Writing process

6 More specific detailes of questionnaires

Designing and administering a questionnaire for research information

The first two questions you need to answer for yourself are: (1) What (kind of) information do you want to get?  And: (2) How do you intend to analyse it?

The type of information you want to get will lead you to the next step, which is identifying your target audience, the persons from whom you want data. Here you could for example distinguish between ‘directly involved’ and ‘indirectly involved’ persons. Do these persons represent the population?

Two terms used in quantitative research:

Population: this refers to the whole group (e.g. all men in the Canton of Jura; all households in Swiss cities with solar PV panels on their roofs).

Sample: this refers to a subset of individuals of a specific population (all households in the city of Lucerne with solar PV panels on their roofs).

Within the scope of this book no information can be provided on more specific issues of quantitative research such as statistical significance of data or statistical methods to evaluate quantitative data.

Considerations for design of questions

1) KISS

Keep the questionnaire short and simple (the KISS rule). The shorter the questionnaire the more people are willing to answer. Best is to estimate the time people need to answer all the questions. It should not take longer than 15 minutes (for some people this is already too long).

Rule of thumb: closed questions take 15-30 seconds to answer; open questions, from 30 seconds to a few minutes.

2) Simple scale

Do not ask for lengthy written answers. Instead, give a choice of possible answers, for example

Circle or X or tick one

always                    usually                   sometimes             rarely                      never

SA           A             NS           D             SD

(Strongly Agree; Agree; Not Sure, Disagree; Strongly Disagree)

5              4              3              2              1 

(5= Excellent; 4= Very Good; 3= Good; 2= Needs improvement, 1= Poor)

3) One question- One idea

In each statement/ question should be one idea only.

Do you use English to communicate orally with clients?                             Yes          No

Do you use English in written communication with clients?                        Yes          No

Not: Do you use English for written and/ or oral communication with clients? (requires more than one answer)

 

Question types

When designing questions for an interview or questionnare, carefully define the type of information you are interested in. This should lead you to the type of question to ask. Your questions should all relate or be relevant for your research question(s).

If, for example, you are investigating ‘the use of English at work’, your questions or statements could include these:

Likert[1] scale questions (useful to assess people’s opinions on a 4-10 point scale)

How frequently do you use English at work? Circle one.

Never                       Sometimes                             Often                      Always

[1] The American psychologist Rensis Likert first modelled the use of a five-point survey scale in 1932

Categorical questions (you can only select one option)

Do you use English at work? Tick the correct answer.

Yes           No

Multiple choice questions (you are allowed to choose only one option)

If English was the corporate language of your next employer, which model would you prefer?

A) English only if somebody is present who does not understand the local language.

B) English only in all cases.

Multiple response questions (you are allowed to choose more than one answer)

In which of the following contexts do you use English? (you  can tick √ more than one answer)

  1. Communicating with clients (oral)
  2. Communicating with clients (written)
  3. At internal meetings
  4. Communicating with colleagues (written)
  5. Communicating with colleagues (oral)
Ranking questions (you place items in a following list in order of preference, writing 1 for most preferred option, 2 for the next and so on)

Open-answer questions

In qualitative research you can give respondents opportunities to comment (in their own words)

a) after options you have chosen, or

b) at the end of the survey because:

some respondents may choose an option you have not considered.

some respondents may prefer to comment in their own words.

Respondents’ comments are strong, credible support when quoted directly.

To avoid weak results because of weak design: Try it out

Before you send the survey ask another student to complete it to see whether your questions are clear, or try it with a ‘friendly user’, if possible a member of your sample. Then the survey can be imroved before being sent to all respondents.

Writing conventions for questionnaires

  • A clear, descriptive title
  • An introduction describing the reasons for the survey
  • Clear directions about completing the questionnaire
  • Demographic data (for example function in a company) about the respondent, but not the name
  • The questions or statements to be completed by the respondent
  • Space for written comments at the end of the survey
  • Information on how to return the questionnaire to the designer
  • An expression of appreciation for the time spent by the respondent.
A useful and very detailed resource for the creation of questionnaires for a Bachelor or Master thesis: Survey_Writing_Guide_NEW

Writing up Survey results

Tabulating survey data

The more carefully the survey has been designed, the easier it is to collect, report and interpret the data. Usually the writer tabulates the results and translate those results into percentages and/ or averages because numbers are easily understood, and averages and percentages have greater impact than individual results. Find an example in Table 3  below.

Table 3: Results of a survey about driving under the influence of alcohol

question yes yes no yes
number percentage* number percentage*
1 18 81% 2 19%
2 10 50% 10 50%
3 3 8% 17 92%
4 10 50% 10 50%
Total number of students surveyed: 20

Total number of responses received: 20

*Note: if you visualize results of surveys that are not quantitative but qualitative you do not use percentages. Some people in social sciences argue that 20 students are too small a sample to count as representative. Ask your tutor;  if in doubt leave percentage away.

Principles to consider when writing up results

Writers use survey results to support their ideas and opinions. Therefore, it is necessary to

  1. introduce the use of a survey in your methodology section,
  2. include a copy of your questionnaire with the responses in an appendix,
  3. present the results in the main body of the report.

There are several ways to report survey results.

Possible structure of results of questionnaires

There is not one way of structuring your results but always move from general to specific.

Some examples of structuring results of questionnaires

Refer to the survey and state whether the results disagree or agree with previous research.

Example:

Results of a survey of 15 university students who were first born children supported previous research (see Appendix A for the survey). It showed that…..

In the background paragraph, describe and refer the reader to the survey in the Appendix.

Example:

A survey (XC, 2012) of 15 international students at Lucerne University was used to determine the study routines prior to exams (see Appendix 3 for the survey).

Describe the survey briefly and give the general results of the survey to support an idea in a main body paragraph.

Example:

The survey (XY, 2012) asked 15 undergraduate students to identify their personality characteristics. More than 9 of the respondents indicated that they fit the profile of first-borns, and more than 7 characterized themselves as independent achievers who were conscious of time and well-organized. Table _ gives the average percentages for the group.

Introduce results, report them in a table, and interpret them for the audience.

Example:

The results of the survey confirmed the author’s belief that a great majority of typical university students wanted to learn more about the culture in Switzerland; 12 respondents chose ‘culture’ as one of their top three choices.  Table 2 depicts the top five choices students made and the percentage for each.

Direct quotations from a survey can be powerful evidence. When a survey provides respondents with a ‘comments’ space, those comments can be used as direct quotations. Use same conventions as explained for interviews in the next chapter.

Licence

Academic Writing in a Swiss University Context Copyright © 2018 by Irene Dietrichs. All Rights Reserved.