[DSRP Evidence](https://dsrpevidence.org/)

# Common Biases in Systems Thinking

Di Marco G, Cabrera D, Journal of Systems Thinking (Preprints), 2024

## Abstract

An independent replication using language network analysis on a different sample, testing whether the predicted structural biases appear in written language.

## What they found (results)

Distinctions appeared in 100% of responses. Element ratios ran 1.7–6.8 and pattern ratios 1.6–6.5, in the predicted direction for identity over other, part over whole and view over point. Two predictions failed: action and reaction came out equivalent rather than unequal, and relationships and perspectives overlapped rather than separating. Training increased use of every element while making the ratios more pronounced, against the stated hypothesis.

## Response

The nearest thing to an independent replication: different author, different method, different sample. It is not arm's length — the author came through the lab's training courses and Cabrera is second author. Two of its own predictions failed and are reported as failures.

[Read the paper](https://doi.org/10.54120/jost.pr000037.v1)