Wielkość próby a istotność wnioskowania statystycznego

• Authors: Kośny Marek , Peternek Piotr
• Languages of publication: PL EN

Abstracts

EN

Commonly used method of determining the quality of research results is the application of significance tests – both parametric and non-parametric. They enable researchers to eliminate low-quality results, when observed differences are the consequence of too small samples. Therefore relatively little attention – also in didactics – is paid to issues of big samples, when the number of observations amounts up to several thousands. Whereas situations, when significance test is applied to big samples requires no less attention than the situations of too small samples. In this context the main aim of the paper is the analysis of results of the significance test for the correlation coefficient (parametric test) and chi-square test for independence (non-parametric test) for samples exceeding 1000 observations. For these tests it will be shown, how imprecise interpretation could lead to the significant distortion of obtained results.

Keywords

EN

statistical inference; tests of significance; the correlation coefficient

• Publisher: Wydawnictwo Uniwersytetu Ekonomicznego we Wrocławiu
• Journal: Didactics of Mathematics
• Year: 2011
• Issue: 8 (12)
• Pages: 71-80

Contributors

author

Kośny Marek

author

Peternek Piotr

References

• Krysicki W., Bartos J., Dyczka W., Królikowska K., Wasilewski M. (2000). Rachunek prawdopodobieństwa i statystyka matematyczna w zadaniach. Wydawnictwo Naukowe PWN. Warszawa.
• Ostasiewicz S., Rusnak Z., Siedlecka U. (2001). Statystyka. Elementy teorii i zadania. Wydawnictwo Akademii Ekonomicznej we Wrocławiu. Wrocław.
• Steczkowski J. (1995). Metoda reprezentacyjna w badaniach zjawisk ekonomiczno-społecznych. PWN. Warszawa-Kraków.