Naslov
Statistics in Physical Chemistry

Autori
Sikirić Dutour, Maja

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

Izvornik
Book of Absrtacts / Itoh, Yoshiaki - Tokyo : Institute of Statistical Mathematics, 2006, 9-9

Skup
ISM Symposium Packing and Random Packing

Mjesto i datum
Tokyo, Japan, 01.03.2006. - 03.03.2006

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Nije recenziran

Ključne riječi
physical hemistry; statistics

Sažetak
Physical chemistry, discipline which connects chemistry and physics, provides molecular-level interpretation of observed macroscopic phenomena. Therefore it is not surprising that today physical chemistry is fundamental to various scientific fields, including chemistry, biology, engineering, agriculture, medicine and materials science. In obtaining maximum information from the physico-chemical measurements, statistics has been an essential tool from the beginning. Statistical methods frequently used in physical chemistry are linear and non-linear regressions, t-test, ANOVA, principal component analysis, etc. Due to the rapid theoretical and experimental progress in physical chemistry, the need for more sophisticated statistical methods is arising, especially for the signal processing and pattern recognition. Strict control of all environmental conditions and experimental design, necessary for obtaining meaningful results, makes statistical data analysis straightforward in most cases. However, in interdisciplinary research the control of experimental conditions ceases to be strict. Small sample sizes, data violating normality, inequality of variances, transformation of data, randomization, sampling, etc. become issues to be addressed with special care. Examples of applying physical chemistry to biomedical materials and agricultural research will be used to illustrate these problems.

Izvorni jezik
Engleski

Znanstvena područja
Kemija

POVEZANOST RADA