Multi Stage Sampling

This is  a modified  form of cluster sampling. While in cluster sampling  all the units  in a selected cluster constitute the sample  in multistage sampling  the sample units are selected  in two  or three  or four stages. In  this system the  universe is first divides into  first stage sample  units from which  the sample  is selected. The  selected  first  stages samples  are then  sub divided into second  stages  units  from  which  another sample  is selected. Third  stages  and fourth  stages sampling  is done  in the same meaner if necessary. Thus for an  urban  survey  a sample  of towns  may be taken first and then for each of the  selected town  a sub sample of households may be taken  and then  if need be  from  each of the selected  household  a third  stages  sample  of individuals  may be  obtained.

Area Sampling 

If  clusters happen  to be some  geographic  subdivisions in that case  sampling  all the units  in a  selected cluster  constitute  the sample in multistage sampling  the samples  units are selected in two or three or four stages. In this system the universe is first  divided into  are then  sub divided into second  stage units from which  another sample  is selected. Third  stage  and fourth  stage sampling  is done it  the same  manner if  necessary. Thus for an urban  survey a simple  of towns may be taken  first  and then  for each of the  selected town  a sub  sample  of households may be  taken  and then if need be from  each of the  selected household  a third  stage sample  individuals  may be obtained.

There  are two merit  of this sampling design viz,

a.It is easier to administer than most  single stage design mainly  because  of the  fact that  sampling  frame  under multi stage  sampling  is developed in partial units.

b.A large  number  of unit scan be  sampled  for a given cost  under multistage  sampling because  of sequential clustering whereas this  is not possible  in most  of the  simple  designs.

Advantages of Probability Sampling

a.Probability sampling is the only  sampling methods  that  provides  essentially unbiased estimates having  measurable precision. If  the investigator requires  this level  of objectivity then  some  variant  of probability is essential/.

b.A second  advantage is that probability sampling  permits the researcher to  evaluate in quantitative terms the  relative efficiency of alternate sampling  techniques  in a given situation  usually this  is not  possible  in non probability  sampling.

c.A third  major  advantage is that  probability sampling  requires relatively little  universe knowledge:  essentially only  two things  need to be known.

a. A way of identifying each universe element uniquely and

b. The total  number of universe element.

If  in area sampling a correspondence can be established between area  units and universe items and area maps  are available then  probability  sampling  can be  used. In  contrast non probability  sampling  methods  require  substantial information.

Disadvantages of Probability Sampling

a. It is relatively  expensive than   non  probability sampling.

The methods  of selection in many  case can be  complex  and time  consuming. Specially in the  cases of  marketing  research