英语论文网

Survey instrument comprises qualitative questions including 5 point Likert scales to give respondents an option to be more expressive. Respondents were instructed in the starting of questionnaire to mark the most suitable answer. The score ranges from 1-5 as follows (Nyengane, 2007):

Strongly Disagree

Disagree

Neutral

Agree

Strongly Agree

The survey tool measures six factors. Each factor is mapped to three questions. The survey tool is available in Appendix B. The question dealing with our variable of interest-success of supplier development program are kept in the starting to keep salient questions in the start. This was done to make the survey logical to the respondent. Respondent can see the intent of coming questions with the existing question. Questions were formulated in a way that respondents can answer them easily and accurately. Response format was standardized so that respondents know their choice of answers and need not to waste time in reading choices question by question. It will help to reduce the time respondent will take while answering the questionnaire. Response format was strongly agree, agree, neutral, disagree or strongly disagree. There were no open questions in survey to avoid misinterpretation or confusing answers, all the questions were closed and as the questions were on ordinal scale, Neutral option was kept in the questionnaire

Reliability and Validity

Reliability is statistical measure of how reproducible the survey instrument data are (Litwin, 1995). Reliability is determined by three different forms: test-retest, alternate forms, and internal consistency. Test-retest needs a lot of time to check the reliability, alternate forms were not given priority because of technical nature of survey thus on recommendation of thesis committee, internal consistency method was selected to check the reliability of instrument. Internal consistency was measured by determining Cronbach's alpha. It is a statistic that reflects the homogeneity of scale (Litwin, 1995). Cronbach's alpha was preferred over split half reliability coefficient because split half requires one test to be treated as two tests. The resulting coefficient is the correlation between two subsets may differ depending on how the initial test is divided (Ekholm & Pashei, 2009). Ideally the value of Cronbach's alpha above 0.50 will be accepted if the number of questions is less (Tamir, 2008).

Validity refers to the degree to which a survey instrument actually measures what it purports to measure. There are four types of validity: Content, face, criterion and construct. This questionnaire has proved the face and content validity at the initial phase of survey instrument design. Face validity refers whether survey asks all the needed questions and does it use appropriate language to do so (Fink, 2003). To take on this face validity, thesis committee and supplier development professionals were consulted during and after preparation of questionnaire. They agreed on the appropriateness of language for all the questions in the survey. Content validation refers to the extent to which a question appropriately assesses the characteristic本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。