5.9 Error control processes
5.9.3 Participant-based error
Participant-based error includes the lack of knowledge of the participants, misinterpretation in relation to the questionnaire, incomplete participation (item nonresponse) and total nonresponse (Cooper & Schindler 2011). The lack of participant knowledge has been addressed via the selection of the participants’ characteristics. The targeted participants were selected from the sample population whose expertise and experience are relevant to the research topic. The possibilities of misinterpretation about the items and items’ instructions in the
questionnaire have been addressed via the two-step pre-testing process. Accordingly, these types of error have been anticipated.
The occurrence of nonresponse error which results in low response rates that influences the quality of findings of a survey research has been widely discussed in the literature (see Borkan 2010; Dennis 2003; Díaz de Rada 2005b; Dillman et al. 2009; Dillman, Smyth & Christian 2009; Fauth et al. 2013; Groves & Peytcheva 2008; Hager et al. 2003; Helgeson, Voss & Terpening 2002; Kawash & Aleamoni 1971; McAuliffe et al. 1998; Reio 2007; Thomsen 2000). Both total nonresponse error and item nonresponse error should be properly addressed in a postal survey (Sax, Gilmartin & Bryant 2003). The total nonresponse error refers to any failure to return the mailed survey and the latter refers to incomplete returned mail surveys (Platek 1985; Reio 2007; Sax, Gilmartin & Bryant 2003).
To address the issue of total nonresponse error, several strategies have been applied for the current research. Larson and Poist (2004) and Schirmer (2009) suggest that, in order to increase response rates in a survey, the topic of the questionnaire should be interesting and relevant to the interests of the intended participants. Schirmer (2009) and Ford and Bammer (2009) suggest that the use of multiple reminders with ethical considerations such as persuasion, ensuring confidentiality, representativeness and not putting a burden on the participant can be useful in order to increase the response rates. The use of stamped addressed return envelopes may also have the effect of increasing response rates (Cook, Dickinson & Eccles 2009; Levy et al. 2012).
The necessary efforts to anticipate the item nonresponse issue in the questionnaire-based postal survey due to the absence of the interviewer have been addressed. Scott et al. (2011) found that item nonresponse rate relates to sensitive questions which impose upon participants’ reluctances or uncertainties. The questionnaire has been designed to make it easier for the respondents to answer all items based on provided responses by filling in the answer to the corresponding box. Furthermore, sensitive questions were moderated by providing a set of responses in Likert-type scales in order to encourage the participants to answer the questions. Some questions encompass ‘Don’t Know’ or ‘Not Applicable’ responses in order to minimise the item nonresponse, and provide a range of responses. Riphahn and Serfling (2005) suggest that these types of responses could accommodate the respondents when they refuse to answer whilst maintaining the nonresponse error at a low level. In all, the questionnaire has been carefully developed in order to address the total survey error issues, which may influence the validity and the reliability of the research findings resulting from the data collection processes.
5.10
Summary
The detailed research design and methodology in undertaking the current research has been discussed throughout this chapter. The research is designed to enable this thesis to achieve its objectives by addressing the primary research question that is constructed with two subsidiary research questions. The nature of the research and the availability of databases about the population have been
considered in the process to select the most suitable and viable data collection method for the research survey.
The generation of the population resulted in 604 shipping companies in Indonesia, from which a sample of 230 were needed in order to obtain survey results with 95 per cent confidence with + 5 per cent. Various survey methods have been reviewed which resulted the questionnaire-based postal survey being determined as the most viable method to accomplish the objective of the research. As the best use of any method can provide optimum results, the inherent weaknesses of questionnaire-based postal surveys have been addressed in order to capitalise on the benefits of such method. In terms of the error control process, several strategies have been applied to diminish the perceived total survey error which influences the quality of the research findings. In the next chapter, the results of the survey are presented and discussed to address the research objectives.
Chapter Six
RESEARCH FINDINGS
AND
DISCUSSION
CHAPTER SIX: RESEARCH FINDINGS AND DISCUSSION
6.1
Introduction
This chapter provides a discussion of the results and analysis of the data collection to address the research objectives. The chapter first provides an overview of the demographic information about the participants of the survey to obtain the context of the discussion. The data analysis investigates current ship maintenance management from the perspective of supply chain management which includes internal readiness of the shipping companies, the external relationship conditions and the supply chain service processes. This analysis should reveal whether supply chain management is applicable in improving ship maintenance performance while at the same time enabling this research to identify the key elements for successful implementation of such an approach to ship maintenance. This research also investigates the benefits that can be attained by undertaking a supply chain management approach to ship maintenance.