Half-hourly flux data were subject to a series of quality control tests46 and filtered for the potential underestimation during nighttime periods of low turbulence47.
These data were subject to a systematic aspect offset of approximately 0.4 arcsec, believed to be the result of a changed thermal environment following the procedure to cool the ACA CCD from -15° C to -19° C that took place between late Nov-2006 and early Jan-2007.
Qualitative data were subject to content analyses by a single researcher.
For the OLS linear regressions, the data were subject to an outliers and influential records removal process.
Once they had been transcribed and translated, the data were subject to thematic analysis using a coding-based approach (Bernard & Ryan [2009]).
To eliminate statistical confounder by the negative values related to the vertical maxillary cant direction, the 3D images and measurement data were subject to be adjusted to let all subjects have the right downward maxillary cant.
While the actual co-localization data were subject to some experimental variability, they grouped, across all our results, into two completely separated clusters (Figure 5, Table 1, Tables S1 and S2).
Data were subject to narrative analysis.
Data were subject to a range of quality checks.
Dose response data were subject to regression analysis and analysis of variance (ANOVA).
All the data were subject to constant comparison in order to facilitate thematic analysis [ 42].
