qualitative research using online forums

An official website of the United States government

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Publications
• Account settings
• Advanced Search
• Journal List

An Online Forum As a Qualitative Research Method: Practical Issues

Eun-ok im , phd, mph, rn, cns, faan, wonshik chee , phd.

• Author information
• Copyright and License information

Address for Correspondence: Dr. Eun-Ok Im, Associate Professor, School of Nursing, The University of Texas at Austin, 1700 Red River, Austin, TX 78701, Telephone: (512) 471-7971/Fax: (512) 471-3688/e-mail: [email protected]

Despite positive aspects of online forums as a qualitative research method, very little is known about practical issues involved in using online forums for data collection, especially for a qualitative research project.

The purpose of this paper is to describe the practical issues that the researchers encountered in implementing an online forum as a qualitative component of a larger study on cancer pain experience.

Throughout the study process, the research staff recorded issues ranged from minor technical problems to serious ethical dilemmas as they arose and wrote memos about them. The memos and written records of discussions were reviewed and analyzed using the content analysis suggested by Weber.

Two practical issues related to credibility were identified: a high response and retention rate and automatic transcripts. An issue related to dependability was the participants’ easy forgetfulness. The issues related to confirmability were difficulties in theoretical saturation and unstandardized computer and Internet jargon. A security issue related to hacking attempts was noted as well.

The analysis of these issues suggests several implications for future researchers who want to use online forums as a qualitative data collection method.

Keywords: Issues, Online Forums, Cancer Patients, Qualitative Research

The use of electronic data collection methods in a research project allows people in different geographical areas to communicate and exchange their experiences and opinions without long-distance travel. Studies that have used electronic data collection methods have supported the feasibility of electronic data collection methods, the richness of electronically-collected data, informative and prompt responses of research participants, and decreased human errors ( Curl & Robinson, 1994 ; Fawcett & Buhle, 1995 ; Lakeman, 1997 ; Stanton, 1998 ).

Among various electronic data collection methods, online forums (web message boards) are prominent in terms of allowing asynchronous interactions, through which participants can join in the discussions at their convenience unlike other methods requesting synchronous interactions (e.g., chat groups). Asynchronous online forums have been reported to be observable, relatively easy to use, accessible, and safe ( Anderson & Kanuka, 1997 ; Hsiung, 2000 ). Online forums have been reported to provide a more comfortable forum for some people to discuss sensitive personal health issues, and have been suggested as a feasible alternative to traditional face-to-face focus groups ( Kramish et al., 2001 ). The effectiveness of the emotional and informational support that online forums provide has been reported ( Hsiung, 2000 ). Also, online forums require that a user register and either create or be given a password so that confidentiality of the data can be easily ensured. Compared with other types of Internet data collection methods that are accessible to the public without passwords, online forums are safer ( Kollock & Smith, 1999 ; Saba & McCormick, 2001 ). In addition, online forums are usually administered over a longer timeframe (e.g., 6 months, 1 year, 3 years), which reduces deadlines, gives researchers the flexibility to modify the forum’s technical aspects, and subsequently reduces the burden of participation and time pressure ( Kollock & Smith, 1999 ; Saba & McCormick, 2001 ).

Despite these positive aspects of online forums, very little is widely disseminated about practical issues in using online forums for data collection, especially for a qualitative research project. This paper describes the practical issues that the researchers encountered in an online forum as a qualitative component of a larger study on cancer pain experience. First, the study will be concisely described according the research process. Then, the method used to record and analyze the practical issues will follow. Third, the practical issues raised in conducting the online forum among cancer patients will be discussed. Finally, the implications for the future use of online forums as a qualitative research method will be proposed based on the discussion on the practical issues. In this paper, an issue means an incident, situation, or discussion in which two or more parties encounter problems working together or disagree on the process or expected outcomes. Some issues are resolved positively and constructively; some are not resolved at all, but are incorporated as a disagreement into the daily protocol of managing the online forum.

The issues in the online forum were viewed and analyzed in light of the evaluation criteria for rigor in qualitative studies. For qualitative studies, sets of evaluation criteria of rigor have been proposed and used ( Burns, 1989 ; Sandelowski, 1986 ; Whittemore, Chase, & Mandle, 2001 ). The evaluation of qualitative studies focuses on openness, scrupulous adherence to a philosophical perspective, thoroughness in collecting data, and consideration of all of the data in subjective theory development phase ( Burns & Grove, 1997 ). The evaluation criteria by Lincoln and Guba (1985) that have been widely used in nursing studies include four criteria for establishing the trustworthiness of qualitative data: credibility, dependability, confirmability, and transferability. Credibility refers to confidence in the truth of data, which can be ensured by prolonged engagement and persistent observation, triangulation, peer debriefing, member checks, searching for disconfirming evidence, and researcher credibility ( Lincoln & Guba, 1985 ). Dependability refers to the stability of data over time and over conditions, which can be ensured by stepwise replication and inquiry audit ( Lincoln & Guba, 1985 ). Confirmability refers to the objectivity or neutrality of the data so that two or more independent people can reach an agreement about the data’s relevance or meaning ( Lincoln & Guba, 1985 ). Transferability refers essentially to the generalizability of the data, that is, the extent to which the findings from the data can be transferred to other settings or groups ( Lincoln & Guba, 1985 ). In this paper, the focus of discussion on the issues is given to these four criteria for qualitative studies, including credibility, dependability, confirmability, and transferability.

The study was a descriptive Internet study on cancer pain experience of four major ethnic groups in the U.S. [Hispanics, non-Hispanic (N-H) Whites, N-H African Americans, and N-H Asians] including two phases: (a) an Internet survey and (b) online forums. The Internet survey was conducted among 438 cancer patients recruited through general and ethnic-specific Internet Cancer Support Groups, and the online forums were administered for 6-months’ duration among four ethnically different groups (about 30 members per group) recruited from among the Internet survey participants. To potentiate group cohesiveness, four ethnic-specific online forums were conducted because relative homogeneity in the composition of any particular focus group is usually advised, especially in the area of ethnic and racial background ( Merton, Fiske, & Kendall, 1990 ). A feminist approach was used to guide the research process. In this paper, only the issues raised in the online forum among 29 non-Hispanic White cancer patients are presented for the convenience of discussion, because each of the four ethnically different online forums indicated different issues. In general, 6–12 participants are thought to be ideal for a focus group discussion ( Stevens, 1996 ), and in qualitative study designs including this online forum method, intensive focus and in-depth data collection are of greater value than a large number of participants. Thus, 29 is an adequate number for the online forum discussion.

The participants were online cancer patients aged at least 18 years who can read and write English and whose self-reported ethnic identity is non-Hispanic (N-H) White. The mean age of the participants was 44.31 years ( SD =10.35), and about 86% were women. More than 80% of the participants were college graduates, and about 41% were employed. More than 50% reported that their family income was sufficient or more than sufficient, and about 58% were Protestants. About 97% were born in the U.S. About 52% perceived that they were healthy. The participants tended to be young, high income, highly educated cancer patients, which reflects other Internet studies that have reported that the online population tends to be a select group of people ( Im & Chee, 2004 ).

For the online forum, the researchers set a secure online forum site that only the participants could log into using their usernames and passwords. All the participants in this forum self-identified as Caucasian, lived in the U.S., and currently have or have had some form of cancer pain. The participants were asked to register on the forum site using a pseudonym to maintain their anonymity. Discussion topics such as verbal and nonverbal expressions of pain, cultural manifestations of pain expression, and gender differences in the expression of pain were posted to the forum site every 3 to 4 weeks. Respondents were asked to post their responses directly to the forum.

The responses entered by the participants were analyzed using a thematic analysis. First, line-by-line coding was done with a supplementary use of N-6. Then, categorization of the codes was completed, and idea categories were identified. Based on the idea categories, themes were extracted. Five major themes were identified: (a) “pain accompanies cancer”; (b) “culture more than ethnicity; (c) “women’s pain not seriously taken”; (d) “highly individualized pain experience with emotional pain”; and (e) “controlled or controlling.” First, the participants perceived that pain accompanies cancer throughout the diagnosis and treatment process. Pain was regarded as normal and a natural process of cancer. Second, the specific characteristics of the participants’ individual culture and its view of pain and cancer could result in different cancer pain experience even among White cancer patients. The participants viewed their culture regionally, labeling themselves as “non-traditional Southerners,” “Appalachians,” or “proper Northerners.” They also defined themselves according to forms of social organization, particularly economic (work-related), such as a blue-collar, steel mill community culture, or a rural farm community culture. Third, the participants complained that women’s pain was not taken seriously by health care providers. Fourth, the cancer patients’ pain experience was highly individualized and ranged from “no pain” to “the most extreme pain that could be imaginable.” Regardless of the level of pain that the participants were experiencing, all of them reported the emotional pain that they had gone through. Finally, the findings indicated that the participants wanted to control their own bodies and lives including cancer pain management rather than being controlled by their health care providers or the disease itself. The findings suggest that researchers and health care providers need to respect cancer patients’ own views and experiences related to pain during cancer pain management process, consider diversities within the culture of cancer patients, and view cancer as a chronic condition.

While conducting the online forum among non-Hispanic White cancer patients, research staff members including the moderator and co-moderators had weekly research meetings. During the meetings, the problems and/or concerns in the administration of the online forum were discussed and the discussion contents were written. For example, when several participants kept complaining about their problems in log-in, the complaints and possible reasons for the complaints were discussed in the research meetings. The discussion was recorded, and the research staff members wrote their own memos on the complaints and possible reasons. In this instance as in others, the research team members recorded the discussion during weekly research meetings, and each of the members wrote memos on the issues as they arose.

These memos, as well as written records of the discussion during weekly research meetings, were reviewed and analyzed using the content analysis method suggested by Weber (1990) . In the content analysis method by Weber, the unit of analysis is individual words. The transcripts are analyzed by line-by-line coding. Then, the codes are categorized according to the contents. The same procedures were used to analyze the issues that are reported in this paper. The transcripts were the memos and written records of the weekly research meetings. The memos and written records were analyzed using line-by-line coding first. Then, the codes were categorized according to the contents, and idea categories were developed from the categorization process. The idea categories that were initially developed were discussed in research meetings, and the idea categories of the issues that all research staff members agreed upon were chosen as the final idea categories. The final idea categories included: (a) a high response and retention rate; (b) forgetfulness; (b) highly individualized responses; (c) difficulty in theoretical saturation; (d) computer and Internet jargon; (e) automatic transcripts; and (f) security issues. As mentioned above, throughout the analysis process, the evaluation criteria for rigor in qualitative research suggested by Lincoln and Guba were considered as important components of research. The findings from the analysis are discussed with these idea categories as follows.

Practical Issues

Credibility issues.

As defined above, credibility means confidence in the truth of data that can be ensured through various ways. In the study reported here, two major issues related to credibility were identified: (a) a high response and retention rate; and (b) automatic transcripts.

First, the study indicated a high response and retention rate. According to the online forum’s ability to track how often individuals came to the site and how long they stayed, the study produced a high response and retention rate. This indicates, albeit indirectly, that the collected data is highly credible, since it implies that (a) the participants were interested in the study; (b) they were engaged in the discussion for a prolonged time; and (c) the interactions could be considered more authentic than in a group with a low response and retention rate.

As described above, the participants of the online forum discussion were recruited from among the participants of the larger Internet study. When a participant completed the Internet survey, she/he was asked if she/he would be interested in participating in an additional online forum discussion for the next 6 months. When she/he clicked “I am interested in additional online forum discussion,” she/he was later contacted by the researchers by e-mail. The response rate of the potential participants was very high: among 30 contacted cancer patients, 29 patients agreed to participate in the online forum discussion. Out of 29 respondents who agreed to participate in the online forum, a total of 25 (86%) were retained during the 6-month period of the online forum. A total of 291 messages were posted to the forum site. 16 participants posted messages to all 10 of the topics spanning the 6-month period.

Before starting the online forum discussion, it was assumed that the drop-out rate of the online forum participants would be around 30%, given the high drop-out rate (66%) in an e-mail group discussion among cancer patients reported by Im and Chee (2003) . Thus, to retain at least 10 participants, 30 participants were recruited. As mentioned above, 6 to 12 participants are thought to be ideal for a focus group discussion ( Stevens, 1996 ). Yet the finding indicated a high retention rate that was hardly expected before starting the online forum discussion.

Second, automatic transcripts directly printed out from the online forum site were an asset of the online forum method, which can increase the credibility of the data. Yet the automatic transcripts were not as perfect as expected. In most qualitative studies, transcription of audiotaped interviews and conversation has become the norm ( Sandelowski, 1994 ). Consequently, in the process of ensuring the rigor of a qualitative study, the form and accuracy of the transcription process play a very important role in determining the data analysis process, which will influence the degree of dependability ( MacLean, Meyer, & Estable, 2004 ). Indeed, the transcription process has been considered to be central to the qualitative data analysis, which frequently influences what the researcher and transcriptionist preserve from the taped speech. Yet as Sandelowski (1994) posited, transcribed text can never totally capture the complexity of the interaction nor be completely error free. Even verbatim transcription that is supplemented with the addition of researchers’ notation of nonverbal behavior has been cited as critical to the reliability and to the validity and trustworthiness of qualitative research ( Easton, McComish, & Greenberg, 2000 ; Seale & Silverman, 1997 ). In the online forum discussion presented in this paper, the transcription process was automatic since the participants themselves needed to enter their answers through the Internet, and the computer server could generate the transcripts automatically from its databases. However, that did not mean that the transcripts were always accurate or could produce dependable and trustworthy findings.

The first issue that researchers encountered regarded which format the transcripts would be coded in. The original plan was to analyze the online forum data only using N-6 (which is the most recent version of the NUDIST qualitative data analysis software), so the data was saved as a simple text file and imported it to the N-6 program. However, during this importing process, some jargon including the “smileys” in the original messages were lost, and some important cues from the data (e.g., space, line breaks, etc) were lost. Subsequently, rather than only using the N-6 program that possibly had changed the original meanings of the transcripts, a line-by-line coding was used with the transcripts printed out directly from the online forum site.

Since the online forum is based on non-face-to-face interactions, the automatically generated transcripts do not contain non-verbal cues that could have been obtained only through face-to-face interviews. Therefore, some important aspects of qualitative research methodology were missing from the study presented in this paper. For example, in the case of face-to-face interviews, the researcher/interviewer can catch the emotional status of the participants through their facial expression and record exactly what happened in the process of the interviews through memos and field notes. However, in the online forum discussion, there was no way to identify the emotional status of the participants or contextual cues. The only way to identify such status was the line breaks, periods, pictures, and written cues that were also sometimes very vague and difficult to interpret. In the near future, cameras and microphones may become the normal tools of Internet interactions, allowing non-verbal cues to be easily obtained through the Internet, thereby minimizing this issue.

Dependability Issues

The study indicated an issue related to dependability of the study: the participants’ easy forgetfulness. For example, the online forum site was designed to be accessed only using user names and passwords that the participants chose when they registered. Throughout the 6-month period of online forum discussion, many participants forgot their user names and/or passwords. One participant needed help to retrieve her username and passwords 3 different times, and she finally wanted to change her username and passwords. Many of the participants complained about their forgetfulness, and linked their forgetfulness with cancer treatments, especially chemotherapy. Some respondents even called their forgetfulness “chemo-brain,” which implied that their brain cells had been negatively affected by chemotherapy to the point that their brains no longer worked normally. This forgetfulness might reduce the stability of data over time and over conditions, which also made stepwise replication and inquiry audit difficult. In addition, this forgetfulness might also decrease the credibility of the study since some of the participants could forget that they had already entered data. In the same way, it is reasonable to expect that patients with other conditions might interact with online environments more or less successfully, based on the cognitive and psychosocial effects of their conditions.

Along with their forgetfulness, the participants also experienced some logistical problems. Sometimes, the participants posted a same message multiple times to the forum site. In two cases, the participants complained that their messages got “lost,” did not show up on the forum site, or needed to be re-written. On five occasions, the participants did start a new thread, which actually related to the topic that they were discussing. To resolve logistical problems, the researcher often needed to access the forum site and sometimes delete the replicated messages. The researchers also needed to frequently remind participants about the topics, clarifying them and the questions asked in each.

Confirmability

As defined above, confirmability means the objectivity or neutrality of the data so that two or more independent people can reach an agreement about the data’s relevance or meaning. Therefore, lack of theoretical saturation in a qualitative study can easily threaten its confirmability, since a different researcher could possibly come up with different idea categories. In many qualitative studies, especially grounded theory studies, researchers aim for theoretical saturation to ensure credibility, dependability, confirmability, and transferability. According to Pandit (1996) , during initial data collection when the main categories are emerging and a full “deep” coverage of the data is necessary, theoretical sampling requires only collecting data on categories for the development of properties and propositions. Thus, the criterion for judging when to stop theoretical sampling is “theoretical saturation,” as termed by Glaser and Strauss (1967) . Theoretical saturation refers to a situation in which no additional data are needed for the researcher to develop the properties of the category. According to Glaser and Strauss (1967) , when the researcher sees similar cases/instances repetitively, she/he becomes empirically confident that a category is saturated. Theoretical saturation is achieved when no new themes or issues arise regarding a category of data, and when the categories are well-established and validated ( Glaser, 1978 ).

The online forum discussion poses a challenge for theoretical saturation. Online forums, such as the one used here, encourage asynchronous interaction; that is, the interaction does not happen immediately but according to the participants’ separate schedules and so may be delayed or spread out in time. Usually, researchers need to be actively and directly involved in a discussion to get saturated data. Thus, real-time data collection is much more appropriate for theoretical saturation compared with later data collection. In other words, for theoretical saturation, the researcher and participant need to schedule a time for the participant to respond in real time so that both the researcher and the participant are available simultaneously. Through this, the researcher can clarify confusing written comments or probe comments that require further explication at the time the participant responds them ( Strickland, Moloney, Dietrich, Myerburg, Cotsonis, & Johnson, 2003 ). This could definitely increase the quality of data and the researcher’s control over the data ( Strickland et al., 2003 ), through which theoretical saturation could be obtained.

In the study presented in this paper, however, it was difficult to achieve such saturation because the participants visited the website only once a month and were required to post only two messages in order to receive reimbursement for their participation. To get saturated data, the participants needed to go back to the online forum site more frequently and more immediately respond to the questions that the researchers raised. Thus, the online forum discussion would be a very good method for content analysis rather than other qualitative approaches that require theoretical saturation. For theoretical saturation, synchronized interactions between researchers and research participants would be ideal, which could be possible by using synchronous Internet technologies (e.g., chat groups, instant messages, etc).

Another confirmability issue was unstandardized computer and Internet jargon. Internet users have always used computer and other technical terms, and as the number of Internet users increases, the size of the population using Internet-related jargon also increases. For instance, a Web search using Google.com with a key word of Internet jargon, 1,720,000 websites were retrieved. There are many websites that are trying to summarize and develop a dictionary for computer and Internet jargon, but the meanings of the various terms are relatively unformalized and uncodified. Even for “smileys”—a “smiley” is a sequence of characters on the computer keyboard, which represents the eyes, the dash represents the nose and the right parenthesis represents the mouth—there are a number of definitions available. A “smiley” usually tells what the sender really means when the sender makes an offhand remark, and smileys are also called “emoticons” because they intend to convey emotion ( Netlingo.com, 2005 ). Some examples of Internet dictionaries for smileys can be found at http://www.netlingo.com/smiley.cfm , http://www.cis.columbia.edu/glossary.html , http://www.stevegrossman.com/jargpge.htm , and http://www.stevegrossman.com/jargpge.htm .

In the online forum discussion reported in this paper, not much computer and Internet jargon was noticeable. However, many “smileys” were frequently used in the discussion to indicate the participants’ emotions. Since the topic of the discussion was pain, it was thought that the “smileys” were important nonverbal cues that the researchers might otherwise get in face-to-face interviews. Thus, each of the “smileys” was assigned a code for the data analysis process based on the suggestions in http://www.netlingo.com/smiley.cfm . This coding was more particular to this study, given the unformalized nature of Internet jargon as a whole. The sample codes of the “smileys” that we set for the coding process are provided in Table 1 . Yet not all of them were used by the participants in the online forum. This might raise questions about the confirmability of the study, if subsequent researchers relied on different definitions of these (and other) Internet terms.

Sample codes of the “Smileys”

Transferability Issues

The participants’ responses to the online forum discussion could be characterized as highly individualized, which could raise a question on the generalizability of the data, which puts into doubt the transferability of the study. Most of the participants of the online forum responded favorably to the forum postings. Several times, respondents posted a second message right after reading messages from other participants. However, some of the participants just posted 1–2 lines in a message, which sometimes discouraged other participants. On the whole, the participants’ messages ranged in length from 1 or 2 lines to more than 90 lines.

Some participants initiated new topics based on their own interests that were pertinent to their cancer diagnoses. Some participants discussed intimate topics such as sexuality without prompting from the moderator, perhaps due to their anonymity and the security of the site. At the same time, some participants raised questions about the topic and posted emotionally charged messages. For example, two participants raised a question on a posted question on sexual identity. They wanted the researchers to clarify why this question was asked and what this question implied. With the clarification on the question (the question was given to explore what are possible factors influencing cancer pain experience), they eventually posted their responses.

Some participants did not post their messages related to the posted topic for the month and were considered inactive, so every month the researchers reminded them by e-mail encouraging them to visit the online forum site and post their messages. With the reminding messages, in two cases, the participants who had not posted their messages for several months came back to the online forum site and completed their messages.

Security Issues

Despite systematic efforts to develop standardized guidelines for human subject protection in Internet research ( Frankel & Siang, 1999 ; Klemm & Nolan, 1998 ; U.S. Department of Health and Human Services [USDHHS], 2001 ), there still exist no standardized guidelines. Yet, in general, computer security usually refers to the following four areas: confidentiality , integrity , authentication, and availability ( Kinkus, 2002 ). Confidentiality refers to whether or not information can be accessed by unauthorized parties. Integrity refers to whether or not information is protected against unauthorized changes; many incidents of hacking can threaten the integrity of databases. Authentication means that users are who they claim to be, and availability means that resources are accessible by authorized parties. With these four areas, privacy is the most important aspect of computer security ( Kinkus, 2002 ). Even when users think that they have nothing to hide when they are registering with an online forum site, privacy on the Internet is important for protecting personal information, whether sensitive or not. The reason is that even small pieces of related information from different sources can be easily linked together to form important information that some industries want to know ( USDHHS, 2001 ).

In the online forum discussed here, maintaining the forum site’s security and confidentiality was a challenge to the researchers during the data collection process. Because several hacking attempts were noted, it was necessary to regularly update the software including the firewall installed in the project server that hosted the online forum site, in order to prevent any potential virus infection or hacking attempts. More importantly, one of the participants emailed the web address of the online forum site with an electronic copy of the postings made in the forum site to all his friends/colleagues in his electronic address book, which was detected by the researchers because the e-mail was mistakenly sent to one of the researchers. The engineering co-investigator of the project also detected that the forum site was once accessed without usernames and passwords by non-participating outsiders, possibly an attempt to hack. As soon as the hacking efforts were detected, the forum site was closed to outsiders with additional programming to block this access.

Conclusion and Implications

This paper has discussed practical issues that arose in using an online forum as a qualitative research method among cancer patients in order to explore their cancer pain experiences. The issues were viewed and analyzed while considering the evaluation criteria for rigor in qualitative research including credibility, dependability, confirmability, and transferability of the study. Based on the discussion on the issues, we want to conclude this paper with the following suggestions for the future use of online forums as a qualitative research method.

First, researchers need to carefully consider if the online forum could be the best data collection method for their research questions. As discussed above, it was difficult to achieve theoretical saturation in the online forum, which means that theoretical sampling was impossible. Also, researchers’ immediate involvement in the discussion was difficult because the participants did not come to the site very often. Thus, researchers who want to use online forums need to carefully consider the costs and benefits of doing so, clarify the purpose for using the forum, and defining how it fits with the most critical aspect of the qualitative approach in their research. Indeed, as Strickland et al. (2003) posited, Internet-based data collection could be used only for some specific qualitative approaches.

Second, researchers need to consider the characteristics of their participants. As the findings indicate, cancer patients frequently forgot their usernames and passwords, and the researchers needed to provide technical support whenever the participants contacted them. If the researchers’ target population had characteristics that might inhibit administering the online forum (e.g., elderly populations, patients with cognitive and/or memory problems), they might reconsider the feasibility of the online forum.

Third, researchers need to define computer and Internet jargon before starting an online forum discussion so that consistent definitions for the jargon can be used throughout the research process. As discussed above, the meanings of the jargon are relatively unformalized and uncodified. Thus, in the study presented in this paper, the “smileys” needed to be defined for the coding process in order to catch non-verbal cues from the participants.

Fourth, researchers need to decide how to deal with the automatic transcripts from the online forum site and/or databases before starting the online forum discussion. In the study presented in this paper, the transcripts were directly printed out from the online forum site in order to obtain more non-verbal cues implied in the online forum discussion. Depending on the importance of the non-verbal cues, the researchers might consider simply generating these transcripts into a text format and importing it into a computerized data analysis program.

Finally, researchers need to continuously consider potential security issues throughout the data collection process and carefully monitor online interactions. As in other Internet studies, the study presented in this paper indicated important security issues involved in the online forum. Also, researchers need to update and upgrade computer security measures and regularly update their skills and knowledge related to Internet research. Internet technologies are changing daily, and some issues that are important today could be minimal in the near future.

Acknowledgments

This analysis was conducted as part of a larger study funded by the National Institute of Health (5R01 NR007900).

Contributor Information

Eun-Ok Im, School of Nursing, The University of Texas at Austin, 1700 Red River, Austin, TX 78701, Telephone: (512) 471-7971/Fax: (512) 471-3688/e-mail: [email protected].

Wonshik Chee, Department of Mechanical Engineering, College of Engineering, University of Texas at Austin, Austin, TX 78701, Telephone: (512) 773-0006/Fax: (210) 458-5589/e-mail: [email protected].

• Anderson T, Kanuka H. On-line forums: New platforms for professional development and group collaboration. Journal of Computer Mediated Communication. 1997;3(3) NP. [ Google Scholar ]
• Burns N. Standards for qualitative research. Nursing Science Quarterly. 1989;2:44–52. doi: 10.1177/089431848900200112. [ DOI ] [ PubMed ] [ Google Scholar ]
• Burns N, Grove SK. The practice of nursing research: Conduct, critique, & utilization. 3. Philadelphia: WB Saunders; 1997. p. 304. [ Google Scholar ]
• Curl M, Robinson DK. Handheld computers in clinical audit: A comparison with established pen and paper method. International Journal of Health Care Quality Assurance. 1994;7:16–20. doi: 10.1108/09526869410059709. [ DOI ] [ PubMed ] [ Google Scholar ]
• Easton KL, McComish JF, Greenberg R. Avoiding common pitfalls in qualitative data collection and transcription. Qualitative Health Research. 2000;1:703–707. doi: 10.1177/104973200129118651. [ DOI ] [ PubMed ] [ Google Scholar ]
• Fawcett J, Buhle EL., Jr Using the Internet for data collection. An innovative electronic strategy. Computers in Nursing. 1995;13(6):273–279. [ PubMed ] [ Google Scholar ]
• Frankel MS, Siang S. [Retrieved March 24, 2004];Ethical and legal aspects of human subjects research on the Internet. 1999 from http://www.aaas.org/spp/dspp/sfrl/projects/intres/report.pdf .
• Glaser BG, Strauss AL. The discovery of grounded theory. Chicago: Aldine; 1967. [ Google Scholar ]
• Glaser BG. Theoretical sensitivity. Advances in the methodology of grounded theory. Mill Valley (CA): The Sociology Press; 1978. [ Google Scholar ]
• Hsiung RC. The best of both worlds: An online self-help group hosted by a mental health professional. Cyber Psychology & Behavior. 2000;3(6):935–950. [ Google Scholar ]
• Im EO, Chee W. Feminist issues in e-mail group discussions among cancer patients. Advances in Nursing Science. 2003;26(4):287–298. doi: 10.1097/00012272-200310000-00006. [ DOI ] [ PubMed ] [ Google Scholar ]
• Im EO, Chee W. Issues in Internet survey among cancer patients. Cancer Nursing. 2004;27(1):34–42. doi: 10.1097/00002820-200401000-00005. [ DOI ] [ PubMed ] [ Google Scholar ]
• Kinkus JF. [Retrieved March 26, 2005];Science and Technology Resources on the Internet Computer Security. 2002 from http://www.istl.org/02-fall/internet.html .
• Klemm P, Nolan MT. Internet cancer support groups: Legal and ethical issues for nurse researchers. Oncology Nursing Forum. 1998;25:673–676. [ PubMed ] [ Google Scholar ]
• Kollock P, Smith MA. Communities in cyberspace. In: Smith MA, Peter K, editors. Communities in cyberspace. London: Routledge; 1999. [ Google Scholar ]
• Kramish CM, Meier A, Carr C, Enga Z, James AS, Reedy J, et al. Health behavior changes after colon cancer: A comparison of findings from face-to-face and on-line focus groups. Family and Community Health. 2001;24(3):88–103. doi: 10.1097/00003727-200110000-00010. [ DOI ] [ PubMed ] [ Google Scholar ]
• Lakeman R. Using the Internet for data collection in nursing research. Computers in Nursing. 1997;15:269–275. [ PubMed ] [ Google Scholar ]
• Lincoln YS, Guba EA. Naturalistic inquiry. Beverly Hills, CA: Sage; 1985. [ Google Scholar ]
• MacLean LM, Meyer M, Estable A. Improving Accuracy of Transcripts in Qualitative Research. Qualitative Health Research. 2004;14(1):113–123. doi: 10.1177/1049732303259804. [ DOI ] [ PubMed ] [ Google Scholar ]
• Merton RK, Fiske M, Kendall PL. The focused interview: A manual of problems and procedures. 2. New York: The Free Press; 1990. [ Google Scholar ]
• Netlingo.com. [Retrieved March 26, 2005];Smilyes and emoticons. 2005 from http://www.netlingo.com/smiley.cfm .
• Pandit NR. The Creation of Theory: A Recent Application of the Grounded Theory Method. [Retrieved March 28, 2005];The Qualitative Report. 1996 2(4) from http://www.nova.edu/ssss/QR/QR2-4/pandit.html .
• Saba VK, McCormick KA. Essentials of computers for nurses: Informatics for the new millennium. 3. New York: McGraw-Hill; 2001. [ Google Scholar ]
• Sandelowski M. The problem of rigor in qualitative research. Advances in Nursing Science. 1986;8:27–37. doi: 10.1097/00012272-198604000-00005. [ DOI ] [ PubMed ] [ Google Scholar ]
• Sandelowski M. Focus on qualitative methods: Notes on transcription. Research in Nursing & Health. 1994;17:311–314. doi: 10.1002/nur.4770170410. [ DOI ] [ PubMed ] [ Google Scholar ]
• Seale C, Silverman D. Ensuring rigour in qualitative research. European Journal of Public Health. 1997;7(4):379–384. [ Google Scholar ]
• Stanton JM. An empirical assessment of data collection using the Internet. Personnel Psychology. 1998;5(1):709–725. [ Google Scholar ]
• Stevens PE. Focus groups: Collecting aggregate-level data to understand community health phenomena. Public Health Nursing. 1996;13(3):170–176. doi: 10.1111/j.1525-1446.1996.tb00237.x. [ DOI ] [ PubMed ] [ Google Scholar ]
• Strickland OL, Moloney MF, Dietrich A, Myerburg S, Cotsonis GA, Johnson RV. Measurement issues related to data collection on the World Wide Web. Advances in Nursing Science. 2003;26:246–256. doi: 10.1097/00012272-200310000-00003. [ DOI ] [ PubMed ] [ Google Scholar ]
• U.S. Department of Health and Human Services. [Retrieved March 22, 2004];Standards for privacy of individually identifiable health information-The HIPAA privacy rule. 2001 from http://www.hhs.gov/ocr/hipaa .
• Weber RP. Basic content analysis. 2. Newbury Park (CA): SAGE; 1990. [ Google Scholar ]
• Whittemore R, Chase SK, Mandle CL. Validity in qualitative research. Qualitative Health Research. 2001;11(4):522–537. doi: 10.1177/104973201129119299. [ DOI ] [ PubMed ] [ Google Scholar ]
• View on publisher site
• PDF (46.5 KB)
• Collections

Similar articles

Cited by other articles, links to ncbi databases.

• Download .nbib .nbib
• Format: AMA APA MLA NLM

Add to Collections

An online forum as a qualitative research method: practical issues

Affiliation.

• 1 School of Nursing, University of Texas at Austin, Austin, TX 78701, USA. [email protected]
• PMID: 16849979
• PMCID: PMC2491331
• DOI: 10.1097/00006199-200607000-00007

Background: Despite the positive aspects of online forums as a qualitative research method, very little is known on the practical issues involved in using online forums for data collection, especially for a qualitative research project.

Objectives: The aim of this study was to describe the practical issues encountered in implementing an online forum as a qualitative component of a larger study on cancer pain experience.

Methods: Throughout the study process, the research staff recorded issues ranging from minor technical problems to serious ethical dilemmas as they arose and wrote memos about them. The memos and written records of the discussions were reviewed and analyzed using content analysis.

Results: Two practical issues related to credibility were identified: (a) a high response and retention rate and (b) automatic transcripts. An issue related to dependability was the participants' forgetfulness. The issues related to confirmability were difficulties in theoretical saturation and unstandardized computer and Internet jargon. A security issue related to hacking attempts was noted as well.

Discussion: The analysis of these issues suggests several implications for future researchers who want to use online forums as a qualitative data collection method.

Publication types

• Research Support, N.I.H., Extramural
• Computer Security
• Cross-Cultural Comparison
• Data Collection / methods*
• Neoplasms / ethnology
• Neoplasms / psychology
• Pain / ethnology
• Pain / psychology
• Qualitative Research*
• Reproducibility of Results
• United States

Grants and funding

• R01 NR007900/NR/NINR NIH HHS/United States
• R01 NR007900-05/NR/NINR NIH HHS/United States
• 5R01 NR007900/NR/NINR NIH HHS/United States