Treating statistical data set with missing values

• Autorzy: Katerina Gurinová , Vladimírá Hovorková Valentová
• Języki publikacji: EN

Abstrakty

EN

Many persons, specialists as well as laymen, sometimes or often have to deal with data sets. They have to solve one important problem which occurs in many data sets very often - how to treat missing values. Firstly, we have to decide whether to impute missing values or leave out. There is a wide spectrum of methods for data imputation. It is not easy to choose the correct method, because their using depends on many factors, e.g. on type of a variable, variability of a sample, the nature of missingness, types of missing values etc. Also, the number of missing values depends on many factors. However, we can influence just some of them. One of the factors which we can influence is a preparation of a survey. Many authors assert that the ratio of failures in excellently prepared surveys is between 4% and 15%.(fragment of text)

Słowa kluczowe

EN

Value analysis; Statistics; Measurement

PL

Analiza wartości; Statystyka; Pomiary

• Czasopismo: Scientific Publications / University of Economics in Katowice
• Rocznik: 2010
• Tom: Survey sampling methods in economic and social research
• Strony: 111--120

Twórcy

autor

Katerina Gurinová

• Technical University of Liberec, Czech Republic

autor

Vladimírá Hovorková Valentová

• Technical University of Liberec, Czech Republic

Bibliografia

• Cyhelský L., Kahounová J., Hindis R. (1999): Elementdrni statistickd analyza. Management Press, Praha.
• Kish L. (1995): Survey Sampling. John Wiley & Sons, New York.