New Study: 67% Of Scientific Papers Can Be Said To Reject The AGW Hypothesis…

…when using the same assumption-based methodology to arrive at the conclusion only 0.5% of scientific papers reject AGW.

In a new study, six scientists (Dentelski et al., 2023) effectively eviscerate a methodologically flawed 2021 study (Lynas et al.) that claims 99.53% of 3,000 scientific papers examined (by subjectively classifying papers based only on what is written in the abstracts) support the anthropogenic global warming, or AGW, position.

Image Source: Dentelski et al., 2023

The Lynas et al. authors begin with the assumption that a consensus on the human attribution for global warming not only exists, it is ensconced as the unquestioned, prevailing viewpoint in the scientific literature. So their intent was to effectively quantify the strength of this assumed widespread agreement by devising a rating system that only assesses the explicit rejection of AGW in the paper’s abstract as not supporting the presumed “consensus.”

Of the 3,000 papers analyzed in Lynas et al., 282 were deemed not sufficiently “climate-related.” Another 2,104 papers were placed in Category 4, which meant either the paper’s authors took “no position” or the position on AGW was deemed “uncertain”…in the abstract. So, exploiting the “if you are not against, you are for” classification bias, Lynas and colleagues decided that the authors of these 2,104 scientific papers in Category 4 do indeed agree with AGW, as what is written in the abstract does not explicitly state they do not agree.

Interestingly, if this classification bias had not been utilized and the thousands of Category 4 (“no position” or “uncertain”) papers were not counted as supporting AGW, only 892 of the 2,718 (climate-related) papers, or 32%, could be said to have affirmatively stated they support AGW. So, simply by assuming one cannot divine the AGW opinions of authors of scientific papers by reading abstracts, it could just as facilely be said that 67% (1,826 of 2,718) of climate-related papers reject AGW.

Dentelski and colleagues also point out that by their own analysis, 54% of the papers they examined that were classified by Lynas et al. as only “implying” support (Category 3) for AGW or Category 4 (“no position” or “uncertain”) actually described a lack of support for AGW in the body of the paper itself. But since this expressed non-endorsement of AGW was not presented in the abstract, these papers were wrongly classified as supporting AGW anyway.

To fully grasp the subjective nature of the methodology employed by Lynas and colleagues, Dentelski et al. uncover the internals of the study indicating 58% of the time two independent examiners did not agree on numerical classification scale (from 1 to 7) for a paper. If two people agree just 42% of the time when classifying papers, it cannot be said that the rating system is sufficiently objective.

The Lynas et al. paper appears to be little more than an exercise in propaganda.