Scribd
Upload
148 views38 pages

Testing of Hypothesis A02

This is a pdf for reading testing of hypothesis in statistics

Uploaded by

তোর বাপ
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
148 views38 pages

Testing of Hypothesis A02

This is a pdf for reading testing of hypothesis in statistics

Uploaded by

তোর বাপ
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 38
3-28 ENGINEERING MATH Gr) So we take the null hypothesis 1o| P = 3) -The alter, tip, hypothesis H,| P + 1) this is both sided test. The ga, 3 les, ; 32; n= 9000 ; Obtained sample proportion, p = son = 0358 P~3 _ p-0-33 Here z= = 12 0-00496 3_3 9000 The best critical region is |z|>2-58 at 5% level (as for the previous example). Here the computed value of z are =5- which is greater than 2.58. -. Zlies in the critial Region. So H, is rejected at 5% level We conclude ‘ the die was not unbiased’. 3.1.10. Two Sample Test for Mean (Test of equality of means) Theorem 1. Let two independent random sample: {1522.4} and {xi,x3,--.x4,} be drawn from two Norm populations (11;,0,) and (117,02) respectively to test tht hypothesis Ho(1; =)12) at o level of significance ; taal HT (nty tty), oles erases). h n Then aes Case (1) : if 6,,6, are known z=—2!—*2_. js the test ste 0 : ees tha Gk . Statistic which is standard normal variate. Accordingly is (i) z<~zq if Hy(4) 0 if Mil > |2) (right sided) git) [21> #02 if Ai(4 #2) (both sided) where a is the area under standard normal curve enclosed yetween the ordinates z=2, and :—= as shown in the adjacent Note. If the two samples are taken from any population and if the two sample sizes n, and n, are large then the above orem is also true.Moreover if the population s.do, and 6, not known then we may think o,=S,,0,=5, where 5,5; are sample standard deviations. But if the sample sizes are small (< 30) and the two samples t-curve are not drawn from normal tions then the following theorem shows the shows the method of testing the Theorem 2. If o is unknown, he nya (m +2 -2) X-% mtn, In S? + nS? is the test statistic which is a students t variate with 1, +n; -2 degrees of freedom where 5?,S? are variances of the two samples respectively. Accordingly the C.R is (i) t, if Hy(ps > 2) (right sided) (ii) |1|> raya if (hs #2) (both sided) where g is the area under t- ta furve with n,+n,-2 d.o.f ‘nelosed between the tes (= 1, and ¢ =~ as shown in the adjacent figure EM-1y.25 3-30 ENGINEERING MATHEMATICS, py Mlustration. meaieeat Uh ty rand ( Let two samples be drawn from tw: 1,9) and (j1;,9) normal populations. The details of the two samples are P .d size mean 8. 6 7.62 0.024 5 7.49 0.032 We are to test the null hypothesis Ho(i) =H2),6 is not given, with the help of the above sample-obsorvations, We choose the test statistic es ning (m +m —2) = au ny +My mis +n;S} which is a student-t variate with m+n, -2 degrees of freedom. Corresponding to the drawn sample the computed value of 6x 5(6+5-2) 152-7.49 t= |————_- = 7 = 1.60496 6+5 6x (0.024)? +5 (0.032) having d.o.f 6+5-2=9. Let the alternative hypothesis be //(j1; > #12), right sided. From the statistical table we see P(¢ > 2.26) = 0.025 corresponding todof 9. So at 0.025 level of significance the OR is (>226 © corresponding to d.o.f 9. Here the computed value 1.60496 does not lie in this region. So we accept Hy and conclude ith two means of the two populations are equal” at 2.5% level o significance. Illustrative Examples. Ex. 1. The means of two large samples of sizes 1000 and 2000 are 67.5 and 68.0 respectively, Test the equality of means of the two normal populations each with s.d, 2.5 at 0,5% level. Given area under st, normal curve is 0,25 enclosed between the ordinates ==0 and ==00987, Let the two population means be j, and j). The two 4. 0, =0) =25. The two sample means ¥,=675 % = 680. We take the null hypothesis A(t; =}12) against the alternative hypothesis //\(\1; #41). CE ce pssTINGOFHYPOTHESIS 3:31 We take the test statistic z= . 2 : whichiis ormally distributed. (25)° , (2.5) i 1000 2000 corresponding to this sample the computed value of 75-680 -05 ~0. ae yi 1 i = aay S642. 25,J—— + —— ‘ 1000” 2000 Since H, is taken, both sided so from given data the critical region is |z|>.0987 at 5% level of significance. Here for the computed value |z|=|-51642|>0987. So Hy is rejected. We conclude the two population means are not same. ° Ex. 2. A college conducts both day and night classes intended to be identical. A sample of 100 days student yields examination results as under: 72724} 0, =148 A sample of 200 night students yields examination results as under : ¥=739; 6, =179 Are the two means statistically equal at 10% level ? Given P(0<2< 1645) =05. —0981 0987 Suppose the populations are normal. [Infact this assumption is not necessary if the sample size is large (>30)] Take the null hypothesis Ho(ui=H2) against A\(i;#H2) where 1;,u) are the two population means. Here, means of the two sample, %,=724, %=739 s.d. of the two samples, S,=148, S) =17.9 size of the two samples, m =100, m =200- Since the sample size are too large so though the s.d of Population-are unknown we take the test statistic 3-32 ENGINEERING MATHEMA TICS.1y which is standard normal variate, Computeq 2 724-739 _ 2 2 (148) 4 value of z= -0.77. V 100 * 200 Since 4, is both sided and since P(01645 at 10% level. Since |-0.77|=0.77 does not lie in this region, Hy is accepted at 10% level. Ex. 3. Two types of batteries are tested for their length Of life distributed normally and the following data are obtained. No. of batteries Mean life(hours) Variance TypeA : 9 600 121 TypeB : 8 640 144 Is there a significanct difference in the life times of the two pes oes ? Value of t for 15 degrees of freedom at 5% level is 2.131. Take Ho(i)="2) against H,(1, #2). Consider the test statistic, as population s.d is unknown, p= [tmln+m-2) __x-% m*m | IuS? +S which is a ¢ variate with degrees of freedom mtm—2- Corresponding to the samples the computed value of (a [oxs(9+8=2) 600-640 9+8 Yoxi21+ 8x44 -40 =797Xx s- Via Si t- curve having d.of 9+8-2=16, From the given data we have P(|¢|>2.131)=5% shown by shade in the adjacent figure. Here the computed 72131 2131 pest OP NYAROER RSIS rety calve Lies 10 UES PegION, So Ayia reiected, We canal jaa signitioant difference in the average iit one eee yes of hattery®, 4. A group ofS pationts treated : Ba & & anc af hee 3 @ sooo ma yr S he saree ogaitad week with medione Bw Ws Gh BL & aNd @ Ag, Do you mh fa thai medicine ‘B* increases the weld it (WALA ig normally @eiridudad significantly than tha that dy AL [Me wale of t ae Sw leow’ of sienigroance Nr 10 degrees of Preedons is Saket) Let = population meanoweight of the Pationts treated with medicine A ve = Population meanweight of the pationts treated with, anedicine A Take null Aypothesis Ay(iy= Hg) against the alternative Aypothes’s 44 (i z, if H,(R > A) right sided (iii) |z|> zap if H{(R = P,) both sided where P(z, 1-96 at 5% level Corresponding to the samples x, = 400, x, = 600, -_ +x) _ 400+600 5 Pem+m 60041000 8 and the computed value of 400 _600 as" es) Which is greater than 1.96 So, the computed value ‘ 2.66 does lie in this CR .We do not accept Hy at 5% level and 4, 18 accepted. We conclude P, # P, i.e. the two cities have tot same percentage of literacy. Ex.2. 4 company has the head office at Kolkata and a branch at Bombay. The personnel director wanted to know if workers at two places would like the introduction of a new Plan of work and a survey was conducted for this Purpose. ut of'a sample of 500 workers at Kolkata 62% favoured the new plan. At Bombay out ofa sample of 400 workers 41% were (Sainst the new plan. Is there any significant difference een the two groups in their attitude towards the neuplon &t 5% level. Given area under standard normal curve ei between the ordinates 2=0 and z=19% is 0.475, 196 0-96 ENGINEERING MATHpy, Unie, Cé.ty Let R= Population proportion of workers who fave ed Kolkata. P,= Population proportion of workers who fay Bombay Office, . “thre is a significance difference between the tw 0 in their attitudes towards the new plan” «> ‘Pp 4 Py = We shall test the null hypothesis H(R= H(A # A). Now, 62% out of 500 =310 and (100-41) % out of 400 - 246 «. Here x, = 310,n, = 500,x, = 236,n, = 400 Oured ay Fi) againg As we know the test statistic is oleias r= nied where p= “it a af L 1 mtn p(t -a(t+4) which is a standard normal variate. Corresponding to the two samples ry 310+ 236 c 546 9 500+400 - 900” 150 and the computed value of plomaa6 eee ESE gist a! 2h) stot as 0328x100 328 150\° 150 )\ 500 "400 From the given data we have P(| lg 196) =05. So the critical region is |z|>196. Our computed value does not lie in it. $2 Mp is accepted at 5% level. We conclude “there is no significant difference”. Ex. 3. Ina factory producing articles, 400 articles out of° sample of 500 articles were found to be of excellent quality. After taking a step of fuel reducing process it is found a articles in a sample of 600 articles are of excellent. si whethere there is a sufficient decrese in the quality 233 _ 2 production. Test at 1% level, Given i J Pde 2088. TESTING OF HY! ‘POTHESIS 3-37 a H(P; = P2)against H,(P, > P,) 1 =400, x =400, P Since H,(P, > P,) is right sided. From the given data we have P(e>233) =05-049=001. The CR is :>233. Here the computed value lies in this region. So Hp is rejected and H,(P, > P2) is accepted. We conclude “The quality of articles is sufficiently decreased”. 3.1.12. Goodness of Fit. In a random sample drawn from a population there are different classes of observations having different frequencies. The test by which we decide whether the observations are in good agreement with a hypothetical distribution of the population is known as Test of Goodness of Fit. Illustration. Let a random sample of 500 students be drawn from the population of all college students of West Bengal. Let it be observed that there are four classes of students among these, which are Hindu, Muslims, Christans and Buddhist. The frequency of these classes i.e. the number of students of these classes are noted as 200, 150, 100 and 50 respectively in the sample. We consider the hypothesis that “The ratio of students of these classes in West Bengal colleges is 3 : 2: 2:1”. If we want to test whether the frequencies of the classes occured in our sample agree good with this hypothesis then this is a ‘test of Goodness of Fit’. Observed Frequency and Expected Frequency The frequency of a class in a sam) le is called observed frequency of that class in the sample. It is denoted by f.. In the previous Illustration the observed frequency of the class ‘Muslim’, “ f, of Muslim” =150- + 3-38 ENGINEERING MATHEMATIOS.1y is (on the population) which is to bo seated oy the ype of fit’ the ‘should pen ny, io a class is called the expected frequency 0 st iat n fe sample. It is denoted by f. In the previous ot Mali he expected frequency of tho class ‘Muslim’, “ fe uslim’ =2x500=125. 8 Theorem. Let A be a null hypothesis regarding the occurence of different classes in the population. Suppose there are k number of classes. If f, and f, be the observed and expected frequency (under Ho) of a class then 2 re yeah) the summation is taken on all classes, is the fe test statistic which has ;2 distribution with k-1 degrees of freedom. The Best Critical region is w>xi at a level of significance where @ is the area under ? curve (of k-1 d.o.f) enclosed between the PA ordinates at y7=,2 and x? =~ as shown in the adjacent figure. x? - curve Proof. Beyond the scope of the book. For illustration and application of above theorem go through the following illustrative examples. Illustrative Examples. Ex. 1. In a sample of 500 students drawn from the college students in West Bengal it is observed that the number of students reading in H.S., 1st Year, 2nd Year and $rd Year are 200, 150, 100 and 50 respectively. Will it support the hypothesis that in West Bengal the number of students reading in H.S., Ist year, 2nd year and 3rd year are in proportion 3 : 2:2: 1? Test at 5% level of significance, Given that 5% value. of x for 3 and 4 d.o.f is 7.815 and 7.416 respectively, pestiNG OF HYPOTHESIS 3-39 The ‘students reading in H.S.’, ‘students reading in 1st year’ are the classes. So here we have k=4 classes. The null pypothesis is Hy) (In W.B. No. of students reading in H.S : st year ¢ 2nd year : 3rd year =3 : 2; 2; 1), The alternate Hypothesis H, (In W.B. No. of students reading in HLS, : 1st year : 2nd year : 3rd year #3: 2:2: 1) Tet fos observed frequency of a class in the sample. er Expected frequency of a class in the sample, under fo - Computation of f, and /, =x 500=1875 8 2 =*500= 125, Ae 2 =x 500= 125 8 1500 = 625 8 * The total of the expected frequencies should be equal to the total of the observed frequencies. The test statistic is re pWexh) , with 4-1=3 degree of freedom. The summation’ éxtends over all classes. Computation of x? corresponding to the sample is shown below : iiss the computed value of x? =1333 with 3 d. ‘upplied data we have P(x? >781 )= 005 3-40 ENGINEERING MATHEMATICg, ry Se ee “= the best critical region is x >7815. Since our compute value lies in this region (as 1333>7815) we reject Hy at 54, level of significance . We conclude “In W.B the number of students reading in H.S, 1st year , 2nd year and 3rd year jg not in the ratio 3: 2: 2:1”. (Note that here the Second supplied data is useless). Ex. 2. A die is thrown 150 times with the following results No. turned up: 1 2 3 4 5 6 Frequency : 19 23 28-5917 32 31 Test the hypothesis that the die is unbiased. Given xioss=1107 and {95,6 =1259. Here the number of classes, k =6- The die will be unbiased if the frequency of all faces in many throws are same, i.e. the ratio of occurence of the classes <1: Lares deeds Let Ho (In several throws the occurence of Face 1 : Face 2 : Face 3 : Face 4: Face 5: Face 6 =1:1:1:1:1:1) Let /, = Observed frequency of a class in the sample f.= Expected frequency of class in the sample under fj. 2 The test statistic is papel extended over all possible class) Which isa ¢ variate with k-1=6-1=5 degrees of freedom. Corresponding to the sample the value of x is computed in the following table : , (the summation is * this should be equal to S/,. ESTING OF HYPOTHESIS 3-41 (-fF =792 with d.of 5. , the computed value of => from the supplied data we have P(x? >1107)=005. So the critical region is x’ >1107 at 5% level of significance. Since our computed value does not lie in this region (as 7.92 + 11.07) we accept Ho at 5% level and conclude “ The die may be piased”. Ex. 3. The number of road accidents per week in a certain area were as follow : 12, 8, 20, 2, 14, 10, 15, 6, 9, 4 Are these frequencies in agreement with the belief that accident conditions were same during 10-week period. Given the following data Statistic: dof value of x’ at 5% level ¢ ; 9 16.919 This is a sample of 124+8+20+2+14+10+15+6+9+4=100 observations. These observations are arranged as Week : Ist 2nd 3rd 4th 5th 6th 7th 8th 9th 10th Noof 132 g 20 2 141015 6 9 4 accidents Here k= No. of classes = No. of weeks =10. The accident conditions are same during these 10-week period means there are equal number of accidents in each week. We consider Hy (there are equal number of accidents in each week) and #, (number of accidents in each week is not same), Let /, = observed frequency of each class in the sample /, = Expected frequency of each classs in the sample under ( Hp =10)for each class. The test statistic is x°= yen th , (the summation being extended over all classes) which is a x ert mi 10-1=9 d.o.f corresponding to the sample the value of x’ is computed in the following table : ENGINEERING MATHEMATICS. ry 1st week 2nd week 3rd week 4th week 5th week 6th week 7th week 8th week 9th week 10th weel 2 .. the computed value of ve ee =266 with dof 9, From the given data we see P(x” > 16919) = 005, -. the CR is ?>16919. Since the computed value lies in this region we reject Hy at 5% level and conclude “The accident conditions are not same during the 10-week period”. Ex. 4. Of 160 off sprir7s of a certain cross between guinea pigs 102 were red, 24 were black and 34 were white. According to genetic model the probabilities of red, black and white are respectively 9/16, 3/16 and 1/4. Test at 5% level of significance, if the data are consistent with the model. For 2 degrees of freedom P(x’ >599)=005, Here k= No. of type of guine-pig =3. Take A (The probabilities of red, black, white are 9/16, 3/16 and 1/4 respectively); H, (the probabilities are not so as prescribed in genetic model). fo= observed frequency of the class in sample Se= expected frequency of the class in sample 2. pop eotl is © variate with 3-1=2 dof F wstine OF HYPOTHESIS 940 Computation of 7” , the compute value of x” = 5 a; #17 with2dof From the giver data the CR is x’ >599. Our computed value oes not lie in this region (as 3.76.99). Hence 4, is accepted at 5% level of significance and we conclude “ the data are consistent with the model”, Bx. 5. Survey of 320 families with 5 children each revealed the following distribution : No.of boys: 5 4 3 2 x: o% Noofgirls : 0 1 2) sls tare oa 6: Nooffamily: 14 66 110 68 40 12 fh the result consletent with the Ayptheste that female births are equally probable ? The 5% value Sofie 11.07. —t No.ofboys : 6 4 No, offamilies; 14 56 Here we think ‘a family with as the classes, So here | on ci 3-44 ENGINEERING MATHEMATICS. yy g 320 families, the expected number of Therefore, amon; ee : 8 =320%55%'G =10X'C,. families having r number of boy: Let f,= Observed frequency in the sample juency in the sample under 4, 2 The computed value of rh Ef -ns with 5 dof. From the supplied data we have the CR as x >1107. Our computed value does not lie in this region. So Ho is accepted at 5% level and we conclude “ the male and female birth are equi-probable”. i Ex. 6. Four dice were thrown 112 times and the number f times 1, 3 or 5 was thrown were as under : a No. of dice throwing : 0 1 2 3 j 1, 3, or 5 1 Frequency ste el0, 25 40-30 jgstiNG OF HYPOTHESIS 245 ind the value of chi-square presuming that all dice were fair. We assume all dice were fair, Let the event “ 1, 3 or 5” be called ‘success’ ., the probability of ‘success’ in a throw of a die is p= then “No.of dice throwing 1, 3, or 5’ means ‘No. of success” jn four trials. Therefore pe Probability of 0 number of success in 4 trials 4-0 5 f (LN (ld paeselds eof) (5) 16 n= Probability of 1 number of success in 4 trials nay “cG}@) -% p= Probability of 2 succes in 4 trials 2. 4-2 3, 4710 Nye lees “ic (LV (1)_ “af (ay _ sdk mH at) Let f,= observed frequency of the success f.= Expected frequency of the success. 2 p=pecel isa % variate with 5-1=4 dof Computation of x -variate Probability fe |fo-fe| (-S-)° =(3)x112| (3) (4) (5) } (6) 7 4 28 Y/16 3 3/8 42 2 V4 28 1/16 7 “1 =1843 with 4 degrees of freedom. EM-AV.26 | ENGINEERING MATHEMATIcg. Vv EXERCISES Snort ANSWERS QUESTIONS ther a coin is iearter shalll togs j imes. I shall reject the null hypothesis of perfectness ip Sony ifl get no head or 6 heads. What is the Probability of Type I error for my test ? . To test the unbiasedness of a die it is thrown six times oF is accepted that the unbiasedness if not more than one sixes are obtained. Find the probability of Type I Error, (W.B.U.Tech 2005 } . null hypothesis Ho(u=7)_ is tested against H,(y = - oee iiaiers mean and s.d. of 4 normal distributes Given o =2. The test is performed by drawing a random sample of size 25 from the given population and using the best cirtical region at 0.16 level of significance. Find the probability of Type II error. 3. The mean lifetime of a sample of 100 fluorescent light bulbs produced by a company is computed to be 1570 hours with a s.d of 120 hours. The company claims that the average life of the bulbs (distributed normally) is 1600 hours. Using a level of significance of 0.05 is the claim acceptable? 1. (a) In order to test whet 196 (Given [¢(*)at=4750, 4. A sample of 100 iron bars is said to be drawn from a large number of bars whose lengths are normally distributed with mean 4 ft. and s.d 0.6 ft. If the sample mean is 4.2 ft. can the sample be regarded as a truly random sample ? (Null hypothesis and assumptions should be stated clearly). Test 258 at 1% level of significance . (Given J¢(z)42=04951. 0 5. When a certain production machine is in perfect adjustment it produces bolts with a mean diameter 0.0600 inches and a s.d of 0.0150 inches. In order to ascertain whether or not the machine is still in adjustmen} a sample of 36 bolts is selected. The sample mean di is found to be 0.0575. Is the machine still in adjustment? (Use level of significance of 0.05) : [Hint : take Aj,( 4.06)] - psTiNG OF HYPOTHESIS 3-47 200 men out of 600 men and 300 men out of 1200 men in & two cities are highly educated. Do the data indicate that the two cities are significantly different regarding the educational standard of men ? [W.B.U.Tech, 2002] In a sample of 600 men from a certain city, 450 men are found to be smokers. In a sample of 900 from another city 450 are found to be smoker. Do the data indicate that the two cities are significantly different with respect to prevalence of smoking habit among men ? Test at .05 level of significance. Given z)5=196. g, Asample survey results show that out of 800 literate people 480 are employed whereas out of 600 illeterate only 350 are employed. Can we tell the scope of employment among literate and illiterate people are same. Test at 5% level. Given P(0) we have to test Ho(m= 2); m=480, x, =350] 9. In an industry A, 20% of a random sample of 900 razor- blades are defective. In another industry B, 15% of a random sample of 1600 blades are defective. Do the two industries produced same quality of blades ? Test at 1% evel ss pee Given Be J © 2 d=0495, 0 10. The mean yield of wheat from district A was 210 lbs. with s.d =10 lbs. from a sample of 100 plots. In another district B, the mean yield was 220 lbs with s.d. =12 Ibs from a sample of 150 plots. Assuming that the s.d of yield in the - entire state was 11 lbs, test whether there is any significant difference between the mean yield of crops in the two districts. Given’ P(|z|>258)=01. {Hint : Here population o is known. Take Ho(};=2) against Hi(u; #112) ) 11, Intelligence test on two groups of boys and girls gave the following results: F Mean S.D. ‘Number Boys: 70 20 250 Girls; 75 / 16 2 150 | 3-48 ENGINEERING MATHEMATICg py eee” 12. 13. 14. 15. Is there a significant difference in the mean scores obtaineg by boys and girls ? (W.B.U.Tech, 2004) [Hint : 6, =20, 6)=15 as the samples are large] A simple sample of heights of 6400 Englishmen has mean of 67.85 inches and a s.d of 2.56 inches, while g simple sample of heights of 1600 Australians has a mean of 68.55 and a s.d of 2.52 inches. Do the data indicate that Australians are on the average taller than the Englishmen ? Test at 1% level. Given Joc t)dt=01, [Hint : Since sample-size are so large we may take 6; =256, 07252. Take H(t: 226)=05] | Two independent random samples of size 8 and 6 having mean 38.4 and 33.7 are drawn from two norm! populations. The sum of the squares of deviation from ink respective sample mean are 20.8 and 15.7. Do you t ved the population mean of the first is the larger ? 01 =2.68 , fo95=3.06 for 12 d.o.f. a-49 ppsting OF HYPOT ‘Ten soldiers visit a rifle range for two consecutive weoks. Their scores are : glider A B CD FE F GH dl Jd ist Week: 67 24 67 566 63 64 66 68 33 43 gnd Week:70 38 68 68 66 67 68 72 42 38 Examine if there is any significant improvement in their erformance, Assume tho riflo-rango is normally distributed with same s.d, Given 19,282 for 9 d.o.f, ‘Two independent samples are drawn. from two normal ae populations with common s.d. The followings aro observed ; size mean variance First Sample: 9 600 121 Second Sample : 8 640 144 Test at 5% level of significance whether the two populations have same mean. Given f25,5 = 2.13. 18, What is the test statistic which is used ta test the equality of two means of two normal populations when their common s.d. is unknown, 19. A sample of size 10 is drawn from each of two normal populations with the same unknown variance and the following result are obtained ; Mean Variance Sample I ui 26 Sample] : 4 10 Test at 5% significance if the two populations have the same mean. Given the following data. Statistic d.o.f. Value at 5% level t 18 2.101 _ 1.96 z 2%. A die is rolled 60 times and the following results are observed ; Face-point: 1 2 8 4 & 6 Total No. oftimes; 6 10 8 13 1 12s «60 occured Are the data consistent with the thesis that the die is unbiased ? (Given x}, «1509 for 6 dogroes of freedom. 60 ak 22, 23. 24, 1 4 1, 8. 10. 1k. RNGINEERING MATHEMATICS. 1y rl f poa-br ing results In an exporimont of pea-breeding the follow! ary obtained, Round ead Yollow —= G16 j wrinkled and yellow, —101, Round and green = 108 ; wrikled and greon — 82, Total —-656, According to theory we should get the frequeney ratio as 913; 3 iL of pe above mentioned varletion. Examine at 6% lovol of significance whether (| rowult obtained in the experiment agree with the theory, i 0 7.816, 6% value of x? for 3 dott n at random from a pee ofa 200 digits from 0 to 9 are taken | a raaresson number table. The frequency distribution of the digit is given : Digit 2150, sedi earr8 S4VNO CP 87 Bie 9 19 18 21 16 26 22 20 21 9 Frequency: 18 Can this be regarded as random ? Given Xbs9 "1692, (Hint ; Random No. means drawing of every digit would bo equiprobable, i.e, probability of the digit 0, 1, 2, .., are Nigga: 10°10"10 ° In 60 throws of a die, face one turned’ up 6 times, face two or three 18 times, face four or five 24 times, and face six 12 times. Test at 10% significance level if the die is honest, it being given that P(x? >625)=0.1 for 3 degrees of freedom. In 360 tosses of a pair of dice, 74 sevens and 24 elevens are observed. Using 0.05 significance level, test the hypothesis that the dice are fair. Given 13, = 384 for 1d.o.f ANSWERS (a) 1/32 (b) 0.268 8. No; |x] =25>196 No. |x/=333 6,0.065 11. z=-10, Ho( = 006) is true =9,69 ; There is significant difference ., respectively } :=074;employment scope issame 9. No; z=32! r=-704, there is significance difference in the mean yields of crops. £=-284, ‘There is difference! at 6% level 12. yes #=-)9 pstiNG OF HYPOTHESIS 3-61 _-228, 1st population’s mean may not be smaller than 15: ond population's mean 14. No. ie. H(i #h2) is rejected ; ;=167 15. yes; 1=4.99 36, No. ¢=204 17. ¢=-674, not same _ (=) 18." TstamspV mm 1215, the two means are equal 20. ?=34; The die is unbiased Ik the t variate 19. gi. 7220515 experimental result agrees with the theory ga, yes; 77=63 23. yes, y2=30 24.x?=407; not fair pou LonG ANSWERS QUESTIONS 1. The proportion of defective items in a large lot of items is p.Te test the hypothesis p=02, we take a random sample ‘of 8 items and accept the hypothesis if the number of defectives in the sample is 6 or less. Find the probability of Type I error of the test. What is the Type Il error if p=03? (W.B.U.Tech.2008} 2, Let p dente the probability of getting a head when a given coin is tossed once. Suppose that the hypothesis Ho:p=05 is rejected in favour of A,:p=06 if 10 trials result in 7 or more heads. Calculate the probability of Type land Type II error. 1 3. Given the density function (%8)=5, Osx =0, elsewhere and that you are testing the null hypothesis f:6=1 against H,:0=2 by means of a single observed value 4. Determine the size (i.e. probability) of Type I and Type Il error if you chose the interval 05 (Assume that the breaking strength of ateing is normaly, distributed), Tost at 6% lovel of significance, gor appropriate atatiationl table, (eRe Dake My(es t8o) and 44(WS186)] HK The wages ofa Metory workers assumed to bo normal) Giatributed with mean m and variance 26, random sample of 28 workers gives the total wages equal to 1959 wnits, Test the hypothesis we 52 against the altornatiys mw=dd at 1% level of signifloance Give SP 00 [Given 7} fe Fqreoor) & A company manufactures car tyres. The lives of the tyres are normally distributed with a moan of 40,000 kms, and sd of 8,000 kms, A change in the production process ig brought. A test sample of 64 new tyres has a mean life of 42,200 kms. Can you conclude that the change produces deter quality of'tyres ? Test at 5% level. Soe statistical table, YA new printerhead is introduced into the market. It is claimed that it has an averago life of 200 hours with sd of 21 hours. The claim came under severe criticism from dissatisfied customers, A customer group tested 49 such heads and found that they have an average life of 191 hours, Is the claim justified at 1% level of significance ? See table, S& The sample mean of a random sample of size 10 is 12.1. Given that the s.d_ of the population is 3.2. Can you conclude that the sample comes from a normal population with mean 14.5 ? Test at 5% level of significance. Given Zag = 196 (Hint : Take Ho(u=145)) 9 Is it likely that a sample of 300 item whose mean is 16.0 is a random sample from a normal population with mean 16.8 and s.d 5.2? Test at 0.01 level of significance. (Given area under st. normal curve enclosed between :=0 == 258 is 0.4951)(Hint : Take o(\t= 168) ) 10. A random sample of 900 members is found to have a mean of 4.45 cm. Can it be resonably regarded as a sample from # large population whose mean is § cms and variance is 4 ¢™- (Given area under # curve from g = 2 to 2 = 1:65 is 0.95) panvnrOR ety AODHARIY ann (nin take HQ @ 188), OR ta fe] 1.08 at 10% Level \ Awad wanutieturiig Company wae dlaleibutiny a par toular rand of aoap Uioowal a laege number ofvetall shop: Helare a heavy advertiaanont oampalgn, the mean gales per \wovk per ehop wae 140 dowena, After the eampalan, & aanupte of BG ahopa War taken and the mean salen waa Araned to be 14 dogene with ad 14, Gan you eonalder the advertiaement eotive? tare tt for 86 dot) (rkint (Qh = 140), AQ 10), 1) fe rejected then 2 will pe acepted Le, “the advertisement ta affeotive”, Here ye do, amall (aa it <0) 1a Acertain diet newly tntrodueed to each ofthe pigs resulted in the following {norese th body weight : @.9,8-9) 8,0, <4, 4, G0, 6 and 4 Can we conclude that the diet ia effeotive in inereaaing the weight of the pigs 2 [Given fy 2220) (Mints Rake f( 20) y(n 0)) 1% A drag iz given to 10 pationt and the inereage in their hlood preaure were recorded to be 8, 6, =, 4, -3, 4, 6, 0, 0, 9, Jat resonable to believe that the drug haa no effect on change of blood preaure ? (5% value of t for 9 dof tg 296) 14 A manufacturer claimed that at least 95% of the equip wents whieh he aupplied to a factory conformed to specifi cation, An examination of a sample of 200 pieces of equip ments revealed that 18 wore fuulty, Test his claim at a significance level of 0,02, [Given Is ol) (Hint : "Take 1u{ poe noss}an(P so) a 18 Ly a hospital, 19 fomates and 168 males were b month, Do these figures conform to the hypot Be Gaey sexes are born in equal proportion? [Given ENGINEERING MATHEMATICS. ry 3-54 — eee 1 1 (Hint: Take n= 5) 0(P* sh 16. In a sample of 400 burners there were 12 whose internal diameter’s were not within tolerance. Is this sufficient ey; dence for concluding that the manufacturing process is turning out more than 2% defective burners. Test at 5% Les e =e level of significance. (Given Jo=-475) > 2 (Hint Take Me( P= 735 -0-02} 1(P > 0:02)) 17. Ina random sample of 400 persons from a large popula tion, 120 are females. Can it be said that males and fe males are in the ratio 5 : 3 in the population? Use 10% level of significance. [Given fo=os] 196 3 int : =—>_ =0-375],H,(P 40-375 (Hint: a ? a ) Al 1 18. A bottle manufacturing process is ‘under control’ ifno more that 1% of the bolltes are defective. A random sample of 120 bottles showed 5 to be defective. Do these data indi cate that process is out of control ? Use 5% level of signifi cance. [Given zo; =1-64] 19. In a sample of 400 parts manufactured by a factory, the number of defective parts was found to be 30. The com pany, however, claim that only 5% of their product is de fective. Is the claim tenable? — [Given P(z>1-645)=-05] (Hint : Take Ho(P=0-05),H,(P>0-05)] ESTING OF HYPOTHESIS 3-55 a — 90. Adie was thrown 400 times and resulted 80 times. Do the data justify the hypothesis that the die is unbiased ? [Give fo=0-05) 196 1 {Hint : Take a{P=2); g1. In a sample of 600 students of a certain college 400 are found to use bi-cycle. In another college from a sample of 900 students 450 were found to use bi-cycles. Test at 1% level whether the two colleges are significantly different with respect to the habit of using bi-cycle. Given area under standard normal curve enclosed between the ordi nates z=-258 and z=-sis 0-005. 22, Before an increase in excise duty on tea, 400 people out of a sample of 500 persons were found to be tea drinkers. After an increase in duty, 400 people were tea drinkers in a sample of 600 people. State at 2.5% level of significance whether there is a decrease in consumption. of tea due to increment of excise duty. Given 1 fe e 2 dz=047 a Je 0.475 | 23, In a year there are 956 births in town A, of which 52.5% were males, while in town A and B combined, this proportion in a total of 1406 birth was 0.496. Is there any significant difference in the proportion of male births in the two towns. Test at 5% level. See appropriate statistical table. 24. A machine produced 20 defective articles in a batch of 400. After overhauling it produced 10 defectives in a batch of 300. Has the machine improved ? Test at 5% level of significance. Given area under standard normal curve between the ordinates z=1645 and z= is 0.05. 25. There are two populations, population I and population IL. 8 and 6 are standard deviation of the two population respectively. Two independent samples of size 10 and 12 are drawn from the two populations respectively with sample means 20 and 27. Test at 1% level of significance whether the two population means are equal. Given P(|z|>258)=01. 3-56 ENGINEERING MATHEMATIC8. ry —————— _ 26. In a certain factory there are two different processes of manufacturing same item. The average weight in sample of 250 items produced from one process 18 found to be 120 grammes with s.d of 12 grammes ; the correspondin figures in a sample of 400 items from the other process are 124 and 14. Is this difference significant ? Seg appropriate statistical table. [ Hint : For large samply assume o,=12, 6,=14 and so the test statistic z.] 27. Two random samples of size 500 and 400 are drawn from two normal populations having same 8.d. 5. It is observed that the two sample means are 11.5 and 10.9 respectively, Gan the sample be regarded.as drawn from the same population. See appropriate table. 28. Intelligence tests on two groups — one group consisting 121 girls and the other group consisting of 81 boys gives the mean 84 and 81 respectively. The intelligence are normally distributed with s.d. 10 for girls and 12 for boys respectively. From these observations can we say the average intelligence of boys and girls are same ? Test at the level which can fit the data All| <196)=95. 29. From two normal populations two independent samples of size 30 and 55 are drawn. The two populations have a common s.d. 4.195. The means of the two samples are seen as 23 and 21.9 respectively. Test at 5% level of significance whether the two populations have also the same mean. Given P(z>196)=0.025. 30. 10 and 25 observations are drawn at random from two populations respectively whose variances are 9.61 and 7.29. It is found that the means of the two set of observations are 23.0 and 20.3 respectively. Test at 1% significance level the hypothesis that the mean of the Ist population is larger. 31. Two sets of ten students are selected at random from & college. One set was given memory test without any training and the other set was given the same test after two weeks of training. The scores obtained are given in below : ; Bet'A’:°108 7 Oe NB NOLO 96) 207908 SetB:12 8 8 10 86 11 9 8 9 9 6 STING OF HYPOTHESIS 3-67 Do you think there is any significant effect of the training 2 (Given fos = 2.10 at 8 d.o.f) two working designs are under consideration for adoption ina plant. A time and motion study shows that 12 workers using design A have a mean assembly time of 300 seconds with a standard deviation of 12 seconds and that 15 workers using design B have a mean assembly time of 335 seconds with a s.d of 15 seconds. Is the difference in the mean assembly time between the two working designs significant at 1% level of significance ? 33. The incomes of a random sample of engineers in industry A are Rs. 630, 650, 680, 690, 690, 710 and 720 per month. The incomes of a similar sample from industry B are Rs 610, 620, 650, 660, 690, 700, 710, 720 and 730 per month. Discuss the validity of the suggestion that industry A pays its engineers much better than industry B. Test at 5% level. (Given fos=2.145 at appropriate d.o.f mentioned by you. 94. Sample of sales in similar shops in towns A and B regarding a new product yielded the following information : FortownA : ¥=345,)x1=38, Dx? =228,m=11 FortownB : %=444,)2=40, )x$=222,m=9 Is there any evidence of difference in the sales in the two towns ? Test at 5% level. The value of ¢ at 5% level for 18 do.fis 2.101 ye ‘ Ex 2 Ss int sie. P(|t|> 2.101) =05; Ss? = [Hint :ite. P(|+| ) =05; 5} i a 35. A company claims that its light bulbs are superior to those of a competitor on the basis of a study which showed that a sample of 40 uf its bulbs had an average life time of 628 hours of contfnuous use with a s.d of 27 hours, while a sample of 30 bulbs made by the competitor had an average life time of 619 hours of continuous use with a s.d of 25 hours. Check, at the 5% level of significance, whether this claim is justified. 36, 5 identical coins are tossed 320 times, and the number of heads appearing each time is recorded as follow : No, ofhesds:(0° @1 2) Sit Oe Coral Frequency :14 46 80 112 61 8 $20 3 32. 79 etc.) 3-58 ENGINEERING MATHEMATICS. ny, 5-saaieaiaidaisiaaeiaaiianalaiaiaiamaaenncaaneaMadesiniiiimiaaae 37. (i)Are the coins unbiased ? Given 3, =1107 and xy = 1509 for 5 degree of freedom. - (ii)Is the sample from a binomial population (given P(x? >1107)=005 for 6 d.o.f) A sample analysis of certain examination results of 9 atudeiits in ayant was made. It was found that 49 students failed, 68 secured third division, 62 Secured second division and rest were placed in first division, Are these results commensurate with the annual examination results which are in the ratio of 2: 3:3: 2 for above various categories respectively ? (The tabulated value of chi-square for 3 d.o.f at 5% leve] of significance is 7.81) 38. Test whether the following distribution fit the Poisson 39. distribution with mean 0.6 = -eanO) 1 2 3 4 5ormore frequency : 140 75 25 6 3 1 Given e-*= 05488 , 1453=781, x}s4=949, Xh55= 1107 A bird watcher sitting in a park has spotted a number of birds belonging to 6 categories. The exact classification is given below. 4 Category: 1 23 4 5 6 Frequency : 6 7 13176 5 Test at 5% level of significance whether or not the data is compatible with the assumption that this particular park is visited by birds belonging to these six categories in the Proportion 1:1:2:3: 1:1. [Given P(x? >11.07)=008 for 5dof] The that occured during the various days of the week. Test whether the accidents are uniformly distributed over the week using 5% level of Significance. Day : Sun Mon Tue WedThu Fri Sat No. ofaccidents ; 13 14 18 12 11 15 4 Given (2>1258)=005 for 6 d.o.f [W.B.U Tech 2005,2008 | 165 4l. ING OF HYPOTHESIS 9-50 Hint : If the accidents are uniformly distributed of over the week then No. of accidents will be equal on every day, 97 97 97 Les pa The random discrete variable has the following distribution x sO tiled san 8 wae f : 80 6246 10 2 Use the chi-square test to determine where X follows pinomial distribution with p=032. Test at 5% level. 42. The following figures show the distribution of digits in the 2. 10, number chosen at random from a telephone directory : Digts : 0 1 2 Sea ene Frequency : 10261107997 966 1075 933 1107 Digits : 7 8 9 Total Frequency: 972 964 ‘ 853 10,000 « Test whether the digits may be taken to occur equally frequently in the directory. Test at 5% of significance. Give P(z2>16919)=005 for 9 d.o.f. (Hint: f,: 1000 1000 1000 ... ... ete.) ANSWERS 0,00008448 ; 0.99870967 Fl 0382= 0618 3. 0.5,0.25 4. No; += -032 2=-2; m=52 is true at 1% level 6. yes 7. No; |z[=3 No; ||=237 9. No; |z|=267, : e|=8-21; the population cannot have méan 5 cm 11, ¢=2.19. My is rejected f 12.1=3.01>2.20,The diet is effective in incroasing body Weight ' ‘ 3-60 ENGINEERING MATHEMATICS. py eee 13. t=2<2-26.The drug has no effect in change of blood presure 14, Computed value of z = -2:597 . Supplier's claim is not valiq 15.Computed value of 7 = 2-08 ‘The sexes are born not in equal proportion 16. Value of z= 1.429, Manufacturing process is not turning out more then 2% defective burners 17. value of z = -3-25.Males and females in the population are not in 5 : 3 ratio 18. Value of z= 1-69; Rejected 19.Value of z=2:27, Company’s claim is rejected 20.2 =1.79. The die is unbiased 21. yes ; |2|=638 22, yes; 7=494 23, :=337; There is difference in A and B 24. not improved, z=108 25. z=~—228, two means are equal 26. -=-39 Reject Ho(\i=H2) at 1% level 27. yes ; ==179 28. 2=186,No 29. yes, z=116 30. yes, 7=241 31. noeffect 32. ¢=637 33. ¢=0099; suggestion not valid 34. ;=-078 ; no difference in sales 36. (i) x2 =1036; coins are unbiased (ii) x? = 1036; the sample is from binomial distribution. 37. 72 =843,No 39. x? = 0.472 ; assumption is correct 40. x2 =233, accident are uniformly distributed over the week. 41. The fit is good 42. 7? =5854; Note equally frequently occurence. (on) MULTIPLE CHOICE QUESTIONS 1. Consider the normal population of life times of tyres manufactured by a company whose mean and s.d are unknowD. Then the assumption average life time =3600 kms is 2 (a) simple hypothesis (b) composite hypothesis (c) alternative hypothesis _ (8) none of these. rs qestING OF HYPOTHESIS 3-61 Consider the normal population of the body weight all ech students in West Bengal whose standard deviation is 5, Then the assumption "the average bodyweight = 50kg" is a (a) simple hypothesis (b) composite hypothesis (c) alternative hypothesis _ (d) none of these. 3. Simple hypothesis does (a) specify the population completely (b) not specify the population completely (c) simplify the calculation of parameters of the population (d) none of these. 4, If m be the mean of a population having t distribution. | Then the assumption m= 20 is to be tested. Then this is a (a) alternative hypothesis —_(b) simple hypothesis (c) Null hypothesis (d) none of these. 5. Analternative hypothesis (a) is same as the null hypothesis (b) different from the null hypothesis (c) may or may not same as the null hypothesis (d) none of these. S 6. If (s=3) be the null hypothesis then which one of the following is alternative hypothesis (a) (6=4) (b) (5=1) (©) (s=0) (d) (6 #3) 1. If u isa parameter and H(t =5) is null hypothesis, then which one of the following is Left sided alternative hypothesis (a) H(u#5) (b) H(u<5) (©) H(u>5) @) Huss) EMIV.27 3-62 ENGINEERING MATHEMATIcg py ee 8. If is a parameter and H(u=7) is null hypothesis, the which one of the following is Both sided alternaive hypothesis (a) H(w#7) (b) H(u=8) (c) H(u<7) (d) none of these. 9. If H,(u>60) be an alternative hypothesis then the Nut hypothesis is (a) H(i < 60) (b) Ho(u 2 60) (c) Ho(u < 60) (d) Ho(u = 60). 10. To test the acceptance of the null hypothesis (population mean =]0) for a population having normal distribution with s.d 4 the test statistic is (a) sample s.d (b) sample proportion (c) sample mean (d) none of these. 11. If x be test statistic and (a,b) is region of acceptence | correspending to 3% level of significance than P(a39 © 0<1<39 (d) none of these. 14. In a test of hypothesis if (~4,6,68) is region of acceptane® then the Null hypothesis is rejected if the computed value of the test statistic is (a) 5 (b) 6 (c) -3 (d) 6. esTING OF HYPOTHESIS 3-63 15. ‘Falling of the computed value of the test statistic critical region’ = complement of the even ‘Falling of the computed value of the test statistic in the region of acceptance. 16. Ifina test of hypothesis (—~»,1.02) is critical region at 01 Jevel of significancé then, if¢ is test statistic, (a) P(0<¢<102)=0.01 (b) P(102a () t2a @) t=a- 18. Ifthe alternative hypothesis is taken left sided then indicate which one of the followings may be a possible critical region @ta © t2a @) t=a 19, Ina test of hypothesis Type I Error is committed when (a) Null hypothesis is rejected though it was really false. (b) Null hypothesis is rejected though it was really true. (c) Null hypvthesis is accepted though it was really false. (d) Null hypothesis is rejected though it was really false. _ 2, Ina test of hypothesis Type II error is committed when (a) Null hypothesis is rejected when it was really false (b) Null hypothesis is rejected when it was really true (c) Null hypothesis is accepted when it was really false (d) none of these. ( W.B.U.Tech 2007) 21, Ifin a test of hypothesis the null hypothesis Hy is accepted at 2% level of significance then the probability that the Statement in H, is false is (a) .2 (b) .02 (c) .98 7 (d) none of these. 3-64 ENGINEERING MATHEMATICS, py 22, If Ho(¢=5) where tis test statistic is null hypothesis they probability of Type I error = ' (a) Probability of acceptance of Hy assuming ;=5 (b) Probability of rejection of Ho assuming ¢ #5 (c) Probability of rejection of Ho assuming ;=5 (a) Probability of acceptance of Ho assuming (#5. 23. If H,(@=2) is null hypothesis then the probability of rejection of Hy though @ =2 is true is probability of (a) Type I error (b) Type II error (c) Probability of #2 (d) none of these. 24, Ina test of hypothesis Probability of Type I error is same as level of significance of the test. (a) Yes (b) No. 25. If H,(@=10) be null hypothesis then the probability of Type II error = (a) Probability of rejection of Hy assuming 6 =10 (b) Probability of rejection of Hy assuming 6 410 (c) Probability of acceptance of Hy assuming § =10 (d) Probability of acceptance of Hy assuming 9 +10. 26. Ina test of hypothesis corresponding to a particular level of significance, among all critical region ‘Best critical region’ has (a) least Type I error (b) least Type II error (c) greatest Type I error (d) greatest Type II error. 27. Ina test of hypothesis the critical region corresponding to a particular level of significance is unique, (a) True (b) False. 28. Ifa null hypothesis is accepted at .05 level of significance then this decision is " (a) 5% correct (b) 05% omen (b) .95% correct (d) 95% correct. ppsTING OF HYPOTHESIS 8-66 39. AS the probability of Type II error decreases the goodness of the test (a) decreases (b) increases (c) does not chang (a) none of these. go. To test the hypothesis Ho(}t=2) where jt is the mean of the normal population with s.d 0.1 samples of size 100 are drawn. If P(2.32232 (b) 7<232 (c) 0 2.33 be the critical region then the null hypothesis is accepted if the mean of the drawn sample be (a) 1850 (b) 1700 (c) 1900 (a) 1820, $2. In ‘Goodness of fit’ which of the following is used as test, statistic (a) normal variate (b) t variate (c) Poisson variate (a) x" variate ANSWERS lb 2a Sa 4c 5b Gd Tb 8.a 9d 1c Ib 1a 18d 14d 15b 16d “la i8a 19b 20c 21b 22c 28.0 24a 2d 26.b 27b 28d 2%b 30a Sid 82d

You might also like