Study Methodology Criticisms

Comments in this cluster repeatedly criticize scientific studies for methodological flaws such as self-reporting, selection bias, small sample sizes, and lack of generalizability or replication.

📉 Falling 0.4x Science
3,618
Comments
20
Years Active
5
Top Authors
#1221
Topic ID

Activity Over Time

2007
6
2008
33
2009
51
2010
85
2011
66
2012
65
2013
113
2014
145
2015
157
2016
236
2017
215
2018
225
2019
299
2020
323
2021
361
2022
387
2023
307
2024
241
2025
281
2026
22

Keywords

princeton.edu OK design.html carcinisation.com SELF UCLA P3 archive.ph biorxiv.org TI study bias research sample selection bias surveys reported data survey profession

Sample Comments

eurasiantiger Nov 17, 2021 View on HN

I would take this study with a grain of salt. It relies on self-reporting.

ggggtez Jan 9, 2018 View on HN

Not convinced. What about the effect of selection bias? That done probably accounts for the entire "study"

rasz_pl Oct 2, 2015 View on HN

same story yesterday: https://news.ycombinator.com/item?id=10309249study relies on SELF REPORTING, need I say more?

dstaley Jun 12, 2020 View on HN

Is there a link to the study? I'm curious if there was some sort of selection bias.

skybrian Mar 7, 2021 View on HN

The study is paywalled, but since it seems to based on surveys I’ll link to an article about why research based on surveys are often bullshit:https://carcinisation.com/2020/12/11/survey-chicken/

johnchristopher Dec 28, 2023 View on HN

This is weird. If you don't mind my asking, why don't you google for studies with larger scale experiments/surveys rather than using a single random data point ?Edit: yeah, that's what I thought...

richmarr Jan 9, 2018 View on HN

With respect, your opinion is anecdotal. These things aren't always obvious, or even visible to respondants at all.I'm not saying there's no effect, just that right now this claimed finding seems contrary to the existing body of evidence.This finding may well be accurate (and the existing studies in error). Alternatively it may be a mistake. Alternatively both could be true and this survey represents a special case, some kind of selection bias.Given that this survey is ye

acdha Mar 28, 2021 View on HN

That study was voluntary and appears to have been advertised in a few Facebook groups. It seems quite plausible that the 95% number is artificially high due to sampling bias.

AstralStorm Jan 9, 2018 View on HN

No idea about the downvote. This is a specific likely skewed sample. The study has to be reproduced on general population.

moxious Nov 2, 2017 View on HN

There's clear sampling bias. I read the article as a data irrelevant advertisement for the app