Useful data collection is one of the primary issues that should be considered when conducting a research study. First (2008) emphasizes that inaccurate data collection affects the research findings adversely, hence the validity of the course. Some of the widely used sources of data include primary and secondary sources. Academicians and research practitioners have developed diverse methods of data collection that can be used during the data collection process. Examples of direct forms of data collection include interviews, observations, and surveys.
Interviewing is mainly used in collecting primary data from the natural setting. Interview methods have been described in a variety of ways. Cohen, Manion, and Morrison (2007) specify interviews as a mutual discussion, which is usually perpetrated by two individuals guided by the interviewer, whose primary objective is to obtain the appropriate information for research. The researcher directs the conversation to focus on the matter specified by the research objectives. Therefore, the researcher is in a position to obtain the information necessary to arrive at an organized description or come up with an explanation for the phenomena being investigated. This information assists the interviewer to delve deeper into the analysis, thus enhancing the research.
Ary, Jacobs, and Razavich (2002) note that among the various tools for collecting information available to the researchers, interviews aid in the collection of reliable information. In addition to this aspect, interviews have the merit of enabling the interviewer to restate the questions to the subject and thus make the meaning of the questions clearer to the interviewee. This method is unlike other forms of inquiry, which do not provide the means for clarification. For example, questionnaires do not give the respondent the chance to seek clarification on what they are being asked. According to Murugan (2013), an interview is a face-to-face interpersonal role situation that is structured specially to extract answers relevant to the research hypotheses.
While interviews might be structured in nature, semi-structured interviews are the most commonly used tools in qualitative research. By definition, semi-structured interviews are a mix of structured and unstructured interviews as elements of both are present in the semi-structured interview (Hall 2008). In the semi-structured interview, the interests of the interviewee guide the path taken by the interviewer. For this reason, even an inexperienced researcher might be in a position to obtain rich data from the informant using this method (Smith et al. 2000). In addition to this aspect, the semi-structured interviews have inherent flexibility that surpasses that of other methods such as structured interviews and questionnaires. Through interviews, the interviewer is in a position to pursue interesting issues as they appear. The researcher can encourage the participants to expound on particular topics of interest. The participant has the opportunity to express personal views on the subject and elaborate experiences, thus giving information that might be of importance to the research (Smith et al., 2000). A semi-structured interview is most suitable for research where the goal is to explore the perceptions and experiences of the participants.
When used as a data collection method, interviews have a number of significant merits and demerits. The first advantage is that these interviews serve as vital sources when obtaining case study information. Another distinction is that the interviews give the researcher the ability to delve deeper into data collection efforts (Cohen, Manion & Morrison 2007). Additionally, the interview is flexible since the questions can be adapted to suit individual situations. Murugan (2013) corroborates this observation by noting that interviews have “adaptability” where the interviewer can add or change the questions in light of new information provided by the participants. Skilled interviewers can pursue ideas expressed by the interviewee and investigate opinions. This aspect would not be possible using questionnaires. Moreover, the researcher has the opportunity to learn differently from the participant through face-to-face interaction. Another significant merit of the interview is that they allow one to obtain additional information about the interviewees. This information includes the interviewees’ attitudes, biases, and interests (Denscombe 2010). Marsh (2014) confirms that there is a higher level of willingness among subjects to offer information in interviews than in questionnaires, thus leading to greater validity when using this method.
In spite of these advantages, some significant disadvantages are associated with interviews. Firstly, this data collection method is costly and time-consuming (Haas & Springer, 2014). The researcher has to plan and arrange the consultation carefully. Once the data has been obtained from the sources, it has to be transcribed before coding. All these steps take time and effort. In addition, in this research, six participants may not be enough to obtain enough information for the study.
Pros and cons of semi-structured interviews
Pros of semi-structured interviews
Flexibility and reliability
Semi-structured interviews are based on an interview guide that includes both open and close-ended questions, and thus this method of data collection is standardized. Harrell and Bradley (2009, p. 64) hold that standardization ‘increases the reliability of the collected data, and the researcher has the discretion of determining the order in which the questions would be asked.’ The incorporation of open-ended questions further increases the flexibility associated with semi-structured interviews. Harrell and Bradley (2009) suggest that if the interviewee has a problem in answering the asked questions, the interviewer can provide cues or prompts, which motivates the respondent to continue participating in the research study. Moreover, semi-structured interviews give the interviewers the discretion to ask a follow-up question to a response provided by the interviewee.
Semi-structured interviews are based on open and close-ended questions. The close-ended questions ensure that the respondents provide answers to similar items, which increases the comparability of responses (Cohen, Manion & Morrison 2007). Furthermore, the semi-structured interviews are very effective in minimizing interviewer effects and in eliminating bias. The comparability of data is further enhanced by the fact that researchers are able to evaluate the data collection instrument effectively, hence facilitating data analysis (Cohen, Manion & Morrison 2007).
Dornyei (2008) defines truth as the success with which a research instrument measures what it is designed to measure.
Semi-structured interviews are based on both open and close-ended questions. Thus, the open-ended questions enable the researcher to gather a wide range of opinions and ideas from the respondents by asking follow-up questions (Harrell & Bradley 2009). This capability originates from the fact that the items do not curtail the respondent to provide additional information related to the subject under exploration, which the researcher might not have envisaged. Using semi-structured interviews offer researchers an opportunity to gather first-hand information, which improves the validity of the study. Therefore, semi-structured interviews increase the probability of gathering relevant information by providing the interviewees an opportunity to respond to the questions asked in-depth (Cohen, Manion & Morrison 2007).
One of the core characteristics of a semi-structured interview is that it incorporates both open and close-ended questions (Cohen, Manion & Morrison 2007). However, only a few close-ended items are included. Tabler (2008, p. 90) emphasizes that one ‘of the core challenges faced by researchers involves predetermining what will or will not be discussed in the interview.’ Cohen, Manion, and Morrison (2007) corroborate that using open-ended questions minimizes the likelihood of interviewers pre-judging what information is essential to the study and which is not.
Ability to elicit a distinctive phenomenon
According to Craighead and Nemeroff (2004), semi-structured interviews provide researchers with an opportunity to draw out a typical phenomenon. This capability originates from the fact that the researcher has a chance to adopt different techniques in synthesizing the data collected from the field, which might be difficult to define concisely. For example, the researcher can incorporate qualitative measurement instruments such as the Likert Scale in managing the data collected from the field. Therefore, the researcher can make concise results regarding the subject under study.
Effective and efficient data storage
The quality of a study is affected by data accuracy. One of the ways through which researchers can increase data accuracy is by ensuring that the data collected is stored virtually.
Considering the fact that semi-structured interviews enable researchers to enrich the collected data by asking additional questions during the interview, the researcher must ensure that the different responses provided by the respondent are appropriately stored in order to add value to the study. This goal can be achieved by using various data storage instruments such as audio and video recorders and other digital mediums (Willing 2013).
Cons of semi-structured interviews
Despite the above strengths, semi-structured interviews are characterized by a number of challenges as evaluated below.
The application of semi-structured interviews is influenced by the interviewers’ competency and skills; semi-structured interviews provide an opportunity to undertake in-depth interviews. Subsequently, the success of such talks depends on the interviewees’ competence and critical thinking (Ary, Jacobs & Razavich 2002). Furthermore, the random or follow-up questions asked during the interviewing process make the quantification and analysis process difficult (Harrell & Bradley 2009).
According to Mitchell and Jolley (2010), the data obtained from follow-up questions cannot be easily interpreted because the interviewer asks the respondents different questions. Cope (2006, p. 97) emphasizes that the ‘extraneous material collected in semi-structured interviews cannot be compared to using questionnaires which involve short written answers were the responses tend to be more focused.’
Collecting research data using semi-structured interviews is a time-intensive activity. Thus, this method is not suitable for gathering data involving a large number of research participants (Cope, 2006).
Using semi-structured interviews increases the likelihood of a research study being biased. Mitchell and Jolley (2010, p. 277) emphasize that ‘even the answers from the standard questions are difficult to interpret because the standard questions are not asked in the same standard way to all the research participants.’ Subsequently, the interviewee may decide to please the interviewer by providing biased answers. Consequently, the likelihood of the researcher collecting socially desirable responses is increased. Lietz (2010, p. 257) contends that socially ‘desirable answers reduce the accuracy of the data collected from the field because the respondents are not able to answer truthfully to the questions asked.’
These types of interviews are also referred to as the standardized interviews, which enable researchers to ask the selected respondents similar questions. According to Hersen, Turner, and Beidel (2011), structured interviews are applied extensively in three main areas, which include clinical training, research, and clinical practice. The researcher adopts a schedule of questions, which are developed prior to the actual data collection. Moreover, the researcher also formulates possible answers that can be provided by the selected respondents (Hersen, Turner & Beidel 2011).
Structured interviews are based on pre-planned questions, which are asked during the actual data gathering process. Additionally, the problems have pre-coded response choices for the respondent to select. Therefore, using a pre-coded format enables the researcher to respond in accordance with the response choice provided (Bowling 2005). Hersen, Turner, and Beidel (2011, p. 33) assert that pre-coded ‘responses are essential in allowing for comparison across all respondents.’ The element of predetermining possible answers improves the effectiveness and efficiency with which data analysis is conducted by reducing the amount of time that the researcher requires to code the responses obtained from the field. Moreover, it becomes relatively simple to enter the data collected into a computer. Therefore, using structured interviews improves the effectiveness of utilizing the quantitative method of data analysis (Bowling 2005).
In order to be useful in collecting data using structured interviews, it is essential for the interviewer to adhere to a number of aspects. First, the interviewer should keep his/her opinion regarding the issue being explored (Fox 2009). Secondly, the open-ended responses in line with the study should be clearly written (Fox 2009).
Pros of structured interviews
Using structured interviews increases the reliability with which a researcher undertakes a diagnostic study on the issue under investigation (Hersen, Turner & Beidel 2011). Researchers should ensure that the data collection instrument used is highly reliable. According to Lewis and Slack (2007), reliability involves the extent to which a data collection instrument, tests, or other measurement procedures generate similar results or outcomes on repeated trials. Lewis and Slack (2007) emphasize that a reliable tool should be free from errors despite it being applied in different environments or conditions, which increases the reliability of a study by ensuring a high degree of consistency in the data collection process. Hersen, Turner, and Beidel (2011, p. 81) are of the opinion that systemizing and ‘standardizing the questions that interviewers ask and the way answers to those questions are recorded and interpreted decreases information variance in diagnostic evaluations.’ Therefore, the likelihood of different interviewers conducting the same study collecting additional information using similar structured questions is significantly minimized (Lewis & Slack 2007).
Structured interviews help in eliminating variability by ensuring that all the aspects considered in diagnosing a particular issue are standardized. Hersen, Turner, and Beidel (2011, p. 82) argue that ‘structured interviews assure that diagnostic criteria are covered systematically and completely, which increases the validity of the diagnosis.’ Using structured interviews increases the quality of the data collected by eliminating bias. This strength arises from the fact that the interviewer is guided by predetermined questions and possible answers during the data collection process (Hersen, Turner & Beidel 2011). Tabler (2008) asserts that one of the most significant weaknesses of unstructured interviews is that the interviewer’s decision may be influenced by personal biases arising from subconscious factors such as the halo effects, the respondents’ attractiveness, and other subjective judgments.
In addition to the above aspects, Tabler (2008, p. 92) asserts that using ‘structured interviews increases the likelihood of researchers collecting data from a large number of respondents.’ The interviews can be administered to respondents who do not have the capability to read or write. Furthermore, structured interviews enable researchers to minimize mistakes arising from the misinterpretation of the questions. Therefore, researchers obtain clear answers.
Reliability is considered as one of the critical tests of consistency of a data collection and testing instrument or procedure (Patton 2002). Conversely, Curry, Ingrid, and Bradley (2009, p. 1443) hold that validity ‘refers to the appropriateness, meaningfulness, and usefulness of the specific inference made from test scores.’ The degree of reliability in structured interviews is relatively high compared to semi-structured interviews because it is possible to obtain consistent results in repeated trials. This assertion is supported by Hersen, Turner, and Beidel (2011, p. 81), who argue that ‘structured interviews decrease the chances that two different interviewers will elicit different information from the same respondent, which results in a different diagnosis.’ On the other hand, semi-structured interviews are mainly based on qualitative research, which means that the respective interviewer develops a measurement framework, which can vary from one interviewer to another (Curry, Ingrid & Bradley 2009).
Cons of structured interviews
Despite the above strengths, structured interviews are characterized by a number of limitations. First, conducting structured interviews can be time-consuming, especially if the research process involves collecting data from a large number of respondents (Hersen, Turner & Beidel 2011).
Secondly, the quality and efficacy of the data collected are highly dependent on the questions asked. The interviewer does not have an opportunity to adjust the questionnaires in order to gather different opinions and views from the respondent based on the answer provided during the interviewing process. This challenge is further increased by the fact that the respondents’ answers are limited to the predetermined questions. Hersen, Turner, and Beidel (2011) argue that structured interviews limit the scope within which the respondent can answer the questions in detail is limited. Additionally, the success of structured interviews allows the researcher to invest a substantial amount of resources during the pre-planning stage.
These types of interviews are also referred to as personal interviews and are the oldest method of consultations. They are conducted through one-on-one interaction between the respondent and the interviewer.
Pros of face-to-face interviews
Personal interviews are mainly suited for conducting lengthy interviews that require a substantial amount of time and effort (Carr & Worth, 2001). However, the interviewer must put significant effort in order to sustain the respondents’ concentration during the interview. However, the amount of effort required to maintain a face-to-face interview is relatively lower as compared to telephone interviews. This element explains why telephone interviews are usually characterized by a short duration (Carr & Worth, 2001).
The likelihood of the researcher gathering actual data using face-to-face interviews is increased by the fact that social and visual cues can be adopted (Carr & Worth 2001). Moreover, the researcher is in a position to identify the respondents’ body language and facial expressions, hence understanding their state of mind (Novick 2008). This aspect increases the ability with which the researcher can adjust the questionnaires in order to make the respondent feel comfortable. Furthermore, the interviewer may further probe the interviewee during the face-to-face interview to provide additional information in case the answer provided is not clear or incomplete, hence achieving a precise solution unlike in a telephone interview (Novick 2008).
When conducting telephone interviews, it is impossible for the researcher to determine the interviewees’ body language and facial expressions towards the asked questions. Consequently, the likelihood of the interviewer misjudging the interviewees’ mood is high, which might cause them to discontinue the interview by hanging the telephone (Novick 2008).
Another major strength of face-to-face interviews is that the interviewer is in a position to develop a representative sample, unlike in other interviewing methods (Carr & Worth 2001). The probability of the selected respondents participating in the research study is higher in face-to-face interviews as compared to other forms of data collection because the researcher has an opportunity to convince the chosen respondents of how research ethics will be observed in order to maintain the respondents’ confidentiality. Furthermore, face-to-face interviews are mainly conducted in an environment familiar to the respondents; for example, their homes, offices, and workplaces (Carr & Worth 2001).
Limitations of face-to-face interviews
One of the significant limitations of face-to-face interviews is that the interviewer and the respondent interact with each other during the interview process, which might affect the interviewee’s conviction of confidentiality of the information provided (Novick 2008). The interviewee may develop the perception that the anonymity of the information provided may be compromised (Novick 2008). Additionally, face-to-face interviews are considered as intrusive. This arises from the fact that the researcher may be required to conduct the interview at the respondent’s premises, hence making the respondent develop a perception of the interviewer being an intruder (Becker, Bryman & Ferguson 2012). Subsequently, some individuals may react negatively to the questions asked and hesitate to participate in the interview process. A face-to-face interview is costly to conduct as it consumes a substantial amount of time (Becker, Bryman & Ferguson 2012).
According to Carr and Worth (2001, p. 512), telephone interviews involve ‘a conversation sequence without the assistance of visual cues’. The significance of telephone interviews in conducting research studies has increased significantly over the past few decades (Carr & Worth, 2001). Its growth has been stimulated by the high rate of technological development and the prevailing social change, which is evidenced by the high quality of acceptance of telecommunication. According to Novick (2008), telephone interviews are mainly used in conducting qualitative research.
Pros of telephone interviews
One of the significant benefits of using telephone interviews is that the researcher is in a position to reach a large number of respondents at a minimal cost as opposed to personal interviews. It is estimated that the cost incurred in conducting personal interviews is twice of what is incurred in executing telephone interviews (Carr & Worth 2001).
Considering the high rate of infrastructural development in most countries with specific reference to telecommunication, it is possible for researchers to interview geographically dispersed respondents. Therefore, the researcher is not limited to source data from the local area, but s/he can also collect data from the global population. Additionally, some individuals may not be comfortable with providing information at their workplaces. However, using telephone interviews enables the researcher to resolve such problems (Novick 2008).
The process of collecting data using a telephone interview is not a one-on-one affair. However, the interviewer and the interviewee do not see each other. Thus, the respondents can answer the questions asked without the fear of being identified, and therefore they are more likely to answer questions that might be considered confidential or embarrassing. Findings of studies conducted by Novick (2008) and Carr and Worth (2001) identify some of these questions to include questions related to health status, income level, and sexual orientation. Previous studies show that questions touching on these aspects are characterized by high refusal rates when other methods of data collection are used. Novick (2008, p. 391) corroborates that ‘telephones may allow respondents to feel relaxed and able to disclose sensitive information and no study has been conducted to show that telephone interviews produce lower quality data.’ This aspect is motivating researchers to use telephone interviews increasingly in conducting qualitative interviews (Novick 2008).
Call-backs to respondents
The likelihood of gathering sufficient data from the field using face-to-face and structured interviews may be limited by the selected respondents being engaged in diverse activities or absenteeism. However, using telephone interviews provides the researcher with an opportunity to call back and conduct the interview (Carr & Worth 2001).
Researchers are faced with the responsibility of ensuring that the data collected can be generalized to the total population. One of the ways through which this goal can be achieved is by constructing a representative sample. Telephone interviews allow researchers to build a representative sample by using random-digit-dialing (Carr & Worth 2001). This technique entails the random generation of the numbers to contact during the research process using a computerized program. The random-dialing capability increases the ability with which the researcher eliminates bias during the data collection process.
Limitations of telephone interviews
Telephone interviews limit the flexibility with which the researcher applies visual aids during the process of collecting data (Novick 2008). The interviewer may not be in a position to direct the respondent using visual aids such as pictures, maps, and other graphic layouts, unlike in face-to-face interviews. Novick (2008, p. 391) emphasizes that the ‘absence of visual cues via telephone is thought to result in loss of contextual and non-verbal data and to compromise rapport, probing, and interpretation of responses.’ Consequently, the researcher might not be in a position to collect data from interviewees with limited levels of education.
The likelihood of gathering sufficient data may be affected by the respondents’ opinions. Some interviewees may perceive the interview to be a lengthy process and decide to hang up. Novick (2008) asserts that most individuals prefer being interviewed for short durations.
Ary, D, Jacobs, L & Razavich, 2002, Introduction to research in education, Thomson Learning, Wadsworth.
Becker, S, Bryman, A & Ferguson, H 2012. Understanding research for social policy and social work themes, methods and approaches, Policy Press, New York.
Bowling, 2005, Handbook of health research methods; investigation, measurement and analysis, Maidenhead Open University Press, Maidenhead.
Carr, E & Worth, 2001, ‘The use of the telephone interviews for research,’ Nursing Times Research, vol. 6, no.2, pp. 511-526.
Cohen, L, Manion, L & Morrison, K 2007, Research methods in education, Routledge, London.
Cope, C 2006, Beneath the surface; the experience of learning about information systems, Informing Science Institute, Santa Rosa.
Craighead, W & Nemeroff, C 2004, The concise Corsini encyclopedia of psychology and behavioral science, John Wiley and Sons, Hoboken.
Curry, L, Ingrid, N & Bradley, E 2009, ‘Qualitative and mixed methods provide unique contributions to outcomes research,’ Circulation, vol. 11, no. 10, pp., 1442-1452.
Denscombe, M 2010, The excellent research guide; for small-scale social research projects, McGraw-Hill, London.
Dornyei, Z 2008, Questionnaires in second language research; construction, administration, Routledge, New York.
First, M 2008, Standardised evaluation in clinical practice, American Psychiatric, Washington, DC.
Fox, N 2009, Using interviews in a research project, National Institute for Health Research, Yorkshire.
Haas, P & Springer, F 2014, Applied policy research; concepts and cases, Routledge, London.
Hall, R 2008, Applied social research; planning, designing and conducting real-world research, Palgrave McMillan, South Yarra, Vic.
Harrell, M & Bradley, M 2009, Data collection methods; semi structured interviews and focus groups, RAND Corporation, Pittsburgh.
Hersen, M, Turner, S & Beidel, D 2011, Adult psychopathology and diagnosis, John Wiley and Sons, New York.
Lewis, M & Slack, N 2007, Operations management; critical perspective on business, Taylor & Francis, New York.
Lietz, P 2010, ‘Research into questionnaire design; a summary of literature’, International Journal of Market Research, vol. 52, no. 2, pp. 249-274.
Marsh, I 2014, Theory and practice in sociology, Routledge, New York.
Mitchell, M & Jolley, J 2010, Research design explained, Wadsworth, Belmont, CA.
Murugan, S 2013, Social work research and statistics, Social Works Department, PSGCAS, New York.
Patton, M 2002, Qualitative research and evaluation methods, Sage, Thousands Oak.
Novick, G 2008, ‘Is there a bias against telephone interviews in qualitative research,’ Research In Nursing and Health, vol. 31, no.6, pp. 391-398.
Smith, J, Michie, S, Allanson, A & Alwy, R 2000, ‘Certainty and uncertainty in genetic counselling: A qualitative case study’, Psychology and Health, vol.15, no.4, pp. 1–12.
Tabler, R 2008, The social construction of a special needs program for hurricanes, ProQuest, New York.
Willing, C 2013, Introducing qualitative research in psychology, Open University Press, Maidenhead.