Sunteți pe pagina 1din 47

1 Data Classification, Tabulation, and Presentation

1.1 CLASSIFICATION OF DATA


Classification of data is the process of arranging data in groups/classes on the basis of certain properties. Classification of statistical data serves the following purposes: 1. 2. 3. ". #. It condenses the raw data into a form suitable for statistical analysis. It removes complexities and highlights the features of the data. It facilitates comparisons and drawing inferences from the data. or example! if university students in a particular course are divided according to sex! their results can be compared. It provides information about the mutual relationships among elements of a data set. or example! based on literacy and criminal tendency of a group of people! it can be established whether literacy has any impact on criminal tendency or not. It helps in statistical analysis by separating elements of the data set into homogeneous groups and hence brings out the points of similarity and dissimilarity.

Basis of Classification $enerally! data are classified on the basis of the following four bases: Geo ra!"ical Classification In geographical classification! data are classified on the basis of geographical or locational differences % such as cities! districts! or villages % between various elements of the data set. &he following is an example of a geographical distribution.

C"ronolo ical Classification 'hen data are classified on the basis of time! the classification is (nown as chronological classification. )uch classifications are also called time series because data are usually listed in chronological order starting with the earliest period. &he following example would give an idea of chronological classification:

#ualitati$e Classification In *ualitative classification! data are classified on the basis of descriptive characteristics or on the basis of attributes li(e sex! literacy! region! caste! or education! which cannot be *uantified. &his is done in two ways:

1. 2.

Simple classification: In this type of classification! each class is subdivided into two sub+classes and only one attribute is studied! for example male and female, blind and not blind! educated and uneducated, and so on. Manifold classification: In this type of classification! a class is subdivided into more than two subclasses which may be sub+divided further.

#uantitati$e Classification In this classification! data are classified on the basis of characteristics which can be measured such as height! weight! income! expenditure! production! or sales. -xamples of continuous and discrete variables in a data set are shown in &able 1.1.

Table 1.1

1.% O&GANI'ING DATA (SING DATA A&&A)


&able 1.2 presents the total number of overtime hours wor(ed for 3. consecutive wee(s by machinists in a machine shop. &he data displayed here are in raw form, that is! the numerical observations are not arranged in any particular order or se*uence.

Table 1.% /aw 0ata 1ertaining to &otal &ime 2ours 'or(ed by 3achinists

&he raw data can be reorgani4ed in a data array and fre*uency distribution. )uch an arrangement enables us to see *uic(ly some of the characteristics of the data we have collected. 'hen a raw data set is arranged in ran( order! from the smallest to the largest observation or vice+versa! the ordered se*uence obtained is called an ordered array. &able 1.3 reorgani4es data given in &able 1.2 in the ascending order

Table 1.* 5rdered 6rray of &otal 5vertime 2ours 'or(ed by 3achinists

It may be observed that an ordered array does not summari4e the data in any way as the number of observations in the array remains the same. Fre+uenc, Distribution 6 fre+uenc, distribution divides observations in the data set into conveniently established numerically ordered classes 7groups or categories8. &he number of observations in each class is referred to as frequency denoted as f. )ummari4ing data should not be at the cost of losing essential details. &he purpose should be to see( an appropriate compromise between having too much of details or too little. &o be able to achieve this compromise! certain criteria are discussed for constructing a fre*uency distribution. &he fre*uency distribution of the number of hours of overtime given in &able 1.2 is shown in &able 1.".

Table 1.- 6rray and &allies

Number of Overtime Hours Tally Number of Weeks (Frequency) 9" 2 9# 9: 9; 99 9< <. <1 <2 <3 <" <# % 2 . 1

" 3 2 2 2 : # 1 3.

Constructin a Fre+uenc, Distribution 6s the number of observations obtained gets larger! the method discussed above to condense the data becomes *uite difficult and time+consuming. &hus! to further condense the data into fre*uency distribution tables! the following steps should be ta(en: 1. 2. 3. )elect an appropriate number of non+overlapping class intervals. 0etermine the width of the class intervals. 0etermine class limits 7or boundaries8 for each class interval to avoid overlapping.

1. Decide the number of class intervals &he decision on the number of class groupings depends largely on the =udgment of the individual investigator and/or the range that will be used to group the data! although there are certain guidelines that can be used. 6s a general rule! a fre*uency distribution should have at least five class intervals 7groups8! but not more than fifteen. &he following two rules are often used to decide approximate number of classes in a fre*uency distribution:

1.

If k represents the number of classes and N the total number of observations! then the value of k will be the smallest exponent of the number 2! so that 2k > N.

If N ? 3. observations. If we apply this rule! then we shall have 23 ? 9 7@3.8! 2" ? 1: 7@3.8! 2# ? 32 7A3.8.

&hus we may choose k ? # as the number of classes. 2. 6ccording to )turgeBs rule! the number of classes can be determined by the formula

k ? 1 C 3.222 loge N where k is the number of classes and loge N is the logarithm of the total number of observations.

or

N ? 3.! we have k ? 1 C 3.222 log 3. ? 1 C 3.222 71.";;18 ? #.;#< #.

%. Determine the width of class intervals &he si4e 7or width8 of each class interval can be determined by first ta(ing the difference between the largest and smallest numerical values in the data set and then dividing it by the number of class intervals desired.

&he value obtained from this formula can be rounded off to a more convenient value based on the investigatorBs preference. rom the ordered array in &able 1.3! the range is <# D 9" ? 11 hours. Esing the above formula with # classes desired! the width of the class intervals is approximated as

width of class interval ?

hours.

or convenience! the selected width 7or interval8 of each class is rounded to 3 hours. *. Determine class limits (boundaries) &he limits of each class interval should be clearly defined so that each observation 7element8 of the data set belongs to one and only one class. -ach class has two limits%a lower limit and an upper limit. &he usual practice is to let the lower limit of the first class be a convenient number slightly below or e*ual to the lowest value in the data set. In &able 1.3! we may ta(e the lower class limit of the first class as 92 and the upper class limit as 9#. &hus the class would be written as 92F9#. &his class interval includes all overtime hours ranging from 92 upto but not including 9# hours. &he various other classes can be written as shown below:

O$erti.e /ours Tallies Fre+uenc, 0Class Inter$als1 92 but less than 9# 2 9# but less than 99 3

99 but less than <1 <1 but less than <" <" but less than <;

< 1. : 3.

2id3!oint of Class Inter$als &he class .id3!oint is the point halfway between the boundaries 7both upper and lower class limits8 of each class and is representative of all the observations contained in that class. &he width of the class interval should! as far as possible! be e*ual for all the classes. If this is not possible to maintain! the interpretation of the distribution becomes difficult. 2et"ods of Data Classification &here are two ways in which observations in the data set are classified on the basis of class intervals! namely 1. 2. -xclusive method Inclusive method

45clusi$e 2et"od 'hen the data are classified in such a way that the upper limit of a class interval is the lower limit of the succeeding class interval 7i.e.! no data point falls into more than one class interval8! then it is said to be the exclusive method of classifying data. &his method is illustrated in &able 1.#.

Table 1.6 -xclusive 3ethod of 0ata Classification

Dividends Declared in er cent Number of !om#anies (!lass "ntervals) .F1. 1.F2. 2.F3. 3.F". # ; 1# 1. (Frequency)

6s shown in &able 1.#! five companies declared dividends ranging from . to 1. per cent! this means a company which declared exactly 1. per cent dividend would not be included in the class .F1. but would be included in the next class 1.F2.. )ince this point is not always clear! to avoid confusion data are displayed in a slightly different manner! as given in &able 1.:.

Table 1.7

Dividends Declared in er cent Number of !om#anies (!lass "ntervals) . but less than 1. (Frequency) #

1. but less than 2. 2. but less than 3. 3. but less than ".

; 1# 1.

Inclusi$e 2et"od 'hen the data are classified in such a way that both lower and upper limits of a class interval are included in the interval itself! then it is said to be the inclusive method of classifying data. &his method is shown in &able 1.;.

Table 1.8 Inclusive 3ethod of 0ata Classification

Number of $ccidents Number of Weeks (!lass "ntervals) .F" #F< 1.F1" 1#F1< 2.F2" (Frequency) # 22 13 9 2

If a continuous variable is classified according to the inclusive method! then certain ad=ustment in the class interval is needed to obtain continuity as shown in &able 1.9.

Table 1.9

!lass "ntervals Frequency 3.F"" 29 "#F#< 32 :.F;" "# ;#F9< #. <.F1." 3#


&o ensure continuity! first calculate correction factor as

and then subtract it from the lower limits of all the classes and add it to the upper limits of all the classes. rom &able 1.9! we have x ? 7"# D ""8 2 ? ..#. )ubtract ..# from the lower limits of all the classes and add ..# to the upper limits. &he ad=usted classes would then be as shown in &able 1.<.

Table 1.:

!lass "ntervals Frequency 2<.#F"".# 29 "".#F#<.# 32 #<.#F;".# "# ;".#F9<.# #. 9<.#F1.".# 3#


Class intervals should be of e*ual si4e to ma(e meaningful comparison between classes. In a few cases! extreme values in the data set may re*uire the inclusion of open-ended classes and this distribution is (nown as an openended distribution. 6n example of an open+ended distribution is given in &able 1.1..

Table 1.1;

$%e (&ears) Ender #

o#ulation ('illions) 1;.9 "".;

#F1; 2<.< 19F2" :<.: 2#F"" "".: "#F:" 2;." :# and above 23"..
&able 1.11 provides a tentative guide to determine an ade*uate number of classes.

Table 1.11 $uide to 0etermine the Gumber of Classes to Ese

Number of Observations (N) (u%%ested Number of !lasses 2. # #. ; 1.. 9 2.. < #.. 1. 1... 11

45a.!le 1.1< &he following set of numbers represents mutual fund prices reported at the end of a wee( for selected ". nationally sold funds.

6rrange these prices into a fre*uency distribution having a suitable number of classes. Solution< )ince the number of observations are ".! it seems reasonable to choose : 72 : A ".8 class intervals to summari4e values in the data set. 6gain! since the smallest value is 1. and the largest is 39! the class interval is given by

Gow performing the actual tally and counting the number of values in each class! we get the fre*uency distribution by exclusive method as shown in &able 1.12.

Table 1.1% re*uency 0istribution

!lass "nterval Tally Frequency 0'utual Fund rices) *s+1 0Number of 'utual Funds) 1.F1# : 1#F2. 2.F2# 2#F3. 3.F3# 3#F". 11 < ; # 2 ".
45a.!le 1.%< &he ta(e+home salary 7in /s.8 of ". uns(illed wor(ers from a company for a particular month was

Construct a fre*uency distribution having a suitable number of classes. Solution< )ince the number of observations are 3.! we choose #72 #A3.8 class intervals to summari4e values in the data set. In the data set the smallest value is 23:# and the largest is 2#".! so the width of each class interval will be

)orting the data values into classes and counting the number of values in each class! we get the fre*uency distribution by exclusive method as given in &able 1.13.

Table 1.1* re*uency 0istribution

!lass "nterval Tally ((alary) *s+) 23:#F2".. 2"..F2"3# 2"3#F2";. 2";.F2#.# 2#.#F2#".

Frequency (Number of Workers) : ; 1. : 1 3.

45a.!le 1.*< 6 computer company received a rush order for as many home computers as could be shipped during a :+wee( period. Company records provide the following daily shipments:

$roup these daily shipments figures into a fre*uency distribution having the suitable number of classes. Solution< )ince the number of observations are "2! it seems reasonable to choose :72 :A"28 classes. 6gain! since the smallest value is 22 and the largest is 9;! the class interval is given by

Gow performing the actual tally and counting the number of values in each class! we get the following fre*uency distribution by inclusive method as shown in &able 1.1".

Table 1.1- re*uency 0istribution

!lass "nterval Tally Frequency (Number of !om#uters) (Number of Days) 22F32 " 33F"3 ""F#" " <

##F:# ::F;: ;;F9;

1" : # "2

45a.!le 1.-< ollowing is the increase of 0.6. in the salaries of employees of a firm at the following rates.

/s. 2#. for the salary range up to /s. ";"< /s. 2:. for the salary range from /s. ";#. /s. 2;. for the salary range from /s. "<#. /s. 29. for the salary range from /s. #1#. /s. 2<. for the salary range from /s. #3#. Go increase of 0.6 for salary of /s. ##.. or more. 'hat will be the additional amount re*uired to be paid by the firm in a year which has 32 employees with the following salaries 7in /s.8H

Solution< 1erforming the actual tally and counting the number of employees in each salary range 7or class8! we get the following fre*uency distribution as shown in &able 1.1#.

Table 1.16 re*uency 0istribution

2ence additional amount re*uired by the firm for payment of 0.6. is /s. 9:#.. 45a.!le 1.6< ollowing are the number of items of similar type produced in a factory during the last #. days.

6rrange these observations into a fre*uency distribution with both inclusive and exclusive class intervals choosing a suitable number of classes. Solution< )ince the number of observations are #.! it seems reasonable to choose :72 : A #.8 or less classes. )ince smallest value is 1" and the largest is 3:! the class interval is given by

1erforming the actual tally and counting the number of observations in each class! we get the following fre*uency distribution with inclusive class intervals as shown in &able 1.1:.

Table 1.17 re*uency 0istribution with Inclusive Class Intervals

!lass "ntervals Tally 1"F1; 19F21 22F2# 2:F2< 3.F33 3"F33

Frequency (Number of "tems roduced) : 19 1# #

3 3 #.

Converting the class intervals shown in &able 1.1: into exclusive class intervals is shown in &able 1.1;.

Table 1.18 re*uency 0istribution with -xclusive Class Intervals

!lass "ntervals 'id,-alue of !lass "ntervals 13.#F1;.# 1;.#F21.# 21.#F2#.# 2#.#F2<.# 2<.#F33.# 33.#F3;.# 1#.# 1<.# 23.# 2;.# 31.# 3".#

Frequency (Number of "tems roduced) : 19 1# # 3 3

45a.!le 1.7< Classify the following data by ta(ing class such that their mid+values are 1;! 22! 2;! 32! and so on.

IMadurai-Kamaraj ni!., ".#om., $%%&J Solution< )ince we have to classify the data in such a manner that the mid+values are 1;! 22! 2;! etc.! the first class should be 1#F1< 7mid+value ? 71# C 1<8/2 ? 1;8! second class should be 2.F2"! etc. 1erforming the actual tally and counting the number of observations in each class we get the fre*uency distribution as shown in &able 1.19.

Table 1.19 re*uency 0istribution with Inclusive Class Intervals

'arks Tallies Frequency 1#F1< " 2.F2" " 2#F2< " 3.F3" 9 3#F3< ".F"" "#F"< #.F#" *:
45a.!le 1.8< 3ar(s obtained by #. students are given below:

" < 3 3

I#alicut ni!., M. #om., $%%'J Construct a grouped fre*uency distribution. Solution< )ince the number of observations are #.! we may choose :72:A#.8 or less classes. &he lowest value is 2 and largest :3! the class intervals shall be

&he fre*uency distribution is shown in &able 1.1<.

Table 1.1: re*uency 0istribution

'arks Tallies

Frequency (Number of (tudents)

; 2F12 1" 13F23 1" 2"F3" 13 3#F"# 1 "#F#: 1 #;F:; 1 #.


45a.!le 1.9< 1oint out the mista(es in the following table to show the distribution of population according to sex! age! and literacy.

I"ombay ni!., M. #om., ())&J Solution< 6ll the characteristics are not revealed in the given table. &he characteristic of literacy are complete and hence table needs to be re+arranged as shown in &able 1.2..

Table 1.%; 0istribution of 1opulation 6ccording to 6ge! )ex! and Kiteracy

45a.!le 1.:< 1. 1resent the following data of the percentage mar(s of :. students in the form of a fre*uency table with 1. classes of e*ual width! one class being #.F#<.

I#S*, +oundation, $%%,J 2. 7b8 6 sample consists of 3" observations recorded correct to the nearest integer! ranging in value from 2.1 to 33;. If it is decided to use seven classes of width 2. integers and to begin in the first class at 1<<.#! find the class mar(s of the seven classes.

I#alicut ni!., ".Sc., $%%-J Solution< 1. )ince the number of observations are :.! we may choose :72: A :.8 or less class intervals. &he class interval is given by

&he fre*uency distribution with inclusive intervals is shown in &able 1.21.

Table 1.%1 re*uency 0istribution

'arks Tallies Frequency : .F1# # 1:F31 1: 32F"; 19 "9F:3 9 :"F;< ; 9.F<# :. 2.


)ince it is decided to begin with 1<<.# and ta(es a classes interval of 2.! the first class will be 1<<.#F 21<.#! the second would be 21<.#F23<.#! and so on. &he class mar( shall be obtained by adding the lower and upper limits and dividing it by 2. &hus! for the first class! the mar(s shall be 71<<.#8 C 721<.#8/2 ? 2.<.#. )ince class interval is e*ual the other class mar(s can be obtained by adding 2. to the preceding class mar(. &able 1.22 gives the class limits and class mar(s of the seven classes.

Table 1.%%

!lass .imits !lass 'arks 1<<.#F21<.# 2.:.#

21<.#F23<.# 23<.#F2#<.# 2#<.#F2;<.# 2;<.#F2<<.# 2<<.#F31<.# 31<.#F33<.#

22<.# 2"<.# 2:<.# 29<.# 3.<.# 32<.#

Bi$ariate Fre+uenc, Distribution &he fre*uency distributions discussed so far involved only one variable and are therefore called uni!ariate frequency distributions. In case the data involve two variables 7such as profit and expenditure on advertisements of a group of companies! income and expenditure of a group of individuals! supply and demand of a commodity! etc.8! then fre*uency distribution so obtained as a result of cross classification is called bi!ariate frequency distribution. It can be summari4ed in the form of a two-way .bi!ariate/ frequency table and the values of each variable are grouped into various classes 7not necessarily same for each variable8 in the same way as for univariate distributions. re*uency distribution of variable x for a given value of y is obtained by the values of x and vice versa. )uch fre*uencies in each cell are called conditional frequencies. &he fre*uencies of the values of variables x and y together with their fre*uency totals are called the mar0inal frequencies. 45a.!le 1.1;< &he following figures indicate income 7x8 and percentage expenditure on food 7y8 of 2# families. Construct a bivariate fre*uency table classifying x into intervals 2..F3..! 3..F"..! L! and y into 1.F1#! 1#F2.! L 'rite the marginal distribution of x and y and the conditional distribution of x when y lies between 1# and 2..

Solution< &he two+way fre*uency table showing income 7in /s.8 and percentage expenditure on food is shown in &able 1.23.

Table 1.%*

&he conditional distribution of x when y lies between 1# and 2. per cent is as follows:

45a.!le 1.11< &he following data give the points scored in a tennis match by two players 1 and 2 at the end of 2. games:

&a(ing class intervals as #F<! 1.F1"! 1#F1<! L! for both 1 and 2, construct 1. 2. Mivariate fre*uency table Conditional fre*uency distribution for 2 given 1 A 1#

Solution< 7a8 &he two+way fre*uency distribution is shown in &able 1.2".

Table 1.%- Mivariate re*uency &able

7ii8 Conditional fre*uency distribution for 2 given 1A1#.

layer & #F< 1.F1" 1#F1< 2.F2"

layer / 01203 45246 1 % 1 % 3 1 1 1 : 2

45a.!le 1.1%< 3. pairs of values of two variables 1 and 2 are given below. rom a two+way table!

&a(e class intervals of 1 as 1.F2.! 2.F3.! etc.! and that of 2 as 1..F2..! 2..F3... etc.

I3smania ni!., ".#om., $%%'J Solution< &he two+way fre*uency distribution is shown in &able 1.2#.

Table 1.%6 Mivariable re*uency &able

T,!es of Fre+uenc, Distributions Cu.ulati$e Fre+uenc, Distribution )ometimes it is preferable to present data in a cumulati!e frequency .cf8 distribution. 6 cumulative fre*uency distribution is of two types: 7i8 more t4an type and 7ii8 less t4an type.

In a less t4an cumulative fre*uency distribution! the fre*uencies of each class interval are added successively from top to bottom and represent the cumulative number of observations less than or e*ual to the class fre*uency to which it relates. Mut in the more t4an cumulative fre*uency distribution! the fre*uencies of each class interval are added successively from bottom to top and represent the cumulative number of observations greater than or e*ual to the class fre*uency to which it relates. &he fre*uency distribution given in &able 1.2: illustrates the concept of cumulative fre*uency distribution.

Table 1.%7 Cumulative re*uency 0istribution

&he Nless thanB cumulative fre*uencies are corresponding to the upper limit of class intervals and Nmore thanB cumulative fre*uencies are corresponding to the lower limit of class intervals shown in &ables 1.2;7a8 and 7b8.

Table 1.%80a1

7##er .imits !umulative Frequency (.ess Than) Kess than " Kess than < Kess than 1" Kess than 1< Kess than 2"
Table 1.%80b1

# 2; ". "9 #.

.ower .imits !umulative Frequency ('ore Than) . and more # and more 1. and more 1# and more 2. and more #. "# 23 1. 2

&elati$e Fre+uenc, Distribution &o convert a fre*uency distribution into a corresponding relative frequency distribution) we divide each class fre*uency by the total number of observations in the entire distribution. -ach relative fre*uency is thus a proportion as shown in &able 1.2<. Percenta e Fre+uenc, Distribution 6 #ercenta%e frequency distribution is one in which the number of observations for each class interval is converted into a percentage fre*uency by dividing it by the total number of

observations in the entire distribution. &he *uotient so obtained is then multiplied by 1..! as shown in &able 1.29.

Table 1.%9 /elative and 1ercentage re*uency 0istributions

45a.!le 1.1*< ollowing are the number of two wheelers sold by a dealer during 9 wee(s of : wor(ing days each.

1. 2.

$roup these figures into a table having the classes 1.F12! 13F1#! 1:F19! L! and 29F3.. Convert the distribution of part 7a8 into a corresponding percentage fre*uency distribution and also a percentage cumulative fre*uency distribution.

Solution<

1.

re*uency distribution of the given data is shown in &able 1.2<.

Table 1.%: re*uency 0istribution

Number of $utomobiles (old Tally Number of Days (!lass "ntervals) 1.F12 13F1# 1:F19 1<F21 22F2" 2#F2; 29F3. 2. 2 : 1. 1: 9 # 1 "9
1ercentage fre*uency distribution is shown in &able 1.3..

(Frequency)

Table 1.*; 1ercentage and 3ore &han Cumulative 1ercentage 0istribution

Self3Practice Proble.s 1A

1.1 orm a fre*uency distribution of the following data. Ese an e*ual class interval of " where the lower limit of the first class is 1.. 1. 1; 1# 22 11 1: 1< 2" 2< 19 2# 2: 32 1" 1; 2. 23 2; 3. 12 1# 19 2" 3: 19 1# 21 29 33 39 3" 13 1. 1: 2. 22 2< 2< 23 31 1.% If class mid+points in a fre*uency distribution of the ages of a group of persons are 2#! 32! 3<! ":! #3! and :.! find 1. 2. 3. the si4e of the class interval! the class boundaries! the class limits! assuming that the age *uoted is the age completed on the last birthdays.

1.* &he distribution of ages of #.. readers of a nationally distributed maga4ine is given below:

$%e (in &ears) Number of *eaders Melow 1" 2. 1#F1< 12# 2.F2" 2# 2#F2< 3# 3.F3" 9. 3#F3< 1". ".F"" 3. "# and above "#
ind the relative and cumulative fre*uency distributions for this distribution. 1.- &he distribution of inventory to sales ratio of 2.. retail outlets is given below:

"nventory to (ales *atio Number of *etail Outlets 1..F1.2 2. 1.2F1." 3. 1."F1.: :. 1.:F1.9 ".

1.9F2.. 2..F2.2 2.2F2."

3. 1# #

ind the relative and cumulative fre*uency distributions for this distribution. 1.6 6 wholesalerBs daily shipments of a particular item varied from 1!1#2 to <!999 units per day. Indicate the limits of nine classes into which these shipments might be grouped.

1.7 6 college boo( store groups the monetary value of its sales into a fre*uency distribution with the classes! /s. "..F#..! /s. #.1F:..! and /s. :.1 and over. Is it possible to determine from this distribution the amount of sales 1. 2. 3. less than /s. :.1 less than /s. #.1 /s. #.1 or moreH

1.8 &he class mar(s of distribution of the number of electric light bulbs replaced daily in an office building are #! 1.! 1#! and 2.. ind 7a8 the class boundaries and 7b8 class limits.

1.9 &he mar(s obtained by 2# students in )tatistics and -conomics are given below. &he first figure in the brac(et indicates the mar(s in )tatistics and the second in -conomics.

1repare a two+way fre*uency table ta(ing the width of each class interval as " mar(s! the first being less than ". 1.: 1repare a bivariate fre*uency distribution for the following data for 2. students:

6lso prepare 1. 2. a marginal fre*uency table for mar(s in Kaw and )tatistics a conditional fre*uency distribution for mar(s in Kaw when the mar(s in )tatistics are more than 22.

1.1; Classify the following data by ta(ing class intervals such that their mid+values are 1;! 22! 2;! 32! and so on:

IMadurai-Kamraj ni!., ".#om., ())&J

1.11 In degree colleges of a city! no teacher is less than 3. years or more than :. years in age. &heir cumulative fre*uencies are as follows:

ind the fre*uencies in the class intervals 2#F3.! 3.F3#! L

/ints and Ans=ers

1.1 &he classes for preparing fre*uency distribution by inclusive method will be 1.F13! 1"F1;! 19F21! L! 3"F3;! 39F"1 1.% 1. 2. )i4e of the class interval ? 0ifference between the mid+values of any two consecutive classes ? ; &he class boundaries for different classes are obtained by adding 7for upper class boundaries or limits8 and subtracting 7for lower class boundaries or limits8 half the magnitude of the class interval! that is! ; 2 ? 3.# from the midvalues. #lass *nter!als: 21.#F29.# 29.#F3#.# 3#.#F"2.# Mid-5alues: 2# 32 3< #lass *nter!als: "2.#F"<.# "<.#F#:.# #:.#F:3.# Mid-5alues: ": #3 :. 3. &he distribution can be expressed in inclusive class intervals with width of ; as 22F29! 2<F3#! L! #:F :3.

1.6 5ne possibility is 1...F1<<<! 2...F2<<<! 3...F3<<<! L! <...F<<<< units of the item.

1.11

1.* TAB(LATION OF DATA

&abulation is another way of summari4ing and presenting the given data in a systematic form in rows and columns. )uch presentation facilitates comparisons by bringing related information close to each other and helps in further statistical analysis and interpretation. Parts of a Table

1.

2.

3. ". #. :.

Table nu.ber< 6 table should be numbered for easy identification and reference in future. &he table number may be given either in the centre or side of the table but above the top of the title of the table. If the number of columns in a table is large! then these can also be numbered so that easy reference to these is possible. Title of t"e table< -ach table must have a brief! self+explanatory! and complete title which can 1. indicate the nature of data contained. 2. explain the locality 7i.e.! geographical or physical8 of data covered. 3. indicate the time 7or period8 of data obtained. ". contain the source of the data to indicate the authority for the data! as a means of verification and as a reference. &he source is always placed below the table. Ca!tion and stubs< &he headings for columns and rows are called caption and stub! respectively. &hey must be clear and concise. Bod,< &he body of the table should contain the numerical information. &he numerical information is arranged according to the descriptions given for each column and row. Prefactor, or "ead note< If needed! a prefactory note is given =ust below the title for its further description in a prominent type. It is usually enclosed in brac(ets and is about the unit of measurement. Footnotes< 6nything written below the table is called a footnote. It is written to further clarify either the title captions or stubs. or example! if the data described in the table pertain to profits earned by a company! then the footnote may define whether it is profit before tax or after tax. &here are various ways of identifying footnotes: 1. Gumbering footnotes consecutively with small number 1! 2! 3! L! or letters a! b! c! L! or star O! OO! L 2. )ometimes symbols li(e P or Q are also used to identify footnotes.

6 blan( model table is given below:

&able Gumber and &itle I2ead or 1refactory Gote 7if any8J

+ootnote : Source Note :

T,!es of Tables &he classification of tables depends on various aspects: ob=ectives and scope of investigation! nature of data 7primary or secondary8 for investigation! extent of data coverage! and so on. &he different types of tables used in statistical investigations are as follows:

Si.!le and Co.!le5 Tables In a simple table 7also (nown as one+way table8! data are presented based on only one characteristic. &able 1.31 illustrates the concept.

Table 1.*1 Candidates Interviewed for -mployment in a Company

!andidate8s rofile Number of !andidates #. -xperienced ;. Inexperienced 12. &otal


In a complex table 7also (nown as a manifold table8 data are presented according to two or more characteristics simultaneously. &he complex tables are two+way or three+way tables according to whether two or three characteristics are presented simultaneously.

1.

6ouble or 7wo-8ay 7able: In such a table! the variable under study is further subdivided into two groups according to two inter+related characteristics. or example! if the total number of candidates given in &able 1.31 are further divided according to their sex! the table would become a two+way table because it would reveal information about two characteristics! namely male and female. &he new shape of the table is shown in &able 1.32.

Table 1.*% Candidates Interviewed for -mployment in a Company

2.

3.

74ree-8ay 7able: In such a table! the variable under study is divided according to three interrelated characteristics. or example! if the total number of male and female candidates given in &able 1.32 are further divided according to the marital status! the table would become a three+way. &he new shape of the table is shown in &able 1.3#. Manifold .or 9i04er 3rder/ 7able: )uch tables provide information about a large number of interrelated characteristics in the data set. or example! if the data given in &able 1.33 is also available for other companies! then the table would become a manifold table.

Table 1.** Candidates Interviewed for -mployment in a Company

Ori inal and Deri$ed Tables 5riginal tables are also called classification tables. )uch a table contains data collected from a primary source. Mut if the information given in a table has been derived from a general table! then such a table is called a deri!ed table. 45a.!le 1.1-< 6 state government has ta(en up a scheme of providing drin(ing water to every village. 0uring the first " years of a five+year plan! the government has installed 3<!::" tubewells. 5ut of the funds earmar(ed

for natural calamities! the government has sun( 1"!.;2 tubewells during the first " years of the plan. &hus! out of the plan fund <2"# and 9:3. tubewells were sun( in 2...F.1 and 2..1F.2! respectively. 5ut of the natural calamities fund! the number of tubewells sun( in 1<<9F<< and 1<<<2... were "#11 and :3;! respectively. &he expenditure for 2...F.1 and 2..1F.2 was /s. 9:3."1 la(h and /s. 119#.:# la(h! respectively. &he number of tubewells installed in 2..2F.3 was 1:!;". out of which "9.. were installed out of the natural calamities fund and the expenditure of sin(ing of tubewells during 2..2F.3 was /s. 1"11.1; la(h. &he number of tubewells installed in 2..3F." was 13!<;3! out of which <9"< tubewells were sun( out of the fund for the plan and the total expenditure during the first " years was /s. #""3..# la(h. /epresent this data in a tabular form. Solution< &he data of the problem is summari4ed in &able 1.3".

Table 1.*- &ubewells for 0rin(ing 'ater for Rillages in a )tate

45a.!le 1.16< In a sample study about coffee+drin(ing habits in two towns! the following information was received:

&own 6 &own M

: emales were ". per cent. &otal coffee drin(ers were "# per cent and male non+ coffee drin(ers were 2. per cent : 3ales were ## per cent. 3ale non+coffee drin(ers were 3. per cent and female coffee drin(ers were 1# per cent.

/epresent this data in a tabular form. Solution< &he given data is summari4ed in &able 1.3#.

Table 1.*6 Coffee 0rin(ing 2abit of &owns 6 and M 7in 1ercentage8

45a.!le 1.17< Industrial finance in India has showed great variation in respect of sources of funds during the first! second! and third five+year plans. &here were two main sources%internal and external. &he internal sources of funds are depreciation! free reserves! and surplus. &he external sources of funds are capital issues and borrowings.

0uring the first plan! internal and external sources accounted for :2 per cent and 39 per cent of the total! and of the depreciation! fresh capital! and other sources formed 2< per cent! ; per cent! and 1..: per cent! respectively. 0uring the second plan! internal sources decreased by 1;.3 per cent compared to the first plan! and depreciation was 2".# per cent. &he external finance during the same period consisted of 1..< per cent fresh capital and 29.< per cent borrowings. Compared to the second plan! external finance during the third plan decreased by "." per cent! and borrowings and Nother sourcesB were 2<." per cent and 1".< percent respectively. 0uring the third plan! internal finance increased by "." per cent and free reserves and surplus formed 19.: per cent. &abulate this information with the above details as clearly as possible observing the rules of tabulation. Solution< &he given information is summari4ed in &able 1.3:.

Table 1.*7 1attern of Industrial inance 7in 1ercentage8

45a.!le 1.18< &he following information about weather conditions at different stations was recorded at 9.3. a.m. on &hursday! 2< 6ugust 1<<.. 6t 6hmednagar station! the maximum and minimum temperatures in 2" hours were 29SC and 2.SC respectively. &he rainfall in the past 2" hours at 6hmednagar was nil. )ince 1 Tune the rainfall was 19# mm which is 1.# mm below normal. MangaloreBs minimum and maximum temperatures in 2" hours for the day were 1<SC and 23SC respectively. It had no rainfall in the past 2" hrs and since 1 Tune the rainfall was 2#2 mm which is #" mm below normal. &he minimum temperature at Edaipur was 21SC and the rainfall in the past 2" hours was nil. )ince 1 Tune it experienced "3" mm of rainfall which is 2" mm below normal. 1anagarhBs maximum temperature in 2" hours was 29SC. It had " mm of rain in the past 2" hours and since 1 Tune it had <## mm of rain. Uol(ataBs maximum and minimum day temperatures were 3.SC and 2:SC respectively. It had 3 mm of rainfall in the past 2" hours. )ince 1 Tune it experienced a rainfall of 1.;< mm which is 1#" mm above normal. 1resent the above data in a tabular form. Solution< &he given information is summari4ed in &able 1.3;.

Table 1.*8 'eather Conditions at 0ifferent )tations 7at 9.3. a.m. on 2< 6ugust 1<<.8

45a.!le 1.19< 1resent the following data in a tabular form:

6 certain manufacturer produces three different products 1! 2! and 3. 1roduct 1 can be manufactured in one of the three plants 6! M! or C. 2owever! product 2 can be manufactured in either plant M or C! whereas plant 6 or M can manufacture product 3. 1lant 6 can manufacture in an hour 1. pieces of 1 or 2. pieces of 31, however! 2. pieces of 2! 1# pieces of 1! or 1: pieces of 3 can be manufactured per hour in plant M. 1lant C can produce 2. pieces of 1 or 19 pieces of 2 per hour. 'age rates per hour are /s. 2. at 6! /s. ". at M and /s. 2# at C. &he costs of running plants 6! M! and C are respectively /s. 1...! #..! and 12#. per hour. 3aterials and other costs directly related to the production of one piece of the product are respectively /s. 1. for 1! /s. 12 for 2! and /s. 1# for 3. &he company plans to mar(et product 1 at /s. 1# per piece! product 2 at /s. 19 per piece and product 3 at /s. 2. per piece. Solution< &he given information is summari4ed in &able 1.39.

Table 1.*9 1roduction )chedule of a 3anufacturer

45a.!le 1.1:< &ransforming the ratios into corresponding numbers prepare a complete table for the following information. $ive a suitable title to the table. In the year 2... the total strength of students of three colleges 1, 2, and : in a city were in the ratio " : 2 : #. &he strength of college 2 was 2.... &he proportion of girls and boys in all colleges was in the ratio 2 : 3. &he faculty+ wise distribution of boys and girls in the faculties of 6rts! )cience! and Commerce was in the ratio 1 : 2 : 2 in all the three colleges. Solution< &he data of the problem is summari4ed in &able 1.3<.

Table 1.*: 0istribution of )tudents 6ccording to aculty and Colleges in the Vear 2...

45a.!le 1.%;< /epresent the following information in a suitable tabular form with proper rulings and headings: &he annual report of a 1ublic Kibrary reveals the following information regarding the reading habits of its members. 5ut of the total of 3;19 boo(s issued to the members in the month of Tune! 21.. were fiction. &here were ":; members of the library during the period and they were classified into five classes%6! M! C! 0! and -. &he number of members belonging to the first four classes were respectively 1#! 1;:! <9! and 12<! and the number of fiction boo(s issued to them were 1.3! 119;! :";! and #9! respectively. &he number of boo(s! other than text boo(s and fiction! issued to these four classes of members were respectively "! 3<.! 21;! and 3"1. &ext boo(s were issued only to members belonging to classes C! 0! and -! and the number of text boo(s issued to them were respectively 9! 31;! and 1:.. 0uring the same period! 12": periodicals were issued. &hese include 3<: technical =ournals of which 3: were issued to members of class M! "# to class 0! and 31# to class -.

&o members of classes M! C! 0! and - the number of other =ournals issued were "1<! 2:! 231! and <<! respectively. &he report! however! showed an increase of ".1 per cent in the number of boo(s issued over last month! though there was a corresponding decrease of :.1 per cent in the number of periodicals and =ournals issued to members. Solution< &he data of the problem is summari4ed in &able 1."..

Table 1.-; /eading 2abits of the 3embers of a 1ublic Kibrary

Note9 &he figures for the month of 3ay were calculated on the basis of percentage changes for each type of reading material given in the text.

Self3Practice Proble.s 1B

1.1% 0raw a blan( table to show the number of candidates sex+wise appearing in the pre+university! first year! second year! and third year examinations of a university in the faculties of 6rts! )cience! and Commerce in a certain year.

1.1* Ket the national income of a country for the years 2...F.1 and 2..1F.2 at current prices be 9.!:#.! <.!.1.! and <.!#3. crore of rupees respectively! and per capita income for these years be 1.#.! 1.#:! and 1.:; rupees. &he corresponding figures of national income and per capita income at 1<<<F2... prices for the above years were 9.!:#.! 9.!92.! and 9.!9#. crore of rupees and 1.#.! 1.#1! and 1."9 respectively. 1resent this data in a table.

1.1- 1resent the following information in a suitable form supplying the figure not directly given. In 2.."! out of a total of "... wor(ers in a factory! 33.. were members of a trade union. &he number of women wor(ers employed was #.. out of which ".. did not belong to any union. In 2..3! the number of wor(ers in the union was 3"#. of which 32.. were men. &he number of nonunion wor(ers was ;:. of which 33. were women. 1.16 5f the 112# students studying in a college during a year! ;2. were )C/)&! :29 were boys! and "". were science students, the number of )C/)& boys was 3<2! that of boys studying science 2.#! and that of )C/)& students studying science 2:2, finally the number of science students among the )C/)& boys was 1"9. -nter these fre*uencies in a three+way table and complete the table by obtaining the fre*uencies of the remaining cells.

1.17 6 survey of 3;. students from the Commerce aculty and 13. students from the )cience aculty revealed that 19. students were studying for only C.6. -xaminations! 1". for only Costing -xaminations! and 9. for both C.6. and Costing -xaminations. &he rest had opted for part+time management courses. 5f those studying for

Costing only! 13 were girls and <. boys belonged to the Commerce aculty. 5ut of the 9. studying for both C.6. and Costing! ;2 were from the Commerce aculty amongst whom ;. were boys. 6mongst those who opted for part+time management courses! #. boys were from the )cience aculty and 3. boys and 1. girls from the Commerce aculty. In all! there were 11. boys in the )cience aculty. 1resent this information in a tabular form. ind the number of students from the )cience aculty studying for part+time management courses. 1.18 6n 6luminium Company is in possession of certain scrap materials with (nown chemical composition. )crap 1 contains :# per cent aluminium! 2. per cent iron! 2 per cent copper! 2 per cent manganese! 3 per cent magnesium and 9 per cent silicon. &he aluminium content of scrap 2! scrap 3! and scrap " are ;. per cent! 9. per cent and ;# per cent respectively. )crap 2 contains 1# per cent! iron! 3 per cent copper! 2 per cent manganese! " per cent magnesium! and the rest silicon. )crap 3 contains # per cent iron. &he iron content of )crap " is the same as that of scrap 3! scrap " contains twice as much percentage of copper as scrap 3. scrap 3 contains 1 per cent copper. )crap 3 contains manganese which is 3 times as much as copper it contains. &he percentage of magnesium and silicon in scrap 3 are 3 per cent and 9 per cent respectively. &he magnesium and silicon contents of scrap " are respectively 2 times and 3 times its manganese contents. &he company also purchases some aluminium and silicon as needed. &he aluminium purchased contains <: per cent pure aluminium! 2 per cent iron! 1 per cent copper and 1 per cent silicon respectively! whereas the purchased silicon contains <9 per cent silicon and 2 per cent iron respectively. 1resent the above data in a table.

1.19 &he N inancial 2ighlightsB of a public limited company in recent years were as follows: In the year ending on 31 3arch 1<<9 the turnover of the company! including other income! was /s. 1#; million. &he profit of the company in the same year before tax! investment allowance! reserve! and prior yearBs ad=ustment was /s. 1< million! and the profit after tax! investment allowance! reserve! and prior yearBs ad=ustment was /s. 9 million. &he dividend declared by the company in the same year was 2. per cent. &he turnover! including other income! for the years ending on 31 3arch 1<<<! 2...! and 2..1 were /s. 1:<! 1<1! and 1<; million respectively. or the year ending on 31 3arch 1<<< the profit before tax! investment allowance! reserve! and prior yearBs ad=ustment was /s. 1<2 million and the profit after tax! and so on /s. ;.# million! while the dividend declared for the same year was 1; per cent. or the year ending on 31 3arch 2...! 2..1! and 2..2 the profits before tax! investment allowance! reserve! and prior yearBs ad=ustment were /s. 21! 12! 13 million respectively! while the profits after tax! and so on! of the above three years were /s. <.#! "! and < million respectively. &he turnover! including other income! for the year ending on 31 3arch 2..2 was /s. 2"3 million. &he dividend declared for the year ending on 31 3arch 2...F.2 was 1; per cent! 1. per cent! and 2. per cent respectively. 1resent the above data in a table. 1.1: 1resent the following information in a suitable form: In 1<<"! out of a total of 1<#. wor(ers of a factory! 1".. were members of a trade union. &he number of women employed was ".. of which 2;# did not belong to a trade union. In 1<<<! the number of union wor(ers increased to 1;9. of which 1"<. were men. 5n the other hand! the number of non+union wor(ers fell to ".9 of which 29. were men. In the year 2.."! there were 2... employees who belonged to a trade union and 2#. did not belong to a trade union. 5f all the employees in 2...! #.. were women of whom only 2.9 did not belong to a trade union.

/ints and Ans=ers

1.1% 0istribution of candidates appearing in various university examinations

1.1* Gational income and per capita income of the country

or the year 1<<<F2... to 2..1F2..2

1.1- 3embers of union by sex

1.16 0istribution of College )tudents by Caste and aculty

1.17 0istribution of students according to aculty and 1rofessional Courses

1.18 &he chemical composition of )craps and 1urchased 3inerals

1.19 inancial highlights of the 1ublic Ktd.! Co.

1.1: &rade+union membership

1.- G&AP/ICAL P&4S4NTATION OF DATA


6ccording to Uing! W5ne of the chief aims of statistical science is to render the meaning of masses of figures clear and comprehensible at a glance.X &his is often best accomplished by presenting the data in a pictorial 7or graphical8 form. Functions of a Gra!" &he shape of the graph gives an exact idea of the variations of the distribution trends. $raphic presentation! therefore! serves as an easy techni*ue for *uic( and effective comparison between two or more fre*uency

distributions. 'hen the graph of one fre*uency distribution is superimposed on the other! the points of contrast regarding the type of distribution and the pattern of variation become *uite obvious. 6ll these advantages necessitate a clear understanding of the various forms of graphic representation of a fre*uency distribution. General &ules for Dra=in Dia ra.s &he following general guidelines are ta(en into consideration while preparing diagrams: Title9 -ach diagram should have a suitable title. It may be given either at the top of the diagram or below it. &he title must convey the main theme which the diagram intends to portray. (i:e9 &he si4e and portion of each component of a diagram should be such that all the relevant characteristics of the data are properly displayed and can be easily understood. ro#ortion of len%th and breadth9 6n appropriate proportion between the length and breadth of the diagram should be maintained. 6s such there are no fixed rules about the ratio of length to width. 2owever! a ratio of : 1 or 1."1" 7long side8 : 1 7short side8 suggested by Kut4 in his boo( ;rap4ic <resentation may be adopted as a general rule. ro#er scale9 &here are again no fixed rules for selection of scale. &he diagram should neither be too small nor too large. &he scale for the diagram should be decided after ta(ing into consideration the magnitude of data and the si4e of the paper on which it is to be drawn. &he scale showing the values as far as possible should be in even numbers or in multiples of #! 1.! 2.! and so on. &he scale should specify the si4e of the unit and the nature of data it represents! for example! Nmillions of tonnesB! in /s. thousand! and the li(e. &he scale adopted should be indicated on both vertical and hori4ontal axes if different scales are used. 5therwise! it can be indicated at some suitable place on the graph paper. Footnotes and source note9 &o clarify or elucidate any points which need further explanation but cannot be shown in the graph! footnotes are given at the bottom of the diagrams. "nde;9 6 brief index explaining the different types of lines! shades! designs! or colours used in the construction of the diagram should be given to understand its contents. (im#licity9 0iagrams should be prepared in such a way that they can be understood easily. &o (eep it simple! too much information should not be loaded in a single diagram as it may create confusion. &hus if the data are large! then it is advisible to prepare more than one diagram! each depicting some identified characteristic of the same data.

1.6 T)P4S OF DIAG&A2S


One3Di.ensional Dia ra.s &hese diagrams provide a useful and *uic( understanding of the s4ape of the distribution and its characteristics. &hese diagrams are called one+dimensional diagrams because only the length 7height8 of the bar 7not the width8 is ta(en into consideration. 5f course! width or thic(ness of the bar has no effect on the diagram! even then the thic(ness should not be too much otherwise the diagram would appear li(e a twodimensional diagram. &he one+dimensional diagrams 7charts8 used for graphical presentation of data sets are as follows:

2istograms re*uency polygons

1ie diagrams re*uency curves Cumulative fre*uency distributions 75give8

0Bar Dia ra.s1 /isto ra.s &hese diagrams are used to graph both ungrouped and grouped data. In the case of an ungrouped data! values of the variable 7the characteristic to be measured8 are scaled along the hori4ontal axis and the number of observations 7or fre*uencies8 along the vertical axis of the graph. &he plotted points are then connected by straight lines to enhance the shape of the distribution. &he height of such boxes 7rectangles8 measures the number of observations in each of the classes 7 ig. 1.18. &e.ar>s< Mar diagrams are not suitable to represent long period time series.

Fi ure 1.1 2istogram for 3utual unds (im#le <ar !harts &he graphic techni*ues described earlier are used for group fre*uency distributions. &he graphic techni*ues presented in this section can also be used for displaying values of categorical variables. )uch data are first tallied into summary tables and then graphically displayed as either bar c4arts or pie c4arts. Mar charts are used to represent only one characteristic of data and there will be as many bars as number of observations. or example! the data obtained on the production of oil seeds in a particular year can be represented by such bars. -ach bar would represent the yield of a particular oil seed in that year. )ince the bars are of the same width and only the length varies! the relationship among them can be easily established. )ometimes only lines are drawn for comparison of given variable values. )uch lines are not thic( and their number is sufficiently large. &he different measurements to be shown should not have too much difference! so that the lines may not show too much dissimilarity in their heights. 45a.!le 1.%1< &he data on the production of oil seeds in a particular year are presented in &able 1."1.

Table 1.-1

Oil (eed

&ield ercenta%e roduction ('illion tonnes) ('illion tonnes) #.9. "3..3 3.3. 2"."9 9.;# 1:.32 ;."2

$round nut /apeseed 1.19 Coconut 2.2. Cotton 1... )oyabean

13."9

1.....

/epresent this data by a suitable bar chart. Solution< &he information provided in &able 1."3 is expressed graphically as a fre*uency bar chart as shown in ig. 1.2. In this figure! each type of seed is depicted by a bar! the length of which represents the fre*uency 7or percentage8 of observations falling into that category.

Fi ure 1.% Mar Chart 1ertaining to 1roduction of 5il )eeds &e.ar>< &he bars should be constructed vertically 7as shown in ig. 1.28 when categori4ed observations are the outcome of a numerical variable. Mut if observations are the outcome of a categorical variable! then the bars should be constructed hori4ontally. 45a.!le 1.%%< 6n advertising company (ept an account of response letters received each day over a period of #. days. &he observations were as follows:

Construct a fre*uency table and draw a line chart 7or diagram8 to present the data. Solution< igure 1.3 depicts a fre*uency bar chart for the number of letters received during a period of #. days presented in &able 1."2.

Table 1.-% re*uency 0istribution of Ketters /eceived

Number of .etters *eceived Tally Number of Days . 1 2 3 " # % (Frequency) 23 1; ; 2 . 1 #.

Fi ure 1.* Gumber of Ketters /eceived

'ulti#le <ar !harts 6 multiple bar chart is also (nown as grouped 7or compound8 bar chart. )uch charts are useful for direct comparison between two or more sets of data. &he techni*ue of drawing such a chart is same as that of a single bar chart with a difference that each set of data is represented in different shades or colours on the same scale. 6n index explaining shades or colours must be given. 45a.!le 1.%*< &he data on fund flow 7in /s. crore8 of an International 6irport 6uthority during financial years 2..1F.2 to 2..3F." are given below:

/epresent this data by a suitable bar chart. Solution< &he multiple bar chart of the given data is shown in ig. 1.".

Fi ure 1.- 3ultiple Mar Chart 1ertaining to 1erformance of an International 6irport 6uthority Deviation <ar !harts 0eviation bar charts are suitable for presentation of net *uantities in excess or deficit such as profit! loss! import! or exports. &he excess 7or positive8 values and deficit 7or negative8 values are shown above and below the base line. 45a.!le 1.%-< &he following are the figures of sales and net profits of a company over the last 3 years.

71er cent change over previous year8

&ear 2..2F.3 2..3F." 2.."F.#

(ales =rowth Net rofit 1# 3. 12 #3 19 D;2

1resent this data by a suitable bar chart. Solution< ig. 1.# depicts deviation bar charts for sales and per cent change in sales over previous yearBs data.

Fi ure 1.6 0eviation Mar Chart 1ertaining to )ales and 1rofits (ubdivided <ar !harts )ubdivided bar charts are suitable for expressing information in terms of ratios or percentages. 'hile constructing these charts the various components in each bar should be in the same order to avoid confusion. 0ifferent shades must be used to represent various ratio values but the shade of each component should remain the same in all the other bars. 6n index of the shades should be given with the diagram. 45a.!le 1.%6< &he data on sales 7/s. in million8 of a company are given below:

Solution< ig. 1.: depicts a subdivided bar chart for the given data.

Fi ure 1.7 )ubdivided Mar Chart 1ertaining to )ales ercenta%e <ar !harts 'hen the relative proportions of components of a bar are more important than their absolute values! then each bar can be constructed with same si4e to represent 1..Y. &he component values are then expressed in terms of percentage of the total to obtain the necessary length for each of these in the full length of the bars. &he other rules regarding the shades! index! and thic(ness are the same as mentioned earlier. 45a.!le 1.%7< &he following table shows the data on cost! profit! or loss per unit of a good produced by a company during the year 2..3F.".

/epresent diagrammatically the data given above on percentage basis. Solution< &he cost! sales! and profit/loss data expressed in terms of percentages have been represented in the bar chart as shown in ig. 1.;.

Fi ure 1.8 1ercentage Mar Chart 1ertaining to Cost! )ales! and 1rofit/Koss 2eights of the 1olygon at each 3id+ point Fre+uenc, Pol, ons 6 fre*uency polygon is formed by mar(ing the mid+point at the top of hori4ontal bars and then =oining these dots by a series of straight lines. &he fre*uency polygons are formed as a closed figure with the hori4ontal axis! therefore a series of straight lines are drawn from the mid+point of the top base of the first and the last rectangles to the mid+point falling on the hori4ontal axis of the next outlaying interval with 4ero fre*uency. 6 fre*uency polygon can also be converted bac( into a histogram by drawing vertical lines from the bounds of the classes shown on the hori4ontal axis! and then connecting them with hori4ontal lines at the hieghts of the polygon at each mid+point. ig. 1.9 shows the fre*uency polygon for the fre*uency distribution presented by histogram in ig. 1.1.

Fi ure 1.9 re*uency 1olygon for 3utual und Fre+uenc, Cur$e It is described as a smooth fre*uency polygon as shown in ig. 1.<. 6 fre*uency curve is described in terms of its 7i8 symmetry 7s(ewness8 and 7ii8 degree of pea(edness 7(urtosis8.

Fi ure 1.: re*uency Curve &wo fre*uency distributions can also be compared by superimposing two or more fre*uency curves provided the width of their class intervals and the total number of fre*uencies are e*ual for the given distributions. -ven if the distributions to be compared differ in terms of total fre*uencies! they still can be compared by drawing per cent fre*uency curves where the vertical axis measures the per cent class fre*uencies and not the absolute fre*uencies. Cu.ulati$e Fre+uenc, Distribution 0O i$e1 It enables us to see how many observations lie above or below certain values rather than merely recording the number of observations within intervals 7see &able 1."3 and ig. 1.1.8.

Table 1.-* Calculation of Cumulative re*uencies

Fi ure 1.1; 5give for 3utual unds 1rices Pie Dia ra.s &hese diagrams are normally used to show the total number of observations of different types in the data set on a percentage basis rather than on an absolute basis through a circle. Esually the largest percentage portion of data in a pie diagram is shown first at 12 oBcloc( position on the circle! whereas the other observations 7in per cent8 are shown in cloc(wise succession in descending order of magnitude. &he steps to draw a pie diagram are summari4ed below: 1. 2. 3. Convert the various observations 7in per cent8 in the data set into corresponding degrees in the circle by multiplying each by 3.: 73:.Y1..8. 0raw a circle of appropriate si4e with a compass. 0raw points on the circle according to the si4e of each portion of the data with the help of a protractor and =oin each of these points to the center of the circle.

45a.!le 1.%8< &he data show mar(et share 7in per cent8 by revenue of the following companies in a particular year:

0raw a pie diagram for the above data. Solution< Converting percentage figures into angle outlay by multiplying each of them by 3.: as shown in &able 1."".

Table 1.--

!om#any

'arket (hare 0 ercent1 $n%le Outlay (De%ree)

3. Matata+M1K 2: 2utchison+-ssar 1< Mharti+)ing &el 12 3odi 0ista Com # -scorts irst 1acific 3 /eliance 2 /1$ 2 )rinivas 1 )hyam 1.. &otal

1.9.. <3.: :9." "3.2 19.. 1..9 ;.2 ;.2 3.: 3:...

Esing the data given in &able 1."" construct a pie chart as shown in ig. 1.11 by dividing the circle into < parts according to degrees of angle at the centre.

Fi ure 1.11 1ercentage 1ie Chart 45a.!le 1.%9< &he following data relate to area in millions of s*uare (ilometer of oceans of the world.

Ocean 1acific

$rea ('illion sq km) ;..9 "1.2

6tlantic 29.# Indian ;.: 6ntarctic ".9 6rctic


Solution< Converting given areas into angle outlay as shown in &able 1."#.

Table 1.-6

Ocean $rea ('illion sq+ km+) $n%le Outlay (De%rees) ;..9 1acific "1.2 6tlantic 29.# Indian ;.: 6ntarctic ".9 6rctic 1#2.< &otal
1ie diagram is shown in ig. 1.12.

:;.1. 1;.9< 11.31 3:....

Fi ure 1.1% 1er cent 1ie 0iagram T=o3Di.ensional Dia ra.s In one+dimensional diagrams or charts! only the length of the bar is ta(en into consideration. Mut in twodimensional diagrams! both its height and width are ta(en into account for presenting the data. &hese diagrams! also (nown as surface dia0rams or area dia0rams, are categori4ed as following: &ectan les )ince area of a rectangle is e*ual to the product of its length and width! while ma(ing such type of diagrams both length and width are considered. /ectangles are suitable for use in cases where two or more *uantities are to be compared and each *uantity is sub+divided into several components. 45a.!le 1.%:< &he following data represent the income of two families 6 and M. Construct a rectangular diagram.

"tem of >;#enditure ood

Family $ Family < ('onthly "ncome *s+ ?5)555) ('onthly "ncome *s+ 65)555) ###. ;29. #1.. :99.

Clothing

"9.. 2ouse rent ";". uel and light "<#. -ducation "9:. 3iscellaneous 3.!... &otal

:"9. :32. ::". :".. ".!...

Solution< Converting individual values into percentages ta(ing total income as e*ual to 1.. as shown in &able 1.":.

Table 1.-7 1ercentage )ummary &able 1ertaining to -xpenses Incurred by &wo amilies

&he height of the rectangles shown in ig. 1.13 is e*ual to 1... &he difference in the total income is represented by the difference on the base line which is in the ratio of 3 : ". S+uares &o construct a s*uare diagram! first the s*uare+root of the values of various figures to be represented is ta(en and then these values are divided either by the lowest figure or by some other common figure to obtain proportions of the sides of the s*uares. &he s*uares constructed on these proportionate lengths must have either the base or the center on a straight line. &he scale is attached with the diagram to show the variable value represented by one s*uare unit area of the s*uares.

Fi ure 1.1* 1ercentage of -xpenditure by &wo amilies 45a.!le 1.*;: &he following data represent the production 7in million tonnes8 of coal by different countries in a particular year.

!ountry roduction 13..1 E)6 "".. E))/ 1:." EU 3.3 India


/epresent the data graphically by constructing a suitable diagram.

Solution< &he given data can be represented graphically by s*uare diagrams. or constructing the sides of the s*uares! the necessary calculations are shown in &able 1.";.

Table 1.-8 )ide of a )*uare 1ertaining to 1roduction of Coal

&he s*uares representing the amount of coal production by various countries are shown in ig. 1.1".

Fi ure 1.1- Coal 1roduction in 0ifferent Countries Circles Circles are alternatives to s*uares to represent data graphically. &he circles are also drawn such that their areas are in proportion to the figures represented by them. &he circles are constructed in such a way that their centers lie on the same hori4ontal line and the distance between the circles is e*ual. )ince the area of a circle is directly proportional to the s*uare of its radius! the radii of the circles are obtained in proportion to the s*uare root of the figures under representation. &hus! the lengths that were used as the sides of the s*uare can also be used as the radii of circles. 45a.!le 1.*1< &he following data represent the land area in different countries. /epresent this data graphically using suitable diagram.

!ountry .and $rea (crore acres) #<.." E))/ 32..# China 1<..# E)6 91.3 India
Solution< &he data can be represented graphically using circles. &he calculations for constructing radii of circles are shown in &able 1."9.

Table 1.-9 /adii of Circles 1ertaining to Kand 6rea of Countries

&he various circles representing the land area of respective countries are shown in ig. 1.1#.

Fi ure 1.16 Kand 6rea of 0ifferent Countries Picto ra.s or Ideo ra!"s 6 pictogram is another form of pictorial bar chart. )uch charts are useful in presenting data to people who cannot understand charts. )mall symbols or simplified pictures are used to represent the si4e of the data. 45a.!le 1.*%< 3a(e a pictographic presentation of the output of vans during the year by a van manufacturing company.

Solution< 0ividing the van output figures by 1...! we get 2..."! 2.<<:! ".21<! and #.32" respectively. /epresenting these figures by pictures of vans as shown in ig. 1.1:.

Fi ure 1.17 5utput of Rans

1.7 4?PLO&ATO&) DATA ANAL)SIS


In this section one of the useful techni*ues of exploratory data analysis! stem-and-leaf displays .or dia0rams/ techni*ue! is presented. &his techni*ue provides the rank order of the values in the data set and the shape of the distribution. Ste.3and3Leaf Dis!la,s It is a graphical display of the numerical values in the data set and separates these values into leadin0 di0its .or stem/ and trailin0 di0its .or lea!es/. &he steps re*uired to construct a stem+and+leaf diagram are as follows: 1. 0ivide each numerical value between the ones and the tens place. &he number to the left is the stem and the number to the right is the leaf. &he stem contains all but the last of the displayed digits of a numerical value. 6s with histograms! it is reasonable to have between : to 1# stems 7each stem defines an interval of values8. &he stem should define e*ually spaced intervals. )tems are located along the vertical axis. )ometimes numerical values in the data set are truncated or rounded off. or example! the number 1#.:< is truncated to 1#.: but it is rounded off to 1#.;. 2. 3. ". Kist the stems in a column with a vertical line to their right. or each numerical value! attach a leaf to the appropriate stem in the same row 7hori4ontal axis8. 6 leaf is the last of the displayed digits of a number. It is standard! but not mandatory! to put the leaves in increasing order at each stem value. 1rovide a (ey to stem and leaf coding so that actual numerical value can be re+created! if necessary.

&e.ar>< If all the numerical values are three+digit integers! then to form a stem+and+leaf diagram! two approaches are followed:

1. 2.

Ese the hundreds column as the stems and the tens column as the leaves and ignore the units column. Ese the hundreds column as the stems and the tens column as the leaves after rounding of the units column.

45a.!le 1.**< Consider the following mar(s obtained by 2. students in a business statistics test:

1. 2. 3.

Construct a stem+and+leaf diagram for these mar(s to assess class performance. 0escribe the shape of this data set. 6re there any outliers in this data set.

Solution< 1. &he numerical values in the given data set are ranging from #" to <3. &o construct a stem+and+leaf diagram! we ma(e a vertical list of the stems 7the first digit of each numerical value8 as shown below:

Ste. # :

Leaf ": "3123

; 9"293 9 <;919:3 < 3


2. /earrange all of the leaves in each row in ran( order.

Ste. # :

Leaf ": 1233"

; 23"99 9 13:;99< < 3


3. ". -ach row in the diagram is a stem and numerical value on that stem is a leaf. or example! if we ta(e the row :/1233"! it means there are five numerical values in the data set that begins with :! that is! :1! :2! :3! :3! and :". If the page is turned <.S cloc(wise and rectangles are drawn arround the digits in each stem! we get a diagram similar to a histogram.

#.

:. ;.

)hape of the diagram is not symmetrical. &here is no outlier 7an observation far from the center of the distribution8.

45a.!le 1.*-< &he following data represent the annual family expenses 7in thousand of rupees8 on food items in a city.

Construct a stem+and+leaf diagram. Solution< )ince the annual costs 7in /s. B...8 in the data set all have two+digit integer numbers! the tens and units columns would be the leading digits and the remaining column 7the tenth column8 would be the trailing digits as shown below:

Ste. 12 9 13 9 1"

Leaf

1;<<<<99:211"2:29 1# 2:2<23# 1: ; 1; 2 19 ... 1< 21 Ste. 12 9 13 9 1" 111222"::;999<<<< 1# 2223#:< 1: ; 1; 2 Leaf

19 ... 1< 12
/earrange all the leaves in each row in the ran( order as shown above.

Self3Practice Proble.s 1C

1.%; &he following data represent the gross income! expenditure 7in /s. la(h8! and net profit 7in /s. la(h8 during the years 1<<<F2..2.

1.%1 'hich of the charts would you prefer to represent the following data pertaining to the monthly income of two families and the expenditure incurred by them.

>;#enditure on ood

Family $ Family < ("ncome *s+ 0@)555) ("ncome *s+ 05)555) "... #".. 29.. 3:.. 3#.. 29.. #... D3..

Clothing 3... 2ouse rent 23.. -ducation 3... 3iscellaneous C 1<.. )aving or deficits
1.%% &he following data represent the outlays 7/s. crore8 by heads of development.

Heads of Develo#ment 6griculture Irrigation and flood control -nergy Industry and minerals

!enter (tates ";:# ::3# <<<# ;.3< 11!3<# 92<3

12!;;. 2<9#

&ransport and communication 12!2.. #12. )ocial services &otal 921: 1"2.

#"!#91 3:!2#2

/epresent the data by a suitable diagram and write a report on the data bringing out the salient features. 1.%* 3a(e a diagrammatic representation of the following textile production and imports.

-alue .en%th (in !rore) (in Hundred &ards) 11:." "2:.< 3ill production 2andloom 1.:.9 production 31<.; Imports
'hat conclusions do you draw from the diagramH 1.%- &he following data represent the estimated gross area under different cereal crops during a particular year.

1<2.9 :".;

0raw a suitable chart to represent the data. 1.%6 &he following data indicate the rupee sales 7in B...8 of three products according to region.

1. 2. 3.

Esing vertical bars! construct a bar chart depicting total sales region+wise. Construct a component chart to illustrate the product brea(down of sales region+wise by hori4ontal bars. Construct a pie chart illustrating total sales.

1.%7 &he following data represent the income and dividend for the year 2....

&ear "ncome er Dividend er (hare (in *s+) (hare (in *s+) 1<<# #.9< 3.2. 1<<: :."< 3.:. 1<<; ;.3. 3.9# 1<<9 ;.;# 3.<#

1<<< 9.3: 2... <...


1. 2. 3.

3.2# "."#

Construct a line graph that indicates the income per share for the period 1<<#F2.... Construct a component bar chart that depicts dividends per share and retained earning per share for the period 1<<#F2.... Construct a percentage pie chart depicting the percentage of income paid as dividend. 6lso construct a similar percentage pie chart for the period 1<<9F2.... 5bserve any difference between the two pie charts.

1.%8 &he following time series data ta(en from the annual report of a company represents per+share net income! dividend! and retained earning during the period 1<<:F2....

1. 2. 3.

Construct a bar chart for per+share income for the company during 1<<:F2.... Construct a component bar chart depicting the allocation of annual earnings for the company during 1<<:F2.... Construct a line graph for the per+share net income for the period 1<<:F2....

1.%9 ind a business or economic related data set of interest to you. &he data set should be made up of at least 1.. *uantitative observations. 1. 2. )how the data in the form of a standard fre*uency distribution. Esing the information obtained from 7a8 briefly describe the appearance of your data.

1.%: &he first row of a stem+and+leaf diagram appears as follows: 2:/1""9<. 6ssume whole number values. 1. 2. 3. 'hat is the possible range of values in this rowH 2ow many data values are in this rowH Kist the actual values in this row of data.

1.*; $iven the following stem+and+leaf display representing the amount of CG$ purchased in litres 7with leaves in tenths litre8 for a sample of 2# vehicles in 0elhi.

< ;1" 1. 9223. 11 #:1;;:;3# 12 3<"292 13 2.


1. 2. /earrange the leaves and form the revised stem+and+leaf display. 1lace the data into an ordered array.

1.*1 &he following stem+and+leaf display shows the number of units produced per day of in item an a factory.

3 9

" % # : : .133##< ; .23:;;9 9 #< < ..1#: 1. 3:


1. 2. 3. ". #. 2ow many days were studiedH 'hat are the smallest and the largest valuesH Kist the actual values in second and fourth row. 2ow many values are 9. or moreH 'hat is the middle valueH

1.*% 6 survey of the number of customers that used 1C5/ )&0 both located at a college gate to ma(e telephone calls last wee( revealed the following information

1. 2. 3.

0evelop a stem+and+leaf display. 2ow many calls did a typical customer madeH 'hat were the largest and the smallest number of calls madeH

/ints and Ans=ers

1.%8 1. 2. 3. 1.*; 1. 2:. to 2:< # 2:1! 2:"! 2:"! 2:9! 2:<

< 1. 11

1"; .2239 13##::;;;

12 13
2. 1.*1 1. 2. 3. ". #. 1.*% 1.

223"9< .2

<1 <" <; 1.. 1.2 1.2 1.3 1.9 111 11311#11#11:11:11;11;122 122 123 12" 129 12< 13. 132

2# 39! 1.: Go values! :.! :1! :3! :3! :#! :#! :< < ;:

. 1 2 3 " #
2. 3.

# 29 F ..2";9< 123:: 2

1: customers were studied. Gumber of customers visited ranged from # to #2.

S-ar putea să vă placă și