| 1 |
Researcher degrees of freedom in phonetic sciences ...
|
|
|
|
Abstract:
Every data analysis is characterized by a multitude of decisions, so-called “researcher degrees of freedom”, that can affect its outcome and the conclusions we draw from it. Intentional or unintentional exploitation of researcher degrees of freedom can have dramatic consequences for our results and interpretations, increasing the likelihood of obtaining false positives. It is argued that quantitative phonetics faces an extremely high number of researcher degrees of freedom due to their scientific object, speech, being inherently multidimensional and exhibiting complex interac-tions between many co-varying layers. A Type-I error simulation is presented that demonstrates the severe false error inflation when exploring researcher degrees of freedom such as measuring more than one phonetic exponent, subsetting the data, and including speech-relevant co-variates in the analysis. It is argued that, combined with common cognitive fallacies, unintentional exploi-tation of researcher degrees of freedom introduces ...
|
|
Keyword:
FOS Languages and literature; FOS Psychology; Linguistics; Phonetics and Phonology; Psychology; Social and Behavioral Sciences
|
|
URL: https://dx.doi.org/10.17605/osf.io/fp4jr
https://psyarxiv.com/fp4jr/
|
|
BASE
|
|
Hide details
|
|
|
| 2 |
Dynamic speech adaptation to unreliable cues during intonational processing ...
|
|
|
|
BASE
|
|
Show details
|
|
|
|
|
