•
RANDOM/PROBABILITY SAMPLING
METHODS OF SAMPLING 2
BUSINESS STATISTICS
RESEARCH METHODOLOGY
MARKETING RESEARCH
METHODS OF SAMPLING 2
BUSINESS STATISTICS
RESEARCH METHODOLOGY
MARKETING RESEARCH
•
RANDOM SAMPLING METHODS
1.
SIMPLE RANDOM SAMPLING
2.
STRATIFIED SAMPLING
3. SYSTEMATIC SAMPLING
4.
MULTI STAGE SAMPLING
•
SIMPLE RANDOM SAMPLING
1.
REFERS TO THE SAMPLING TECHNIQUE IN WHICH EACH AND EVERY ITEM OF THE
POPULATION IS GIVEN AN EQUAL CHANCE OF BEING INCLUDED IN THE SAMPLE
2.
SELECTION OF THE ITEMS ON CHANCE ,THIS METHOD IS ALSO KNOWN AS THE METHOD OF
CHANCE
3.
FREE FROM PERSONAL BIAS BUT CONSCIOUS EFFORTS IS MADE TO ENSURE THE
OPERATIONS OF CHANCE FACTORS
4.
ALSO KNOWN AS REPRESENTATIVE SAMPLING AND IF THE SIZE OF THE SAMPLE IS LARGE
IT WILL BE REPRESENTATIVE OF THE GROUP
5.
ALSO KNOWN AS PROBABILITY SAMPLE BECAUSE EVERY ITEM OF THE POPULATION HAS AN
EQUAL OPPORTUNITY OF BEING SELECTED IN THE SAMPLE
•
METHODS OF OBTAINING SIMPLE RANDOM SAMPLING
1.
LOTTERY METHODS
2.
TABLE OF RANDOM NUMBERS
•
LOTTERY METHODS:- ALL ITEMS OF THE POPULATION ARE NUMBERED OR
NAMED ON SEPARATE SLIPS OF PAPER OF IDENTICAL SIZE,COLOR AND SHAPE. THESE SLIPS ARE THEN FOLDED AND MIXED UP IN A
CONTAINER OR DRUM. A BLINDFOLDED SELECTION IS MADE.
• SUITABILITY OF LOTEERY METHOD:- POPULAR WHERE A
DECISION ABOUT PRIZES IS TO BE TAKEN IN LOTTERY DRAWN
• CHECKING OF SLIPS
POSSIBILITY OF PERSONAL BIAS
•
METHODS OF OBTAINING SIMPLE RANDOM SAMPLING
•
THE TABLE OF RANDOM NUMBERS
1.
TIPPETT’S TABLE OF RANDOM NUMBERS
2.
FISHER AND YATES NUMBERS
3.
KENDALL AND BABINGTON SMITH NUMBERS
•
TIPPET;S NUMBERS ARE QUITE POPULAR. THEY CONSISTS OF 41600 DIGITS TAKEN FROM
CENSUS REPORT AND COMBINED BY FOUR TO GIVE 10,400 FOUR FIGURE NUMBER.
•
EXAMPLE USING RANDOM NUMBERS
•
THE TABLE OF RANDOM NUMBERS
1.
SUPPOSE WE HAVE TO SELECT 20 ITEMS OUT OF 6000, THE PROCEDURE IS TO NUMBER
ALL 6000 ITEMS FROM 1 TO 6000. A PAGE FROM TIPPET’S TABLE MAY BE THEN CONSULTED
AND FIRST TWENTY NUMBERS UP TO 6000 ARE NOTED DOWN, ITEMS BEARING THESE NUMBERS
WILL BE SELECTED IN THE SAMPLE
2. POPULATION SIZE LESS THAN 1000: SAMPLE OF 10 ITEMS OUT OF 400, ITEMS WILL BE
NUMBERED FROM 1 TO 400 AS 0001 TO 0400
3.
FISHER AND YATE TABLE CONSISTS OF 15,000. THESE HAVE BEEN ARRANGED IN TWO
DIGITS IN 300 BLOCKS AND EACH BLOCK CONSISTS OF 5 ROWS AND 5 COLUMNS.
4.
KENDALL AND SMITH CONSTRUCTED 10000 IN ALL BY USING A RANDOMIZING MACHINE.
•
EVALUATION OF SIMPLE RANDOM SAMPLING
•
MERIT
- NO POSSIBILITY OF PERSONAL BIAS
- MORE REPRESENTATIVE AS COMPARED TO JUDGMENTAL SAMPLING
3.
CAN EASILY ACCESS THE ACCURACY OF THE ESTIMATE
DEMERIT
1.
NECESSARY TO MAKE A LIST AND IT IS VERY DIFFICULT TO MAKE
2. PREPARATION OF THE SLIPS IS EXPENSIVE AND TIME CONSUMING
3.
SAMPLE SIZE IS REQUIRED VERY LARGE
4.
SAMPLE SELECTED ON THIS METHOD MAY BE SCATTERED GEOGRAPHICALLY
•
STRATIFIED SAMPLING
1.
THE PROCESS OF STRATIFICATION REQUIRES THAT POPULATION MAY BE DIVIDED INTO
HOMOGENEOUS METHODS GROUPS OR CLASSES CALLED STRATA
2.
SAMPLE MAY BE TAKEN FROM EACH GROUP BY SIMPLE RANDOM METHODS
3.
IT MAY BE PROPORTIONAL OR
4.
IN CASE OF PROPORTIONAL STRATIFIED SAMPLING PLAN,THE NUMBER OF ITEMS DRAWN
FROM EACH STRATUM IS PROPORTIONAL TO THE SIZE OF THE STRATA
•
PROPORTIONAL STRATA:- FOUR STRATA ARE CREATED, AND THEIR RESPECTIVE SIZE IS
10,25,15 AND 50 PERCENT FROM THE POPULATION
•
FIRST STRATA= 1000X10/100=100
•
SECOND STRATA= 1000X15/100=150
•
THIRD STRATA=1000X25/100=250
•
FOURTH STRATA= 1000X50/100=500
•
TOTAL = 1000
•
IN ORDER TO MAINTAIN MAXIMUM EFFICIENCY,GREATER REPRESENTATION TO A STRATUM
WITH LARGE VARIATION.
• DISPROPORTIONATE STRATIFIED INCLUDES PROCEDURE OF TAKING AN EQUAL NUMBER OF
ITEMS FROM EACH STRATUM
•
EVALUATION
•
MERITS
1.
MOST EFFICIENT SYSTEM OF SAMPLING AS THE POPULATION IS DIVIDED INTO
DIFFERENT STRATA
2.
GREATER ACCURACY
3.
MORE CONCENTRATED GEOGRAPHICALLY
DEMERIT
1.
DIFFICULTY IN CREATING STRATA
2.
SKILLED SUPERVISOR IS REQUIRED FOR RANDOM SELECTION FROM EACH
•
SYSTEMATIC SAMPLING
1.
SUITABLE WHERE A COMPLETE LIST OF THE POPULATION FROM WHICH SAMPLING IS TO
BE DRAWN IS AVIALABLE
2.
THE METHOD IS TO SELECT KTH ITEM FROM THE LIST WHERE KTH REFERS TO SAMPLING
INTERVAL
3.
IF WE HAVE A COMPLETE LIST OF 1000 STUDENTS AND WE WANT TO DRAW SAMPLE OF
200 STUDENTS,1000/200=5 TH
4.
WE MUST TAKE EVERY FIFTH ITEM. THE FIRST ITEM BETWEEN 1 AND 5 IS TO BE
SELECTED AT RANDOM. SUPPOSE 4 TH ITEM IS SELECTED,SECOND WILL BE 9,14 AND SO ON
•
EVALUATION
•
MERITS
- MORE
- TIME AND WORK IS VERY LESS AS COMPARED TO OTHER METHODS
- IF POPULATION ARE QUITE LARGE IT WILL PRODUCE THE SIMILAR RESULT TO
PROPORTIONAL STRATIFIED SAMPLING
DEMERIT
1.
NOT SUITABLE WHERE THE POPULATION HAVING HIDDEN
2.
BIASEDNESS IN SUCH CASES
• MULTISTAGE SAMPLING
1.
SAMPLING PROCEDURE WHICH IS CARRIED OUT IN SEVERAL STAGES
2.
THE MATERIAL IS REGARDED AS MADE UP OF A NUMBER OF FIRST STAGE SAMPLING
UNITS,EACH OF WHICH IS MADE OF NUMBER OF SECOND STAGE UNITS ETC
3.
FIRST STAGE UNITS ARE SAMPLED BY SOME SUITABLE METHOD SUCH AS RANDOM
SAMPLING
4.
THEN A SAMPLE OF THE SECOND STAGE IS SELECTED FROM EACH OF THE SELECTED
STAGE BY SOME SUITABLE METHOD
5.
FURTHER STAGES MAY BE ADDED AS REQUIRED
•
EXAMPLE
1.
LIKE WE WANT TO TAKE A SAMPLE OF 5000 FROM THE STATE OF U.P
2.
FIRST STAGE THE STATE MAY BE DIVIDED INTO NUMBER OF DISTRICTS
3.
FEW DISTRICTS WILL BE SELECTED AT RANDOM
4.
AT THE SECOND STAGE SELECTED DISTRICTS WILL BE DIVIDED INTO NUMBER OF
VILLAGES AND SAMPLE OF VILLAGES WILL BE TAKEN AT RANDOM
5.
OUT OF SELECTED VILLAGES, A NUMBER OF HOUSEHOLD MAY BE SELECTED
6.
IN THIS WAY SAMPLE SIZE GOES ON SMALLER AND SMALLER
•
EVALUATION
•
MERITS
1.
FLEXIBILITY
2.
PERMITS THE FIELD WORK TO BE CONCENTRATED
•
DEMERITS
1. LESS
ACCURATE
No comments:
Post a Comment