Scribd
Upload
37 views

Unit 2

Uploaded by

Dynamic Clothes
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
37 views

Unit 2

Uploaded by

Dynamic Clothes
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/ 15
‘An infinite population is one whose size is unlimited and, therefore, its number fant be counted. (Example: no. of fishes in river, no. of microbes inthe sol is termed infinite population ( Population i) Population (i) Population a ae ee en Finite Infinite Countable Uneounible Existent Hypothetical Existent Population: The population of concrete existent objects willbe called a existent population. Hypothetical Population: The population may aso be hypothetiealy constructed Ex coin tossing HL Sample survey method. Sampler A sample isa selected part of observations that is taken rom some souree of ‘Observations for getting observation about the source Somplings isthe acl, process of technique cf selecting a suitable sample or Topresenative part of 2 population for the purpose of determining parameters average o posses the characteristis ofthe 04D, means if a sample is drawn from a population i porulation {b) Principle of ‘Inertia of Large Numbers’ eooeetatey that, other thing being equal large the size ofthe sample, more accurate Tenet ae likely tobe, This because large number are more sable as compare to small one. vantages of samplin {treduced cost {iy Takes less times (ii) Greater speed (colletion and summarization quicker) is) Greater scope (less no. of trainee require) (6) Greater accuracy the same characteristics as in {i) A sample survey must be eareilly planned and executed: atherwise the reslis tbfained may be iniceurate and misleading (i) In the absence of qualified and experienced persons. the information obiained fiom sample survey cannot be relied upon. (Gi) If the sample size i larg proportion of toal population. the sampling plan may the complicated and. it requires more time, labor and money. than complete is) The information is required foreach and every unit inthe domain of study compete enumeration is necessary bias and error in sampling Generally. a simple isthe representative of the population from which i comes. But there fs to guarantee any sample willbe precisely representative ofthe population. In practice, there isan unfepresenative ofthe sample and a discrepancy always exits Fetwcen sample and its parent population. This diference Js termed ax “error”. Two ‘uses of eros {i Sampling error and {iy Non-sampling error () Sampling error sampling eroris the difference between a parameter value and sample estimate of the character which occurs when sampling is dane a population, (The eror arising due to drawing inferences about the population on the hasis of few observations {Gample) is termed as sampling eror:) There are to cases of sampling errors ~ (2) Chance or unbiased error and () Bias or non-sampling eo. (@) The fist type errors arse due to chance diff population includ inthe sample and those not include alte second type errors arise from any basin selection, estimation et. 11 the fesalt of poor sampling plan. Bias may arisen several ways: {i) Bias may arse selecting sample itself, {id Bins may arise hecause of substitution of selecting unit by another Bias may arise [rom incomplete investigation: (Gu) Bias may arise in the process of estimation of prameters (0 Agtin bias may arse due 10 negligence ot carcessness om the part of {Restantors in asking qvestons or inaeeuraies in recording or false resorting by Aishonest investigators without atually visiting the simple ronces between the members of inpling error Mi du Tctrs other dan the inductive proces of inferring boat the population from a sample. Such errors can arse 4 to defective method of data PeMention and tabulation, faulty definition. inomplet coverage of the population oF imple et. It arses in case of complete enumeration method Tin sample survey non-sompling errs may also arise due to defective frame and faulty selection of sampling wits) =. Subjective ims Ay type oF sampling depends ‘upon the persona judament Aiseretion of the Simpler Method of Sampling Onjecive Sampling method whichis Fixed by sampling rule ors indepewent ofthe samplers cn juement ee Mixed: the selection of fist unit made in such Probabilistic: 1° tor || Non-probabitistc cach unit there is a] | Thereis «fixed sampling | | manner tht cach 10 gets Getinte presssigned | | rvle but there no | | equal chance of being elected. | probabiity of being | | Pebbliyataied tothe | | becomes case of mixed Selected trode of selection sampling partly probabilistic a and paily on- probabilistic | = + yee Quote Porpive | Doub Stating Simling | Sampling Siratfied | | Chs Malt: Random || Sampling || staze Sampling Sampling {ses ] SRSWOR ][Pronoctioat | [_Nov-wrportinat 1 Prabal tie Sampling: Te Simple Random sampling (SRS): Iter the sm vein which each and aoa eet the population fs given an equal chance of being included in the sample random samp! Simple random sole are drain from the Sopulaton one by one, and afer each drawing the selected population unit Pofed and then returned tothe population before the next one is drawn. Suppose, N'=3 the sizeof population (numberof individual) ‘nd => the sizeof sample Sine kame individual ofthe population may oesur more than once inthe sample. Sampling is std fo be "SRSWR” w (in which 1 there are N’ possible samples and each has the probability W samples nit appear) ‘Ranin, each drawing members is returmed to te population, then number of the renin ete oily tba ed Example, elements of population af€2, 12,22, 32; e.,N=4. Suppose n = 2. total umber of samples 4° ~ 16 and the possible samples are: (2,2); 2.12): (2.22): 2.32), $(8232)Total = 16, {zai since each drawing member is returned othe population then numbers of eso gee oaiy oti let ce pois 0 individual in &* drawing is (b) Simple random sampling without replacenent (SRSWOR Simple random ae ese SRSWOR when ier sample member are drawn all Sampling am one B) one in such a manne tet afer each ding the aa pt ie not retuned to the poplaion when the next one fs ran Sees N dened the ic of population (nib of inivdual) and n denoted the sive of sample (amber of sample tobe dawn, then there are ("msi ae mee ‘Recall example 1 and explain. ‘Method of drawing SRS. {a) Drawing chips from bowl {hy Using random sampling sumbers: random sampling number series is an ammangement which may be looked upon ether as ined of as rectangular in which ‘each place has filled in with one ofthe digits 1.2.3.4.5.6.7,8.9. Advantages: {We need not construct a miniature population. = i) Randomization of the numbers beng dane once fo all the tious proves randomization of the miniatare population each time before the next dev Ismade fs no necessary, (Gi) Arandom sampling number series can be used for any enumerable population so that series of random numbers has a wide range of applicntion. Example: Draw random sample of size 10 without replacement Irom a population of 121 bays numbered fom Io 12 Solution: We shall take thre digit numbers from the table of random number (since Population size is thre digit number) and we shill take the numer from Oe ‘968 (908 isthe greatest three digit umber mulipleof 121) and shall ignore the ether lee digit number. We shal divide the numbers by 121 and ake the remainder. The ‘remainder of cours, varies from 000 to 120 The remainders O01 to 120 wil be ten ‘o correspond tothe boys with the same numbers, where asthe remainder 000 wil correspond tothe 121 bo The selection is dn in the following lable Table 1: Showing the seletion of a random sample of size 10 without replacement from population of'size 121 boys. ‘iimber taken" Remainder when | Serial of he le | divided by 121-| bay selected _| —102 102 1, Stratified Random Sampling: Here the population subdivided nto several called strata and then a sub-simple is drawn from each of them. All the subsample combined together gives the strafed sample. If the ceston from stita ts done by ‘and sampling the mes is Known as ratified rom sampling = Be [LM Se] Between strata dissin abt HPO P| ses onan Irmay be of two types io) proportionate Stratified Random Sampling: tn this method the resercher eons Folation according to Known characteristic of the population and siasenty randomly draws the individuals in a similar propotio from each see eT the population. For example, al no of students of an Institue is 1000 aera hich 600 are B. Se, 300 are M. Se. anc 100 are Ph.D. Ifthe sample size ym B. Se. are 160 = 96and 50 on 1000 is 160, then students to selee () Non-proportionate Stratified Random Sampling In his sampling method the Simpler drawn from each stratum are not nebessarily distributed according 10 FP toportion inthe population from which they ae randomly selected. For Serle tata no. of stents of an Insitute is 1000 out of which 700 are male SAI'¥oO are female. IT any researcher selects 100 students both the males and females randorsly with an equal number (ie. 50 each) it will constitute = non proportionate stratified random sample {ML Claster Sampling: Ths is random sampling method in which the sampling unis ire not individual slements of the universe, but groups of the elements or cluster Example: Sau area problem of rural school going children. Selection: He children of rural schools in the dstiets them the researchers sleet at random 10% ofthe rural schools inthe district and use as sample al the children attending tose schools, ccnitd_[ Adults Parent Pupils Between strata similar attribute Within strata dissimilar attributes 1V Mutstage sampling. This is foam of strated random sampling method in whieh ‘ampling is done from diferent strata at different stages Peample ina statewide std. small bu repeseniative sample may be selected as Stage - Sile Stage IL Several istries Selected at random (or purposively) Stage IIT fixed nos. of blacks selected from the districts at random (oF purposive) Si IVA Fixed nos. “i from the locks at andom (or purposively) Stage-V A fixed nos, of households selected from the villages a random (or purposively) 2. Non-probabititc voting ‘This is sampling procedure in which no way of wssewing the probability of the clement. ‘of population being inchade nthe sample ‘There ie severl technique of dressing the non-probabstic sampling methods: i Quota sampling: In quota sampling, sampling is done ws per the fixed quota In this method. a sampling quota is fixed oprior! for diferent strat of » heterogencous population and the enumeraors are instructed to continwing sampling unl the noeessn 100) clang to upper case, 7008 a 2000 individuals belong to middle and lower class respectively. If the investigoior wants to. selest 1000 individuals and finally selects 100 vidas from upper-class. 700 from middle elas and 200 from lower class, according to conve not random). tis constittes quota sampling Purposive sampling: Purposive sampling is done ith a purpose. which "means tht slestion of sampling units i purposive in nature In this sampling Procedure (also called judgment sampling) the sampler ses his om” siseretion in selecting a represeniative sample Irom the population. Snowball sampling: Snowball sampling is generally used in the case of ‘explorative ressarch study/design, where researchers do not have moch led information, This is essentially a socio metric sampling technique for study of | Sal ups, All the penons ina grap or ogaiaton ine thee ends ho int en th i dasa te P coverage into some type of a definite social pat ef thf sec hangs son rma tr sooty or social organization, the other members ofthe sample are fixed depending onthe chee ofthe first member isis als called uasi-rndom samplings Fundamental concepts nee anal measure calculated on the bass of sample abservatins is called a statistic. (py Parameters Any sisal measure cafelted on the asi of al nis inthe population scaled parameter pop proportion aw moment 1g) = ray moment {o Standard error. The standard deviation of the sampling distiution of any ae crc a speial name of Sandan err 11s the measure of warily sine sttetie fom sample to sample. For ecample, standard error of mean, 0, = ,SRSWRand 3, = S. Nn Sampling Distribution: Sa ir the sampling sty isin the distribution of statistic across all possible Reps stn servations from population The cone of sampling sistbuion must he examined so tat probabil seek ample ba selected prt of observations fat is taken fom some source of Shsetvations for getting obseration about the sour, seine Wis the. ak, proces or teewique of selecting» suitable sample of Serr separ of a poration forthe purpose of determining parameters of ‘Characteristics ofthe whole population Sprig Disuion: A sompling dsb i a theoresicldstibuion tat sho ar pal elation between the possible values ofa given stats based on a sample aren the probability (density) associated with each valve, for all posible ‘amples of sizen dass from a particular population. Expectation and standard error of sample meat: Suppose N => the population size identifying from 1 N > a member of the population. = 1.2.3. clraven from the population, X23" member ofthe sample (i= 1.2, ‘Thus the population members are Xi, No, ‘The population mean a Xsan sample members ae. avlance are defined. respecively as allows: Under random sampling — Vail and Yar(s,)= Els, - 1 Ser. —p'abs = yar, 1 x 5 puting the value o (in 4) ve Et a= SH 00) ante vale of )in() Vert) =| S070) ml ay Case II-SRSWOR teaeae of SRSWOR, the drawings are fot independent and $0), -¥ovu%,8T6 Rot independent citer. For instance, we hove th <3) Pls, =X): = X_]=0 foram a ile pe sfe = kale O tore! To find the expression fr Bs }ndFar(3) note that 1 Pl, = X]*5y Pha =X) Srl = Koon ed ara Since in ole tha the 2 drawing may veld theat meniber the 1 dang must yield one tthe ther members. Simi Pls, =X.) N. a2) hE im. Bi etude Pl = Ply ote Varts,) = Bt, = D(X, — Ph = Kad er x, Ny =a foreach >) * ant Cov, .#,) = Elle, Ne, — 19 SUX, 1g PLR = [Exe mer -9-¥ op.) wna 4 Dice - ny -ad=[ 0 }-m oie andar ervor af sample proportion: ‘with N members. ffs the pepalation proportion with character he population with character A is NP and if isthe

You might also like