SAMPLING

- RESEARCH Studies give insights and inferences that are crucial for decision making - Size of sample is quiet small as compared to the total group of units of interest - Sampling is the starting point of inferential process involved in MR studies

Sampling Vs. Census

Sample, Sampling, Population or Universe, Sampling frame.

A university offering a post-graduate program in business wants to conduct a study to determine the criteria prospective students use to evaluate different post graduate business programs. Population ³ALL INDIVIDUALS WITH AN INTEREST IN PURSUING A POST GRADUATE DEGREE IN BUSINESS WITHIN THE NEXT YEAR´

 Cost effective  Time effective ACCURACY????

TOTAL ERROR ASSOCIATED WITH RESEARCH MAY CONTAIN ONE OR BOTH OF THE FOLLOWING: Sampling Error Non-Sampling Error

Sampling Error

 It is difference between a statistic and parameter  Present only in sample study Statistic: value generated through a sampling procedure. Parameter can only be determined through accurate census of the population.

NON SAMPLING ERRORS

 Any error in research study other than sampling error.  Non sampling errors can arise because of:
» » » » Poor questionnaire construction Ill trained fieldworkers Errors on part of respondents Errors in coding responses

 Best way to control non sampling errors is to have adequate control over the entire process of
» Data gathering » coding » analysis

Types of Sampling

 The type of enquiry and the nature of data fundamentally determines the technique or method of selecting a sample
± Non-Probability Sampling Methods: Subjective or Judgment Sampling ± Probability Sampling ± Mixed Sampling

 Non-Probability Sampling Methods:
± Convenience Samples ± Judgment Samples ± Quota Samples ± Snowball samples

Advantages of Non-probability Samples:  Less expensive  acceptable  Less research time  produces samples quite similar

 cannot calculate Sampling error.  do not know the degree to which the sample is representative of the population  The research results cannot be projected (generalized) to the total population of interest with any degree of confidence.