설문지는 어떻게 작성해야 할까? (1) : 설문의 목적, 워딩(Wording)을 어떻게 할 것인가

설문의 목적(Objectives in designing questionnaires)

(1) 응답에 응하는 사람을 최대한 확보한다. : 

응답률을 높이는 것. 어떻게 설문을 수행할 것인지, 어떻게 라뽀를 쌓을 것인지, 설문의 목적은 무엇인지, 응답하지 않는 사람은 누구인지 생각해야 한다. 설문의 길이도 적절해야 한다.

(2) 조사를 위해 필요한 정보를 얻는다. : 

질문을 통해 어떤 것을 물을 것인지, 어떻게 물어볼 것인지, 어떤 순서로 물을 것인지, 설문지의 전반적인 레이아웃은 어떤지 생각해야 한다.

There are two main objectives in designing a questionnaire:

- To maximise the proportion of subjects answering our questionnaire—that is, the response rate.

To maximise our response rate, we have to consider carefully how we administer the questionnaire, establish rapport, explain the purpose of the survey, and remind those who have not responded. The length of the questionnaire should be appropriate.

- To obtain accurate relevant information for our survey.

In order to obtain accurate relevant information, we have to give some thought to what questions we ask, how we ask them, the order we ask them in, and the general layout of the questionnaire

무엇을 물어볼지 정하기(Deciding what to ask)

독립변수

종속변수

교란변수

there are three potential types of information: 

Information we are primarily interested in—that is, dependent variables. 

Information which might explain the dependent variables—that is, independent variables.

Other factors related to both dependent and independent factors which may distort the results and have to be adjusted for—that is, confounding variables.

Box 1: Advantages of open or closed format

열린 형식(Open format) : 다양한 범위의 응답. 

- Allows exploration of the range of possible themes arising from an issue

- Can be used even if a comprehensive range of alternative choices cannot be compiled

닫힌 형식(Closed—that is, forced choice—format) : 응답하기 쉽고, 기록하기 쉽고, 보고하기 쉽다.

- Easy and quick to fill in 

- Minimise discrimination against the less literate (in self administered questionnaire) or the less articulate (in interview questionnaire)

- Easy to code, record, and analyse results quantitatively

- Easy to report results

설문 문항의 구체적인 워딩(Wording of individual questions)

짧고 간결한 문장을 사용하라. : 

Use short and simple sentences

간단한 팁을 주자면, 하나 혹은 두 개의 절(clause)로 구성되도록 하라

Short, simple sentences are generally less confusing and ambiguous than long, complex ones. As a rule of thumb, most sentences should contain one or two clauses. Sentences with more than three clauses should be rephrased.

한 번에 한 가지만 물어라

Ask for only one piece of information at a time

For example, “Please rate the lecture in terms of its content and presentation” asks for two pieces of information at the same time. It should be divided into two parts: “Please rate the lecture in terms of (a) its content, (b) its presentation.”

부정문은 피하라.

Avoid negatives if possible

Negatives should be used only sparingly. For example, instead of asking students whether they agree with the statement,

정확한 질문을 하라.

Ask precise questions

하나의 단어도 여러 개의 의미를 가질 수 있다. (medicine : general medicine? specialty medicine?)

Questions may be ambiguous because a word or term may have a different meaning. For example, if we ask students to rate their interest in “medicine,” this term might mean “general medicine” (as opposed to general surgery) to some, but inclusive of all clinical specialties (as opposed to professions outside medicine) to others.

질문의 프레임 : 얼마나 자주 책을 빌리나요? => 지난 6개월간 몇권의 책을 빌렸나요?

Another source of ambiguity is a failure to specify a frame of reference. For example, in the question, “How often did you borrow books from your library?” the time reference is missing. It might be rephrased as, “How many books have you borrowed from the library within the past six months altogether?”

물어보는 사람이 질문과 관련한 지식을 갖추고 있는지를 확실하게 하라

Ensure those you ask have the necessary knowledge

For example, in a survey of university lecturers on recent changes in higher education, the question, “Do you agree with the recommendations in the Dearing report on higher education?” is unsatisfactory for several reasons. Not only does it ask for several pieces of information at the same time as there are several recommendations in the report, the question also assumes that all lecturers know about the relevant recommendations.

세부사항의 수준을 구체적으로 정하라

Level of details

정확한 세부사항의 수준을 정해야 필요한 정보를 다 모으면서, 불필요하게 세세한 내용은 거를 수 있다.

It is important to ask for the exact level of details required. On the one hand, you might not be able to fulfill the purposes of the survey if you omit to ask essential details. On the other hand, it is important to avoid unnecessary details. People are less inclined to complete long questionnaires. 

개인의 재정적 문제나 결혼과 관련한 민감한 정보는 기밀로 다뤄야 한다.

This is particularly important for confidential sensitive information, such as personal financial matters or marital relationship issues.

민감한 사항들

Sensitive issues

민감한 사항에 대해서 진실된 대답을 얻기는 쉽지 않다. "컨닝을 한 적이 있는가?" 와 같은 문제. 좀 덜 직접적으로 물어볼 필요가 있다.

It is often difficult to obtain truthful answers to sensitive questions. Clearly, the question, “Have you ever copied other students’ answers in a degree exam?” is likely to produce either no response or negative responses. Less direct approaches have been suggested.2 

"어쩌다 보니 학위 시험에서 다른 학생의 답을 배낀 적이 있는가?"

Firstly, the casual approach: “By the way, do you happen to have copied other students’ answers in a degree exam?” may be used as a last part of another decoy question.

"다음 중 학위 시험에서 답을 맞췄던 방법을 골라보라" => 보기에 "다른 사람의 답 배끼기" 넣기

Secondly, the numbered card approach: “Please tick one or more of the following items which correspond to how you have answered degree examination questions in the past.” In the list of items, include “copy from other students” as one of many items. 

"의대생이 학위 시험에서 다른 사람의 답을 배낀다는 것은 공공연한 사실이다. 너도 그런 적이 있는가?"

Thirdly, the everybody approach: “As we all know, most medical students have copied other students’ answers in degree exams. Do you happen to be one of them?” 

"존은 진의 답을 배꼈다. 존이 잘못했다고 생각하는가? 그렇다면 어떤 벌을 줘야 할까?"

Fourthly, other people approach. This approach was used in the recent medical student survey.3 In this survey, students were given the scenario, “John copies answers in a degree exam from Jean.” They were then asked, “Do you feel John is wrong, what penalty should be imposed for John, and have you done or would you consider doing the above?”

비뚤림 최소화하기

Minimise bias

사람들은 사회적으로 바람직하다고 여겨지는 방향으로 답을 하는 경향이 있다. 중립적으로 보이는 문제도 이러한 비뚤림에서 벗어날 수 없다. "저녁 일 때문에 수업을 빠진 것이 몇 번이나 되는가?" => "저녁 일이 수업시간과 충돌한 적이 몇 번이나 있는가?"

People tend to answer questions in a way they perceive to be socially desired or expected by the questioner and they often look for clues in the questions. Many apparently neutral questions can potentially lead to bias. For example, in the question, “Within the past month, how many lectures have you missed due to your evening job?” students may perceive the desired responses to be “never” to the first question. This question could be rephrased as, “Within the past month, how many times did your evening job commitment clash with lectures? How many times did you give priority to your evening job?”

Take another example. The question, “Please rate how useful the following text- books are. Please also state whether they are included in your lecturer’s recommended reading list?” There is a risk that the students may perceive that they should rate books recommended by lecturers more favourably than those not recommended by their lecturers. This risk may be minimised by putting the second question later on in the questionnaire

---

Leung, Wai-Ching. "How to design a questionnaire." student BMJ 9.11 (2001): 187-189.