Development, validation, and reliability of a new instrument of observation and technical evaluation in karate athletes

The present study aims to develop and establish the reliability of the System for the Technical Assessment of Kumite (STAK) in karate. Twenty karate fighters aged between 10 and 22 years, consisting of novice and expert athletes participated in the present study. The reliability process focused on three points: i) analysis and evaluation by experts of the content of the observation instrument; ii) analysis and evaluation by expert evaluators of the athletes' performance in each of the chosen techniques; and iii) reliability of the evaluators' observations. The technical evaluation system in karate evaluated the Mawashi Geri, Ura Mawashi Geri, Kizami Zuki, and Gyaku Zuki techniques, described in the literature as the most common techniques in kumite competitions, which were adjusted to different levels of proficiency. Substantial to almost complete intra-rater reliability was observed for both expert and beginner athletes' evaluations, with ICC values ranging from 0.774 to 0.959. These values suggest that the evaluation conducted at different times did not cause significant interference in the assessment and that the estimate of this agreement is within a 95% confidence interval. Overall, the inter-observer and intra-rater reliability results suggest that the STAK evaluation process is reliable for assessing the success criteria of different techniques to help coaches in the promotions to Kyu and Dan ranks. The present study also indicates that novice athletes may require more specific and consistent evaluation criteria to improve their observation technique and refine their ability to interpret success variables in each motor action.