Scribd
Upload
24 views

Chapter 03

Uploaded by

Fantasy Potraits
Copyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
24 views

Chapter 03

Uploaded by

Fantasy Potraits
Copyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 9
CH-03 Measure of Central Tendency Summary Statistics + Inchaoter 2, we constructed tabes and graphs om raw ta, Te esting pletres of frequency dtibuionslusvated ends andpattersin the at In mest cases, however, we need more crit measures. nese ass, we cn se single numbers ale summary sales to ‘eserbe characteris of dala set, + Two ofthese characteristics ae party important to dean makers: central tendancy and dspersion| (Cental Tendency |» Central tendency the mie point oF airibaton ‘Measures af central tendency ate aso called measutes cf eeation (COMPARISON OF CENTRAL LOCATION OF THREE CURVES Dispersion * ieparson the spread fhe data ina dstsbuton tats, the extent to which the ‘Skewes Curves representing the dat points inthe dataset may Be ether sme o skewed Symes curves, tke the one in Figure 33, fe uch tal aerial line awa or he center ofthe curve tothe horizon! is deste are ofthe cure into wo equ part Each part the itor iage ofthe eter. curves Aand in Figure 3-4 ae skewed curves They are skewed because ales in thee froauaneyestbutons are concentrate at athe he low en ote ign end of easing sae onthe hoes ans Te vakies are not qual dst buted Cre Ais Skewed to sera (or ostvely sewed because tals of tower the highend of the scale, Curve 8st the oppose, skewed tothe elt (negate skewed) becae ta | {toward the ow end ofthe sae LAN 7" Kertose ‘When we measure the kurtosis ofa distribution, we are measting ts peakecnss In ieure 35, forexample, curves A and 8 fer only in that one it more peaked than the er They have the same canta locaton and depeson, and both re sya. Statistician say hat the two curves have ferent grees of kos. Measure of Central Tendency ‘Mean / Arithmetic Mean [Mean rom + Average of data \Unerouped ata | When we ine all te obervaton of the data ocaleuste mean, we called it unerounes data sample ean ma pa Population Mean pa ‘wean trom | uppase we lave very ug ata awe drt want oli each for apoio we have vcs to Srovpedoata | oi eavency srbulon deta then we can en se eqn ato ahh 9 ovae2 6). Sone inci cone Orevpe ot Bon ba ‘Advantages of Mean Tes Sg numberrearesentng whole datasets ery deta sets ns mean Ieisungue, sine every datasets have only ane mean, {Cane compared wih other datasets. Disevantages of |» Eay affect by ousiers Mean 1 For groupes cata The wird dsadvontage stat we are unable to compute the mean for a dataset hathas ‘openenaedlascr tether the igh orlow end af th ele Weighted Mean ‘The weighted mean enables us to caleuate an average that takes inte account the importance of ch valuetothe overall oa xampie- Tar Hoar ern o One SGredeoter Henry Wage) Print Ped? Unkle ‘S$ T 7 1+ Horeif we want to know tha average wage cot of aber pr hour for each product 1 Asimpe arthmeticmean would give (5170313 «7 Using this average rat, we would compute the labor cost of one unit of product Yo be S7(1+2-+ 5} $88 and of one unt of product 2tabe $7 «3+ 3)» S70 But hes rewers are incorrect ‘Tobe corect te answers must ake nto account that fran amounts ofeach grade of abr are use We an determine the carect nnwers inthe folpwing manner, Fr produ 1, he ttl bor cost er units (951) +(§7»2} +5935)» $64, and sine there are 8 hours oaber input, the average [shor cost poor $648 = $8.00 per hou. For product 2, the ol lbor cost per units ($5 » 4) +57 3} (99>3} 868, for an average abreast per hou 68/10, oF $5.0 per hour Anat! way to {aleult he correct average eos er hour for te prods to ake 3 weignted average ofthe est ofthe thee grades of abot ‘wea ton ‘This wolhted mean concepts similar to the concpt of arithmetic mean cleulsted or roupe data Distinguish berween distin values and indidua cbsersatins no dota Set, since sever observations con hove he same vale f volves ocurwi efferent frequencies the ated mean ofthe values fs ‘opposed tothe tmete mean ofthe sbsevetiens| may not bean accurate measure of ena) tendency m such oss we nees ous the weighted mean ef the voles. fyov are using an averone Value to make odecion ase how wes cleulte. the values inthe sample do ot onze th the ome frequency, inst on a weighted mean asthe corer baie for our deion Geometric Mean + Sometimes when we ate dealing wth quantities that change over a period af tune, we need > know an average rate of change, ich 2 an average growth ate over a period a several yas. i Las 168.00 on GM. = gprodactofallevaloes Bo, = (LOTTO TOLLE = eT 1093 <——— Average growth factor (he geometric mean ofthe 5 growth factor) eimai ign in pind ine i sgreraticerrine Hyer ain fe nen eer an eam Median 1 the madian sa si value from the dat set that measures the central tam inthe dats 1 Thissinaleitem ithe middlemen or most entra tern the et of number Half ofthe Res i shove ‘The median nt affected by tier [Median for erouped dete ‘nee on ‘Median for sea when we ave data group na equencyasvbuton ngrouped data 7 {Ce gti 21 cm mi of or dine i Seat Sant Aaa (RHE, co {=p “(ipo 751535 — td area eel a ny dng espana ‘Advantages of median 1+ Notatfete oy outers The medion ey to understand ane canbe cleats ram any kine of data-even for grouped data wen open-ended dases suchas the equeney dstrbuton unless the mean fas in an ‘openended css Disadvantages of mean The data >must be sorten order to give te carect mean which can be tecious we are dealing it arger set of daa Dincsensuarrors peesenemnseses ‘natn tential pc aoe Te pteo a y ac n Shana tar enn mas ht EBC Stma eee wg Mode + Moses the vlue thats repeated most often inthe dataset 1 Modes ely used san measueo etal tendancy oF ugrouped at NONE 70-04" ERO Tip Are neatng Order ‘The above datas ungrouped data anak shows the numb of delivery tris er day made by ReMi one plant. The modal value 525 because tac’ more often than any ote vale (hee tes). ‘mode of 15 imps hatte plant acy ish than 6.7 [6.7 sth answer wee get fe ened ‘he mean) The mode tls us tat 5 the mos Hequent eumber of is, but il fas 1 et vs know tat ‘mos ofthe valver ar under 30 Now we group te data we st : Now es group these dataine aequencydtibutin, as we have done in Tole 3-14 we selec the ‘lath te mort observation, which we can athe modal das, we wos chook «7 ripe. This ‘lass sore representative ofthe acy ofthe plat than the mode of 15 sper dy Fo hs reason, wenever we se the mode a 3 measure of he central tendancy of acta set, we shoud Caleta tne mode om grouped da Muttinodal Distribution \nen we have more than one modes na étaet ‘Advantages ofmode + uke mesian, mode can also be ved for both qualitative and quantitative data + Modes not atlected by cuties 1 Athiedadiantage ofthe mode tat we can use even when ane ormore athe dasa are open ended Disadvantages of mode Most of the time ther snot mode nthe datasets, in such caes made cao be sea central tendency Sometimes athe ue present in dataset occas the same numberof Ses, such ease modes cess ‘Comparing Mean, Median and Mode + Risveryinporantte decide whetherto use man, madianormode athe measur oferta tenceney symmetrical astibutions thet contain ony one mode aways have te same valve for he central tendency because the choice hasbeen made for us. + Postvty howe als ale as right shewed - Modes at highest int, medians to Isriht ane mean stright of both mode ad median + Negatively Stewed sz cle as let skewed Mode sat hghes point, mean isto inleand mean ile to both medion and made *+ hen the populations stewed negstiely or poste, the medians often the best measure of + Otherwise, there ae no universal gueines or apphingthe mesa, medlan, or mode asthe ispersion +The mean ofall hee cures isthe same but curve Aha les spread (or vara than cove B are curve has ess vray tan curve He measure oly he mean ofthese vee “stibutors, we wi miss animporant erence among the hee cures, + Utewise fr any dca, the mean, the medan, andthe mode ell us only part of what we eed to now about the characteristics ofthe data, To erase our understanding othe ptt of the ata, we must also measures ispersion™s Spread or vanity Need toknow about dipesion: + Dispersion ges us adeitonal formation that enbles uso joe the elablty of our measure ofthe cortaltenseney. I data are wid period such a those curve Cin Figure 39, the Central lesion [les rereseeative ofthe das 5 whole thn t would be for te more casey + Second, bcaut thereat problems pect widely sapersed data, we must be ae to recogate tat data te widely persed before we can take thoseprblens 1 Th, we may wh 2 compare dspetsions of various samples. fa wide spread of values away ‘rom the contr sundasrabl or presents an uracceptable sk, we need tobe abl a recognize ana avod choosing theastrbutons with the pests: dspesion Useful measures of dispersion Range + Therng tthe erence beeen te highest and lowest tioned vate + Throng tent undertand na tofnd ut te tues ao mens of pens + Terorge conserson te gest andowet ns of ttn ad also tte cout + Kignosthe rnc ofthevaratonamongalte other bsvtons andi hen ced by + Suserrnemures oy two vals the ange sey chong asta fom ove sample 2 Interfractile range + Ina requency dtroution, en fraction or proportion ofthe dale tor below a Frac. The media or examples te 05 Hace, beause hal te data set ss ano egal oth value +The race are simi to percentages + Teinterractie anges a measure he spreod between two frctles ino frequency dstbution, that the difference between the value of he two rte, interquartile range +The nteruartle ange messures approximate how far rom the mean we must goon ether side before we can ince one half the vals of the data et, + Tocompute ns range, we dive our ati four pas, cach of which contains 25 percent of +The quartes are then the gest vaesin each of tes four pars, and thentrquartl anges 1s Wotan cust ga 11 quasnies cert nnd opal ee snr peas Mase ‘sitet i Attain atv a ‘Average Deviation Measure ‘The most comprehensive desriptions of pesion are thas ha dea wih the average deviation rare some masse of canal tendency, rumberinthe st /Ameasurement of how far each number fom dataset som themean and thas fom every other Population Variance Sample Variance Standard |» Squoerootaf Variance Deviation |+ The standard devaton enabesusto determine, wih 2 gest dea of accuse, where the values of frequency stbution re wth respect the mean, | = ‘aber osiaton (olues wl win £3 standard devotions rm the mean Standard Score 1+ measure calle! he standard score ves us the numberof standard deviations a partelar ‘beeration lies blow or above the mean ‘A per chebsey theorem no mater what the shape ofthe dstibution, a east 75 percent of th values ‘io within 12 standod eins from te mean of he crown an atest 8 percent ofthe 1+ The sanard deviations alo vefulndeseng how ar indies in a detrbution dear from ——_—_—_—_——— Relative Dispersion : The coefficient of variation +The standard deviation san absolute asi of dispersion hat expresses variation nthe same unis asthe erga ta. 1+ However, the standard deviation canst be the soe bass fr comparing two dstibution + har weneedisaelatve measure at wl ge us fee for the magnitude ofthe deviation relativto te magntue ofthe mean, The coeficient of vation one such relative measre of

You might also like