# Definition:Relative Frequency Model

## Definition

The **relative frequency model** is a mathematical model that defines the probability of an event occurring as follows:

- $\map \Pr {\text {event occurring} } := \dfrac {\paren {\text {observed number of times event has occurred in the past} } } {\paren {\text {observed number of times event has occurred or not occurred} } }$

That is, the probability of an event happening is defined as the **relative frequency** of a finite number of events of a particular type in some finite reference class of events.

Symbolically:

- $\map \Pr \omega := \dfrac {f_{\omega} } n$

where:

- $\omega$ is an event
- $f_{\omega}$ is how many times $\omega$ occurred
- $n$ is the number of trials observed.

## Also defined as

Many sources adopt a slightly different perspective as an underlying constant that such a frequency converges to, were the number of observations to approach infinity.

According to this approach, it may be impossible to find out the "true" probability of an event.

The model as defined above, then, is an approximation.

Either way, the assumption is that were we to conduct a large number of trials, the frequency of events occurring in the new experiments should be roughly equal to the frequency of events occurring in the observed cases.

## Application to Statistics

In the context of statistics, the **relative frequency model** is the relative frequency of a variable.

The probability of a variable occurring is then the probability of the variable being satisfied by some individual.

## Also see

## Sources

- 1970: Michael C. Gemignani:
*Calculus and Statistics*: $\S 1.3$ - 1995: Merrilee H. Salmon:
*Introduction to Logic and Critical Thinking*: $\S 6.1$ - 2001: Michael A. Bean:
*Probability: The Science of Uncertainty*: $\S 1.2$ - For a video presentation of the contents of this page, visit the Khan Academy.