Ian Hussey
@ianhussey.mmmdata.io
Meta-scientist and psychologist. Senior lecturer @unibe.ch. Chief recommender @error.reviews. "Jumped up punk who hasn't earned his stripes." All views a product of my learning history.
I don’t think we have good evidence for this change in average patterns. It’s hard to speculate on the true population of untrustworthy work, I try to think only of what a given test can and can’t detect.
November 9, 2025 at 4:39 PM
I don’t think we have good evidence for this change in average patterns. It’s hard to speculate on the true population of untrustworthy work, I try to think only of what a given test can and can’t detect.
I’m not saying it’s appropriate, but it’s a semi plausible mistake if you’re doing all by hand. Either way the results are wrong, but it could speak to the data generating mechanism (bad math va there being no underlying data)
November 9, 2025 at 2:24 PM
I’m not saying it’s appropriate, but it’s a semi plausible mistake if you’re doing all by hand. Either way the results are wrong, but it could speak to the data generating mechanism (bad math va there being no underlying data)
Even more reasonable is “please share data upon request that you previously stated you would upon such request, in addition to the journal, funder, university, and professional body requirements that you do so”
November 9, 2025 at 11:32 AM
Even more reasonable is “please share data upon request that you previously stated you would upon such request, in addition to the journal, funder, university, and professional body requirements that you do so”
Their thinking was perhaps it’s just those GRIM inconsistent cells, but yes the 5.18 is then beyond the scale bounds.
November 9, 2025 at 11:31 AM
Their thinking was perhaps it’s just those GRIM inconsistent cells, but yes the 5.18 is then beyond the scale bounds.
I presented this to my students two weeks ago, and one of them pointed out that if the denominator for calculating the mean was 60 (total N) rather than 30 (N per condition), these cells become GRIM consistent. I haven’t gone back to see what the ‘real’ means would be on this basis. @steamtraen.eu
November 9, 2025 at 9:48 AM
I presented this to my students two weeks ago, and one of them pointed out that if the denominator for calculating the mean was 60 (total N) rather than 30 (N per condition), these cells become GRIM consistent. I haven’t gone back to see what the ‘real’ means would be on this basis. @steamtraen.eu
Reposted by Ian Hussey
Call me old-fashioned, but I think if you have to correct one of the key results cited in your *abstract*, then the correction actually does affect the overall conclusions of your study!
November 6, 2025 at 11:21 PM
Call me old-fashioned, but I think if you have to correct one of the key results cited in your *abstract*, then the correction actually does affect the overall conclusions of your study!
This is the full crossref dataset for all psychology journals, using their spring 2025 dataset release - it’s very comprehensive and the results are reproducible! This really is the distribution of citations!
November 6, 2025 at 3:23 PM
This is the full crossref dataset for all psychology journals, using their spring 2025 dataset release - it’s very comprehensive and the results are reproducible! This really is the distribution of citations!
November 6, 2025 at 3:02 PM
Actually_to_be_read_THISWEEK_new
November 6, 2025 at 11:14 AM
Actually_to_be_read_THISWEEK_new
This page has been very useful for me, I’ve even been able to correct editors that their journal doesn’t allow it because their publisher requires it.
October 24, 2025 at 2:01 PM
This page has been very useful for me, I’ve even been able to correct editors that their journal doesn’t allow it because their publisher requires it.