CONTENTS

1) Measures of Central Tendency (Introduction) Average (Meaning) Objectives/Advantages of Averaging Requisites of a good average Limitations of averaging

2) Arithmetic Mean

3) Median

4) Mode

5) Geometric Mean

6) Harmonic Mean

7) Use of Average in Science

8) Use of Average in Business

9) Some other Applications of Measures of Central Tendency

Bibliography

Measures of Central Tendency

Average Value
Definition An average value is a single value within the range of the data that is used to represent all of the values in the series. Since an average is somewhere within the range of the data, it is sometimes called the measure of central value. - Croxton & Cowden Since it represents the middle or typical value of the data set, it is also called the central tendency/value.

Objectives/Advantages of Averaging

1) Deriving a single value representing the entire data Averaging, by using different measures of central tendency, condenses a huge mass of data in one single value, giving us a birds-eye view of the entire data. For example, the average marks of a class give us a fair view of how the class is performing. 2) Facilitating Comparison Reducing the mass of data to one single value facilitates comparisons. For example, for comparing average marks of one section with the other section. However, while making comparison one should take into consideration the multiplicity of factors that might be affecting the data. For example, absolute increase in per capita income should not lead one to think that standard of living is necessarily improving because the prices might be rising faster than the rise in per capita income so in real terms people might be worse off. Moreover, the same measure should be used for making comparison between two or more groups. For example, mean wage of one factory and median wage of another should not be compared for any interpretation about wage levels.

3) Ease of further calculation Since averaging compresses the data in a single value, it provides the ease of performing further operations/ calculations for deriving further conclusions.

Requisites of a Good Average

1) Easy to understand 2) Simple to compute 3) Based on all observations of the data 4) Not affected by extreme observations 5) Rigidly defined 6) Capable of further algebraic treatment 7) Sampling stability (difference between samples average and populations average should be least possible)

Limitations of Averaging
1) May not be properly interpreted 2) Average value may not exist in data 3) Absurd results (fractional values) 4) Does not give idea about formation of series 5) Fails to represent the concentration of observations 6) Not meaningful in case of bi-modal, U-shaped and J-shaped distributions

Types of Averages
1) Arithmetic Mean 2) Median 3) Mode 4) Geometric Mean 5) Harmonic Mean

ARITHMETIC MEAN
The arithmetic mean is the central tendency of a collection of numbers, taken as the sum of the numbers divided by the size of the collection. Arithmetic mean may either be simple arithmetic mean or weighted arithmetic mean. The arithmetic mean of sample space is defined via the equation

Properties of Arithmetic Mean 1) The sum of deviations of the items from the arithmetic mean, taking signs into account, is always zero, i.e., (X Arithmetic mean) = 0. 2) The sum of the squared deviations of the items from arithmetic mean is minimum 3) If and are the arithmetic mean of two samples of sizes n1 and n2 respectively then, the of the distribution combining the two can be calculated as

arithmetic mean

Suitability 1) When the data does not include extreme values 2) When the average needs to be calculated taking into consideration all the values present in the data 3) When the data is not skewed, i.e., the frequency distribution for the data is not skewed) 4) The data is normally distributed, i.e., continuous and symmetrical

Merits 1) Simple 2) Affected by every observation in the data 3) Rigid 4) Facilitates further algebraic treatment 5) Sampling Stability 6) Calculated and not a positional value

Limitations 1) Affected by extreme values 2) Not suitable for open-ended data 3) Not suitable for skewed distribution

MEDIAN
Median is described as the numerical value separating the higher half of a sample, a population or a probability distribution, from its lower half. In other words, it is the middle value in a distribution. Derivation of Median from Ogives

Less than ogive

More than ogive

Intersection of Less than and More than Ogives

Property: The sum of deviations of items from median, ignoring the signs, is minimum.

Suitability

1) In case of open-ended classes 2) In case of unequal classes 3) In case of skewed distribution 4) In case of qualitative data

Merits 1) Useful for data with open-ended and unequal classes, skewed distribution and qualitative data 2) Can be determined graphically 3) Represents the middle value of the distribution

Limitations 1) Array is must 2) Does not consider every observation 3) Not capable of algebraic treatment 4) Sample fluctuation 5) Erratic if number of items are uneven

MODE
The mode or the modal value is that value in a series of observations which occurs with the greatest frequency.

Suitability It is most suitable in case of normal/nominal data.

Merits 1) Not affected by extreme values 2) Can be determined in case of open-ended intervals 3) Can be used in case of qualitative data 4) Can be determined graphically

Limitations 1) Not capable of algebraic manipulations 2) Not based on each and every item 3) Not rigidly defined 4) Cannot be always determined (in case of two modal classes) 5) May be misleading

In this case, there are two modal classes

In this case, mode is not representative

GEOMETRIC MEAN
The geometric mean is a type of mean or average, which indicates the central tendency or typical value of a set of numbers by using the product of their values. It is defined as the nth root of the product of n number of items or values. If the numbers are , the geometric mean or equals

Properties 1) The product of values of series will remain unchanged when the value of geometric mean is substituted for each individual value. For example, the G.M for 2, 4 and 8 is 4 so 2*4*8=64=4*4*4 2) The value of the G.M is such as to balance the ratio deviations of the observations from it

Suitability 1) G.M is most suitable to find average per cent increase such as in sales, population, investment or other economic or business series. 2) G.M is considered the best average in the construction of index numbers. It satisfies the time reversal test and gives equal weight to equal ratio of change. 3) G.M is most suitable when large weights have to be given to small items and small weights to large items.

Merits 1) Based on each and every item of series 2) Rigidly defined 3) Gives less weight to large items and more weight to small items 4) Capable of algebraic manipulations

Limitations 1) Restricted Application 2) Cannot be computed with both positive and negative values 3) Cannot be computed when one or more values are zero

HARMONIC MEAN
The harmonic mean is based on the reciprocals of the numbers averaged. It is defined as the reciprocal of the arithmetic mean of the reciprocal of the individual observation. Thus,

Suitability Harmonic Mean is useful in problems in which values of a variable are compared with a constant quantity of another variable, i.e., rates e.g., time, distance covered within certain time and quantities purchased or sold per unit,etc.

Merits 1) Value based on every item 2) Tends itself to algebraic manipulations 3) Problems relating to time and rates give better results using H.M

Limitations 1) Complicated/ Not very easy to understand 2) Gives largest weight to smallest item 3) Cannot be computed with both positive and negative values 4) Cannot be computed when one or more values are zero

RELATIONSHIP between Arithmetic Mean, Geometric Mean and Harmonic Mean

1) a b A.M G.M H.M

2) a = b

A.M = G.M = H.M

3) G.M = A.M *H.M

, i.e.

Square of G.M = A.M * H.M

Use of Average in Science

Microbiology
Use of Geometric mean In Bacteriology

Example: Mean Rate of Growth of Bacteria

If a strain of bacteria increases its population of 100 by 20% in the first hour, 30% in the next hour and
50% in the next hour, we can find out an estimate of the mean percentage growth in population. In this case, it is the geometric mean, and not the arithmetic mean that is relevant. - After the first hour, they grow to 120 bacteria, which is a grow rate of 1.2 (100*1.2) - After the second hour, they grow to 156 bacteria, which is a grow rate of 1.3 (120*1.3) - After the third hour, they grow to 234 bacteria, which is a grow rate of 1.5 (156*1.5)

To find the mean growth rate

Geometric Mean = (a1 a2 . . . an)^1/n

Geometric Mean = (1.2 1.3 x 1.5) ^1/3

Geometric Mean = (1.2 1.3 x 1.5) ^1/3

Geometric Mean = (2.34) ^1/3 Geometric Mean 1.3276

if we find the geometric mean of 1.2, 1.3 and 1.5, we get 1.3276. This should be interpreted as the mean rate of growth of the bacteria over the period of 3 hours, which means if the strain of bacteria grew by 32.76% uniformly over the 3 hour period, then starting with 100 bacteria; it would reach 234 bacteria in 3 hours.

Acoustics
Calculation of Spectral Flatness

Spectral Flatness or tonality coefficient, also known as Wiener entropy, is a measure used in digital signal processing to characterize an audio spectrum. Spectral flatness provides a way to quantify how tone-like a sound is, as opposed to being noise-like.

The spectral flatness is calculated by dividing the geometric mean of the power spectrum by the arithmetic mean of the power spectrum, i.e.:

Spectral Flatness is one of the many audio descriptors used in the MPEG-7 standard, in which it is labeled Audio Spectral Flatness.

In birdsong (bird vocalization) research, it has been used as one of the features measured on birdsong audio, when testing similarity between two excerpts.

Use of Average in Business

Average Salary
Median can be used to determine the average salary across different parameters such as type of industry, job, and years of experience since it is found by arranging the observations from highest o lowest or vice-versa and then picking the middle value as the average.

Example: Median Annual Salary in India across different Industries

(A report updated on 23rd January, 2013 by PayScale.com that provides data and insights around salary and career topics for various publications such as CNN Money, Time, Bloomberg, The Washington Post, etc)

INDUSTRY
Information Technology (IT) Services

NATIONAL SALARY DATA (in Rs)

550,602

Software Development

515,868

Banking

534,696

Software Services

571,314

Information Technology (IT) Consulting

633,690

Telecommunications

598,584

Financial Services

571,493

*In the above example, average salary of each industry has been calculated by arranging the data in ascending order and then finding the middle observation depending upon the number of observations (odd or even).

Some other Applications of Measures of Central Tendency

Arithmetic Mean
1) Calculation of Per Capita Income 2) Calculating average marks of the class for judging the performance of a class.

Median
1) Calculation of the average production in a day. 2) Calculation of average wages given to the employees with same skills in a company.

Mode
1) Calculation of the average shoe/ garment size or average weight/height in a population. 2) Finding out the most watched TV show or the most saleable color in a particular season.

Geometric Mean
1) Calculation of average per cent increase in sales/ production/ population, etc 2) Computation of average daily exchange rate of a currency.

Harmonic Mean
1) Calculation of the Price/Earnings Ratio. 2) Calculation of average speed over the distance travelled.

BIBLIOGRAPHY
(a) Statistical Methods by S.P. Gupta and Archana Gupta.

(b) "Transform coding of audio signals using perceptual noise criteria" Article by J. D. Johnston (1988) (Spectral Flatness)

(c) http://recherche.ircam.fr/equipes/analysesynthese/peeters/ARTICLES/Peeters_2003_cuidadoaudiofeatures.pdf

(d) http://soundanalysispro.com/manual-1/chapter-4-the-song-features-of-sap2/wiener-entropy

(e) http://www.payscale.com/research/IN/Country=India/Salary#by_Industry

(f) "Fairness Opinions: Common Errors and Omissions" The Handbook of Business Valuation and Intellectual Property Analysis, McGraw Hill, 2004.

(g) University of Toronto Mathematics Network http://www.math.toronto.edu/mathnet/questionCorner/geomean.html

(h) http://explorable.com/geometric-mean.html (Microbiology)