Some Missing Cannabis Attention Scores Were Replaced With Group Means
The smoking study replaced some missing divided-attention observations with group means. Those values were analytic substitutions, not additional measurements of the missing participants’ actual performance.
The report identified the missing observations
Divided Attention Task data were absent for two occasional users at approximately 5.6 hours and for a frequent and an occasional user at the late assessment. The authors described replacing these observations with mean group values. The task combined tracking and peripheral-target responses.
A completed table can contain different evidence types
An observed score reflects what a participant did during testing. A mean replacement supplies a value based on other available observations. It does not recover the missing individual trajectory. How missingness and replacement affect uncertainty depends on the data and analysis; the method alone does not establish whether a reported effect is too large or too small.
Why the detail matters
The paper reported selected attention and psychomotor differences, not uniform impairment across all domains. A review should retain both the measured outcome and its missing-data handling before applying the findings to an individual. Small group sizes, unequal cannabis histories and the absence of a placebo smoking session further constrain causal interpretation. A substituted score is not evidence that a particular participant had normal or impaired function at that time.
Related analysis: blood THC and task effects across use groups.
Research cited
Desrosiers and colleagues (2015). Smoked Cannabis’ Psychomotor and Neurocognitive Effects in Occasional and Frequent Smokers. DOI: 10.1093/jat/bkv012.
