I agree! The vaccination data by age from NYC surprised me. Fully vaccinated a/o 12-7-21:


35-44: 86%

45-54: 86%

55-64: 88%

65-74: 86%

75-84: 76%

85+: 60%


Almost the polar opposite of what I expected.


Source: https://www1.nyc.gov/site/doh/covid/covid-19-data-vaccines.page


Me gusta
Jeffrey Morris
Jeffrey Morris
10 dic 2021
Contestando a

Interesting! That is unusual compared to most other places. The rates for middle aged not so unusual -- higher than one would expect, but the very low rate for 75% is quite shocking. I wonder why that is?


Me gusta

Nice talk. I'm glad you are encouraging statistics students to engage in important practical studies and communicate results in ways that can have great impact.


While I agree with much of what you said in the talk, I'd recommend additional emphasis on the importance of randomized experimental designs and the profound difficulties of relying on observational studies to answer efficacy and safety questions involving causation.


We need far more rigorous trials to answer these questions for newer treatments and vaccines. Statisticians should lead efforts to demand data be collected from well-designed, rigorous and well-controlled clinical trials. Study participants will be more than willing to have the necessary physical examinations, blood draws, and other testing needed to provide many convincing measur…


Me gusta
Contestando a

Each participant had at least 2 blood samples drawn to look for Covid antibodies during the 3+ months of the study. One before the 1st jab, another a month after the 2nd jab. If the study hadn't ended in December, others were scheduled. However, no efforts were made to test for diseases or organ damage as I'm suggesting. Here's a pdf of the Pfizer study protocol (see pages 85-94 ) 4.

https://cdn.pfizer.com/pfizercom/2020-11/C4591001_Clinical_Protocol_Nov2020.pdf

Me gusta
Post: Blog2_Post
bottom of page