UNIT-V: |_| OMITSN: TESTS (OF sIgNIFICANCE |? ~ SMALL SAMPLes The Kquore of a Atandard neamal variete is brown a6 a cthi-cquare variate with Vdegsce ol predom TY thee i: a nowmal dikibrd pooand standend deviatton o, then DWT mean 2: xm eS ; ns (zat) iota gtiatale with df se Cat. dagace ef fecclom) cen genval y Xi Cte 2. Mm) ann independent nomal vertetes with man py and standard deviatton oj for i=1,2,...,1, Then . x. 2 ( ) a Wvortake with on dt bri Deciace o freedom (A.£): Pegace | frudom aejs to the maxinourn humbea of logically independent values of a Atatiatic that haa the fecdom to vouy in the data satple. * een TY oo by fhe sarnple ize) te degae of frudom fe O-1). ZX’ probability eunwe: Lak in denote Hu’ degases of freedom. We probability of C.ditsibubten ‘hilak denote the der 4 sth that P(x*> we *'P6e) , ; : m \ a at. 4 : wpe fF . / _— Rajectton avg t on (oe) Ny . ’ 1 Gitzo! voli: a wikieal valve of Notes |. . oo The X* probability usve depends on * Lawl sf significance ., at deeince of freedom Applications of i” dicri bution To teat pth goodnuse ge, #70 the independence. of vari allnibubes.. +7 tak He homegenity of independent Sel4 the population Voniance. and comelatioo eosfticdent 4 Te wrmbing arate probabilities obtained hom indapendenE ax puiments to ge a dinglt best e( aignigicance: . xr teat Good ness ef Fit: This tut t used for trating the signi feance of 4he discrepancy between they and experiment» the deviatton of the iat Tt enables us to find v] cepeaioatot gor Rey uw (ust by scaly dus +5 the ‘(nadequacy’ ef the theowy to git the observed data, ° ay on Cizt/2,-- (exparirototel) frequen ccs dig Act ef axpected (thenrebical) pequenctes chance obs iF op) boa at of observed and. Ej Ciet2,--0) bs the CAALE POD then . . : oa = (ae with (OV), af. mL eg; Problems: L The demand for o particular Bpose part ina factory was found to vory from day te day. Th a sample stacy, the following infarmatron was obtained . | Poys | Mon Tue Wed thu | Fat Sak | 1 1 +} 0. ‘F Be lie | ls | (to [ise li26 | lies | Tast the hypothesis that the number of parts demanded does not depend on the day the week given the valurs of x* signypicanee at §,6,7 degnacs of freedom one seapechivaly U.07, 12.59/17 07pO.IT9 eu 24 h1ns +110 +ULZO+ I 264ULE 6 E= \Jo0o . Sol olitisetec Uo1nsX, cat «rs Ye lave . t of signlficance ~ Newmber of parts damanded does not capes 20 the doy “ef Fhe wreck’, 2. Th following figuras show the dishibutton of digits in number, chosen at random from a tele phone directo. Lest Whether the digits may be token to occu equally fequentty in the ise ctory . det Hot The digits Oca aqguclly frequentty. E= tv000 =toeo to db = 9a ae yi el tatte le 2 | = vig o | € | o-e | @-e)"| O# ° 1026 hooo 26 6G o.6 | | | (tot tooo | lor (14449 11.444 are aa | 2 | 997 lows | 23 74 0/004 | 3 966 love | ~34 use | 1456 | | | | ] | 4+ Lois tooo 15 Sb25 5.625 | 1 | —t —— | [933 | tooo | 61 | 489 | 4409 | 1 - | | 6 | tet | tooo (01 \ideq | N44 | | | 9712 loco -28 | 7184 0.184 . | | aa | 3 e¢ | too | -36 1296 | 1296, 4 83 | tooo | -1e1 | 21609 | 21-609 | 58. $42 Tetal | tooo | (e000 xe 58.542 516.919 me ab x=s% level of sig nfcance ee | we The digits do nob occtn 181s -Xx at K= 57% level of Significance with 3 Af HE Agjeded The gen data docs not cormmensurate with “the genial sxam suult # A survey of 80 familie with 4 childaen fach Agvealed the fetlowing distrib ubfons y ! 2 $ ° No. of boys | No. of gir | 4 3 2 | Me of familia) 39 tts 290 | ase ee | that male and female births ore zqually probate (Given, My = 4488 at Ko% level of significance with + Af) — iw]Lal Hoe Mate and female binths ove eq nally Ieee Jeb «8 7% level of significance af{-4 To find expected values , / ; . Wie awe esata oa ret rales and ferme ee axe equally probable tef p= Paobability of the ehitd born being o wale. ot el-pe Reteoc ees | Assuming That this avant follows Pinorat dishib ution, Fo. o boys, fc, (£ Fa 1 bey , *C, (E)(4Y *#ee = 22° Fo 2 boys, #c, (2) (4) RPO = Bee Fo 3 boys, *ey(4) (4) Kevo = 200 Fo. + boys 9 7G (Ly (LY xe" ve yay meer ee (Ned 45 € | boys | | eo | 32 | bo C | (1a | ee | 2 | 240 | 800 [2 | ase | 22 | 4 64 | Soeote 12.65 > 4.488- acy at 0 SH bevel o| “orifices: with + df | , acento oy & Male and fumale diaths are not Aqeatly _ probable io he given data.Oy : . « ONIT-Vi TEST - ye ie” ees RR ES CFOR ELSNE Gansce -SMALL SAMPLES yo teak i atti bates: Lit A and B be two ‘odin’ bute, whee ¢ A is aivided Unto a closers AN Aa JA and Bois divided into 4 classes BBs, Bg. The various call faquencts “ea be sxpesed in the fotiowving a x ae ae at] PA “Total | | 8, | (a) fon | x8) | (8) laa oe Teo tiny | e.(A Be) (aBa) e+ | Aaa) |) L - - [tc ekgane r ___}*_+ 8 CBa) (AaB) ove | AnBa) | (80) | [total tay) | CAL) | a I (a0 N The problem js t tut whidhoa A and 8 ore independent a nob. We take Ho: A and B ast independent The theoutt al dee quencits one coleulated ag follows: PCAR) =, CAL), LEN awh. ihe ; PCBi) = ei) i Jrlre& PCALBj)= PCAL) PCBI)~C*A and B an assumed to be indaptndent)PCA Bj )= CAi)(By) s(irne,. ANT GEM b) N? “The anpected vate of (18j), (MB), = NEP CALBG) (AiBj), = canta) > oN (AE BU) 5 = (ADC) N (istry Was) | The value of X” i given by eid (Gu tainir) CArBi)o | 5 “we (48) which hw distributed with Gs Dacula Paoblume: : | Two oun ple pols. of votes qo 2 candidates A and B fe a public office om taken, One for from among tha Accidents ct Aural aro, The Acwults are QIAN in tHe table. Examine whether the nabere of FR ona is Related to Voting piyeianee C@iven X"x= 8.24) fa x5 7% oo A significane with de) SL yy Nata | Sada ‘oundk voling pane on indktendent 4 each othr , K= 5 Yo lowel of signaieance Ate (2-1) (2a). 1Rural 620 | 3a0_ lo00 | Vaban| Sso | 420 tooo | | | | | 1 Total WT | 835 2000 | _ oo ae : oO 6 UE o-€ (0-E) ozn | (AR) | 620 | 58S BS ms) bong 380 tis eS ies: 2.452 (A, vy -_ ss 73s lors 2.09 F (8,0) 450 FIs as hos 2.as2 Le - 10,092 x2 to, 012 > a.ea1-X, at ele laval | significance with 1 df # te te aufeeted. The nature of He oma is Aslabd to vating prejrence. 2. Out o go0o grodusits in a town, 80 are feroales. Ouk of (ooo grad unk zanployees 120 ane dumales. Use Kulak to debamine 4 any skin ckton bs made in appointment on tHe boas cf gunda. Ca@iven x22 3.841 faoss % Lval of signi{tcane with tdtLer Ho: Thur is no distin chion made in appointonent bead on gender Ke Fo level o| significance a= Co-1)Ca-1) 27) | °, © ©-€ (-€)* coe) 04,4) | 5120. S1Gbo -40 tooo 0.2718 (F, 6) beoe 640 to l600 2s (ae) {f80 | I¢40 | Fo Wooo any | (FG€) tro l660 =4 6) l60o to 13.89" | x. B.89> 3.841 XY at cot H level of significance with tat He Ww Aefecked “ Thiae is Afi nitely a ginder bias In appointenent shel aalepe Re given 2x2 contingeney table , prove that test fs indepencln ce gins Xs NM Cad-be)? Casey (b+d) Cast) (cad) Whee Nz atb+eyg“| eS | |. (arb) C bra) bo CarbXbrd) = aoa J} > eR (atey (ead) | ¢- taredtetd) = be-ad le an.) N No d (bid) (ead) d- (bid) Cc4+d) | Nn N \ 2 % 2 o a X= (ad-be) Cad-be) @d-be (ad—be) a + No + Ne + N™ (atb) Case) (arb) (bd) tO Cord) (baa) (er) N N N N \ x (ad—be)™ 4) +t aces | S nN : an Carsbid) (arererd) (bid ced) pas (ae be) toed \a a x: Cod-be) boil Gr ar) (ac nes (ad-be) * (estaced )Coabrerd N Carb) Ceaartars) Cora, 2 xen (adebo™ 16 Carb) Cet Care) (b4d) cab sa ( ate bid) Crd (a: a) =) ‘Sa> oN Cad= bey Carb Ce: tA) Cares b +l) Xe teat fo populetfon vor tence Suppose we want to x; Ciz1,2, treet 4 @ Aando tm europe »0) has been drawn froro ® Dame) population with a spedficd Vortane 6 =o, Undar Hyi o*. o;* Hyi o> 0 , with (Haj Af Pro Rie Ob is believed that He pacision (ax measured by Voriane).e| on insbuamnent ls no mon tan 0.16. Waite down the nul oral allanakive hype thes is ci Yating this belig - Cosay out the teat at 1% lwvat > aign icone given M rpaccurtmnents oA the seine dubjeck ois the clowtrument : . 7 1S 2) 2.3) P-4, 2.3), 2.9, 2.1, 3.5, 2.6, 26,2-7,,> (Given 2% 223.2 at ‘oxen’ Level of cignificane With to df . ' let tgs or = O16 . His oP So. 6 ici der wsry, bet > sig fi ‘ af = N-yetox X-X (x-% > a — 0.004 0, 00008) 2.3 —0.201 0.043681 2-4 0.104 oO.o118 2) 0.0F368) | el 2-5 — 0.009 9,000 081 { 27 | oa) | ©. 08648) 2 ~ 0-004 | e-eecce) 2-6 0.041 0.00828) 2.6 0.04) 0.00828) 2-7 ena 0.03648) 2-5 | -— 0.004% ©. 00008) | Te o.1s2s Exe 27-6 | ne = 21.6 = 2.509 aNgonerng o.isss € 2382 aint rd teva) { stgrif ficeence With ‘ Af oO X(x-%)* = 0.1525 = net = Ux> 7a _pobbues: 1. TRe following toble gives the mur be of ahaaft accidents Hak oecumnd during the 1 days of te week. Find whether the accidents ore uniforcaly dishi buted throughout the week Day cecal € o-€ | (© ea co- | Mon \¢ 14 o i et © Tue \e 14 + we | a wed | Io 14 -2 | 4 | oe Thun | ou 1¢ | -3 a | Obs mi | ois efor fo al Sot | Iq | 14 eS a | . - \ Total | &4 ‘ | 23 Lit Hoi The accidents ax uniformly distribubed Hough out dhe week Lut = 5% bevel of significance dfe 6-t=5 Ea = SF = \q 6 HPs aaa Level of aignigicance wits 6 af S Hy can be acapted : 1 The aceiclents are unifernaly distri buted Heroughout - the weele ssh ; : , 2. Arnong 64 od prings of a, certain . cross between guinea pigs, 34 Were Aed,to were black and 20 wee white. According to Qemete model thee numbershould bein the aati 9:3:4-De the data consistent with He model ak 5% Level of significance. Let Hoi The’ number of of pring. ie la the ackto a3 4 von Set oe 5% level of signifi cence [44 df = 3-le2 i ~ 2 ot o-e€ | (0-€) e-8 | Red 34 36 2 4 oul Black | to 2 ~2 4 (0:33 | Whi | 20 | te 4 | te | a | 4 144 25.44 x at 57% Level -of significance with 2 dif Jette can be accepted The, data, jo consistent with the moete| 3.Dn an 2xpuiment on immunisation ef catte from Tuberculosis the foUowing sausuls Were obtained. en [Onafpcted Ino cutee ta | 28 we | ed ; inowtabd | S| 7 Tetol | 25 al Examine the affect of vaccine in controlling the ise case.Let Hot Vaccination and aquiring TE is todependent ef zach otter (Le, vaccination Ww infective In con Le 5% bevel of signipican dfe QA) =1 ho ling the disease) ce 2 | ° | e | (oe oer |S | AW | [es ~4.67 | 21-8089 | 1-208 | a eae ey | (¥v) | 28 | 23,33 | 4.67 | 21.8089 | 0.935, (ANY) 12 8.33 4.69 j2uecss | 2.618 | @NW] 7) U6 -4.67 [at aosa | 1.869 «| x2 6,13> B.8F-Xy at oe % level o| one with 1 df Hy bs payected The vaceind affective io controlling the Alar ade.oy. ypol® 4, Two groups of 100 people goeh vveac taken far testing the use of a vaceine. 15 pasons cwonbacted the dicose ouk 4 the inowlated pasons whi 25 contacted the diceace in tte other group: Test the sificeney of Hie vaccine using chi-square tet | Affected | ajjected | Total, Treculeded Is ss | loo | | | Roe | | | inocwlakd 25 | ats? | tee Total +o . (bo 200 | Let Hot Vacetnakion and aquising +h Aisecse ane unslated Chath, vaceine is not affective) dia hs 5% level ot signi Leon oe df = (2nd =1 | og o-e (o-e* Get L E (4,2) (rs | 20 == as teos (waa 8s | fg | > 2s! | ovaias | CAND | as | 3 | s | ae | vag | (NANT) Is | eo “sls oa | | Saas | O02 Bas < Bar ER ot te % level of dignificans witb td. f 2 “Ho con be aecepted a + The veceine te inaffective.5A random sample of employers of a lorge Company wos delected ancl thre szrnploy cx wus asked to comple a qusstionnain. One gquistion asks Whether the arnployer was in favour of the introduction Of fleri ble working hous. The folowing table clasfics Ha employees by thelr response and Hetr arco | nak. Maa of wak . Res ponas —~ Production Non-poductton Totat | In feveur 429 mn 300 Not in favour a 64 Joo | Tote: loo 240 400 Lat Hot Ansa of Wak and Acwponses are independent ef tach obhy det fos K% Level of signgt cance df= (2-2-1 } oO) € o-€ (0-=)* @-e)* I € 9 120 4 8 Ape Mr 18 “4 8\ 0.45 3) 40 -9 & 2.025 69) bo 9 8) 1.36 | ers} A 3.84 X= 45> RING = Keg ok KaT%o level of cignificance with 1d : 2 He bb Agjected vw The Aecponees oe dependent on theta arta + weak.6 Tn an experiment on pea- breeding , He gellowing deqrendis of deede wae observed. gis aound and yellow ,101 Wainklid and yellow 10g asund and green, 3a Wainkted and QMen. Theory predicts that the frequencies whould be th the propatton 9:3°3%! Aapectivaly. Sek up a proper hypothesia and tat ak 10% level e| Aignificance. | O-E | (0-€) (o-m" 2 | 4 | eo | -3 | 9 | oe Ga P I o.ls4 =5 9 6.257 ost) Lat Ho: The frequencies of ders matches with theorelttcal propations oe 5% level 2 significance df= 4-153 x= ost <7! eX, at f= 57% level of Aigniyican wih 3 4 o He dan ‘be xeeepted OTe frequencha | stads mottches with theoretical proporttons :ope - 2 Suldea Go pact ; \. The standard deviation of a ceatain group of leoo high school grades wos WU % With a mean Brad: 18%. Assuming the distribuitton to be Aral : a) How rrany grades wos above 90%9 b) What wes the highest grads of Lowest 10 7 © Within what Limits does the middle 10% grades tind 2 A darmple of (00 ibm daawn fiom an tuniverse had pecs and o=3 It has rman valu 63.6. Te the Aifsenc in rean ign icant ? What will be your inference 4 Aarmple has 200 eee . 3. XN (30,57) a P(26¢% 240) b)P(l x-30)} > 5) ©) P(x> 42) A) P(xe2e) ; 4A bined coin Wot Hom foe aed Hands Was’ Obletimd oo bine with od ago deduce the probability o{ gelling heads in a single throw of a win certainty Wing betwen 03 and 0.64Hetoog, me T8, ‘eet @) P(x>40) = P( 251.09) Ze REM or Haq oc on oT iW PCx>40) = os— 0.2621 = 0.1379 Num ber | graces above Fo%A= O1379Xto00 = 137.4 . 2138 be RG 3s -30 =! ia Plas cx c3s) = 2X 0-343 5 OGG P(lx-30l>s) = 1-0.6826=O0.311F |. ©) P(x 242) = P(z>2-4) Zs 2-30 =2.4 s ‘ P(w2qa)y % aS. 0-498 5 D, 006 2'« . a) p(x 28) = P(2c-0.4) Zp 28-20 2 +04 Ey P(x cre) = 0-5 O.1SSF 20, 3446 WwSG) 2 [2s = fosxes . 0.025 Vow 400 Probability of gettin, Jeng \. + 3(s. ©) heads in a single throw # . 0.64.3 (0.025) =(o.s25,0,615) UNIT~-4: Test fer Significane for Differne deviations: Tp 4) aod 3, an the sfandard deviatfons of in standard two independant amples, Hhen ‘tendley nil hype thesis, Heo eon, Sinx oj) and TL on acually” Unknown, We Use 4) and 3s in theta place P 2 4t- 4d oP tay Qn) 2a, Problems: I.Random samplis drawn ftom two counkis gow Hh following data rdabing to He huights of adult amales :A 8 Mean height Cin inches) 6142 612s Standosd deviation AeA aise Ci; inches) . Neo. fo samples loco (200 Te the diffana belwin tet oott [ole vor] bat two Independant ratimates of fhe population Voatanee arc homogenous, The tect atabishte F ls given by With Cr-rny-2) dfProblems: Tn 4 sample of 8 obseavaktom the sur of the squoas of deviations of acumple values from sample means Was 84.4 and in another cample © (© obirvabion jit was 102-6 -Tart whether these Aareples axe drawn from populations with carne Voniance af 7% level. n= 8 n,=to Zlgxj-K)= 84.4 | Slyj-H=tor-k kao % bevel » dig nificance afs (t0-1,8-1) = (4,4) Let tor The damples ome drawn rom populabtons WH same voalance Mii The samples one mot from pop lations With dame Vvaatence. Sy = Z(xj-%) ©) x 84.42 ta.057 n-i 8-1 2 = Sy*=_) = Cyp- = $x for.G ath >) lor Be Sx" Sia Fe 12.057 Wet Fe tose <3,26 = Fy at ke 5% leyel of signifi cence with C4,1) af pomcie sus “The samples axe from populotions with same Variance.