Joar Øveraas Halvorsen
LightNews LightNews
banner
joarhalvorsen.bsky.social
Joar Øveraas Halvorsen
@joarhalvorsen.bsky.social
970 followers 300 following 710 posts
Consultant clinical psychologist & associate professor of clinical psychology. Mainly #ClinicalTrials, (post)traumatic stress, #PsychotherapyResearch, & #HumanRights. May contain traces of sarcasm.
Posts Replies Media Videos
joarhalvorsen.bsky.social
Presis og god kommentar!
November 26, 2025 at 8:31 PM
joarhalvorsen.bsky.social
It's about the interpretation of ESs, isn't it? Can we interpret ESs without considering control conditions, quality of the evidence etc.? An effect size can't be interpreted in a "vacuum", e.g. SMD=0.8 considered a large effect size without taking the factors mentioned above into consideration.
November 23, 2025 at 9:43 PM
joarhalvorsen.bsky.social
It's quite telling of the state of affairs when some of the most prominent psychotherapy researchers can publish a meta-analysis claiming psychotherapy is effective relying on passive control conditions without any assessment of risk of bias or quality of evidence. www.cambridge.org/core/journal...
Is psychotherapy effective? A re-analysis of treatments for depression | Epidemiology and Psychiatric Sciences | Cambridge Core
Is psychotherapy effective? A re-analysis of treatments for depression - Volume 28 Issue 3
www.cambridge.org
November 23, 2025 at 9:25 PM
Reposted by Joar Øveraas Halvorsen
maxprimbs.bsky.social
And just because I more often see people using the Aspirin example: They used the wrong effect size and that is why the effect seems to small.
To further illustrate the claim that small effects can be consequential in large samples or at the population level, Götz et al. (2022) presented the correlation between aspirin and the prevention of heart attacks (r = .03). Ferguson (2009) pointed out the flaw in using this effect size to make the generalized argument that small effects matter through an analogy with wearing a bulletproof vest: The effect size of wearing a bulletproof vest on the probability of dying is large if we examine people who get shot but very small if we include the millions of people who never get shot. Likewise, the causal effect of aspirin on the chance of a heart attack is substantial, but there is only a small effect in the reduction of heart attacks if a large group of people, many of which would never suffer a heart attack, regularly take aspirin. The important difference between medicine and psychology is that in psychology researchers rarely include a large majority of individuals in their studies that are not expected to benefit from an intervention. For example, when we examine the effectiveness of a new treatment for depression, we usually do not conduct the study on a sample in which only a small minority of individuals are depressed. Therefore, when a small effect is observed in psychology, it may not matter at the population level, and any claims of why it would matter need to be theoretically justified and/or empirically supported.
November 23, 2025 at 7:20 PM