Naslov
Pattern recognition, evaluation and generalization of kinesiological phenomena from heterogeneous sample via method of partialization of variance

Autori
Gruic, Igor

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Book of Abstracts of the ISCCRO - International Statistical Conference in Croatia / Dumičić, Ksenija ; Erjavec, Nataša ; Pejić Bach, Mirjana - Zagreb : Hrvatsko statističko društvo, 2016, 73-73

Skup
International Statistical Conference in Croatia "New Challenges of Official and Applied Statistics in European Union

Mjesto i datum
Zagreb, Hrvatska, 05.03.2016. - 06.03.2016

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
heterogeneity ; kinesiology ; partialization

Sažetak
Knowledge, abilities, morphology etc., are organized trough function, form&content. Understanding allometry (McMahon, 1984), method of partializations (Milas, 2009) etc., may improve power of insight in practical research&applications - here presented in two diametric effect of partializations on interpreted results in research examples – 1) Gruić, 2011, and 2) Gruić et al, 2015. In 1) evaluation of performance within specific motor skills – with&without influence of certain motor abilities (Gruić, 2011 ; 2 experimental groups ; 80 students ; year 2008/09, age:19 ; 9 specific handball motor skills, 13 basic and specific motor abilities), differences in evaluated knowledge in two groups prior partializations were statistically significant (Wilks' Lambda=0.74, χ2=21.68, p< 0.01). Differences in evaluated abilities were not statistically significant (Wilks' Lambda=0, 87, χ2=9, 51, p=0, 73). Differences in knowledge partialized by results in abilities (residual variance - excluded contribution of abilities), were not affected, compared to previously defined differences in evaluated knowledge (Wilks' Lambda=0, 75, χ2=21, 09, p< 0.01). In 2) evaluation of pedobarographic features of gait – with & without influence of longitudinal dimensionality (Gruić et al, 2015 ; experimental set: 63 variables, 22 participants: 13male&9 female ; year 2014/15 ; age:14-35, average 172, 56cm/73.11kg body height/mass (19.47 BMI)), tested differences between genders were affected by partializations. After partialization of results (excluded influence of longitudinal dimensionality), differences were significant in 10 standard variables, 10 less compared to analysis before partialization. It can be concluded that statistical tool&method of partialization of the results (by excluding influence/contribution of specific characteristic which tend to increase heterogeneity), may standardize&improve the power of the protocol&method.

Izvorni jezik
Engleski

Znanstvena područja
Biologija, Računarstvo, Biotehnologija

POVEZANOST RADA