• No results found

A questionnaire was decided to use as a method for collecting primary data. The following section clarifies the process from developing the questionnaire to the final result of data collection.

4.3.1 Questionnaire development

Churchill (1999, 329) expresses a nine-step procedure of how to develop a questionnaire.

This is illustrated in figure 4.1. This procedure was followed, and in this section we will describe each step and how the questionnaire was developed from start to end.

Figure 4.1: Procedure for developing questionnaires. Adapted from Churchill (1999)

Step 1: What information will be sought

It was decided that SCRM considering inbound and outbound logistics would be the main topic. Additionally, a questionnaire would be completed among Norwegian manufacturing firms for adding new research to the literature as we found a lack of research concerning inbound and outbound SCRM. As this would be an exploratory research and was aiming to discover ideas and insight, such questionnaires could be more loosely structured with an idea of what type of information that should be collected (Churchill 1999). SCRM was decided to be the main dependent variable throughout the research. The independent variables are lean, delays, information sharing, collaboration, relationship distance, product variety, deliveries and performance measures.

Step 2: Determine type of questionnaire and method of administration

Fink (2003) lists up four types of survey instruments: self-administrated questionnaires, interviews, structured record reviews and structured observations. For collecting data, it was decided that a questionnaire completed over telephone would be the best opportunity for having a highest possible response rate. As a result of this, the survey was a

Step 9: Pretest questionnaire and revise if necessary Step 8: Reexamine steps 1-7 and revise if necessary Step 7: Determine physical characteristics of questionnaire

Step 6: Determine sequence of questions Step 5: Determine wording of each questions

Step 4: Determine sequence of questions Step 3: Determine content of individual questions

Step 2: Determine type of quesionnaire and method of administration Step 1: Specify what information will be sought

unclear questions. Using interviews could increase the quality of the data as Gillham (2008) lists up quality as a problem related to misunderstandings of questionnaires.

Step 3: Determine content of individual questions

At this stage the latent variable was decided based on previous research, and it was

ensured that most concept was covered by several questions or items for ensuring it would be possible to measure after the data collection would be concluded (Churchill 1999).

Step 4: Determine response form to each single question

Most of the questions were to be answered in a five-point likert scale. As the questionnaire both includes questions and statement, the scale differs. When using such a scale it will be important that the scale are fulfilling the requirements for equidistance (Hair et al. 2014).

The lowest possible alternative in the scale is number one, and five is the highest possible alternative. When it comes to the statement the scale mostly used was ranging from disagree (1) to totally agree (5). An advantage of this type of question is that statements could be listed, and instructions is only needed to be given in the beginning of a series of statements. These types of questions with statements listed up are used several times throughout the questionnaire.

All questions in the questionnaire were closed, except a few open alternatives where the respondent could write their answer as other. Open questions are characterized by letting the respondent use its own words for answering. Fink (2003) explains that closed questions are more difficult to write compared to open questions, as you must know all the possible answers or response choices. However, when it comes to analysis and interpretation of results, this is easier for closed questions.

Step 5: Determine wording of each single question

When a new topic was introduced, the interviewer was explaining the main purpose of the section. A verbal description was formulated in from of the data collection. Part of the research is based on earlier research, performed by Husdal and Bråthen (2010).

To avoid misunderstandings questions should be formulated simple, in a way that people with different degree of education could understand the formultaion. According to

Churchill (1999) questions should be unmistakable. However, question using simpler wording could also be misunderstood.

It was expected that some of the respondents not would understand the concept of lean management, which was asked in two questions. Because of this, the interviewer was given instructions in how to explain lean management in front of the data collection.

Step 6: Determine sequences of questions

The questionnaire is divided into five different parts, and as a new topic was introduced, the interviewer explained the focus. These five parts was included in the questionnaire:

1. General information 2. Supplier questions 3. Customer questions 4. Transportation questions 5. External risk statement

According to (Churchill 1999) sensitive or difficult questions should be placed late in the questionnaire. Such questions could make the respondent feel threatened and they will turn off. In front of the data-collection, we got access to a database including firm’s financial results. By this, we could avoid sensitive questions about financial performance, which could make the respondent not feel threatened from the interviewer.

Step 7: Determine physical characteristics of questionnaire

As the survey was mostly done by telephone interviews, introduction was short. However, an introduction was written in front of the interviews as this is the point where the

respondents are convinced to answer the questionnaire and understands the importance of the research (Churchill 1999). The respondent was informed about the topic, and purpose of the research, and that the answers would be treated with full confidentiality.

Step 8: Re-examine steps 1-7 and revise if necessary

As a questionnaire takes time to develop, many drafts were created before the final questionnaire was finished. First, the variable to be measured was developed. After this, questions for measuring these variables was written down. For being sure that all variable

single variable. These questions would be the indicator of the variables when analyzing the results. In addition, formulation of each single question was considered to ensure that there would be no misunderstandings during the data collection.

Step 9: Pretest questionnaire and revise if necessary

Pretesting a questionnaire is important for having the possibility to see how it actually performs under the right kind of conditions of collection. Churchill (1999, 265) states,

“data collection should never begin without adequate pretest of the instrument”. In this research, the questionnaire was tested through an earlier research using some of the same questions (Husdal and Bråthen 2010), and during the yearly career day at Molde

University College where several companies was represented. The final questionnaire is presented in appendix 3.