Analyzing the Time Course of Pupillometric Data.
Jacolien van Rij, Petra Hendriks, Hedderik van Rijn and 2 others
PMID 31081486WHAT IT FOUND
Pupil size over time can show false effects if the analysis ignores each trial's individual pattern.
For speech research, check that the analysis followed the real pupil curves before trusting a result.
Key findings
01Analyzing pupil size over time can produce effects that are not real when the analysis does not account for the fact that measurements within one trial are linked.
02In the example language data, making the analysis follow each individual pupil trace removed the picture-sentence match difference when the actor was introduced second, while the actor-first difference remained.
03The authors found that making the analysis follow each individual pupil curve is more important than adding a correction for linked measurements.
STILL TO COME
How it was doneWhat they found
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The paper is a methods tutorial, not a clinical study, so it does not test treatment outcomes or patient populations. The pupil-dilation example used native Dutch participants and existing experimental data, so it does not show what pupil size means in clinical speech or hearing assessment. The analysis depends on specialized statistical modeling and software, and the authors note that the best model may be computationally demanding for larger data sets. The authors state that the problem of linked measurements is not limited to their method and applies to other time-course analyses.
Declared interests
The supplied text lists funding from the Alexander von Humboldt-Stiftung and the Nederlandse Organisatie voor Wetenschappelijk Onderzoek. It does not report a competing-interests statement.
The easy way to misread this
Do not treat the pupil-size differences as evidence for a clinical test or therapy. This paper is a methods tutorial, and the example results changed after the analysis was corrected.