Abstract
Anticipation in sports is commonly investigated using perception-action uncoupled methods, thus raising questions regarding transferability of findings to the field. The aim of this study was to investigate the influence of different degrees of perception-action coupling on anticipation in handball goalkeeping. Advanced, intermediate and novice handball goalkeepers watched videos of throws on the goal and were asked to anticipate throw direction via key press (perception-action artificial condition) and via natural movement response (perception-action simulated condition). Results reveal overall superior performance in the artificial compared to the simulated condition. Skill-based differences, however, were descriptively more pronounced in the simulated condition compared to the artificial condition. The findings further highlight the importance of more representative research methods to unravel perceptual-cognitive skill in sports.
Acknowledgments
We thank Simone Lotz, Nils Bender, Helge Bräutigam and Wiebke Anheier for helping to prepare the stimuli and collect the data, and Nils Hendricks for programming the experimental software.
Disclosure statement
The authors confirm that there are no relevant financial or non-financial competing interests to report.
Data availability statement
Data supporting the findings of this study are available as Microsoft Excel file within this article’s supplemental online material.
Notes
1 Ethical approval was not obtained for the study as at the time of testing, there was no ethics committee at the university where the data collection took place. However, the study was conducted according to the ethical guidelines of the Declaration of Helsinki and the American Psychological Association valid at the time of testing (years 2006 to 2007).
2 Note that another display condition was also included where throwing actions were shown vertically mirrored. These mirrored clips (15 per response condition) were intermixed with the normal sequences in the conduct of the experiment. Initially, inclusion of vertically mirrored clips was inspired by research outside the sporting domain, which highlights that inverting visual displays of biological motion removes the facilitating effects of visual experience on motion recognition (Pavlova & Sokolov, Citation2000). With regard to our experiment, inclusion of a mirrored condition was intended to test whether the skill and perception-action coupling differences hypothesized for normal, upright actions maintain when the same actions are shown upside down. On hindsight, however, we felt that the approach was rather inappropriate to test that question and that inclusion of the mirrored condition would make this report unnecessarily complex. Therefore, we decided to focus the main analyses on anticipation performance in normal, upright videos. For the sake of completeness, the results for the vertically mirrored clips are reported in the Online Supplementary Material but these are not elaborated on further in any part of this manuscript.
3 Due to technical issues, in 4.25% of 2,070 trials overall, data on response side was not registered and stored. We therefore decided to determine response accuracy according to two alternative approaches. In approach one, participants’ percentages of correct predictions were calculated based on trials with available data (i.e. trials with missing accuracy data were ignored). Approach two involved several steps including multiple imputation of missing values, determination of accuracy estimates for missing data trials and the calculation of percentages of correct predictions based on the originally recorded and imputed values (please see the Online Supplementary Material for details). In the main text, results are reported on data obtained from approach 1 (i.e. exclusion of missing data). The results from the analyses of imputed accuracy data (approach 2) are reported in the Online Supplementary Material. Of note, the results differ only marginally between approaches and therefore do not suggest different interpretation. The dataset made available with this contribution includes the raw and summary data underlying both approaches.