Very often, the data available is vast, leading to a situation where dealing with individual numbers becomes laborious and messy. In such circumstances, it is neater and more convenient to summarize results into what is known as a frequency table. The data in the display is called a frequency distribution.

An

- Each interval has a lower limit and an upper limit.
- Intervals must be all-inclusive and non-overlapping.

A

- Each observation can only lie in one interval.
- The total number of intervals will incorporate the whole population.
- The range for an interval is unique. This means a value (observation) can only fall into one interval.

It is important to consider the number of intervals to be used. If too few intervals are used, too much data may be summarized and we may lose important characteristics; if too many intervals are used, we may not summarize enough.

A frequency distribution is constructed by dividing the scores into intervals and counting the number of scores in each interval. The actual number of scores and the percentage of scores in each interval are displayed. This helps in the analysis of large amount of statistical data, and works with all types of measurement scales.

**Absolute frequency**is the actual number of observations in a given interval.**Relative frequency**is the result of dividing the absolute frequency of each return interval by the total number of observations.**Cumulative absolute frequency**and**cumulative relative frequency**are the results from cumulating the absolute and relative frequencies as we move from the first to the last interval.

The following steps are required when organizing data into a frequency distribution together with suggestions on constructing the frequency distribution.

- Identify the highest and lowest values of the observations.
- Setup classes (groups into which data is divided). The classes must be mutually exclusive and of equal size.
- Add up the number of observations and assign each observation to its class.
- Count the number of observations in each class. This is called the class frequency.

Data can be divided into two types: discrete and continuous.

**Discrete**: The values in the data set can be counted. There are distinct spaces between the values, such as the number of children in a family or the number of shares comprising an index.**Continuous**: The values in the data set can be measured. There are normally lots of decimal places involved and (theoretically, at least) there are no gaps between permissible values (i.e., all values can be included in the data set). Examples would include the height of a person and the time to complete an assignment. These values can be measured using sufficiently accurate tools to numerous decimal places.

There are two methods that graphically represent continuous data: histograms and frequency polygons.

- The class frequencies are shown on the vertical (y) axis (by the heights of bars drawn next to each other).
- The classes (intervals) are shown on the horizontal (x) axis.
- There is no space between the bars.

From a histogram, we can see quickly where most of the observations lie. The shapes of histograms will vary, depending on the choice of the size of the intervals.

- Absolute frequency for each interval is plotted on the vertical (y) axis.
- The midpoint of each class (interval) is shown on the horizontal (x) axis.
- Neighboring points are connected with a straight line.

Unlike a histogram, a frequency polygon adds a degree of

It is helpful, when drawing a frequency polygon, first to draw a histogram in pencil, then to plot the points and join the lines, and finally to rub out the histogram. In this way, the histogram can be used as an initial guide to drawing the polygon.

The

Let's look at an example.

The following table shows the holding period returns of a portfolio of 40 stocks.

The highest HPR is 32% and the lowest one is -27%. Let's use 6 non-overlapping intervals, each with a width of 10%. The first interval starts at -27% and the last one ends at 33%. Therefore, the entire range of the HPRs is covered.

Kapil: Lets denote Relative Frequency by RF and Absolute Frequency by AF, then RF of a interval is computed using below expressionRF=(AF x 100)/(sum of all AF) |

surob: Range/k - width of the interval. Good to remember |

JKiro: what other non-overlapping interval can we use? is 6 the only one in this particular problem. |

smartwealth: but how do we find the absolute frequency? |

Seemorr: Just count |

cfatime21: Can someone please explain how to find absolute frequency? |

jaswant: for finding absolute frequency just count the number of observation lying within the the given range of interval... |

johntan1979: surob is right and he's referring to the Sturges' formula.In the example above, to get the number of intervals, k, it is k = log n / log 2 = log 40 / log 2 = 5.3219 = 6 Always round UP And the width of each interval = Range / k = 59/6 = 9.8333 = 10 |

mcbreatz: One of my biggest questions from this reading was calculation of k so thank you! Do you always divide log n by log 2? Thank you. |