Counting data

In statistics , counting data is a data format in which observations can only assume non-negative integer values ( i.e. 0, 1, 2, 3, ...). These values are created by counting, and not by evaluating and arranging (as is the case with the Likert scale , see ordinal scale ). In addition, count data is different from binary data , which can only have two values (0 and 1). It is therefore a matter of discrete frequencies, such as how many days someone was absent in the last year.

Estimation of counting data models

Common methods such as the least squares method are not suitable for estimating count data models. Therefore, there are special statistical methods for analyzing count data ( regression for count data ). The methods refer to the underlying distributions, such as the Poisson distribution or the negative binomial distribution . The use of the Poisson distribution is the simplest and most widely used variant.

Individual evidence

↑ Windzio, M. (2013). Estimating the number of events: models for counting data. In regression models for states and events (pp. 193–207). Springer VS, Wiesbaden.
^ Ludwig Fahrmeir , Thomas Kneib , Stefan Lang, Brian Marx: Regression: models, methods and applications. Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2 , p. 293.

literature

Winkelmann, R. (2008). Econometric analysis of count data. Springer Science & Business Media.

[1] Windzio, M. (2013). Estimating the number of events: models for counting data. In regression models for states and events (pp. 193–207). Springer VS, Wiesbaden.

[2] Ludwig Fahrmeir , Thomas Kneib , Stefan Lang, Brian Marx: Regression: models, methods and applications. Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2 , p. 293.