engleski

Analysis of different sources of measurement error in determining second-to-fourth digit ratio, a potential indicator of perinatal sex hormones exposure

Brain structures change in a consistent way due to perinatal and pubertal sex hormones exposure. Data on these changes and their influence on human behaviour, called organisational effects of sex hormones, are difficult to obtain. One potential indicator is the second-to-fourth digit ratio (2D:4D). Measurement of 2D:4D has recently gathered support based on evidence concerning its validity. Available research on its reliability indicated that software methods of determining finger length are more reliable than direct measurements and that inter-rater reliability varied .60–.90. The present research focused on both of these aspects simultaneously, the retest reliability of software measures, and the comparison of software methods using a balanced design. Two scans, each of left and right hands of 213 participants, were collected following methodological recommendations from previous research. Length estimates followed a balanced design with at least six ratings per participant, varying with regard to scan, software method (GIMP or AutoMetric) and the number of raters. At least two raters were involved for each participant. Reli-ability analysis mostly indicated strong concordance between raters, methods, and scans. We discuss implications for assessing multiple sources of measurement error on continuous rating measures in general, and for measuring 2D:4D in special.

2D:4D measurement ; reliability ; rater effects ; software

nije evidentirano