Reliability, Validity and Representative Sample in Research
RELIABILITY, VALIDITY AND REPRESENTATIVE SAMPLE IN RESEARCH
John and Webb (2002, p.148) distinguishes between validity and reliability, arguing that the first one is the extent to which a research is capable of measuring what it is supposed to be measuring and the second one is the extent to which a research delivers consistent results. Validity and reliability measurement instruments are free of bias and random error.
Haynes and Heiby (2004, p.47) propose some questions addressing sampling adequacy such as: Is the simple sufficiently large to minimize sampling errors? Are the contents of the simple representative? Are the contents of the sample relevant? Does the way the sample data are gathered lead to a representative sample? They suggest that the first question is related with reliability and the other three are validity questions falling under the nomenclature of content-based evidence and process-based evidence.
According to Ritchie, Burns and Palmer (2005, pp.122-129), further consideration should be given to how the sample size of the research can better match the type of project conducted. If the research is to be qualitative, then the sample size can be reduced and the respondent can be seen more truly through eyes not blinkered by a strict questionnaire. By contrast, if the aim is to use the research as a quantitative measure, then the sample size needs to increase such that the reliability of the results can be significant.
Reference list:
- John R. and Webb, J. R., 2002. Understanding and Designing Market Research.2nd ed. London: Thomson Learning.
- Haynes, S. N. and Heiby, E. M., 2004. Comprehensive Handbook of Psychological Assessment, Behavioural Assessment. New Jersey: John Wiley & Sons.
- Ritchie, B., Burns, P. and Palmer, C., 2005. Tourism Research Methods: Integrating Theory with Practice. Cambridge: CABI
John and Webb (2002, p.148) distinguishes between validity and reliability, arguing that the first one is the extent to which a research is capable of measuring what it is supposed to be measuring and the second one is the extent to which a research delivers consistent results. Validity and reliability measurement instruments are free of bias and random error.
Haynes and Heiby (2004, p.47) propose some questions addressing sampling adequacy such as: Is the simple sufficiently large to minimize sampling errors? Are the contents of the simple representative? Are the contents of the sample relevant? Does the way the sample data are gathered lead to a representative sample? They suggest that the first question is related with reliability and the other three are validity questions falling under the nomenclature of content-based evidence and process-based evidence.
According to Ritchie, Burns and Palmer (2005, pp.122-129), further consideration should be given to how the sample size of the research can better match the type of project conducted. If the research is to be qualitative, then the sample size can be reduced and the respondent can be seen more truly through eyes not blinkered by a strict questionnaire. By contrast, if the aim is to use the research as a quantitative measure, then the sample size needs to increase such that the reliability of the results can be significant.
Reference list:
- John R. and Webb, J. R., 2002. Understanding and Designing Market Research.2nd ed. London: Thomson Learning.
- Haynes, S. N. and Heiby, E. M., 2004. Comprehensive Handbook of Psychological Assessment, Behavioural Assessment. New Jersey: John Wiley & Sons.
- Ritchie, B., Burns, P. and Palmer, C., 2005. Tourism Research Methods: Integrating Theory with Practice. Cambridge: CABI