Why some squares are bigger in a quitting-review forest plot

Orient the plot before interpreting its shapes

Read the comparison, outcome, follow-up, axis label, and legend. Identify which direction favors which group and where the no-difference value lies. Ratio measures and difference measures do not use the same null value. A cropped plot can hide the very labels needed to interpret its apparent direction.

Cochrane explains conventional forest-plot symbols and weighting. This guide is an introductory reading exercise, not an audit of a review’s calculations. Actual plots can vary; the displayed legend takes priority over assumptions about colors or shapes. No figure from a named vaping study is reproduced here.

Cochrane: Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses

Separate weight from study quality

Weight concerns how much a result contributes under the chosen method. It often reflects statistical precision, but the model and available data matter. A large square does not by itself establish a relevant population, unbiased measurement, or low risk of bias. Find those assessments elsewhere in the review.

For a hypothetical example, one large study has a prominent square while several small studies have tiny squares. Counting them as one vote each would ignore the calculation shown. Calling the prominent study “the best study” would answer a different question without the necessary methodological information. Both shortcuts can misrepresent the review.

Read intervals instead of tallying threshold crossings

Look at each estimate with its confidence interval and consider how the studies fit the question. Do not classify every line crossing the null as evidence of no effect and then tally a majority. A study can be imprecise without proving absence of an effect; the pooled calculation is not a vote among significance labels.

The diamond’s center and width summarize an estimate and its uncertainty under the reported model. A narrow diamond does not erase study limitations, make different interventions identical, or establish that a particular app was tested. If the image omits study identities or the pooled method, return to the original review rather than guessing.

Write the minimum defensible explanation

A useful note states the effect measure, the most influential result under the shown weights, the pooled estimate and uncertainty, and the key relevance or bias question that remains. You do not need to recalculate the weights to understand that square area and credibility are distinct.

If a promotional image enlarges or highlights one study without preserving the original legend, verify it against the publication. Keep your interpretation limited if the full figure is unavailable. This visual reading task cannot determine treatment suitability or prove an individual outcome. It helps you explain what the picture represents before relying on the headline attached to it.