3 ms·

Also adding to the critique of this paper: "Participants were a convenience sample recruited at the University of Alabama at Birmingham" [0] _Convenience samp

by gls2ro 5y ago

Also adding to the critique of this paper: "Participants were a convenience sample recruited at the University of Alabama at Birmingham" [0] _Convenience sample_ being the keyword here. As far as I understand they first had their sample and then tested the hypothesis. I would expect that they choose a sample that could invalidate/validate their hypothesis and not the other way around. I think we really need to replicate this multiple time. I think that 72 people is a very small sample to draw a conclusion about a disease that we don't understand that much how it works (example: why young people with apparent no predisposition are in ICU ...) we should take this as a good step toward understanding more, but it needs a lot of replication. [0] https://wwwnc.cdc.gov/eid/article/27/9/21-1042-t1 https://wwwnc.cdc.gov/eid/article/27/9/21-1042-t1

pdpi 5y ago
Convenience sampling just means “we took a bunch of people that were readily available (in this case, probably college students)”. It’s not a sample built to be representative of the general population, just enough for a quick-and-dirty initial assessment.
- gls2ro 5y ago
  Yes, I agree, that is also my understanding. But convenience sample is sometimes a bad choice for assessing a hypothesis. So I think these studies are very important for pushing the understanding futher but also they should be treated as you say "initial assessment". In general I think science communication needs to be fixed or improved. We are now in a time where general public can access them. And this is very good and more should be done for this. But also the structure of how we present these studies should be adapted. Just publishing open a paper following this legacy structure and wording is not enough. I think the structure of a published paper should be changed to include quality assessments like for example (and this is only an example): After Abstract we should always have a Method Summary with the following information: a. Original Article | Secondary Article | Replica of Z | b. Replicated | Not replicated yet | Replicated by Y,Z,T ... c. Small Group Sample | Large Group Sample | d. Convenience Sample | (Simple, Stratified, Clustered) Random Sample | ... e. Observational | Experimental | Self-Reporting f. Limitations of this paper: .... Or something like this. Few people read (or have the know-how) to read the papers and also understand how to use that into a decision making process taking into consideration limitations. Edit: added clarify about what I meant with my critique of convenience sample.

by gls2ro 5y ago

pdpi 5y ago

Convenience sampling just means “we took a bunch of people that were readily available (in this case, probably college students)”. It’s not a sample built to be representative of the general population, just enough for a quick-and-dirty initial assessment.

gls2ro 5y ago
Yes, I agree, that is also my understanding. But convenience sample is sometimes a bad choice for assessing a hypothesis. So I think these studies are very important for pushing the understanding futher but also they should be treated as you say "initial assessment". In general I think science communication needs to be fixed or improved. We are now in a time where general public can access them. And this is very good and more should be done for this. But also the structure of how we present these studies should be adapted. Just publishing open a paper following this legacy structure and wording is not enough. I think the structure of a published paper should be changed to include quality assessments like for example (and this is only an example): After Abstract we should always have a Method Summary with the following information: a. Original Article | Secondary Article | Replica of Z | b. Replicated | Not replicated yet | Replicated by Y,Z,T ... c. Small Group Sample | Large Group Sample | d. Convenience Sample | (Simple, Stratified, Clustered) Random Sample | ... e. Observational | Experimental | Self-Reporting f. Limitations of this paper: .... Or something like this. Few people read (or have the know-how) to read the papers and also understand how to use that into a decision making process taking into consideration limitations. Edit: added clarify about what I meant with my critique of convenience sample.