When office editing and field editing should be used? Explain problems faced by editors during Editing of field data.
When office editing and field editing should be used? Explain problems faced by editors during Editing of field data.
Ans.
Stages of Editing of Data
The activity of inspecting, correcting and modifying the collected data can be carried out in two stages:
(1) Field Editing: When personal interviews are involved for collection of data, the field editing becomes important. Because, it is not possible for the interviewer to fill up the entire questionnaire during the interview. The interviewer uses brief notes or symbols to record the answers. Just after the interview, he/she reviews, corrects, and makes answers more specific on the questionnaires. This type of field editing, where the interviewer himself/herself carries out the editing work, is called field editing by the interviewer.
The other form of field editing is called field editing by the supervisor. When teams of interviewers are employed for conducting some major surveys, the supervisor with each team carries out a quick field editing to remove any mistakes in the records of answers made by the interviewers working under his/her supervision.
(2) Office Editing: The office scrutiny and correcting of data is more exacting. The office editing is carried out at some central place where all interview schedules, mail questionnaires and telephone survey responses are brought together. The office editing is more important in the case of mail surveys. This is because interview schedules and telephone survey responses may be edited at the field level also, but the mail survey responses are not edited or processed atm all before they reach the central office.
Problems faced by Editors while Editing the Collected Data
Following are some of the major problems faced by editors while editing the collected data:
(1) Fictitious Interviews: Sometimes in personal interviews and telephone surveys, some interviewers may fabricate the data and fill these out in the questionnaires or forms of interviews, without interviewing the respondents practically. Such fraud practices can be detected by an alert editor. Interviewers who try to give “phony” returns or fake data can easily be detected by an alert editor.
Cheating of this type can be identified when the answers show unusual inconsistencies and uniformity. For example, while surveying a Muslim dominated area and showing average number of Muslims.
To minimize such fictitiousness in the interviews to a certain extant, it is advised that occasional checks may be carried out with randomly selected respondents by telephone or post cards to authenticate whether the actual survey was conducted. It may be brought to the knowledge of interviewers beforehand that such checks will be conducted.
(2) Inconsistencies or Contradictions: There may be inconsistencies in the answers given. For instance, it may be given in the beginning that there is no children in a specific house. In the latter section of the questionnaire, ages of children might be listed. Still in the third section, the likings or disliking (or attitudes) of the children might be indicated. In such a case, the editor may adjust the first answer to the latter two.
(3) Illegible Responses: The handwriting of the interviewer might be illegible in some cases. In such cases, the interviewer may be recalled to decipher or interpret the responses. However, if the handwriting is unreasonable in the mail surveys then there is no alternative but to throw out the particular questionnaire. The handwriting problem is more acute when open-end questions are used.
(4) Incorrect Answers: Suppose in a TV programme study, the answer gives the time a respondent watched a TV programme. However, it is found through evidence that the respondent indeed watched the TV programme but the data given is wrong. The editor must adjust the data himself/herself because it is obviously noted wrongly by the interviewer due to his slight mistakes.
(5) Incomplete Answers: Suppose in a TV brand preference study a respondent gives the names of various brands available but forgets to give his own TV brand. But in the latter section of the Questionnaire, he says that Samsung is the best brand. It may be assumed by the editor that respondent possessed a Samsung TV set and answer may be adjusted accordingly.
(6) Don’t know and No Answers: When the respondent shows his inability to respond to a question because of the lack of knowledge, that answer is called “don’t know answered. On the other and, when the respondent is reluctant to answer a particular question because his/her personal life is linked with it, the answer is called “no answer”. The “no answer” mostly occurs in the case of mail questionnaires when the respondent leaves a question blank.
Don’t know’ answers fall into three categories:
(a) Legitimate don’t knows: They exist in the above example pertaining to a political election. At the time of interview, many voters may not really have any opinion.
(b) Confused don’t knows: The respondent may not know the meaning of the terms used. For example, in a study of pricing, various terms used may confuse the respondent (viz., cost discount, trade, discount, quantity discount, etc.) In such a confusing state, he she (respondent) may check the “don’t know” choice or leave the question blank on the questionnaire.
(c) Reluctant don’t knows: Sometimes, when the respondent’s personal life is involved or his/her organization is involved in the question, he/she may hide himself/herself in the shelter of ‘don’t know’ answer. For example, in a survey where a question is asked: “What do you think about your company’s policies regarding allocation of sales territories?” The salesman may hide himself/herself under the “don’t know answer” particularly when the survey is being conducted by his/her sales manager.
