Interviewing is a professional activity that encompasses heavy responsibilities. It is most important to remember that the survey in which you are participating is an extremely serious matter being used to take systematic observations from which to build knowledge that will help people live healthier lives as well as to inform government agencies about health issues being studied in the survey. The overall results will depend on the quality of your work. Several types of interview are commonly used, including:

1) Unstructured interviews: Interviews in which the person asking questions has a general idea of topics to be covered by questions but adapts the specific questions asked depending on what the respondent says and is free to word questions in ways that the interviewer believes will best indicate the information that is wanted. Examples: Journalistic interviews, some kinds of psychiatric interviews, many exploratory interviews.

2) Structured interviews: Interviews in which the wording and order of questions to be asked is completely determined. Interviewers have little or no latitude to alter wording or question order. Examples: Public opinion polls, many scientific surveys, standardized testing.

3) Partially structured interviews: Combination of unstructured and structured interviews.

The Introduction.

The interview that most polling firms conduct is called completely structured, since you are to read all questions as written and in the order written. Such interviews are designed to insure comparability among the responses to different interviewers by different respondents. Words which appear in normal print are to be read as written. Words which appear entirely in CAPITAL LETTERS or BOLDED LETTERS are instructions to interviews and should not be read to respondents.

If a respondent cannot understand a word or phrase, then repeat the sentence. If the sentence is still not understood, an interviewer may modify the wording to aid respondent in answering, but must make note of the changed wording in the margin of the survey adjacent to the item changed.

The first portion of the interview, the introduction, is designed to initiate an interview with the appropriate respondent in a household contacted through the sampling procedure being used. The introduction is to be delivered as worded and functions to:

1) Introduce the interviewer,

2) Introduce the institution,

3) Give a very briefly introduces the survey purpose,

4) Explain that the survey is confidential and that responses will be used only in statistical analyses of groups of respondents,

5) Select the appropriate respondent to be interviewed in the household,

6) And gain passive informed consent to conduct the interview.

According to the sampling procedure used, interviewers are guided to the correct respondent, for instance, the adult (person 18 years or older) in the household with the most recent birthday. The interviewer then either proceeds as rapidly as possible to do the interview once the targeted respondent is contacted or finds out when the targeted respondent will be available for an interview. In the latter case, the interviewer will arrange a specific date and time to contact the targeted respondent.

Representative Objectives, Comments, and Suggested Responses.

Many questions arise concerning the survey when household contacts are initiated. Representative questions and possible answers include:

"Where did you get this number?" Your number was generated by computer using a scientifically developed procedure to represent all Californians with residential telephones.

"Why are you doing this survey?" We are trying to find out what people in this area really think about a number of important political issues so that we can provide information that may improve our ability to teach students what is going on in the state.

"Who do you represent?" I am an interviewer in a class at San Diego State University. We are trying to find out about citizens' political and social opinions and you can help me a great deal.

"What are you going to do with what I say?" Your response will be changed to numbers and put into a computer along with all other persons so that no one will be able to match your answers with you personally. We will then do statistical analyses to see what general trends exist in the state on our questions.

"I don't know anything about issues (or anything else)!" That is all right. Everyone's views are important to us even if they have no opinion on some issues, since it is important for us to know how many people do not have opinions on some issues.

"Who is going to see what I say?" No one except staff persons on the project. Your identity will be kept in the strictest confidence so that no one will ever match what you tell us with you personally.

"Why do you need my views?" OR "I just do not want to participate!" It is very important for the scientific validity of the survey that we get responses from all subjects. Our sample is accurate only if all people surveyed agree to be interviewed for the short time it will take. Other people have been very helpful, and I certainly need your help for the success of our class project.

"Who is paying you to do this?" No one, I am a student doing a class assignment at San Diego State University.

"What are you selling?" Nothing! This is a SDSU research project, and your help will make this a more successful project and help us to provide more accurate information about Californians' political views.

"I cannot continue!" "I have to go now!" The survey just will take a few more minutes of your time. It will be most helpful to give me a few more minutes of your time. We have finished most of the interview; it will take just a few more minutes. I'd really appreciate your help for just a few more minutes. IF ALL ELSE FAILS, My instructions permit me to finish this at another time. What would be the best time for you? When may I finish this up? (Make a specific appointment to call back and to complete the interview.)

In general, the more interviewers emphasize positive symbols, the higher the rate of cooperation likely. Positive symbols include: "science," "San Diego State University," "help to improve the what we teach students about people in California," "help the state of California do a better job in making laws to protect the our nation,""do a favor for me," and "help me personally to do well in my job."

The Body of the Survey.

The first few questions are extremely important since most respondents will agree to complete interviews once they have been asked several questions. The first few questions reinforces the legitimacy of the survey and raises the interest of respondents. Thus, once interviewers get the correct respondent, they should move directly and quickly to ask the rest of the questions.

In telephone interviewing, it is also important to move deliberately and quickly through all questions (without over-pressuring the respondents). Do not allow respondents to wander or to become overly verbose. Focus questions, if necessary with probes (neutral questions to focus attention on what you want to find out). If necessary, wait until a respondent takes a breath and then quickly re-ask a question or move on to the next question.

Be sure to follow the proper sequencing of questions in the survey (the order in which items appear on the protocol). Surveys will sometimes use filter questions to guide the sequence in which questions are to be asked. For instance, if a respondent says that he/she has never smoked cigarettes the interviewer will be instructed not to ask whether that respondent has quit smoking or whether that respondent feels that he/she is able to quit smoking. The latter questions are not relevant to persons who have never smoked and may lead to irritation and loss of rapport during the rest of the interview. In contrast, all respondents will be asked how many of their friends smoke cigarettes since both smokers and non-smokers may have friends who smoke.

Occasionally a respondent may not want to respond to an item. Interviewers must judge whether to repeat the question or not in order to finish the protocol. It is more important to obtain responses to all the questions in the remainder of the survey than to lose an interview. If a respondent appears to be on the verge of terminating over an item, move directly on to the next question in the protocol and complete the survey. When all of the items have been asked at the end of the interview, just say something like the following: "Now I need to check to see whether I have correctly asked you all of the questions on the survey. Oh, yes! Here is an item." Then repeat the question as if no problem had existed, thank the respondent, and terminate the interview normally. In many cases, the respondent will provide an answer to the omitted question.

There are two kinds of questions in most surveys:

1) Closed-ended questions: Questions that include the proper responses for the question. Example: "Would you say you agree strongly, agree, disagree, or disagree strongly that smoking cigarettes is bad for your health?"

2) Open-ended questions: Questions that do not enumerate the proper responses for the question. Example: "What would you say are the most important health problems facing most people of Korean descent in California today?"

Closed-ended questions are items with all responses read to the respondents or for which responses are provided but not read. Simply ask the item, and mark the appropriate response by circling it. If respondents do not give the correct type of answer, then repeat the question along with response categories, pause, and wait for the response. If the respondent cannot give the correct type of response, note what was said in the margin of the protocol and go on to the next item without circling one of the categories. It is the interviewer's responsibility to insure that coders understand unambiguously what the specific response is to all questions!

Open-ended questions are free response items, questions that are asked and respondents give spontaneous answers. They are typically questions followed by a blank for writing the response. Note that the response is to be written, with the meaning preserved to your best ability. Usually you should write exactly what the respondent says. Responses, however, may occasionally require paraphrasing what the respondent actually said when responses are long or complex. It is the interviewer's responsibility to annotate the questions in margins so that complete and accurate meaning of responses is preserved. Annotation should be checked and clarified as soon as interviewing is complete.

Probes are quite commonly employed in order to clarify the meaning of responses to open-ended questions, obtain more complete information, and to obtain second and third responses to the same question where requested. A probe is a neutrally worded prompt designed to elicit further responses. Probes might include: "Can you tell me more about that?" "Is there anything else?" "Anything else?" "What else?" A brief, silent pause also frequently worked to elicit further response.

Matrix or battery questions involve a series of closed-ended questions using the same responses and asked sequentially. An introduction includes a general question, followed by the response categories, followed finally by items. Sometimes the response categories are forgotten so that interviewers must repeat them. Repeating the response categories is an appropriate prompt whenever required. Most respondents learn appropriate responses in battery formatted sets of questions after the first one or two items.

Forms and Materials.

Immediately when an interview is completed and calling terminated, interviewers should check every item to make sure that each is marked appropriately. Be especially careful to insure that:

1) One and only one response is marked for each item,

2) Written information is easily legible,

3) No item has been omitted (if a question has been omitted, call back the respondent immediately, apologize, ask the question, and thank the respondent). Of course, if a respondent has refused to answer a question after appropriate follow up, the interviewer should mark "NA" for no answer and not re-contact the respondent.

4) All questions have a response marked appropriately.

The disposition of each and every call is to be recorded using codes at the bottom of calling sheets in the row corresponding to the number called (SEE ATTACHED BLANK CALLING SHEET). One mark must be made for each call to each number. If an interview is completed on the first call, for instance, then mark"CO" under the heading "Initial Call" for the number. If no one answers on the initial call, the number is busy on the second call, and the interview is actually completed on the third call, then the call sheet would be marked: NA, BU, and CO under the columns labeled "initial call," First Call back," and "Second Call back," respectively. Once completed, the calling sheets will provide the history of interviewing attempts for each number in the sample. Outcomes and codes include:

NK: Ineligible. No eligible persons are in the household,

CO: Interview completed,

REF: Interview refused, respondent hangs up without replying,

LA: Cannot determine language, but not English,

MA: Answering machine. Leave a message with your name, number, and time for respondent to contact interviewer, but interviewer should also call back at next available chance (since most respondents will not return calls),

BU: Line is busy,

CB: Call back,

APT: Appointment to call back,

NA: No answer, refused to answer question,

DIS: Number disconnected, or

IN: Ineligible number (Business or Institution), not a residence.

Interviewers should cycle through numbers from the top of each calling sheet to the bottom, dialing one number at a time and moving on to the next as an interview is completed, refused, a busy signal received, a machine answers, an ineligible number is contacted (e.g., a business), a household with no eligible persons has been called, or some other disposition has been made for the number. Note in small print beside the code, the day and time of the call. After the end of the list is reached, simply cycle through the list a second time, on a different day and time of day, calling each number again. Do not redial numbers at which respondents have refused, are numbers not belonging to residences, or numbers which do not exist (but nonetheless ring).

The best interviewer will be used as a "conversion interviewer." This person will re-contact respondents who refuse to be interviewed initially following a two week interval. The conversion interviewer will make a specific attempt to complete an interview for those marked REF.

It is imperative that no materials are lost during the interviewing process. All interview protocols, uncompleted as well as completed, calling lists, instructions, and any other materials associated with the survey should be returned to the supervisor as soon as possible. All items belong to the project and are to be returned at the end of interviewing.

Return to top.
 

Supervision: