How do I write a good questionnaire

3.3 Structure of the entire questionnaire

"A questionnaire should be viewed as an overall concept (introduction, main part, end part, design, presentation; author's note), in which the sequence and structure of the questions are important influencing factors for obtaining correct data." (Gräf et al . 2001, internet source)

A coherent overall concept must be implemented with great care, since deficiencies in only one component can considerably impair the quality of the data collected in the entire questionnaire. As an example, if a return date is not specified, there may be a number of questionnaires that have not been taken into account and that only arrive after the end of the examination.

The construction of a questionnaire for the written survey therefore requires more care and preparatory work than for an interview, for example, since the respondent cannot normally be asked during the survey. There is no tool available in the form of an interviewer, whom Dillman (1978, p. 119) describes as a “traditional crutch-poorly constructed questionnaire”.

Therefore, it must be clearly communicated to the respondent what who intends to do with this questionnaire and how the results will be used. Likewise, mundane questions can be asked of the respondent, for example about the time required, his benefit from the survey and whether he can answer the questions at all or is overwhelmed by it. Answers to such questions should be anticipated in the introduction so that nothing stands in the way of answering them.

In the main part now come the real questions. These should be divided into thematic blocks. A wide variety of criteria must be taken into account when creating the questions; otherwise the result will be distorted and possibly completely unusable. Different types of questions are used to determine the respondent's opinion and attitudes. Chapter 3.4 will deal with the creation as the main part of this description.

Last but not least, the deal should come. Here, too, the respondent can be motivated to send his answers back to the interviewer. A short thank you formula at the end is good form.

The basic structure of the questionnaire briefly outlined here should give an overview of the following pages. In the following, the introduction, conclusion and main part are examined in detail.

3.3.1 Introduction

In the introduction, either as a cover letter or on the first page of the questionnaire, general information should be given in order to provide answers to any questions the respondent may have.

The respondent could ask the following questions - these should be answered in advance:

  • Who conducts the survey?
    State the institution, address and position of the interviewer.
  • Who is responsible, who can I contact?
    Specify responsible persons with their email address and, if applicable, telephone number.
  • What is the purpose or aim of the investigation?
    State the reason for the survey and the benefits for the surveyor.
  • Who is interviewed?
    Specification of the group of people surveyed and its scope as well as the criteria for selecting the respondents.
  • What will happen to the results?
    State the consequences of the results.
  • Can the results be viewed?
    Indication of the possibilities of viewing the results for the respondent with effort, location, circumstances, time period. Here can be special
  • Trust and goodwill are won.
  • How much time do I have to invest?
    An estimate of the average time required and the number of questions should be given.

(see Batinic - help texts from the software WWW questionnaire generator)

However, the preliminary information must not be too detailed, because "a too detailed description of the organization or people who hide behind a survey can have a negative influence on the informative value of the answers" (Sassenberg / Kreutz 1999, p.73). For example, information that is too detailed can create social desirability.

According to Schnell et al. (1999, p.339) three components are essential in the introduction:

  • the indication that the survey is useful and relevant,
  • the assertion that the respondent is important for the success of the study,
  • the promise that all data collected will be treated confidentially.

Based on the ranking of Bosnjak and Batinic, the following tendencies in the assessment of the importance of the prior information by the participants can be recognized:

Prescribed items: Participant-relevant preliminary information Rank over all participants (middle rank corresponds to Friedman test; chi2 (4,346) = 283, p<.01)>
Information about access to the email address 1 (3.48)
Feedback on the overall result 2 (3.44)
Information about the exact subject of the investigation 3 (3.29)
Complete anonymity of my answers 4 (2.86)
Personal appeal of the researcher 5 (1.92)

Table 7: Ranking of the participant-relevant preliminary information; Extract from Bosnjak and Batinic (1999, p. 149, tab. 3)

In this study, participants were selected based on their email. Therefore, in my opinion, the first point with “Information about access to the e-mail address” in normal written surveys is comparable to the question of the selection of the participants in an evaluation, provided that the participants do not necessarily result from the measure. As a negative example, e-learning should be mentioned here, which is open to any number of participants.

In my opinion, the ranking positions are important, as they give conclusions about the favorable order of the information within the introduction.

Thus, according to the investigation by Bosnjak and Batinic, the structure of the introduction in the following form would be ideal:

  1. Information on the selection of participants (if necessary),
  2. Opportunities to inspect the result (if possible),
  3. Notes on the object of investigation,
  4. Assurance of complete anonymity (if possible),
  5. personal appeal of the researcher.

Furthermore, if a survey is used for scientific research, it should definitely be labeled as scientific, as this has a significantly positive influence on acceptance (Bosnjak / Batinic 1999, p.151). Using the semantic differential, Bosnjak and Batinic (ibid.) Worked out significant differences between scientific and commercial questionnaires. The willingness to participate is much higher in scientific surveys. There is also a more positive attitude among participants than in commercial surveys. However, cultural characteristics must be taken into account. The results of Bosnjak and Batinic refer to a study in Germany - in other countries science can have a different status.

Furthermore, knowledge of how to use the questionnaire should not be taken for granted. Therefore, a brief instruction manual should be offered in the introduction:

  • How answers should be marked (tick, circle, underline),
  • how to deal with questions that are inadvertently incorrectly marked
  • and whether multiple answers are possible and how they can be given.

The introduction should be short and clear. The trick here is to prepare the information requested above in such a way that the interested respondent can find answers to his questions about the investigation. The beginning of the main part should be easy to find so that the survey can be entered directly if necessary without having to read additional information.

Excursus: reasons for participating in investigations

From the investigation of the determinants of willingness to participate in Internet-based questionnaire studies, Bosnjak and Batinic obtained the following results:

Participant-relevant motives Rank over all participants (middle rank corresponds to Friedman test; chi2 (3.353) = 467, p<.01)>
curiosity 1 (3.22)
Contribute to research 2 (3.02)
Self-knowledge 3 (2.25)
Material incentive 4 (1.51)

Table 8: Ranking for motifs relevant to participants; Extract from Bosnjak / Batinic (1999, p. 148, Tab. 2)

The first three motifs, which are immaterial in nature, seem to predominate. However, the results are based on self-assessments by the test persons and “do not necessarily have to agree with the actual behavior in a specific survey situation” (Bosnjak / Batinic 1999, p.148). However, these self-assessments at least show tendencies.

It is interesting to note the high priority that the item 'Contribution to research' has. This is particularly important for the introduction of the questionnaire.

3.3.2 Conclusion

After the survey, the questionnaire should be concluded at the end. This can be kept very short.

The following elements should be included:

  • Thank you formula,
  • Address for the return (if necessary),
  • Invitation to comment.

A short thank you formula for honestly answering the questionnaire is good form. This point can also increase the motivation to return the product as soon as possible, if this is necessary.

If a return is necessary, as is the case with postal or e-mail surveys, the return address should be listed here again. This prevents the case that, for. B. the respondent misplaced the reply envelope that was sent and therefore cannot return anything.

The end can contain an "invitation to comment on the study and the corresponding free space for comments" (Schnell et al. 1999, p.339). In this way, suggestions and thoughts from the respondents can be sent to the investigator, which may be useful for further examinations. In my experience, respondents appreciate it when they have the space and the opportunity to comment on the questionnaire.