Measure of Central 7 Value INTRODUCTION AVERAGE DEFINED OBJECTIVES OF AVERAGING REQUISITES OF A GOOD AVERAGE TYPES OF AVERAGES GEOMETRIC MEAN HARMONIC MEAN RELATIONSHIP AMONG THE AVERAGES MISCELLANEOUS ILLUSTRATIONS WHICH AVERAGE TO USE? ooo CT HTT eH He GENERAL LIMITATIONS OF AVERAGETRODUCTION . INTRODY wee of statistical anatyes nto Comparison Measiites of renteal vale. try eng hjeeth Is é defe ¥ teu, doe of the moet important TENT acteristic of the yy! 4, a1 ane singe gure. enable comparison ar eine the One of the 4 me ibes: tHe » eoritral vei 4 7 0318 he made either ata point af a ye mare ingle value that desert < called the cent Cop can ime or ever a nd ur ted value of the Va oy example, we often CUT ny Ber, exe jy a certain examination say py as for 2005, and. tae Pe day-terday carersallon. VT oF an Indian, aver Ik yy I int (Ge which college Is the best or we. ea and used in day:to-da eM nau in compare the pass conf the same college for different time perioce we therety 4 pi@el, to whether the results are improving of deteriora Mg. Such oy ina class When we sa¥ neither very good ‘wha h e student’ ‘what it meang” Me, age jst a mediocre type of id i Mt any bad. US nor very bad. Ju gifferent. meaning a ff immense help in framing suitabs by fi hye ude Se are of uitable and timely polict ‘etice the term average 6 raeone “if the, pass percentage of students in College Ate, Rise in sta avERAGE DEFINED BiamlMGqoa and 75 In 2005. the authorities have sufficient veseo fy 4 diterenly by variou, Gp in 20 re possible cause of the deterioration in resulta 1d ‘average’ has, een, deine 8 hen Bouin while making comparisons ‘one should also take into The word. 0 below & Geet. it fo ha be ia nant definitions are give fn Foor ion the multiplicity of forces that might affecting the data find one single figure to describe whole ot py Fierallel if per caplta Income is rising in absolu ny GF camtanather. it should not lead one to think th: at the stand; a selected from 8 GtOUP Of valugg . ~Y, 0 ly improving because the prices might be qndad of stipe nals ee os re 8 og Mo pecessat e rising faster yalue wt le 0 is r capita income and so in real terms people d eT pe people might be seme as typical of all he values in the group, ee oats ne Toreover. the same measure should be used tor making which iti ‘typical value in the sonse that i is someti + "An average is mes amt Bie O% Thetween two OF more groups. For example we should not TT the indwvidual values in a series oF of a variable.” ong) atte mean wage of one factory with the median wage of another ~ pare I swing any inference about wage levels, ed As a nuMbEr Which is typicg, eo, iy for ; So it OF A GOOD AVERAGE » ‘average is an attempt 10 te terms from one + "An avorage iS & thom in some way—a V the average is sometimes describe gow" within the range of the di hey 4 is a single value lata that i + "An average vite values in the series. Since an erate IS somentertet, an average 1S a single value representing a group of values. it i the range ofthe data rs ls called a measure of conta qakes TE go fhat such a valve salisfies the following propertes 4 —Croxton ce gasy 10 Understand Since statistical methods are designed to It is clear from the above, definitions that an average is Aecomplexity. it is desirable that an average be such thar can U2 tis a i i y that represents a group of a Pe 2 value oot wgeat ae iy ‘understood: otherwise, its use is bound to be very limited because it deplete Oe ta its value lles somewhere In beet ala” sanle to Compute An average should not only be easy to understand represents the Ole Gest and the smallest items. Foc tre" ty Osea te to compute so that it con be escn widely. However, though extremes, Le; the largest Tht oa measure of central tendepctta8" computation Is desirable, it should not ne sought at the ext f average is frequently refer Magy, "Ge “vantages. Le. If in the interest of greater accuracy. use ofa mare OBJECTIVES OF AVERAGING Malt average is desirable, one should prefer that ) Based on all the Items. | The average should depend upon each item ot ae series = het tf pay of the items 18 dropped tie aie it ic is altered. For example, the arithmetic mean of 10, 20.30, 40, : (| To get single value that describes the garectercsns of the ent». 30 +40 + 50 150 30, 40, 50. is. Measures of central value, by condensing the mass of data in ome uc, enable us to get a bird's-eye view of the entire data Thy \ha youla be = 25 value can represent thousands, lakhs and even millions of yaiye example, it is impossible to remember the individual incomes of sn.) Not be unduly Affected by of carning, people of India and even if one could do it there is hagh wy ltem should influence the value of ake average, none of the items use. But if the average income is obtained by dividing the total id influence it unduly. If one or two very small or very large Items Income by total population we get one single value that reprean ably allect the average. ie. either Incrente ite value oy cee entire population. Such a figure would throw Nght on the ‘stanjillt. the average cannot be really typical of the entire series. In other living of an average Indian. Beh, extremes may distort the average and reduce its useflulnees There are two main objectives of the study of averages; = 0. If'we drop one item, say, 50, the arithmetic ‘Yuleand Kendall: An introduction to the Theory of Statistics,STA ae Ic iN! 183 a age sould be properly average Sy yt should pre deny waned AY og potato Preferay ed | Ww pail Py one ingore people comnts the 2 1 ye nod op tnat ame answer ( ting arity ag tally, ett ea the personal rej Re ng? ye lt computing mean Is et > of x= UX N Feely get ormcy all fet popend upon ame gure Hg not dep an be mislea \ The avert snot se ihe results © ading, “°S ay) Me f i nvestiator eatment We bi ae ee of PE, ge fae H atitical Bre 4 x = The variable in (oy Capa at could user ample. A we aft given tne lay? in question; N* = Total number on averade “nance Far er employees of two or po tay Hon, “ its willy ne am aun mbined average: More ft ay arenes er te ot scan ne frequency of each row with the variable and obtain th should be ° ut not the least, we sh 4! " . ‘ ul f ing Seal, “gasses call. ‘samplin, ld | ‘¢ total obtained by step (0 by cig samo Spat tne SMecups of college aninlentes uy ete [P pie yal frequency: pil. by the number off ebsevatians, sf ram tne folowing data ofthe marks obtained by 60 students of a cl class, ich as 970" Io asres Fie expect (0 GEL aprox cont No 2 nS gna a: vate, q tet ge of eh 08 9, Have tnt ie 6 Beno i ae “There may be 50! ren fe \ Gs of ater oe jeannialy oa Ee Tucnet Soe Ve ins Nowe Stents Marks No.of Std o mn wd rior tan 03 oe erence ts more) e 8 50 o 4 12 60 craves oF AVERA averas 4 2 % 70 i types of averages: wing are the importa . vs bo denoted by X and + Median . sven | Geometric mean wage No. of students Ferman MA. important a? * ides these. tere are less impo) averages lil 8 Beep average ‘rhese averages have a Coe MoMing 2 2 pateaton and essfore, not s0 Popular. Uted 40 20 ° ‘Arithmetic Mean 5 10 widely used measure of ¥ # & representing yy) 70 j ‘the most popular ar “na by one value 1s what ‘most laymen call an ‘average’ ¢ arithmetic. mean Its value ts rage’ and yy by dividing this total by the oe it ber “y ‘athmetie mean may either be arithmetic mean, OF perce the average marks = 41 (y simple fi) weighted arithmetic mean: ori sot-cut Method Acc ' Tend be raed thatthe siatsuclans do mot Mike the term ' rig this peed a Gilerent meanings; for example 2 ¥ Kany ne eee hias too loose a connotation. ae ee an average wage, on average height, el c , etc. It can refer F 10 ely vsian, mode, geomeiric mean, harmon , \ Mic mean or a rN mean is 30 popular that the word ' any other average hp - {qualification is implied (o denote this parti ine te ota trate setae ere =e serv walla of tie mea or the average cular type of average, Mal eas ' a oar number of observ Jeries of observations, lm’ Te reader chould note carefull “ene fully that in discrete te and continuous fr 's frequency distributions the rule, be assuined that the at hal ihe arithmetic mean is meant unless otherwise shud bl umber of auniber of observations, Le. N= the sum of ft frequency or N= Df.1s Steps: tof each class and denote It by m. @ Take an assumed mean -poin ne MAP paints by the respective frequency of each class (a) Take the 3 ns ty a variable X fom the aun s “pl hese total 2m. denote the deviat i - m. a (ua manta ne devas wih IME TESPEGLVE TreQU Ney Sy ye x oan we obtained In step 0) by the sum of the frequency, fe in \ Pom total Lfd. " a (w) Divide the total obtained in third step by the total fre, tay pw, hy ‘Que, pie he following data compute arithmetic mean by direct method: t! nog From nustration 3, Caloulate arihmetic, mean by Ma SRO Hig, joo cio 1020-2080 30-9 4085050460 sinbution of ustation 2 hg 6 10 8 30 20 10 ‘Solution. ‘ Y et CALCULATION OF ARITHMETIC MEAN Me ae No, of Studen's Km ; on CALCULATION OF ARITHMETIC MEAN BY DIRECT METHOD yo a a — =20 e Midpoint Na, of Students 30 12 10 was f im 40 20 0 He — 5 5 25 Fal i 50 "0 ° v 18 10 150 60 so-20 70 x 25 25 625 Dy. i 0-80 45 20 50-80 55 10 Calculation of oO N= 100 in continuous series, arithmetic mean may be computed by of the following methods: APD fy (Diet aethoe od When short-cut_method d, arithmetis i . . ext Meth en short-cut method is used, arithmetig mean is (1) Short-cut method. ae De appving te following formula thod is used “ = Direct Method When direct met is XaAt afd xan ypere A= assumed mean; d = ‘deviations of mid-points from assumed mean, N mera A), N = total number of observations. where m = mid-point of various classes; f= th Steps: N= the total frequency. F= the frequency of eachy, Take an assumed mean. i) From the mid-point of each class deduct the assumed mean. Note, For the sake of clarity and ease of understanding we have fi) Multiply the respective frequencies of each class by these deviations Inc se cake of eu, and ease of undertanding. we ave wl te (Ty Gotain the a weve the reser soul be iam wy thea Ow Ga iso, Ungrouped data refer to the individual observatibns wherens gogeic (iv) Apply the formula: Keay 8 refer to continuous series and the discrete series. N Caloulate arithmetic mean by the short-cut method from the data of point = Lomer Limit + Upp "Mid ~ point = Lower - ‘Upper }irnit Ilustration 4,- ww pove that all the three methods of finding arithmetic Ise z rom es give us the same answer. The direct method, on. on OF ARITHMETIC MEAN \eP eet, nue Saves more calculations when mid-points and solution: cacu 15) cMeiPles Terge in magnitude, For example, observe the if re ve 3 Sat * 5 No. of Persons Income in Rs. No. of Persons ie 368 700-800 587 6 : 500 10-20 6 mi ° ty go 472 800-900 304 00 a = ao 908 0 30-40 6 ; or? 0-50 , jation method would be far simpler. In fact, step : ase step GOVT adopted wherever possible because It 0-60 __ yw ce nod shoul met tions. 0". the calculatione continuous series the mid-points of the .— 200 _35-2= a es H mee al #95 799“ $5-2= 38 pele computing Fien as representative of that particular class. The a $ aren the data are grouped, the exact frequency with 08's that the variable occurs in the distribution is unknown, ons, we can divi ‘ to simply the cae alii and then aa dey peach AME Mlimits within which a certain number of frequencies in order 1 le, calculate (ula becomes: ly yin, ayy Knew hen we say that the number of persons within class inten@™ ng mean The Uy) we por exami 400-500 is 50 we cannot say as to how many formula for 8 xear xt PM ipeome B55 are getting 401, 402, 403, etc. We, therefore, make cA wars out ey wile caleulating arithmetic mean that the frequencies sumption WI, spread evenly over the range of the class interval, nay be pointed out that whe? mon factor are neauay i cach Gi'be as-many items below the mid-point as above it, anit calito, a tm — AV/t while making och ce 4a et hn assumption is made the value of mean cannot be " lc ie should 2 ifustation 4, compute arithmetic yaton, MBpite aouon is likely to lead to some error. AS a result thereof the From th nod. yg "mis aes P umber of observations calculated from a _freqnency deviation 8 at ot will generally be only an approximation to the mean uorfrom the original data. However, the possibility of compensating Solution CALCULATION OF ARITHMETIC MEAN sopultgt be considered. Some of the mid-points err by being too low by being too high. In general, then mid-points of the Shes - ors err Traore Woo Suen (= 95) Sap a alow the class containing the arithmetic mean tend to be too low ars Types Peiapoints of the classes above the class containing the the — 5 “30 Myneie mean tend to be too high. It is quite possible, therefore, that 0-10 “By ei ihe errors are assumed, those which are (oo low will offset, in part 40-20 15 10 20 2 ay less those which are too high, so that the arithmetic mean for the - 7 8 -10 “f ‘ie distribution will be approximately of the same value as Is obtained a + % — jyma list of values. y 30-40 8 30 0 0 é 20 +40) at 0 gstatin §. Calculate arithmetic mean from the following data: 0 ‘ 50-60 55 10 +20 a " (vers 0-10 10-30 30-60 60—100 SS N= 100 asa (4 5 12 25 8 Swition, The class intervals are unequal but stil to simplify calculations we can take 5 as ‘a common factor.1s 189 4000 57 3943, 75 4018 4018 100 40.18. on of ‘arithmetic Mean in Case of Open-end Classes = 45 Dd=- 44, N=50, Cog ist sjasses are those in which lower limit of the first class and the “a 4 cas Maat clas ate KAGE, TH Such cae ne a Fo 5 Gy XS M544 = 406 wi thmetic mean unless we make an assumption about the “ é fimits, The assumption would naturally depend upon the dass on the first class and preceding the last class, For, exa t same. 8 e St median and mode adjustment ne Hata, 8 asks No. of students Marks No, of students . Seay, q elow 10 4 30-40 is ect Values. It sometimes happens that 10-20 6 40—50 8 comectng me a COONS tg find ont they se 30-30 * ‘Above 50 i, t oF roblem 15 © core tke owrrating mean. Te Pom incorrect EX deduct wrong jt met she above case since the class interval is uniform, the appropriate acess 1s YOY SIMPLE | qivide the correct EX by “yes agi) | ction would be that the lower limit of the first class is zero and the pimreer vcems and TI obtained, will give the value of Munya *UPfimit of the last class is 60. The first class thus would be 0-10 and Gpservations. THe Tesi Ore a spit class 50-60. Observe another case ' ¥ 100 students were found to b ‘he musttin 6. The meat, MANS read as 83, Find the correct mee, Le oy Marks No. of students Marks No. of students discovered that @ SCOre OF Coney Below 10 4 60-100 7 theconec eae y= 100, X= 40 10-30 é ‘Above 100 3 Solution : We are gi 2 30-60 10 7-2 , Since N In the above case since the class interval is 20 in the second class, 30 3x = NX= 100% 40= 4000 rite third class, 40 in the fourth class, ie., it is increasing by 10. The : ppropnate assumption would be that the lower limit of the first class is, Bur this isnot conect 2X gi and the upper limit of the last class 150. In other words, first class EX = Inconect EX- wrong item + correct item Pg ‘0-10 and the last one 100-150. Gorecte™ a : ITithe class intervals are of varying width, an effort should not be made = 4000-83 +53 = 367 jo determine the lower limit of the lowest class and upper limit of the conect 2X _ 3970 aq 7 highest class. The use of median or mode would be better in such a case. Comet ¥ = SAT = "t9g = 39. Because of the difficulty of ascertaining lower limit and upper limit in | pen-end distributions it is suggested that in such distributions arithmetic Hence the correct average = $9.7 { mean should not be used. servations is found to be 40. If at the time of cma Mathematical Properties of Arithmetic Mean 7 instead of 3 and 72. Find correct Mean lo = i erties of the (B.Com, Bombay ti. lowing are a few important mathematical props Illustration 7. Mean of 100 ob: two ilems are wrongly taken as 30 and 2————————— — Sr "Wy, m ww of the Hlems from vations is always zero, te° a the dO unt) fs always Zero, Yo." Any a zx x ne cum a into ae owing exampl Eat, ve ay or NK= x taking SS from pe clea would if we replace cach item In the serles by the mean, then wore substitutions will be equal to the sum of the individual ~ yn 1c, in the discussion of first property EX = 150 and the tg oe eamP If for each Item we substitute 30, we get the same mit ic mean + 30 + 30 + 30 = 150, 0 Mite, 90 4 “15 of great practical value. For example, if we know the rope" factory, say. Rs. 1,060 and the number of workers - wage 200, we can compute total wages bill from the relation rex ya rye" a total wage bill in this case would be 200 x 1,060, te, pf wn is equal to ZX. ; 14% 2,000 whine arithmetic mean and number of items of two or more {150 39, When the SUM Of the doy, give Ped groups, we can compute combined average of these = Xie K : nee I-75 en He eames OU t0 be ze, 08 4 ie ping the following formula i is characterised 85 point of ff ‘ss 3 bY - from it 18 equal to the guy” rae Ny of ee the e. 30: N actual meahs the meant Xi2= + this proper sitive deviations: shes eas Xi2 = combined mean of the two groups deviations {ro of the squared evens of the items fr ‘i %X = arithmetic mean of first group 2 is minimum, that ay other value, <0 ot tt Xo = arithmetic mean of second group feviations of aa INE folonga s N, = number of items in the first group would verify d ° : Na = number of items in the second group. ee oa ~¥ sowing example shall Illustrate the application of the above formula: 2 a 4 The mean height of 25 male workers in a factory s 61 inches and the mean height 3 srt orkers in the Same factory is 58 inches, find the combined mean height of 60 i (site Yactory. _ (M.A. Econ., Punjab Univ., 1995) 5 ees Kin = NiXt + No Xe 6 4 \puien "ite aun Ea-X)=0 a, Ni = 25, X1=61, Ne =35, X= 58 Foo = CSXGN+G5x58) _ 152542030 _ 9555 _ ‘the sum of the squared deviations is equal to 10 in the apy Xe +05 BO gOS taken from any other value the sum Gy cai atrsuld be greater than 10. For example, let ye gt Mg) pstcombined mean height 6 workers is 60.25 inches, deviations woul lcukh® TE havo to find out the combined mean of threo sub-groups the above formula can be Stuares of te devialons of ftem Grom a value less than’ yt Ine as follows > mean, sy ge Fad = XUN Xe No X x (x= 3) (K-9P Te Ne Ne Ne TOO 3 o 0 Neits and Limitations of Arithmetic Mean 4 +1 5 4a | eits Arithmetic mean is most widely used in practice because of the é 5 # owing reasons: 2s tt is the simplest average to understand and easiest to compute BX ~ 3)" = 16 Nelther the arraying of data as required for calculating median nor gE = * it f data ii is vl Its clear that £(X ~ X}? ts greater. ints property that the sums groupi ing : act as required for calculating mode is needed while Squares of items is least from the mean is of immense use in regret analysis which shall be discussed later.= ee oe 193 —————~- Mansy ew A, the value of eves ae in Sere, \ of 4 «the ¥8 - ’ wats feted OF ath ets the sam thy Paps ee welded, ariel, means: represent: the "ie is defined PY utes the a Ang yet Fon ben Hi MB Ws Be COM gad formu Mt Tends iggip Sy Met the. welghts attached to varlable values, ic. vg aermitee Facer than the median or moggs! to. pe ten -espectvely: te, * Bera HHT in the, Sener he Ses not, oe oso weights by the varlable X and obtain the total EX. vely Fe fro! ane we ics canoe samples ee ine ee san ta, V ry vg total by the sum of the welghts, Le, EW. chen Fel yas i al i wi east not 8s much in the sense that it is the seca a pM - frequen distribution, if fi, fy,..., fy are the frequencies of ical ™ either side of it. ny yr ge OF ues Xi Xa0-+ +. Xn respectively then the ceith oa yal ly welghted « The mean is values 09 Op Np 7 ple " he and not based 0M position jn the aia? gn is gives by: palancing © ue. v .itisae stated Ya ‘Wj EWU) IW imitations onli id j ds, UN Oe creas At form since the YaE TMA land very lange Hers, unduly aly gle expanded Ww te, wey Sg ttonial TOU there are 4 54 Ky = Wii Xi) + Wo Ufa X0) +... + Wa Ui Xe) ct Stu he the average: FOC OT go, 70, 10 and 80 the average Ee Wi+Wo+..+Wa in a te it i Np math 10280 200.58. One singe item, ie. 10, yy, veil sant problem that arises while using welghted mean is 4 fraby. The smaller the number of i memizouon of weights. Weights may be either actual or arbitrary, marks const pact of extreme value. er pated. Neciless Sy if actual ‘weights are available, nothing ter is likely t0 _ However, absence of actual weights, arbitré en a ye ee toe makin Ie & wit it is better no weights at all, In practic Prnpoted witout making ees. If such classes SZC of gue ey n practice, it is found jnterval of the Ln a a es Sa ee a abees con : the |h @ { weights ae logically eed keeping the phenomena in view, the rion of the values. : ubject tg a Wp volved will be so simi at it can be easily over looked. propor values of the median and mode can ta ‘ oI md be noted that: f er, the ¥ : cert i at main seat sie of class itera: thmetic mean is not always a ood mag ree Te aee mean pons etaracenstc value Te eR i

Xy if weighted arithmetic mean is: ~~ ww Hf (wy - wy) (X) - Xo) <0. 2 any & 0m mole arlthmetic mean shall be equal to the weighted arith Seat SER the weights are equal. Symbolicaly, ated acltinete I may be noted that wei i sa file hag to: ighted arithmetic mean ts specially useful in : (9 Construction of index numbers, and “the mean of a fequeney distibution X= 226 is, im fact, a mean of te iM Standardized birth and death rates. mic-points) where each X is weighted by its importance. This is only 2 soecdl gy ibe tol uration 9, he more general notion of the weighted mean it an tae ” ean ere fave, are abs workae mal lial aa cae 3 8 lene . 40 per day, is. 32 por day and to a chil W" 15 per day. What is the average wage per day paid by the contractor? y=RAL ™ swt so arithmetic mean, ig 40, c an cog The Her? wee aut ol ma, re peta wey BY \e" Caleutia Madras Seton aca LM al, HE SACH cag! thy a No.of Pass =o of += Pass No. of a NTN pe coe thay ah eit” Students = %~—Students.«—«%~=—— Students a 21018), 400+820+ 159 We % (in (in (in vrage woud ay 110 BEX 30 Rs, 29 ‘Wea hundreds) hundreds) hundreds) gosta oet0 +a woe . Wo 2 ale and chitd WOFKETS Employ 3 82 2 81 2 at ca. ype a amployed by the ca he oH roves My noAES.% on Os WE Lake aSSUmeg yerTFEES 8 4 1% 3 7% 35 1 we a PON ee and hd voters wean Ie 26, aye 1h it ge ra RS ed eat SC ou A Vit os 5 3 6 ™ 45 that them wou ey ‘The avert NOE = ne af workers ‘\ rf 1 2 7% 7 58 2 ea) yy 0 65 3 70 ee 20 [ee 6 : 7 i 3 60 7 73 2 15 F 66 480 ? 5 aes 6 smi WS 8 Ra 3 {aLQULATION OF SIMPLE AND WEIGHTED ARITHMETIC MEANS a TOS oe mene Fae te 388 239.875 oF 39.88, Bombay Calcutta Madras es! 26 mies al 2 spaed of 30 mph, another 50 miles iia No. of Pass No. of Pass No. of (yA wns 25 mS A eck raves for 6 Minutes al a soeey A tea |e FESS gents % Students % Students mph, hen we Ting distance of 24 mos at 8 speed of 24. pnts Gn hundreds) (in hundreds) (ia finally a Me nies pot Rou? (8.Com, Punjab Univ.; B.Com,” Wha, Wo! ‘undreds) average , dof 30 a j 9 25 mikes at @ spect Mh. = 59 solution, Time taken i COVERNG 49 mop.h. = 75 minutes, Digi? itu <7 Ww OWK wo WK taken in coveing 50 mies 21.2 SPE Tine taken in covering 24 nance coe wo eA minutes al a speed of 10 mp! miles ay gy, = 60 minutes. 2a mpl = ing De ime taken as weights we have the woighted mean ay a 213 2 1648 20 162 =——_ =e Ee, 4 332 76 3 228 76 35 266 Seed 4 met ad nme) Wx 5 365 73 6 438 74 45 333, egg 2 148 76 7 532 58 2.0 116 “0 6 3.000 3 195 65 18H 10 6 60 we «6 3 198 60 7 420 73 20 146 a ______ 4 Cy 1440 ~~ ~EW OSWX «SX EW OEWX EX = oDW EW E 24920 -=20 = 1451 = 492-28 = 1,977 =432 0 = 21 = 1,519 2W= 191 Se a and Weighted Arithmetic Mean : Average speed « 8000.34 Aimph. fey lustration 10. Comment on the performance of the students of the three univers below using simple and weighted averages: CabataN session, ta: oot "RB, _gneian—ineual Observations mn ere Wake, Bul tis wit 72 @ng ae! yo! « meat ince 0! Se weighted arithmetic mea? wrth in ascending or descending order of spe etn api te MEgIEG STE Mean ton tne data saconiing oie: of magpie + te pes tre pevtormance ee ied hg ctl reangererss ‘wari eve the same anewer| pecause WHat Tore eS pay Uni lonte ise REN 1 composed of an odd number of values such as 7, add 1 woattes say a" SAY 8 our combi ovata ane by 2 Mus 7 | wal . ine aivded by 2 eves 4 the number of the value Moti iedian } ‘dof the numerically arranged groups will be the “ aefinition refers, he ctibt 2 iste BF elles sige. Ina large aoup the same method may be followed a re iF oe al of the ler and one-half have A have by jell oof 199 stems the middle value would be 100th value, is Value in @ ined by h : She median is fust the Sot etue gig ald be determined by “5. In the form of formula : ize of the yr target es in the samy Dergg he & yt is 7 value OF Font of the vale mle rane ™ an ee Jt . " N+ ba tien ra oye mean which IS calculateg igi! Med.” = Size of “5-*th item. en inn te an es, the median is what is calle Jo Uy u sear nem 3 I eters to the place of a value gt @ poly? sa. rom te foloving dala of the wages of 7 workers compute the mean eam mh Hak gel . of the fy ve 1 on oe side ae income of five employees Is Rs, 999 te . gars) 1100 eH: = wat) ge te EN For ABI fe pel ‘would be 1020. 1200 and 1: % a 900 CALCULATION OF MEDIAN 960 iddle position of the array 7 ae di — arranged in yo value at mi y ie Wages arranged in SI No. Wages arranged in ee 2 ascending order ascending order ° 080 5 1160 128 m 2 1100 6 1200 ition of median i ror ae abe ml rer an even umber of ake s ime ’ 400 odd number of see middle position value andthe Teena 4 ne Isted, there is no single : lan i ee listed, treTrmetie mean of two middlemost items. For example. 4) ize ot Nth tem = Et : Bbove case we are given the income of six employees as 999, ggy41y__yndan = $8 7 ain An 1200, 1280, 1900, the median income would be: : * TO ging ot th item = ue Hones the nate wage = As. 1150 ‘gus find that median is the middlemost item : 3 ss = is nd number, i@., 3, get moresthan Rs. 1150. Pare gala Wage: Res a 1020} ethere are two middle position values ‘he procedure for determining the median of an even-numbered group 1200) dems is not a8 obvious as above. If there were, for instance, different a0) res in a group, the modian 18 realy not determinable since both the fi and 6th values are in the centre, In practice, the median value for a youp composed of an even number of items is estimated by stan «1201200 2220 5549 Her of the tao auiddle Yeluce’ that ts; addiag the to taking = oe . \ddle and dividing by two. ; 7 jin the mid . Expressed in the form of formula, it Hence, in case of even number of observations median may be ay 70'S ' by averaging two middle position values. Nel cis hen Wis ad he median is an actual value, with the rae Medlenseipiae of th item ofthe ses in two equal pris on eter side of. IE NS even, te figure, Le., half the sum of the middle values. * The abbreviation Med. represents median.——_—_— tt ST 198, is odd. ea = s odd as that it is both sient value, an Thus we find dded to determine n {one) has to be at median (ror the following data; 4 ra CALCULATION OF MEDIAN. uo of \ 42, Obtain the v4! 277 MWustration ‘ot 672 52% 783, ef. Income ‘No. of persons Cf. aot ‘arranged in 1 ascending order a CALCULATION OF MEDIAN 16 ee Data aranged in S.No. * a“ = " ‘SI No. ascending order 66 2500: 6 122 ni _ — 7m 6 m= EE 61 5th tem, ; oot z se ti 3 .dian—Continuous Series ; 405 bod ™ son on te a J ¢ the particular class in which the value of median 1 4 522 1,490 ‘ peter rank of the median and not (N + 1)/2. Some writers hee , “Feet! 185 SD Med. = Size of pth itom = —> a — * 175 \ ” se y m s 30-40. es 185 a gyn en the clas at ys ‘own {9 US are 2 . 198 % 8 ar vane HS ch = (10 + 20 +h), i= 10, f=40 : ° a mgs b= pete jn the median formula : is 7 Menara 155, pvalues, we should find out the se we ae ven #° rmid-val MPF ang 95 = 90+ MOH 20 +I 49 Sees ae LCULATION OF MEDIAN © ~~ 0 = 35 = 904254 Zh _Olass group 7 4 soe ‘tRo-170 6 90+ S495 oF 55-1 =20 or A= 85, wo ee 8 sf 80 2 a) «60, hall be 60 ~ 95 = 25. Thus the missing frequencies are f, = 95, = 26. 120-190 z 79 180-190 22 gual oO 130-140 : 181 490-200 3 ® _ CALCULATION OF ARITHMETIC MEAN oie a " f Ep: (m - 3610 fa 7 3 a rote ns ia weve ee “ore vos = Sie ot ih tem = Size of SP sae 5 _——, 2 4. = 0 ‘es oo 20 18 -2 -40 Iedian fas in the cass 160 ee 38 oe 4 a 35 Medan = L+7 0 40 35 ° 25 45 rs] 425 = 195, cf = 181, f= 116, j= 0-20 & = iene eed 0 25 55 +2 +50 Median x 10= 180+3.79= 15379, 30-70 15 65 +3 +45 — N= 70 d= 15 ee NON , : aphid Mlustration 18. An incomplete cistibuion is given below: X= aes Variabie: 0-40 10-20 20-80 30-40 40-80 50-60 697 A = 35, Bfd = 15, N= 170, i= 10 Frequency: 0 OF 2 2 5 f X= 35+ x10 You are given that the median value is 36. Find out missing fecueipe = 38 + 862 = 35.862 the total frequency =_170) Calculate the arthmetic mean of the completed table, (B.Com, Hey m #2? Me aPthmetic mean of the completed table is 36.882.5 . hg, . on ™ Intervals are 1) wen cinst Meaney \ f edie” unequal, the freq, é ae of observations, the median is a more satisfactory interval ea equal and the’ lig, ¢ 1n 8 ental tendeney than the mean the a8 gcussed above Samy ee ON ewedd distributions such as income distributions or a aed ony athe where the arithimetic mean would be distorted by trom the flowing data ne, et butions. median is especially useful. Consequently, the _ 10-30 90-60 Vee ett for some purposes be regarded as a more 010 ‘6 s "4 oo nc ure, for half the income earners must he recetving at 8 cnt an income. One can say as many receive the median ne : (8.Com, oI ibe tgs many do net sist Cary ve ynoqual, 61 US frSt CONVENT it i, eval at the MequeNcies are equ; Sis impwion that WUally ey", ea et appropriate average in dealing with qualitative data ks are given or there are other types of items that are Mr measured BUI are scored median can be determined graphically whereas the value 1 tye sraphically ascertained if vin cannol ia reatest advantage of median is, however. the fact that CALCULATION OF MEDIAN y a aps, Me Eetually does indicate what many people incorrectly oa =i edlan thmetic mean indicates. The median indicates the value § @—~ ise (Pde item in the distribution. This 1s a clear-cut meaning ~ 2 a ve ane We the median a measure that can be easily explained am cs » z 60 { size of Mth itor = > = 30th it ons Mod 2 2 ahem yp renlating median it Is necessary to arrange the data: other © por os not need any arrangement, in tne cass 90-60 + yerages dO seca bas week | nee ig a positional average, Its value Is not determined by each Med = b+ 7, sine ory observation oy —— — - = 20, 1= 30(7= 9) apable of algebraic treatment. For example, median cannot L = 30, N2 = 90, ef = 20, f= 80(7 = 39) nist (aFGctermining the combined median of two or more groups 0-20, 99 . we used Sibi in case of mean, Similarly, the median wage of a Wed. = 904% Gq * 89-904 10= 40 s 5 Pojsinbution times the number of workers will not give the evetayroll, Because of this limitation the median is much less thematieal Property of Median wemular as compared to the arithmetic mean. Mat P : on sayons of the items from median, Ui value of median is affected more by sampling fluctuations than Thc sun of te deve te eG. By 10, 19 RING cpg * Ie value of the arithmetic mean, ieast. For example, the i60l0 0 9" gan the total ices deube sre median, in some cases, cannot be computed exactly as the from & ignoring Sieh ar ed if deviations are takes ‘2 Tey)’ wean When the number of items included in a series of data is woBier than the one ol from 7, values ignriygt™ amy; Men, the median is determined approximately as the mid-point of the wig deviations ae ken from 7, values ignoring giggle te Mell \ 4 ‘ tal 13. A 4s eB De | jis erratic if the number of items is small. Merit Lim of Median and Tielletioe swfmess The median is useful for distributions containing open-end Merits finals since these intervals do not enter its computation. Also since the milan 1S allected by the number rather than the size of items, it is + It is especially useful in case of open-end classes since qipjqently used instead of the mean as a measure of central tendency in position and not the values of items must be known, The maja ase Where such values are likely to distort the mean, also recommended if the distribution has unequal classes, size easier to compute than the mean. ied Positional Measures Extreme values do not affect the median as strongly as the tigeskies median, there are other measures which divide a series into equal mean. For example, the median of 10, 20, 30, 40 and 150 wodbpats. Important amongst these are quartiles, deciles and percentiles. 30 whereas the mean 50. Hence very often when extreme vamiQuarules are those values of the variate which divide the total frequency26 Stay 28 divide the total fre Net into four equal paris, deo The total frequensy ASESY ty, WW? z sO Min item (in individual and disci, Parts and te percentivt series into WO parts, (100 Ig Vf, SI” = Just as one poin § 3 thee : Givi it into four parts, 9 Points (0 09 pasts and oocp Pomel y PF of {hin stem (in continuous series) — parts. Consequently. tHe quartiles are denoted by ge, eats Inythhy pu? sm 1 percentiles for a series, TY Aebscripts 1, 2. 3, ety, YMbo cs Mays GO + Vir, item (in individual and di D and percentiles by P. Tay value that we want {) Deere dat! gue f~ 100 crete series) would refer (0, We Pare. Q> second quartile, Qs thyco™PutyO. Wei ps” N would denote first quartile. Qe oe erand Pag Guth rd qua’ a) f GON: item (in continuous series} decile. Dy 8th decile. P, Rest Peecenarauon vaines aneentlle, eqtle, My) size OF 100 Gepost teal pars, Tt ary gg) gn A SEY US, ied dest ency é : ; ‘ oper ec feaird quartiles. for example. the a the histo; nea SS ~ calla ea 8nd 20th percentile By these lines into four equal parts, The 9 deciles divige Mil A qh 4 lui 39! 7 Tisfogram or fequency cure tte, 10 caval parts ang the ant 28 s 125, 175 225 Givide the area into 100 equ : oh Pale 7 18 25 wi in coor ae gurl ae the TY . Sand percentiles. co : . an eile the distibution into four equal parts, "ops on fey \ (B.A, Hons, Econ, Kerala Univ.) Mreke “quartiles, the second coinciding with the meqiay °° yet Stated, the lower quartile Q; is that point on the x More her ve a0 9 wen mricepoints, we will first find the lower and sod for finding these limits is to take the diforence'temoce yc! oS . js ™ r x ference between the No one-fourth of c ‘ah eae, les than 91 and “tn such pees pe by 2, dea he vais So cisned tom he ower rt an ad he er than Qi, The + 2: Poi foun a ae 75-25 8 eh that three-fourths of the total frequency is beta fy ett . vpnegien case 7 =2=26. The fit dass sha be 0-5, cond 510, ee above it. F ong sil The deciles and percentiles are important in Se et ___catoutarion OF 0,, Oy Dy ig educational satisties concerning grades, rates, ranks, ei /Mlogey ae a rn fav ceonomics and business statistics im personne! yi? mela) 58890 | ratings and other such situations. prog t os , Tt should be noted that quartiles, deciles, ete., are not a, 540 are measures of dispersion and as such shall be diseuca Ya pri es the next chapter. Here only a passing reference is made, ‘tna de 1520 we computing these partition valves Is the same as discussed (@°® tein’ = = Just as quartiles divide the series \yato 4 equal parts, “penta | 20 7 into 5 equal parts, septiles into 7 equal parts and’ octied MU Gy! = 100 parts, However, these partition values are rarely used in pracqnt? § 4 . N, 100. Sry. cuarile = Size of Mth itom = 12° = ast item Computation of Quartiles, Percentiles, ete. p 4 : ae jyesin the class 5-10, The procedure for computing quartiles, deciles, ete., is the median, While computing these values in individual nd discrete (2% a = cM of add | to N whereas in continuous series we do not add 1, th Thus @ = Size of S*4th item (individual observations and series) Q= 54S x Q) = Size of {in continuous series) ecm = fain he coss 15-20, Os = Size of N+ Deh item (in individual and discrete see (Mt H* Coss 1520 4 = 14 SN/A= et = 14 BWAn a, { @: = Size of SM item (in continuous series) L = 18, 3N4 = 75, of = 50, f= 90, /=518 5-50 5154417 = 19.17 e ae vg 8 3Np tem = 2%100 py = Size ot Gg! 10 i less than cS) — 6 1500 Se 16 1600 : 38 1700 10 68 @ wages are between Fs. 1965 and RS. 1490 is 16 as Ghows gs whos AKERS GETTING WAGES BETWer Le Rs. 1365 and 1430 es 120 109 yx 4 20-7 y5-543.61-8: 3 Po = 5+>4g *5 5+9.61=8.61 3 73 2 60] ermination of Median, Quartiles, etc., Graphically ; | ‘Det , can be det eemined graphically’ by applying any o¢ the 4 | 3 4 Median ol is: “ies : two method: ogwes—one bY Jess than’ anetiod and the "oy, al + Draw So id. From the point snes s : sess oe Integy) | than’ ’ yndicular on axis. The other. ray pers the X-axis, gives the value of Tete ° mos Peony one ogve bY ess than’ method. Take the vara, 1100 1200 1800 1400 1500 1600 1700 + Draw only euency on the Yoaxis. Determine the median yi tq WAGES (Rs) ae i Hin item. Locate this value on yy formula: median = siz¢ © 1 of workers whose wages are less than Rs, 1430 = 73 ‘he Frag Wer of workers whose wages are less than Ris, 1965 = 57 the cumulative fr draw a perpendicular on equency of workers Whose wages are between 1365 and As, 1490 = 73 — 57 = from tt dra a Me meets the ogive, draw another perpenycr ty, rats wate and the point where it meets the X-axls is the median ay CALCULATION OF MEDIAN ‘tion values like quartiles, deciles, ote, § The other partition vee method No. 2. CAR ay ages (No. of St Wages No. of 1 determined graphically by following . a) baie ——,—, Workers . for the following distribution. How man en JOer eo ie yu Mlustration 21. peat Reh ‘Tago? Aloo calcula the imodlan ‘inane Worker sy 96-1200 10 16 1500-1600 4 98 wages between 1900 22 38 1800-1700 12 110 Wages No, of workers Wigee No. of wertag | sf-1400 3. 8 s, SS SS (Rs) 4 " enw : 1400-1500 6 | Median = size ot item = 12. th tem 1100-1200 10 1500-1600 4 flee median lies in the class 1900-1400, 1200-1900 2 1600-1700 2 Nedian = 1+ M2=eh Lo = 1800, N2 = 55, ef = 38, f= 90, i= 100 * ‘The value obtained graphically will be the same as obtained algebraically cox flees « 55-98 errors in plotting and reading the scale. 1200 + = x 100 = 1300 + 86.67 = Rs. 1356.67Stagg, 1, VALUE by 100 stters of ant in statistics ot nia oer hal Ob = 175 pe yuan iter = wal y= 2953 c* \,etge OS it 15 ca ta , 4 bs 0 got! x i % : is that ind out Median and Quarg rd modal value Is value In a series of ol ‘suka, Univ, Bon fomyt ge et A the greatest ffequency. For example, the an " : wu! Ds Sh save toqueney CONE by the Tess thas Uy on 5, 8, 3 a oe bere. since this value occurs more ve cumu c he 1, Sos Sy thant said to be that value which f ang. oon often Occurs most often in the sont as rae Ss No. OF Sudeme ‘et seth. the Highest Requency. While this statement ts quite oe ea | Lee Nye jaterPrene i vagaries of camping wong eye pled to any mi oe 0, ite Wah in mpecause of the agaries oe pling. Even fairly large samples Ve (an —— lo. 20 $° p¥fpaloh g"statistcal population with a single well defined mode may De 30 3) gf trom 2 vatic fluctuations in this average if the mode is defined a6 te 55 % wnt be alue in the ungrouped data of each sample which occurs most oer —= ie 7 20 % exact “Rather it should be thought as the value about which the Lees than 35 —% *G ent ee closely concentrated. It is the value, which has the greatest Lest tan 40. 100 27 Mis BE Mepalty in its immediate neighbourhood.” For this reason mode sare gy es) 7 the most typical or fashionable value of a distribution. dan = szo of Nin em = T00V2th tem, Jef called Ue fagram shows the modal value: iG MEDIAN Q, & Q, ‘53% following * LOCATING APHICALLY | ‘he “yd L v4 «tT 1 ) b iw \v ol > of ; MODE | 2 2 3 a z te value of the variable at which the curve reaches a maximum is 3 ahi yin) the mode. It is the value around which the items tend to be most E gl -_ + pully concentrated. ‘though mode is that value which occurs most frequently, yet it does xt fol that its frequency represents a majority out of all the total 20] amber of frequencies. For example, in the election of college president te wes obtained by three candidates contesting for presidentship out of itotal of 816 votes polled are as follows : ae Mr. X 268; Mr. Y 278; Mr. Z 270 : Total 816. | $G, = j|Mediaha, = its Wy 2428s ‘ 3 yo “The mode of a distribution fs the value at the point around which the items tend 5 10 15 20 25 90 a5 40 ‘abe most heavily concentrated, It may be regarded as the most typical of a series MARKS: values,” —Croxton and Couxien Fonte gas a ar atlas eho 224 ‘Ava M, Tuttle: Elementary ‘Business and Economic Statistics. A set of data may Qi=size of N/4th item = 1004 = 25th item. ‘ave single mode in which case itis sald to be unimodal, it may have two modes ‘hich makes it bimodal or it may have several modes and be called multimodal,m2 sident becaust he ha wat be, ete 82 Bi say that he represents Me XH wall be SOME pn (268 + 270 = 596) they factions in. whieh arithmetic mean ay Hoge oh oS y iw yotes. Bi ‘more votes tions: en are situatiotie of data. For example, wig wee om jumber of limes it Size of tem 7 oh rar oF Haber areas gow ga “G7 24 : : 2 there many are ceristi : charact e, Most comm: income on height’ tale e true on Stone, yen neem gre Mad sae oreie mean OF MECon of the data due to the pre aN 1 accurate rou" rove to be quite unrepreseniqeetee a at 1 27 } re aah 1 30 i Pation of the serles. For provide tan may al stribt ae range from, say, Rs, 1g a 3, hence the modal marks fems. Median ri vo all Of et tne upper half of the’ 3, c : es the same G0 oath eS wl geet ihe a7 occur the maxim number of times, 1 1 q " le tt stribution the ta. of determining mode in in Ge true nature of the Oe ay be overcome by the . ht als ¥en case of individual observations Both these shorteone rs, “occurs most frequently "S* of nq Say Aad moves 2 refers to the value SMCE test to compute singe Ma quis, Yai two or more values having th mode {8 . Cet estat, ere are i the same maximum Moreover, mode Ie est frequency. For example, if it ls oh yea ME cannot say which is the modal value and hence smo correspon 6 ? 8 2 © aot Wee, Oy defined. Such a series is also mown as bimat ae 25 40 2 Wg 3 10 hd For example, observe the following data of income: Bg peso 110 120 130 120 110 140 130 120 130 140 since Ht appears maximum number op |e St aes | Ones each a Nie ooo 0 120 330 140 2 3 3 a Calculation of Mode the mode of recise value of the mode of a feateny — "volves 120 and 190-have the same maximum frequency, ie., 3, mode is Determining the Pp’ calculation, E: Jementary ‘ssentlally by no means an él ite type of freque ny me appropriat leney curve since is CASE. mathematially so7on of the value on the X-axds belay (0, De gi hres 2 Mie cs ‘are several elementary methods ‘of © eee « jode—Diserete Series ety data and determinati ‘been discussed for in aid ae Ob ace just by inspection, (i fooking to that value of the variable around wi ¢ oslation of Mi ‘hich the items are curve. However, there methods ea i ite series quite often mode can be determined ese ode, Tesges and continuous Sere. Cateuation of aren Observations te Mealy concentrated. For example, observe the following data : poo ae 1) eT mode count the number of times the vari. For determining id the value occurring maximum number % fepeat themselves ften the modal value appeny ea ‘ value. The more 0} fhe modal valvhe measure 15 an average to represent data, bio, of persons wearing © from the above data we can clearly say that the modal size is 31 Tecause the alee oh has fee the maximum number of times, ie, ‘ iB However, where thé mode is determined just by inspection, ani error of IMtustration 23. Caleulate the mode from the following data of the mats thingy | encmum frequency and the frequency preceding it or succeeding is yery small and the items are heavily concentrated on either side. In such (ases lt is desirable to prepare a grouping table ‘and an analysis table ‘students: ers ‘Marks obtained ‘SI No. ‘Mars kg Or ese tables help us in. ascertaining the modal class. A grouping table has six columns. In column 1 the maximum frequency 's marked or put in a circle; i column 2 frequencies are grouped in maa Si. No. > 10 2 27 7 3 24 8 € s I : 4g | 98. in column 3 leave the first frequency and then group the remaining 4 12 ‘n two's; in column 4 group the frequencies in three's; in column 5 leave — Ste first frequency and group the frequencies in three's; and in column 6 (B.Com, Mysoe Un, “eave the first two frequencies and then group the remaining in three’s. linr means of a inst ij against Manes entered ey represent. ping table, prepare an an, e gemmjuma number on the lefty cqunode on the right-hang “a si re. the highest Fe marke SAN BH, ne mn . axdmom total and mark 4 4, a une (an mn as the maximum total 5, the value of the variable ‘able ts 25, att grouping table and anal al es t ; ireponding OF Neparng BFOUPIE TIC Grd ana, | ONS ee nda BSS sis able oF by inspection swing &xé S hy in ear fromm te followin © ” be Ya ‘ie Borat ihe value of mode by applying the following formu ata tne ve of ode forthe flowing data eee a ila nystration 24 28" ot DB | Mo = bea ya xt 7 Marks rr a ’ ida Cg: : Frequorey : amie mit of the modal class; A, = th 42 lover al lass and the frequency difference between the 1 getgay f ESS gnoking SIEMS), te. | fifo]; ayo tne ee i Gticut to say by inspection as 10 Which ig Sun, reeds Cruency of the modal class and the fe the difference Solution. Since analyse tables. the inf s une, Mea, succeeding class (ignoring EBs fequency of the apes Vg jade Orval of the modal class. } be. | fi-f|: 30 28 wv | : a Ng r\ | os ANALYSIS TABLE Col, No. 25 CALCULATION OF MODE ‘nt class of this formula is: Mo= i) = jower limit of the modal clags; fj = yb MGuency of the dass preceding retency of the modal Ys abe lass succeeding the modal class, ass: fp = acy Of Mg the above formula for calculatin, it is nee Fife apPiNG SS intervals are uniform throughout Wh Mi ecessary |, ole fit ee, mad, cau on the assumption that the reunites yo ey distribut oughout the class. oth 6 Y esults. herwise we will get ss be two values which occur with ere AY equal frequency. ‘The fae. then called bimodal. The following is a graph of bimodal jp ' NUMBER OF WORKERS Mone MODE WAGES (Rs) fs bimodal disttbution the value of mode cannot be determined with Mt ct formula given above. If plotted data produce a bimodal Ceubuton, the data themselves should be questioned. Quite often such a Gaim. $s caused when the size of the sample 1s small; the dficulty remedied by increasing the sample size. Another common cause is. In instances eneous data. the use of non homo sy be done t0 change it, the and io rntral tendency. bimodal 2 of cel Bi ‘as a me: ned, its value may be asce; wee mode we ned jonship between mean, 2° formula Das Mode = 3 Median - 2 Mean ang . ‘ / Le asure 1S called the empirical mode, NN = : This me: "site ani the folowing data: \ ode = 1075+<2 545-1075 +0108 110605 (ustrat - No. of students Marks aga ws 1 110825 KB vem : ne fon the following data of weight of 122 persons determine the modat ait solution, Since this is cumulative frequency distribution, we frst eg frequency distribution Wars vet Hy a (B.Com., Sri Ve No, of students Marks ‘y é /enkaleshwara Univ, Osmania Univ, 1998) ‘on it is dificult to say Which is the modal en 2 ara an ‘analysis table. 8 a GROUPING TABLE 30-40 80-90 a No. of persons SS 40-50 12 fo $0100 No. of persons mis ce 060. By inspection the modal class is 50-60. fobs) | a Mt wv v ‘ass. Heres we propare a ee AL ‘110 4 » beth shs- 10-120 6 30 : 425 N15 ~ 121-3249 26 = 804 q2x10= 50452 20 . My = S04 575% 10 50+5=55. 120-130 i 58 7 2 Illustration 26. Find the value of mode from the data given below ww ad 8 65 Weight (kg.) ‘No. students »__ Weight (kg) 40-150 33 82 93-97 2 WSN 4 50] 58 98-102 5 118-122 eye | 25 a 103-107 2 128-4127 2 Dim 6 108-112 128-192 10 ma 2VALUE ‘ real o® a9 ose intervals are Unequal inode 1s expected 10 le" seh chase in whe ANALYSIS TABLE es ulating the value of mode a 90-190 ip 06 te equal class intervals, ite poi '8 applicable oil ere ree make them equal before we as’, imeFals thet, Mine class Interval should be made equal aca fe f ode, gssumption that they are equally distributed thrvencs oof 1 roughout 1 ically 8 : 4 oe id the value of _— - 9 ‘bution the value of mode Total - = ry ogra, ee in calculation are: can also be determined es. Hence mode has ‘ ‘This is a bemotl seri 0 be fa be, pstgram of the ven cat Jagonally e insid anpbnk Mode = 3 Median - 2 Mean, ne fos ios Ore 2 the modal an ve each upper corner of the bar to the upper cca ae sendicular line from the intersection of the two di A Me X-axis (horizontal scale) which gives us the tiodal wae ine raw a histogram for the following distrbution a omy nee and find the modal wage ang ve 1% 4 on } pevae PPHO-18 15-20 2025 25.00 g005 ag.ag @ te 404s so nas Fe se 40 1101501201009 ve a id Pred weg (8.Com,, Osmania Unw., 1998) 160-170 165 a . 8 a 's oh ae ee 3 "i ga toa of a gen Belo, —3!! LOCATING MODE GRAPHICALLY 2 atin 160) 150 x A = 135, 2d = 55, N= 122, 140 140 x 55 = 195+ 555 x 10= 195+ 4.51 = 139.51 Nn, 122 0 Med. = Size of 5th tlom = Size of 2 ~ is ten | 100 Hence median lies in the class 130-140. Na Median Le xt 6 ~ i 190, M2 = 61, cf. = 90, f= 32, i= 10 (61-30) 310 130 = = 198.69 +e 10= 190+ = 198 L Median NUMBER 60. OF 60} 60 WORKERS 40 Mode = 9 Median - 2 Mean Mode = (3 x 199.69) - (2 x 199.51) = 419.07 - 27900» i 20) Hence modal weight is 140.05 tb. —_——— Ql * While preparing analysis table all the classes need not be written | 10 15 20 25 2890 35 4045 classes be considered in which mode is expected to lie, that is why in ths MODE = 28 classes 100—110, 150—160, ete., have been left out WAGES (N Rx) i