Sampling methods

Sampling methods

1. define the population of interest

   • determine whether to sample or census
     • select the sampling frame
       • choose a sampling method
         • determine sample size
           • implement the sampling procedure
             • sampling and non sampling error
               • non sampling errors - occur in a study that do not relate to sampling error 3 main categories of these
                 • sampling frame error, non response error, data errors
           • this factor relates to financial, managerial and statistical issues
             • budget available
               • sample sizes in similar studies
                 • if there are significant differences in the views or behaviours of the population, likely response rate, the rate if the characteristic being investigated ( a study on tvs would approach less households) The number of subgroups of data that are to be analysed - individual subgroups will have more sampling errors
                 • statistical methods
         • Probability sampling a set of sampling methods where an objective procedure of selection is used meaning every member of the population of interest has a known probability of being selected
           • advantages - unbiased, survey results are projectable to the total population, sampling errors can be computed, the researcher is getting a representative view from a relevant population of interest
             • Disadvantages - respondent selection and sample design increase costs time and effort for both researcher and interviewer
         • non probability sampling a set of sampling methods where a subjective procedure of selection is used resulting in the probability of selection for each member of the population of interest being unknown
           • advantages - cheaper, less stringent procedures meaning that they can be conducted quicker, sample sizes are smaller, can target the most important respondents
             • disadvantages - bias is likely Indicative rather than definitive results, sampling error cant be computed, researcher doesnt know if the sample is representative of the population from which it is drawn, the researcher needs to make assumptions within the population of interest
       • The sampling frame is a list of the population of interest from which the researcher selects the individuals for inclusion in the research e.g a list of addresses for mail, numbers for telephone surveys
         • a telephone directory for example would have a sampling frame error - a bias that occurs as a result of the population implied by the sampling frame being different from the population of interest
           • use of access panels have started - a database of individuals who have agreed to be available for surveys of varying types and topics. This saves time and money
     • census is when you obtain data from every member of the population of interest. Censuses are occasionally used for research. A sample on the other hand is a subset of the population of interest
   • the total number of people that the researcher wishes to examine, study or obtain information from, the population of interest will normally reflect the target market or potential target market for the product or service being researched
     • accurate delineation is the key foundation step in the sampling process