Criticisms
Bias "the interviewer effect"
When the researcher is present, there is always a risk that s/he may influence the interviewee’s responses. “Interviewers might consciously or unconsciously lead respondents towards preferred answers. Personal beliefs could alter the integrity of the interview . . .” (Bryant, n.d., para. 11). Denscombe (2007) suggests that “in particular, the sex, the age, and the ethnic origins of the interviewer have a bearing on the amount of information people are willing to divulge and their honesty about what they reveal” (as cited in Newton, 2010, p. 5). Both verbal comments and non-verbal cues can influence the interviewee’s responses, and interviewers can sometimes inadvertently introduce bias in this way. (see Abbott, 2010).
Validity
Some critics suggest that the interviewer effect introduces bias that may cause the validity of the interview results to be questioned. As Newton (2010) suggests, "interviewees may say what they think the interviewer wants to hear or the opposite of what they think they want to hear" (p. 5). Some sociologists, however, refute this claim, arguing that “interviews work best when there is a relationship of trust between researcher and interviewee” (Abbott, 2010, para. 5).
Time
Interviewing, whether structured or semi-structured, can be time consuming depending on the sample size since the interviewer must be present for each interview. In addition, the preparation work (including formulating questions, identifying the sample, and gaining access to the interviewees) can add to the time required. Transcribing interviews is also time-consuming, as are the processes of coding and analyzing which follow. According to one source, “it is best to allow around five to six hours for transcription for every hour of speech.” (as cited in Interviewing, n.d., p. 322).
Expense
Though interviews conducted online, by mail, or on the telephone are not as costly, face-to-face interviews can involve added travel and accommodation expenses. Such costs may “limit the size and geographical coverage of the survey” (Phellas et al., 2012, p. 183).
When the researcher is present, there is always a risk that s/he may influence the interviewee’s responses. “Interviewers might consciously or unconsciously lead respondents towards preferred answers. Personal beliefs could alter the integrity of the interview . . .” (Bryant, n.d., para. 11). Denscombe (2007) suggests that “in particular, the sex, the age, and the ethnic origins of the interviewer have a bearing on the amount of information people are willing to divulge and their honesty about what they reveal” (as cited in Newton, 2010, p. 5). Both verbal comments and non-verbal cues can influence the interviewee’s responses, and interviewers can sometimes inadvertently introduce bias in this way. (see Abbott, 2010).
Validity
Some critics suggest that the interviewer effect introduces bias that may cause the validity of the interview results to be questioned. As Newton (2010) suggests, "interviewees may say what they think the interviewer wants to hear or the opposite of what they think they want to hear" (p. 5). Some sociologists, however, refute this claim, arguing that “interviews work best when there is a relationship of trust between researcher and interviewee” (Abbott, 2010, para. 5).
Time
Interviewing, whether structured or semi-structured, can be time consuming depending on the sample size since the interviewer must be present for each interview. In addition, the preparation work (including formulating questions, identifying the sample, and gaining access to the interviewees) can add to the time required. Transcribing interviews is also time-consuming, as are the processes of coding and analyzing which follow. According to one source, “it is best to allow around five to six hours for transcription for every hour of speech.” (as cited in Interviewing, n.d., p. 322).
Expense
Though interviews conducted online, by mail, or on the telephone are not as costly, face-to-face interviews can involve added travel and accommodation expenses. Such costs may “limit the size and geographical coverage of the survey” (Phellas et al., 2012, p. 183).
Structured (Quantitative)
|
Semi-Structured (Qualitative)
|
Limitations
- Quality of data is dependent on quality of questions asked
- Social factors
Structured (Quantitative)
|
Semi-Structured (Qualitative)
|