Hacker News new | ask | show | jobs
by snuxoll 958 days ago
Contrarian views on many topics are not backed by facts but by emotion. A collection of knowledge is not obligated to let contrarian viewpoints with no scientific basis share the same space as factual coverage of a topic.

This is the problem I see far too often with people who criticize Wikipedia for their NPOV policy.
1 comments
silent_cal 958 days ago
In some cases sure, but on these issues it doesn't matter if you have facts or not. The contrarian view simply isn't allowed on Wikipedia.
link
sodality2 958 days ago
What points of views are scrubbed from WP, with credible scientific evidence supporting them?
link
silent_cal 958 days ago
Take this passage from the vaccination page for example:

> Some studies have claimed to show that current vaccine schedules increase infant mortality and hospitalization rates;[103][104] those studies, however, are correlational in nature and therefore cannot demonstrate causal effects, and the studies have also been criticized for cherry picking the comparisons they report, for ignoring historical trends that support an opposing conclusion, and for counting vaccines in a manner that is "completely arbitrary and riddled with mistakes".[105][106]

If you go to the vaccination page and look at the cited studies, you'll see they are peer-reviewed studies in real scientific journals. But a Wikipedia editor went in and modified the language to cast doubt on it (see here: https://en.wikipedia.org/w/index.php?title=Vaccination&diff=...).

All statistical analyses are "correlational in nature", so there's no reason why that should weaken the findings of this study, and the source used to accuse the study of "cherry-picking" is a blog post from 2011. In my estimation this kind of spin happens all the time on Wikipedia.

link
nvy 958 days ago
Goodness, are you not familiar with the phrase "correlation does not imply causation"?

There are ways to establish causal relationships, the gold standard of which is to conduct a double-blind controlled trial.

To claim as you have that a correlational study establishes a causal relationship is deeply misguided, irrespective of whether or not it is peer-reviewed.

link
silent_cal 958 days ago
Yes I am - you don't seem to know that all a double-blind controlled trial produces is (at best) a very strong correlation. I'm not claiming that a correlational study implies causation. I'm claiming that ALL statistical analyses can only prove correlation, and never causation. You're wrong about double-blind controlled trials proving causation.
link
nvy 958 days ago
You are factually incorrect.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7081045/

>In clinical medical research, causality is demonstrated by randomized controlled trials (RCTs).

https://www.sagepub.com/sites/default/files/upm-binaries/236...

https://www.statisticssolutions.com/dissertation-resources/r...

https://www.abs.gov.au/statistics/understanding-statistics/s...

https://www2.stat.duke.edu/~jerry/Papers/causal.pdf

If you argue this further I'm just going to conclude you're trolling and won't respond again.

You are wrong. Face it.

link