Short answers, pulled from the story.
Statistician Gene Glass coined the term meta-analysis in 1976 to describe the analysis of analyses. The statistical roots of this method stretch back to 1904 when Karl Pearson published a paper in the British Medical Journal that collated data from several studies of typhoid inoculation.
The first model meta-analysis on psychotherapy outcomes was published in 1978 by Mary Lee Smith and Gene Glass. This work was met with derision by prominent psychologist Hans Eysenck who dismissed their work as an exercise in mega-silliness and later referred to the entire methodology as statistical alchemy.
The number of published meta-analyses grew from 334 in 1991 to 9,135 by 2014. This growth transformed the method into a cornerstone of evidence-based medicine and psychology despite early hostility.
The file drawer problem occurs when studies showing negative or non-significant results are tucked away in cabinets and never published. This phenomenon creates a biased distribution of effect sizes leading to a base rate fallacy where the significance of published studies is grossly overestimated.
The fixed effect model assumes that all included studies investigate the same population and use identical variable definitions while larger studies dominate the weighted average. The random effects model acknowledges that studies differ in their methods and sample characteristics introducing variability known as heterogeneity.
A 2011 study reviewing 29 meta-analyses found that 69% of the underlying randomized controlled trials received industry funding. Only 7% of the meta-analyses reported these funding sources creating a lack of transparency that allows agenda-driven bias to flourish.