Continuing Qualitative Data Collection During the COVID 19 Pandemic: Lessons Learnt from A Study Among Tribes in Manipur, India

Key


BACKGROUND
During the COVID-19 pandemic, researchers devised strategies in conducting research by following appropriate COVID-19 norms and regulations. Due to travel restrictions and lockdown imposed to control transmission, data collection for qualitative study, which requires prolonged engagement with respondents, was a challenge. Researchers decided to explore alternative data collection methods/tools, which could be both feasible and effective such as the use of telephones, WhatsApp Messenger for audio/video calls, Skype, zoom, etc. 1,2 Although face-toface interviews are the most preeminent way for collecting data in qualitative research, studies have also reported the feasibility and effectiveness of telephonic interviews or use of mobile phone for inter-viewing study participants. 3,4 During the current study entitled, "Development and Pilot testing of intervention strategies for smokeless tobacco and areca nut cessation among tribal women in Manipur, India", in mid-March 2020, all field activities were suspended due to nationwide lockdown that got extended till the end of May. 5 Due to severity of the pandemic and contagious nature of the virus, physical collection of data was not considered a plausible option but after consultations with experts it was decided to resume the project activities by implementing alternate strategy. This brief article presents the experiences of mobile phone usage in conducting Key Informant Interviews (KII) and In-depth Interviews (IDI)among tribal population, and the lessons learned during the implementation of the project amidst the lockdown.

STUDY CONTEXT
In the present study, qualitative data collection through KII, IDI, and focus group discussion (FGD) were planned to understand the socio-cultural context and nuances of smokeless tobacco (SLT) use among tribal women in Manipur, India. Key informants included those who were from the study area and had good knowledge of the use of SLT by men and women in that area. For IDI, participants included women who used SLT regularly.
Three villages located in Imphal (west) district in Manipur were selected for the study. There were approximately 210 households in these villages and the total population was around 1,025. The tribal groups inhabited in these villages are Rongmei, Kabui and Liangmaiand the majority were educated till 10 th standard.
Considering the widespread ownership of mobiles/telephones in the study sites 6,7 , interviews via telephone was suggested in conducting the interviews. This was followed by seeking approvals from the scientific as well as the ethics committee. The team decided not to conduct FGD as organizing meetings over the phone was not possible.
As part of the preparation for the interviews, mock interviews were carried out to reorient the investigators, and pilot testing was done among three respondents to assess the feasibility of conducting interviews over the phone. The interviews were recorded. Investigators also noted down their experiences such as participant interest, informed consent taking process, attention span, recording, and convenience of conducting telephonic interviews. Investigators found difficulty in recording the interviews on smart phones without an inbuilt voice recorder, which was sorted out by installing a thirdparty application.
A total of 7 KIIs and 5 IDIs were conducted through phones from September 2020 to February 2021. Verbal informed consent was taken and recorded with permission from the participants. Due consent was also taken from the participants to record the interview, which was later transcribed and translated for analysis purposes.
The following section is organized in terms of (I) feasibility of conducting the interview; (II) content of responses, and (III) challenges faced and strategies implemented.

(I)FEASIBILITY OF USING MOBILE PHONE FOR DATA COLLECTION
Majority of the participants reported that they were comfortable being interviewed over the phone and were willing to be engaged in long conversations. Few participants mentioned they preferred talking on the phone rather than participating in a face-toface interview as they felt they could talk without interference from others. Adhering to COVID-19 protocol and concerning safety of the participants and re-searchers, conducting a telephonic interview was a practical and viable option.

(II) CONTENT OF RESPONSES
The investigators did not perceive any difficulty in exploring and eliciting information on the study objectives. Similar findings were reported before 8 that there was no difference in the number, nature, and depth of responses of telephonic interviews when compared to face-to-face interviews. Another study found that in-depth telephone interviews took a shorter time to complete than face-to-face interviews 9 . Contrastingly, in our study, interviewing over mobile phones took slightly longer as most interviews lasted for more than an hour while face-to-face interviews lasted for about 40-60 minutes 9 . The interviews helped in yielding rich information regarding use of SLT among women. A detailed analysis of the responses, however, is beyond the scope of the present article.

(III) CHALLENGES AND STRATEGIES IMPLEMENTED
The challenges faced are presented according to the different phases of conduct of the telephonic interviews.

Pre-interview phase
Recruitment of participants for telephonic interviews was a challenge, and it was primarily done with the help of our previously contacted key informants. This method took more time as the key informants had to do phone calls for the recruitment process and schedule appointments. Persistent efforts were made to retrieve prospective participants through our key informants. Other studies suggested that recruitment done through face-to-face interaction were more successful 8 . Nevertheless, rescheduling over telephone was relatively hassle free, and some interviews were rescheduled as per the convenience of the participants. A short informal chat in advance was set with participants to establish rapport and briefly introduce the research topic before the interview. Similar findings were also reported in a qualitative study conducted among women alcohol users. 10

During the interview
The major limitation of the telephonic interviews was the technical glitch that occurred during the interviews. For example, network fluctuations due to which the respondents were asked for clarifications or to repeat their responses. There were pauses in between the conversations due to lack of comprehension from the other end. In the few initial interviews, distractions were also a constrain. Continuing the interview at a later appointment was hardly achievable as it disrupted the flow of the interview, and participants did not respond to the follow-up calls. As a result, one such interview was dropped from the study for being incomplete. After conducting few initial interviews, procedures were reviewed by the study investigators to implement strategies for more effective interviewing. The lack of visual cues like facial expressions and body language was also a challenge. An alternative to providing nonverbal reciprocity was to use words such as 'umm' (acknowledging the response) and repeating the participants' responses in short phrases. This helped to ascertain that the participants' words were being valued and listened to attentively. Collecting contextual data was not a part of our study and hence was not considered during the interview. However, for some information, we had to rely only on verbal responses, such as staining of teeth relating to SLT/areca nut use.

Post-interview phase
Recording the interviews using a third-party app was not very reliable. Two such interviews had error playing and could not be transcribed and translated and were dropped from the study. As a result, the researchers only used android phones with an inbuilt call recorder to record the interviews. Also, using a third-party app is not recommended due to the issue of data privacy. Hence, it is advisable to use mobile phones with inbuilt voice recorders for conducting the interviews. Use of digital recorders designed for telephone interviews could be helpful in avoiding technical errors. 10 The following similarities and differences in telephonic KIIs and IDIs were observed 1. Preparing a list of participants and duration of interviews: The ability in getting a contact list of both KII and IDI participants and the duration of the interviews were observed to be similar.

2.
Response: Number of women who denied taking part in the IDIs was more (8/13) compared to that of KIIs. This may be because IDIs required women to share personal experiences rather than provide a community perspective.

CONCLUSION
Our experience affirms that mobile phones can be considered as before a practical tool for successfully collecting in-depth qualitative data. This is more significant during pandemic, when researchers must comply with the protocols without compromising data quality. It further saves time and money, which is one of the advantages of telephonic interviews. The limitations include network failure, especially in places like Manipur that has harsh geographical terrain and weather (long and severe monsoon), call drops, technical issues in call recording and transcription. Face-to-face interaction provides a certain level of human non-verbal assurance and trust which remains absent in telephonic interviews. Moreover, telephonic interviews cannot retrieve contextual data, and the absence of visual cues deprives the re-searcher of valuable data. Despite these limitations, telephonic interviews can still be considered as one of the most effective and alternative methods to faceto-face interviews for collecting qualitative data.