Surveyresearch is usually associated with inaccurate information. This isbecause of the different errors that area associated with surveys.These errors include selection error, sampling error, populationspecification error, non-responsive error, and measurement error. Inorder for an organization to rely fully on information obtained fromsurveys, there is a need of finding solutions to these errors thiswill make the information from surveys accurate. This paper willdiscuss the problem of inaccuracies associated with surveys andprovide solutions to the problem. The population under considerationis the customers of International Textile Group (ITG), while thegeographic location under consideration will be North Carolina.
Thereare different errors that can affect the accuracy of informationcollected by ITG in a survey these errors include selection errors,sampling errors, population specification errors, non-responsiveerrors, and measurement errors (Groves,2004). Measurement error describes the extent to which a statisticwill differ from the true value because of the imperfections in themanner the statistic is collected. A non-responsive error can occur,when an obtained sample tends to differ from an original selectedsample. Sampling error describes an error that occurs the moment asampling technique is used in selecting a sample, but the sampleselected is not a representative of the population beinginvestigated. Population specification error is a kind of an errorthat occurs, when a researcher chooses an inappropriate population orarea in which to collect information. On the other hand, Selectionerror entails a sampling error for a sample, which is chosen througha non-probability method.
Theseerrors can be eliminated by ITG through various interventions.Measurement errors may be caused by poor questioning of the surveyquestions. This error can be eliminated through ITG ensuring that itdesigns its questions to the target population appropriately suchthat the questions are easy to comprehend and answer (Biemer &Wiley InterScience,2004). At times, the questionnaires’ wording and content may bemisleading to respondents. This may make it difficult for respondentsto provide responses accurately. This can be eliminated by ensuringthat the questions designed for research should not be misleading orambiguous the research questions must be relevant to the aims of thesurvey. Non-responsive errors can be mitigated or eliminated throughdifferent interventions. The ITG should ensure that there is goodsurvey design through asking short and simple questions. Besides, itis good to ensure that survey uses and purposes are well explained torespondents in order to encourage respondents to give out accurateinformation and participate fully in a survey. Follow-ups and callbacks for respondents not available during a survey can aid inincreasing the response rates for the responds that were initiallyunable to provide responses (Marsden &Wright,2010). In addition, it is critical to explain the importance ofproviding responses to respondents as representative sample.Furthermore, it is critical to ensure that confidentiality of theresponses provided by respondents is adequately assured. Moreover,the error can be eliminated through increasing public awareness byholding discussions with interest groups or media as this helps inkeeping the target population informed about an incoming survey andits importance.
Samplingerror can be eliminated by ITG through establishing a sample, whichis free from bias and that is capable of acting as a representativesample of the target population. In case the sampling error is due tochance, probability sampling and randomization process can beutilized in minimizing the error (Sapsford,2006). On the other hand, selection error can be avoided through theexamination of the correlations amid a treatment indicator andexogenous variables (Groves,2004). In regression models, selection error can be eliminatedthrough examining the correlation amid the unobserved determinants ofselection and unobserved determinants of the outcome. In addition,ITG can be capable of avoiding population specific bias through thecompany having a marketing problem, which specifically describes thetype of population it desires to study. It is good for theorganization and the researcher to discuss and concur on thepopulation definition. Besides, ambiguous descriptions must beavoided, when specifying the population under consideration.
Despitethe different errors that affect the accuracy of information from asurvey, there are different solutions to the problem of inaccuracy.Selection error can be avoided through the examination of thecorrelations amid a treatment indicator and exogenous variables. Inregression models, selection error can be eliminated throughexamining the correlation amid the unobserved determinants ofselection and unobserved determinants of the outcome. Populationspecific bias can be eliminated through a company having a marketingproblem, which specifically describes the type of population itdesires to study. Measurement errors may be eliminated through anorganization ensuring that it designs its questions to the targetpopulation appropriately such that the questions are easy tocomprehend and answer. On the other hand, non-responsive errors canbe eliminated by ensuring that survey uses and purposes are wellexplained to respondents in order to encourage respondents to giveout accurate information and participate fully in a survey. Besides,it is critical to ensure that confidentiality of the responsesprovided by respondents is adequately assured.
Biemer,P. P., & Wiley InterScience. (2004). Measurementerrors in surveys.Hoboken, N.J: Wiley-Interscience.
Groves,R. M. (2004). Surveyerrors and survey costs.Hoboken, New Jersey: Wiley-Interscience.
Marsden,P. V., & Wright, J. D. (2010). Handbookof survey research.Bingley, UK: Emerald.
Sapsford,R. J. (2006). SurveyResearch.London: Sage Publications.