Henrietta O'Connor and Clare Madge (2001) 'Cyber-Mothers: Online Synchronous Interviewing using Conferencing Software'

Sociological Research Online, vol. 5, no. 4, <http://www.socresonline.org.uk/5/4/o'connor.html>

Abstract

The potential of the Internet as a valuable methodological tool for social science research is increasingly being recognised. This paper contributes to the debate surrounding virtual synchronous interviews and the value of online research. Specifically it introduces the use of a software conferencing technique - Hotline Connect - and discusses the implications of using the technique for Internet- based research. In particular issues of interview design, developing rapport, the role of insiders and outsiders in the research process, language use and the virtual interface are considered. The paper draws on the experience of a recent research project entitled 'Cyberparents' and concludes that the use of conferencing software holds great potential for synchronous online interviewing. However, this must be combined with sensitive, ethical handling of both the research process and the data to overcome both the weaknesses of this particular method and those inherent in any interviewing situation.

Keywords:

Conferencing Software; Cyber-mothers; Internet Methodologies; Online Parenting Community.; Virtual Synchronous Interviews

Introduction

1.1 The potential of the Internet as a valuable methodological tool for social science research is increasingly being recognised (Mann and Stewart, 1999). Cyberspace provides a virtual social arena which is not bound by temporal and spatial restrictions where researchers can, therefore, interact with participants in ways which may not be possible in the real world. For example, the Internet offers potential for interfacing with groups difficult to reach via conventional research approaches (Coomber, 1997; Hodkinson, 1999) and for interaction with participants who are widely geographically distributed (O'Lear, 1996). Internet methodologies offer interesting possibilities for administering electronic surveys (Comley, 1996; Smith, 1997), reconsidering sampling strategies (Smith, 1997) conducting qualitative research (Gaiser, 1997; Smith, 1998; Chen and Hinton, 1999; Mann and Stewart, 2000) and virtual ethnography (Hine, 2000). Qualitative research techniques such as participant observation and discourse analysis have also been used in a virtual setting to study online communities, specialised websites and list servers (Sharf, 1997). Virtual ethnographic analysis of the role of the Internet has been conducted by Hine (2000) and ward (1999) uses the term 'cyber-ethnography' to describe the technique used in her study of two online communities.

1.2 The suitability of cyberspace as an interview venue has also been explored. Asynchronous interviews, characterised by the fact that they do not take place in 'real time', have received the most attention to date. These are usually facilitated by email (Bennett, 1998; Christensen, 1999), bulletin board services (ward, 1999) or in a listserv environment (Gaiser, 1997). In contrast, according to Chen and Hinton (1999) there has been little academic assessment of the advantages and limitations of synchronous or 'real time' online interviewing. Indeed, with the exception of the work of Gaiser (1997) and Mann and Stewart (1999) on carrying out online focus group interviews and Smith (1998) and Chen and Hinton (1999) on virtual interviews, there have been few empirical studies.

1.3 The paper then, aims to address this research gap and to contribute to the debate surrounding virtual synchronous interviews and the value of online research. Specifically it will introduce the use of a software conferencing technique - Hotline Connect - and will discuss the implications of using the technique for Internet-based research. The paper draws on experience from the Cyberparents research project which used a web-based survey and virtual synchronous interviews to collect data. The web-based survey was used to try and identify general patterns of use for one specified website. More in-depth data was gathered through semi-structured interviews which developed themes introduced in the questionnaire.

1.4 The paper is divided into four sections. First, the Cyberparents project and its research context is described. Second, the research process is outlined with particular emphasis placed on the interview procedure. The third section analyses the implications of carrying out virtual interviews using conference software with particular emphasis on interview design, developing rapport, the role of insiders and outsiders, language use and the virtual interface. Lastly, we conclude that while the use of conferencing software holds great potential for synchronous online interviewing, the process has limitations which present challenges to the virtual researcher. However, in certain specific contexts, such as in the study of online communities, the method, if adapted appropriately, has much to offer.

Cyberparents Research Project

2.1 It is now widely recognised that the use of the Internet as a source of health information is increasing (Rippen, 1997; Silke et al., 1998). Parents form a high proportion of those who are using the Internet to seek health care advice for themselves and also, increasingly, for information about their children's health and well-being. The benefits of using the Internet for disseminating parenting skills have been recognised by health professionals, aware that the resources available for teaching new parents about the day-to-day aspects of child care are limited (Lamp and Howard, 1999). A study by Moran et al. (1997) found that nearly 90% of first time mothers who had recently given birth felt that they would have benefited from receiving more information on parenting skills. Whilst there are a range of ante-natal parenting classes available, it is often not until parents are at home and experiencing difficulties that they are receptive to learning about day to day child care. It seems that at this point parents are turning to the Internet for information and support; a trend reflected in the growth of the number of parenting websites (Lamp and Howard, 1999; Williams, 1999; Moorhead, 2000).

2.2 The Cyberparents project arose out of the recognition that whilst the number of parenting websites has increased, one aspect of this type of Computer Mediated Communication (CMC) that has received little research attention is the impact that such websites may have on parenting practices. The role of CMC in forming virtual communities is now well-documented (Rheingold, 1994; Jones, 1995, 1997, 1998; Loader, 1997; Baym, 1998). Jones (1997) illustrated how certain disadvantaged or marginal groups are using the Internet as a means of social support. CMC is resulting in the formation of 'virtual communities' based around common interests such as an illness or disease, for example, breast cancer (Sharf, 1997), a common experience such as alcoholism (King, 1994), or a new experience (becoming a parent) rather than geographical proximity.

2.3 Babyworld's mission is to support a community where: 'new and expectant parents can share experiences and support, women can learn about their bodies, their baby, and childbirth and parents can celebrate the joy of a new life' (<http://www.babyworld.co.uk/> March 2000). Although membership of the 'community' is encouraged it is possible to drop in or 'lurk' and post messages without formally joining the community, although certain areas are restricted to members only. When originally launched the site provided a range of four main 'services' to its online visitors: an online shopping facility, a discussion forum, an interactive 'ask the experts' section and a reference section providing information on topics such as pregnancy.

2.4 During the time period in which the research was carried out (June 1998 - June 1999) the site expanded and later relaunched in April 1999. It now offers further facilities and the areas outlined above have been refined and improved. For example, the discussion forums are now organised in to a series of popular themes (see: <http://babyworld.co.uk/bbs>). There are also new areas including online ante-natal classes, a photo gallery where members can post photos of their babies and an area for personal birth stories.

The Cyberparents project^[1] focused on one pioneer UK parenting website: <http://www.babyworld.co.uk> (see Figure 1). Babyworld was selected as the case study site because it was the first UK based parenting website, launched in 1995. At the time the research project was set up (June 1998) this site had a high profile amongst parents as it was advertised in parenting handbooks distributed by hospitals to new parents (Rodway, 1997). Current usage figures show that the site receives an average of 160 000 visitors and four million page impressions per month.

The Research Process

3.1 Recent surveys show that there are around 19 million Internet users in the UK, a dramatic increase from the estimated one million users in June 1997 (NUA surveys and Nielsen Net Ratings, September 2000). Although access to the Internet is constantly widening, considerable socio-economic barriers to use persist. Graham and Marvin (1996) show how access to a computer is linked with household income and socio-economic background. Gender is another significant factor, more men than women were online initially although this balance has been changing as the total number online has increased. Ethnicity and age also have a role to play, those online have tended to be predominantly white and young, the majority under 35 years old (Mann and Stewart, 2000). With increasing access to the Internet this bias is changing and the overall user group moving towards a more representative cross section of society. The Cyberparents research was carried out in 1998 when estimates of UK net usage showed approximately 4 million Internet users. The majority of our survey respondents were indeed young and white and had access to a computer at home or at work. In contrast to the user trends at the time, our sample was predominantly female (94%) attributable to the 'mother-oriented' nature of the website.

3.2 The research focus for the Cyberparents project was narrowly defined, targeting users who accessed one parenting site during a particular time period. Whilst users are encouraged to join the community officially, the number of visitors is higher than the number of official members. As we did not have access to the email addresses of members and we wanted to encourage wide participation, hoping for responses from both members and 'lurkers', the survey was accessible to anyone who was using the website.

3.3 It is recognised that the research centres around a small group of parents not representative of parents on the whole. This raises a range of methodological issues many of which are common to all social science research. We have not yet, for example, carried out comparative research of those parents who do not have Internet access and neither have we been able to access users of the website who did not answer our survey. Indeed, it must be acknowledged that the research may have only appealed to particular types of user - those parents who are highly computer literate or have a particular interest in (cyber-) parenting issues.

3.4 Despite these limitations, Comley (1996) and Coomber (1997) have suggested that the Internet is most suitable as a methodological tool in cases such as this, when researching a particular group of Internet users. Gaiser (1997:136) is in agreement, stating that: '...if the research question involves an online social phenomenon, a potential strength of the method is to be researching in the location of interest'. As our primary aim was to examine trends of use amongst those parents who access one particular parenting site our sample represented '... a valuable source of indicative as opposed to easily generalisable data' (Coomber 1997:1).

The Web-based Survey

4.1 The first stage of this research project involved setting up an online survey (see O'Connor and Madge, forthcoming) and a project website. The Cyberparents website provided a brief introduction to the project and the researchers and established our credentials through links to the University of Leicester website. Further pages gave a more detailed account of the research and included the online survey (http://www.geog.le.ac.uk/baby).

4.2 The survey was created using the .html compiler 'Adobe GoLive 4.0'. A response database was set up on our own departmental server. The survey ended with a short message to thank the respondents and a request to email us through the direct link if the respondent was willing to participate in a further detailed interview. Overall 155 responses were received with a further 16 email responses^[2] from people expressing their interest in a detailed further interview.

Online Interviewing

5.1 The second stage of the research process involved in-depth interviews. Our first task was to find a convenient way to carry out these interviews. It was immediately apparent that face-to-face interviews would be impractical, costly and time consuming because our respondents were geographically widely dispersed. Apart from the distance factor both the researchers and the respondents had young children and/or were pregnant, making the traditional interview unfeasible. As this research focused explicitly on Internet usage, our interviewees were already, by definition, Internet users and likely to be familiar with virtual communication methods, therefore, an Internet based interview forum seemed to be a logical, low cost, convenient and innovative research method.

5.2 Whilst there is a vast body of literature concerned with qualitative interviewing (Moser and Kalton, 1971; Oakley, 1981; Burgess, 1984; McDowell, 1992) the online approach to interviewing remains a new and innovative research method. The work that has been published has mainly been concerned with asynchronous or non real time exchanges usually conducted via email or a listserv facility (Gaiser, 1997; ward, 1999). Gaiser's (1997) online focus groups were conducted in a listserv environment. One advantage of this is that all participants are regular listserv users and have a high degree of familiarity with the technology. It also eliminates the need to set up mutually convenient chat times. However, it is not a real time facility, respondents can post their reply at anytime and as such the facilitator cannot play an active role in moderating the interview. The level of group interaction is reduced and the sense of immediacy removed. We were keen to encourage group interaction with a high level of immediacy and engagement with the topic. To some extent we wanted the interview to resemble face-to-face group interviews where respondents would answer immediately and not have time to carefully consider their replies. For these reasons the real time interview seemed to be more appropriate to our needs

5.3 At this time we were involved in developing a series of web-based teaching initiatives for distance learning students. An Internet conference software package called 'Hotline Connect'^[3] had been successfully used to facilitate real time chat between lecturers and students (Hughes, 2000) and we believed it could work in the same way as an interview forum.

5.4 Hotline Connect is a user- friendly application, available for both Apple Macintosh and Microsoft Windows based platforms. It enables users to chat, either in groups or one-to- one, to others simultaneously logged on to a specified server address. It does not have high power requirements and can be installed and used easily without the need for sophisticated hardware or a high level of technical ability. This was important to us because as Gaiser (1997) has commented, the use of alternative electronic spaces requires participants to learn new technological expertise. The researcher is already reliant on the goodwill of the interviewees and so the process needs to be as simple as possible. For our project we were, to some extent, relying on an assumption that our interviewees were competent users of computer technology and sufficiently motivated to install the software. Using Hotline meant that we could be relatively confident that they could log on without trouble. Only two participants dropped out at the installation stage, one because her baby was due and the second because, unusually, her computer was not able to run the software.

5.5 Another potential advantage of 'Hotline Connect' is that the facilitators have a high degree of control over proceedings. It is not possible for anyone to 'lurk'; users must identify themselves and the facilitators have the ability to disconnect those who do not. Neither is it possible to 'drop in' to the sessions because they take place at specified times known only to those invited. This level of confidentiality and control over the research process was important to us for ethical and practical reasons.

5.6 The screen shot (Figure 2) illustrates the virtual interface as seen by participants. The screen consists of a number of different windows and a tool bar. There is a large 'Chat' window in which the dialogue is displayed, beneath this is a smaller window where users type their text, and press return, seconds later the contribution is displayed, prefixed with their name.

5.7 We sent each interviewee the software^[4] and a set of guidelines on installation and use. If problems arose email or telephone advice was dispatched. The interviewees were then emailed a list of possible dates and times for interview and we set up a series of 4 group interviews with between 2 and 4 respondents at different times of day and week to suit both the participants and researchers.

5.8 Group interviews were chosen for several reasons. As Fontana and Frey (1994, p.365) explain, group interviews are '...inexpensive, data rich, flexible, stimulating to respondents'. These factors all suited our needs We also wanted to encourage interaction between the respondents on topics selected by us. We hoped that the group interaction would lead to '...relatively spontaneous responses from participants as well as producing a fairly high level of participant involvement' (Morgan, 1988, p.18). As Burgess (1984, p.107) reports, the group situation provides participants with the '...potential power to re-define the topics of the conversation'. However, we were also aware of the weaknesses of the group method over the individual interview, for instance having less control over the data generated (Morgan, 1988).

5.9 We allowed one hour for the interview but asked the interviewees to be online a few minutes early to ensure that everyone was successfully connected. Each interview began with a welcome message and an explanation of who was 'present'. We went on to introduce ourselves and provide guidelines about the interview, explaining the format of the questions and highlighting potential technological difficulties (see Extract 1).

Extract 1: Guidelines used to start the interviews

Guidelines

We want the interview to flow as much as possible and for you to feel that you can contribute exactly what you want to the discussion - almost as if we were having a conversation. However, we think it might be worth mentioning a few guidelines prior to starting the discussion.

1) as this is an `interview' we do have some topics that we would like to cover and we will probably use these to guide the discussion. However, please feel free to ask questions yourselves and to raise any topics that you think are relevant that we have not mentioned- but do try and stick as much as possible to the theme of the Internet and parenting;

2) it may take a while for the response you send to appear on screen - a good technique to speed the process up is to press return frequently, i.e. send the text every few words - don't wait till you have a complete sentence. Because of this the discussion may get a bit 'jumbled'. If this happens we may need to intervene;

3) this virtual interview is an 'experiment' and we anticipate there may be teething problems - we apologise for this in advance!

Do you have any questions before we start the discussion?

5.10 We then went on to begin with a general question, which like the following eight questions, was linked to the themes raised in the original survey. At the end of each interview we thanked the participants, invited them to request a copy of the transcript and explained the publication process to them.

Implications

6.1 The final part of this paper looks in more depth at some of the implications of conducting synchronous virtual interviews. In many respects online and face-to-face interviews are similar processes for the researcher and both can be considered '...conversations with a purpose' (Burgess, 1984, p.102). However, there are also important differences to be considered and, as Selwyn and Robson (1998) have highlighted, the electronic interviewer requires different skills.

Interview Design

7.1 We prepared the interview schedule in advance (see Extracts 1 and 2) and tried to ensure consistency by asking all respondents the same questions using the same wording. Preparing chunks of 'standard' text saved a great deal of time during the interview as it allowed us to cut and paste text as and when it was required. This drastically reduced the amount of typing needed during the interview, which was important when working to a time limit and freed us to be able to concentrate on thinking rather than typing. Thinking, incidentally, was also aided by having two researchers: a 'typist' and a 'moderator', working synchronously. The typist was able to cut, paste and type the questions whilst the moderator worried '...about the script of the questions' and was '...sensitive to the evolving patterns of group dynamics' (Fontana and Frey, 1994, p.365).

7.2At some points in the interview it became impossible to use the prepared text. As with face-to-face interviews, there were times when we needed to probe for more information and times when we wanted to follow up in more depth certain points raised by the interviewees. However, like Burgess (1984) we found that the conversation tended to develop naturally and often topics we had included later on in the schedule were covered earlier than we had planned. This did not matter; we were able to re-order the topics as we progressed, ensuring that all the important areas were covered. At times we found that when we kept a low profile and allowed the respondents a free rein they would move away from our core topics. We occasionally needed to intervene to re-direct the conversation when it digressed too far from our agenda. For example, in Extract 2 we wanted to find out if respondents ever shopped online. The conversation moved rapidly into a discussion which we did not want to pursue. We tried to curtail this by making the comment 'that's useful info' and then introducing a new question in an attempt to redirect the mothers back to a more general area.

Extract 2

Hen and Clare: Rowena^[5] - do you do much shopping online?

Henrietta and: and Barbara - have you ever shopped online?

Rowena: not much - usually books and nappies!

Hen and Clare: I didn't even knmow you could get nappies online!

Barbara: I can't say I have. It is useful to see what is available but I'm paranoid about security!

Rowena: from canada - they are cheaper - reusables

Hen and Clare: that's useful info

Rowena: most credit card payments are through a secure server

Hen and Clare: Do you find that you use the discussions for info. or for support from other parents?

Rowena: both really

Hen and Clare: and do you ever contribute to discussions or initiate them?

Rowena: both again!

Barbara: same here, it is nice to be able to offer advice too

Hen and Clare: yes that's true

Hen and Clare: are you happy with the info. you receive there?

Hen and Clare: does it compare well with info from elsewhere (gp, health visitor, mother etc.)

7.3 In the virtual setting the interviewer cannot make any assessment of the socio-demographic information which may have an impact on the interview. Indeed, ward (1999) found that as a consequence of this, interviewees asked her questions about her own socio-demographic profile, changing the power relations of the interview and giving the interviewer less control. It is perhaps, necessary, therefore, to find other ways of obtaining socio-demographic information and to adapt conventional techniques accordingly. In our research the text used to introduce the interview was carefully designed to allow for the loss of face-to-face interaction and in the hope that participants would follow our 'model' and provide similar profile information, such as age, number and age of children and ethnicity. After cutting and pasting our own information we invited each interviewee in turn to introduce herself. If we felt some information had been omitted we probed for this. As Extract 3 illustrates, our method proved successful; we gained the socio-demographic information required. This technique also proved useful in helping to develop rapport between the group members and the interviewers. This effect is discussed in more detail below.

Extract 3

First of all we thought it would be a good idea to introduce ourselves.

Hi, I'm Henrietta. I have a daughter called Alicia who will be 2 in July. I am 30 years old, white and I work full-time at Leicester University as a lecturer on a distance learning course. I came back to work when Alicia was 3 months old and she has gone to the nearby nursery full-time since then. At work we rely on the Internet a great deal. Many of the students live in different parts of the world and use email to communicate and we also use Hotline to talk to them. When I first came back to work I found that I also used the Internet a lot to look for information and advice about being a new parent.

Hello everyone. I'm Clare and I have a daughter called Isabelle who is nearly 2 as well. I work in the geography department of Leicester University on a 3-day contract and Isabelle is in a local community nursery on those days. I am 35 years old, white and my partner is a psychiatric nurse. I don't use the Internet very often because I am so busy at work and I don't have computer access at home.

Hen and Clare: Hi Amy, Hi Kerry, welcome to the chat - could you tell us a bit more about yourselves?

Amy.: I'm Amy. I have a son called James who was 1 last month. I went back to work as a computer programmer when he was 4 mths - he started with his aunt, but now goes to a nursery full time. I tend to surf the net at the weekend (well more before James, really), but find it really useful for finding people in the same boat!

Amy.: Oh - 33 years old, white and my husband is a draughtsman - workign^[6] at the same company as me.

Kerry: Hello. I'm 22 and have 2 kids. Lisa is 3 and Marie is 8 months. I am a full time mum. My partner Tom is a computer programmer.

Hen and Clare: Thanks for that.

Developing Rapport

8.1 We prepared the interview schedule in advance (see Extracts 1 and 2) and tried to ensure consistency by asking all respondents the same questions using the same wording. Preparing chunks of 'standard' text saved a great deal of time during the interview as it allowed us to cut and paste text as and when it was required. This drastically reduced the amount of typing needed during the interview, which was important when working to a time limit and freed us to be able to concentrate on thinking rather than typing. Thinking, incidentally, was also aided by having two researchers: a 'typist' and a 'moderator', working synchronously. The typist was able to cut, paste and type the questions whilst the moderator worried '...about the script of the questions' and was '...sensitive to the evolving patterns of group dynamics' (Fontana and Frey, 1994, p.365).

8.2Visual pointers may reveal differences or similarities in class, ethnic origin, gender, age and status which can all affect rapport, with shared characteristics likely to contribute to a greater immediate feeling of rapport (Robson 1993, p.237). Researchers such as Oakley (1981) and Finch (1993) have explored the impact that shared characteristics can have, concluding that when women interview women rapport will often happen naturally. Finch (1993, p.167) articulates it in this way:

'Women are almost always enthusiastic about talking to a woman researcher, even if they have some initial anxieties about the purpose of the research or their own 'performance' in the interview situation'.

8.3 Smith (1996, p.64) reports that she was '...expecting to have to work at "something called rapport" and was both surprised and pleased to find that this was not the case'. She, like Oakely and Finch, comments on the '...warmth and hospitality...' with which she was treated (Smith, 1996, p.63) and Finch (1993, p.167) comments that she was '...startled by the readiness with which the women talked' to her.

8.4 Had we met in person it is likely that the similarities between us and the participants would have had a positive impact, fostering immediate feelings of rapport. Significantly, however, we would not be meeting our respondents in person during the research process. It became important to us to find a way of creating a relaxed environment without the benefit of a face-to-face meeting in which we could establish ourselves as '...friendly female interviewer(s) ... with time to listen...' (Finch, 1993, p.169). We therefore experimented with an alternative way of building a relationship by means of self-disclosure at an early stage.

8.5 Initially on our project website we posted photographs and brief biographies of both researchers (<http://www.geog.le.ac.uk/baby> ). This gave the interviewees a visual image and contextual information to which we hoped they would relate (i.e. that we were also new parents). We then established further links through email communication, used to arrange interview dates. Arranging mutually convenient times and dates for the interview sometimes proved complex due to the ways in which the women (interviewers and interviewees) had to organise their lives to fit the demands of childcare, home and work. Often emails were exchanged for sometime, sharing personal information as to why suggested times were not convenient. In these cases it proved easy to empathise and build a relationship because, like the interviewees, we were also facing relatively complex childcare arrangements, for example, when arranging evening interviews.

8.6 Once the interviewees arrived at the virtual meeting place they were quite familiar with us, although we had not met in person. At this stage we built upon the earlier element of self- disclosure to help create greater rapport. We provided personal information to establish common ground and to replace the visual appraisal of other group members that would have occurred in a face-to-face scenario. This mechanism also allowed group members, who had not been in previous contact, to 'assess' each other. Overall this technique seemed to work well and, as Extract 3 illustrates, each of the women shared this personal information with the group.

8.7 We could have requested this information by different means but as Oakley (1981, p.41) has noted:

'...in most cases, the goal of finding out about people through interviewing is best achieved when the relationship of interviewer and interviewee is non-hierarchical and when the interviewer is prepared to invest his or her personal identity in the relationship'.

8.8 We were keen to create a non-hierarchical relationship with our respondents and as such our approach to interviewing did not subscribe to the traditional 'text-book' style criteria for carrying out interviews which encourages the maintenance of distance from the interviewee (McCracken, 1988;, see Oakley, 1981 and Burgess, 1984 for discussion). Rather we believed that as women interviewing women, or indeed as mothers interviewing mothers, it would be difficult and not necessarily desirable to maintain a detached relationship in which we '...elicited and received but did not give information' (Oakley, 1981. p.30). Further to this, Internet based interactions are, by their very nature, frank, because as Michaelson (1996, p.58) comments: 'The relative anonymity that information technology provides also changes the rules of discourse'. Similarly, Kitchin (1998, p.394) suggests that communication on the Internet '...provides an unrestricted freedom of expression that is far less hierarchical and formal than real world interaction'. Maintaining the traditionally detached and hierarchical interviewer/interviewee relationship would have seemed unnatural in this setting and been difficult to 'impose'.

8.9 In order to create the rapport which we hoped would occur naturally in a face-to-face situation we built in mechanisms which reflect what we would have done in a non-virtual environment. These practices had to be adapted to suit the text-based nature of the interaction and to compensate for the fact that we could not see each other. We felt that rapport was established effectively in this way. However, it may be that like Oakley, Finch and Smith we did not need to go to such lengths to create rapport. Perhaps the fact that we were all women with young children and the added element of the freedom of expression offered by the Internet negated the need for the artificial manufacture of rapport.

Insiders and Outsiders

9.1 In establishing our relationship with the mothers, we certainly relied heavily on our own identities as new mothers and as users of the Babyworld website. Paccagnella (1997, p.3) has suggested that particularly in the context of research into virtual communities, a '...stranger wanting to do academic research is seen as an unwelcome arbitrary intrusion'. In this case we were seen as insiders on both counts: parents and Babyworld users. Significant, perhaps, is the fact that the ease with which we built rapport was based not only on the female relationship but also upon our further level of mutuality (Riessman, 1991). Smith (1996, p.64) refers to this further level as a 'shared universe of meaning' whereby the women involved also share other life experience, namely that which the interview focuses upon. In her case this centred on the experience of being a mature student. For Oakley (1981), Ribbens (1989), Finch (1993) and Miller (1998), like us, the shared experience with the respondents was one of motherhood. Ribbens (1989, p.588) argues that motherhood, as a 'classless' topic of conversation and '...a core identity for many women...' plays a crucial role in facilitating mother-to-mother interviews.

9.2 However, even with shared commonalities, such as being a mother, issues can arise over power relations. Ribbens (1989) has highlighted the impact that the age of the interviewers children can have on interviews between mothers. Mothers, as respondents, may feel less inclined to voice their own experiences to an interviewer who they may perceive as being an 'expert' mother due to her 'advancement in terms of a mothering career' (Miller, 1998, p.63). This was not an issue for us as we made it clear from the beginning that we too were new parents. During the interviews we emphasised this, illustrating to our respondents that we were as inexperienced as they and found it just as difficult. Miller (1998, pp. 62-3) found this strategy useful in her interviews as she was aware that the new mothers she interviewed may have perceived her to be 'superwoman' able to cope with the demands of motherhood and work. She made it clear that her own life had only become easier once her children were at school, thus identifying herself again with her respondents' own situations.

9.3 We certainly found that our experiences of being mothers, in this case to children the same age as most of the interviewees' children, allowed us to empathise and respond knowledgeably when the mothers were discussing concerns about their own children. See for example Extract 4.

Extract 4

Kerry: When I first became a mum I found the enormity of it all very overwhelming

Hen and Clare: yes, we would agree with what you have both said ... but add trying not to laugh at their tantrums

Hen and Clare: us too

Kerry: Or trying not to scream and cry along with them

Amy.: Yes, it's so tricky not to laugh when they've done somethign you don't want to encourage!

Amy.: I spent ages worrying that I didn't have the "telepathy" all mums are supposed to have. I did in the end, but I was too busy panicking to realise!

Kerry: There are times when you have got to laugh or you would cry

Amy.: Especially when you're covered in gloppy food!

Kerry: I hope all new mums panic cause I still do sometimes

Hen and Clare: we still do too

Kerry: I know that feeling

Amy.: God, yes!!! Every sniff is a major illness, if he's sleeping noisily I'm worried that he's uncomfortable and if he's sleeping quietly - it worries me even more!!!

9.4 Linked to this idea of shared identity is that of the 'perceived relevance', which like Smith (1996) we saw as key to the success of our interviews. All our respondents were experiencing 'becoming a new parent' and using the Internet to find out more about this role. Hence at the time of the interview the research topic was highly relevant. These shared experiences all contribute to the creation of a high level of reciprocity in the interview situation.

9.5 The apparent ease with which we were able to build rapport was undoubtedly helped by the fact that we were to a great extent seen as 'insiders'. Indeed, participants indicated that the fact that we were new parents ourselves influenced their decision to agree to be interviewed (see Extract 5). Had we been complete 'outsiders' it is unlikely that the research process would have worked so effectively.

Extract 5

Hen and Clare: did the fact that we were both new mums influence your decision to take part? or were there other reasons?

Kerry: I like doing anything that involves parenting although it was nice to know that you would understand my experiences

Amy .: It did help - talking to people who haven't experienced it themselves is never satisfactory really. Also, I wanted to do my bit to help, as other people have helped me in this new experience.

9.6 From our own perspectives, however, we felt that we were both insiders and outsiders (see also: Dyck, 1993; England, 1994; Gibson-Graham, 1994; Madge, 1994). Initially we were insiders, using the website to get information about parenting, we then became interested as 'outsiders', researchers with an academic interest in the topic of cyber-parenting. Throughout the research we have played insider/outsider roles. At certain points in the research project it has been crucial that we have been insiders, for example, in showing initial interest in the topic and understanding issues that are important to new parents. But, we have also been outsiders, using our academic work environment and credentials to gain access to the website providers. Also, during the interviewing we have variously directed the research agenda and taken a back seat in discussion, as 'detached' researchers. This detachment was not always easy to attain, nor was it necessarily desirable. At points in interview discussion it was not possible to remain 'neutral' and not get involved in really interesting personal stories and anecdotes about pregnancy, birth and parenting.

Language Use

10.1 It has been suggested that in cyberspace there is a tendency to be more open with others, often complete strangers, than in real world communication (Nguyen and Alexander, 1996; Wellman and Gulia, 1999). It seems that: 'Individuals appear to enjoy relating narratives to those they have never met and probably never will meet. The appeal is strong to tell one's tale to others, to many, many others' (Poster, 1995, p.90). He goes on to suggest that the lack of visual clues plays an important role in encouraging candid interchanges:

'Without visual clues about gender, age, ethnicity and social status conversations open up in directions which otherwise might be avoided. Participants in these virtual communities often express themselves with little inhibition and dialogues flourish and develop quickly' (p.90).

10.2Similarly, Kitchin (1998, p.394) explains that: 'Individual representation in cyberspace is not based upon biology, birth, social circumstance or geography'. Nguyen and Alexander (1996, p.104) argue that because of the visual barriers which exist, people '...can better control the presentation of self' and as a result are more sociable, friendly and open.

10.3 It is possible, then, that a conscious effort to build rapport is not necessary. Cyberspace, by its very nature:

'...provides social spaces that are purportedly free of the constraints of the body, you are accepted on the basis of your written words, not what you look like or sound like or where you live' (Kitchin, 1998, p.387).

10.4 Mann and Stewart (1999) suggest that this freedom of expression arises because the virtual setting allows users to say exactly what they think without fear of reprisals, making the '...virtual venue ... a safe space in which to interact'. Certainly, several of our interviewees stated that the anonymity provided by the Internet enabled them to ask embarrassing or 'unimportant' questions without feeling self-conscious (see Extract 6, emphasis added).

Extract 6

Amy.: I feel better askign BW^[7] than my health visitor as they're not goign to see how bad I am at housekeeping!!!

Kerry: I feel the same. Like the HV^[8] is judging even though she says she isn't

Kerry: Although my HV has been a life line as I suffer from PND^[9]

Amy.: Also, there are some things that are so little that you don't want to feel like you're wasting anyone's time. Askign the HV or GP might get in the way of something mroe important, whereas sending an e-mail, the person can answer it when convenient

Amy.: My HV is very good, but her voice does sound patronising. I'msure she doesn't mean it, but it does get to me...

Kerry: Being anon means that you don't get embarassed asking about a little point or something personal

10.5 Participants in our cyberparents interviews were all familiar users of information technology who often 'natter on the net' (Spender, 1995). Consequently they seemed at ease with this method of communication and were clear about the 'rules' of online chat and their own roles before commencing. In fact, the respondents were more at ease with the environment than the interviewers. The use of a particular type of language presented some difficulties to us. As Fontana and Frey (1994, p.366) explain '...respondents may be fluent in the language of the interviewer but there are different ways of saying things ...'. In this case it was the interviewers who were not always fluent. We were not familiar with specific Internet language, for example, the 'emoticons' such as :-) (representing smiles) or electronic paralinguistic expressions such as 'lol' meaning lots of laughs or 'rofl' (rolling on the floor laughing) that are used in online chat and crept into our transcripts. It took us a while to learn this new language. This meant that often the empathy we held with many of the women had to be explicit rather than through the use of supportive electronic 'utterances' and 'gestures'.

10.6 This degree of abstraction that a virtual interview involves impacted upon the nature of the interview 'conversation'. There were occasions when we were 'lost for words', taking some time to decide on our appropriate response, because we felt like our written comments sounded banal or our questions too direct and leading. We 'policed' each other on this and feel that we very much benefited from working together as we could agree our approach. In other ways the degree of abstraction was helpful as it was a means by which we could keep the interview flowing along the key themes and avoid being side-tracked too much. Interrupting a conversation about the price of nappies felt somehow more acceptable in the written word than in the spoken face-to-face context (see Extract 1).

10.7 Overall this meant that the interviews flowed well, although it did also mean that the researchers dominated the setting of the research agenda (this was an active decision which could be altered given a different research remit) although not necessarily the content of each interview.

10.8 We concluded that the virtual interview went some way towards bridging the oral/written divide. Although clearly in written format, the type of interventions were very oral in nature. The researchers and participants paid little attention to spelling and grammar, as the nature and meaning of the conversation took precedence over the correctly written word. As such, the transcript very much resembles a 'written conversation'.

10.9 In a virtual interview, the speed of typing dominates the interaction rather than the most vocal personality, which changes the rules of engagement and has the potential to disrupt traditional interviewer/interviewee power relations. This represents an important advantage of virtual interviews, particularly in the group context. Those individuals who are shy and reticent to speak in face-to-face group interactions may find the virtual environment a liberating one in which they can 'speak'. There are many people who, as Rheingold (1994, pp.23-24) suggests:

'...don't do well in spontaneous spoken conversation but turn out to have valuable contributions to make ...These people ...can find written communication more authentic that the face-to-face kind. Who is to say that this preference for one mode of communication - informal written text - is somehow less authentically human than audible speech?'

10.10 However, a difficulty maybe that those with slower typing speeds, or participants who prefer more time to consider their replies may find themselves lagging behind, still preparing an answer to an earlier question and finding the main discussion has moved on. This may result in the loss of valuable interview data as the respondent deletes the reply and moves forward to join the continuing discussion.

10.11 The nature of the typed word can impact upon the interview format in other ways. Sometimes we used a snippet of conversation/question, because we had typed it and therefore wanted to use it rather than delete it and retype it later and this sometimes impeded the flow of the conversation. Because of this, in tracing the genealogy of the interview, both the interviewers and the participants followed the main thread of conversation and ignored conversational side-tracks probably more effectively than would have been the case in a face-to-face encounter. 'Silences' took on an added poignancy, as we needed to consider whether the silence was because the participant was thinking, typing, or had declined to answer the question. No subtle visual clues were available and so direct questioning often replaced subtle probing.

The Virtual Interface

11.1 Mann and Stewart (2000, p.25) argue that CMC '...offers women ... the potential to communicate in a familiar and physically safe environment'. However, whilst this may be true of cyberspace there is still the physical need for access to a computer. As with a face-to-face interview there is a need for the cyber-researcher to be sure that the interview is taking place in a '...secure, private and familiar environment where personal issues might be explored' (Mann and Stewart, 2000, p.79). This requirement can present challenges to the face-to-face researcher. For example, in some situations the interviewer may feel uncomfortable entering the respondent's domestic space (Lee, 1997), equally, the respondent may not feel comfortable inviting the interviewer into their home. Home interviewing may not always be a viable option (the homeless, prison inmates, hospital patients). Smith (1996) overcame her anxieties about interview location by asking her respondents where they would like the interviews to take place. The majority chose their own homes, where they felt that they would not be interrupted. Similarly Bergen (1993, pp.206-7) found that she was able to build a special relationship with the respondents interviewed at home because, in their own space the women, felt comfortable and in control of the situation.

11.2 Although our respondents were taking part in a group interview they were all located separately. The choice of location was their own although this was restricted by the need for Internet access. We found that the majority chose their own home as the venue, only one respondent was located away from home, taking part from work. Importantly the venue represented what we would hope to be an anonymous, safe and non-threatening environment for the interviewee, factors which could take on an added significance when interviewing particular groups of people, for example, abused women (Dillon, 2000).

11.3 Smith (1996) found that a major concern for her respondents was that they did not want to be interrupted by their partner or children whilst taking part. For this reason they arranged interviews at times when the children were at school or asleep. In some cases the children were there but were '...dispatched ... to another room with instructions not to interrupt' (p.63). Our interviewees were able, if necessary, to address these interruptions in a different way and to continue with real life activities such as eating, drinking and conversing, for example with their children or partners, while taking part in the interview (Mann and Stewart, 2000, p.24). In this case, rather than excluding their partner and/or children the interviewee was able to simultaneously do other things which required their attention, in this case caring for their babies (see Extracts 8 and 9).

Extract 8

Rachel: my replies will be a bit slower from now on as E-J has just woken for a feed so i'm typing one handed!

Extract 9

Rowena: i always have Nicola with me - she's here now - asleep on my other arm - that's why I'm on the net a lot I can hold her and type

11.4 This may not always represent an advantage to the interviewer. It may be that distractions result in respondents not being fully engaged with the interview whilst dealing with interruptions of which the interviewer may not be aware. In the face-to-face setting the interviewer can respond immediately to external factors perhaps by suspending the interview, whereas the virtual interviewer, unable to see what is happening may carry on regardless, not aware that s/he does not have the full attention of the respondent.

11.5 There are other disadvantages of virtual interviewing which should be considered. Chen and Hinton (1999) argue that participation in the virtual interview requires a far higher level of motivation and interest from the interviewee than would be the case in a conventional interview. The interviewee has to provide the relevant equipment (the computer), bear the financial costs of being online for the duration of the interview and be prepared to take part in a physically quite demanding interview involving typing and reading. There is a need to think, type, look at the screen, read the text and maintain a logical thread of answering. The same is true for the interviewers who also have to cover all relevant questions, probe unclear answers and ensure that everyone is still taking part, whilst under considerable time pressure to get a response on the screen. For us, the fact that we worked together made this easier as one of us was able to keep the other up-to-date with the responses on screen and dictate the next question or probe whilst the other was typing or thinking. For the lone interviewer this would be far more difficult.

11.6 The disembodied nature of the virtual environment does, however, offer certain advantages to both the researcher and the respondents. Firstly, there is no need to dress up or worry about appearance, because in this environment neither party can see the other. Secondly, it is not necessary to arrange transport or allow for travel time. This can be crucial to those participants who may be spatially restricted, for example, the elderly, those with restricted mobility and in this case parents of newborn babies. Thirdly, the virtual venue is an inexpensive one; there are no costs related to travel or venue hire. Fourthly, the virtual interview places fewer organisational demands upon the researcher who no longer needs to seek out a venue which meets the multiple recommended requirements (Morgan, 1988 and Kruguer, 1994). Finally, all the practical difficulties of data recording are avoided. A transcript of the interview is created as the interview progresses and there is no need to transcribe the interview once it has been completed. This transcript can be converted in to a word document and easily manipulated at the analysis stage.

Concluding Remarks

12.1Cyberspace technologies are without doubt '...transforming space-time relations and creating new social spaces that lack the formal qualities of geographic spaces' (Kitchin, 1998, p.386). This can provide several innovative potentials when considering using Internet based methodologies. In particular, the use of Hotline Connect Conferencing Software enables researchers to synchronise interviews in a virtual space and to interface with groups who may be difficult to reach through conventional research approaches. Although we found that the advantages to be gained through the disembodied interview were many, some difficulties were also encountered in transferring a traditional methodological approach to a new setting.

12.2 Using other traditional interviewing devices such as probes is also problematic, periods of silence and pauses have different connotations online and there is no possibility of using an enquiring glance or verbal prompts such as '...mmhmm...' in order to encourage participants to expand on certain points (Robson, 1993). Alternative mechanisms need to be employed by the researcher.

12.3 As with traditional interviews, being an 'insider' was certainly of great benefit to us. The online researcher faces something of a double challenge here as it may not be enough to be an insider only on one level. In this case it was important not only that we were new parents but also that we were familiar users of the website we were researching. An outsider to the online community being researched may face greater challenges.

12.4 Difficulties associated with language use, although not peculiar to virtual interactions, can also present problems for the online interviewer. Specific symbols such as emoticons may be used by some whilst more novice users may find this level of expertise intimidating. The virtual interface itself presents new challenges to all involved. The researcher is reliant upon participants having access to a suitable computer, some technical knowledge and a considerable level of motivation. Not least the physical aspect of interviewing online requires consideration, as it is more demanding than it would appear. A final yet perhaps most important consideration lies in the fact that Internet access remains restricted to certain groups of people; therefore, the potential for research outside these user groups is limited.

12.5 The Cyberparents project was a pilot study that intended to explore the possibilities which virtual technologies have, and continue to, open up to researchers. Clearly the virtual interview as a methodological technique is still in its infancy and has limitations, some of which may be solved over time and some which may not ever be remedied. As Featherstone and Burrows (1995, p.5) have argued: '...the face and the body are the only 'true' sources which can reveal the character of a person...' and as such, purely disembodied, textual encounters will never be a satisfactory alternative to a face-to-face meeting.

12.6 In conclusion, whilst the data collected through virtual means can be as rich and valuable to the researcher as that generated via traditional face-to-face meetings, the potential of the Internet for research should not be exaggerated. As Kitchin (1998, p.395) comments '...the vast majority of social spaces on the Internet bear a remarkable resemblance to real world locales' and as such it must be remembered that many of the issues and problems of conventional research methods still apply.

Notes

¹The research was funded through small internal grants from Leicester University and, as a result, the study is of a pilot exploratory nature.

²Of the 16 respondents who agreed to be interviewed only 1 was male. However, he did not 'arrive' at the arranged time of interview, hence all interviewees in the end were female.

³ More information about the Hotline Connect Software which was used is available from the Hotline website. This can be found at: <<http://www.bigredh.com>>

⁴Details of the installation guidelines may be obtained by writing to the authors.

⁵All names have been changed for confidentiality.

⁶Typing and spelling errors have been retained in the transcripts in the same way that transcripts of face-to-face interviews are recorded verbatim.

⁷BW was the abbreviation used to denote Babyworld.

⁸HV is the abbreviation used for health visitor.

⁹PND stands for Post Natal Depression.

Acknowledgements

We would like to thank Wayne, Jason and Kate for technical advice and support and our partners for minding the girls.

References

BAYM, N. (1998) The Emergence of Online Community, Chapter 2 in JONES, S.G. (Ed) Cybersociety 2.0: Revisiting Computer-Mediated Communication and Community. Thousand Oakes: Sage.

BENNETT, C.M. (1998) Men Online: Discussing Lived Experiences on the Internet. Unpublished thesis, James Cook University, North Queensland, Australia. Available at: <http://www.hotkey.net.au/~car olineb>

BERGEN, R.K. (1993) Interviewing survivors of marital rape: Doing feminist research on sensitive topics. In RENZETTI, C and LEE, R. (Eds.) Researching Sensitive Topics. London: Sage.

BURGESS, R. (1984) In the Field. An Introduction to Field Research. London: Allen and Unwin.

CHEN, P. AND HINTON, S.M. (1999) 'Realtime Interviewing Using the World Wide Web'. Sociological Research Online, vol. 4, no. 3, <http://www.socresonline.org.uk/4/3/chen.html>

CHRISTENSEN, N.B. (1999) 'Inuit in Cyberspace: Practising and Constructing Computer-mediated Space', paper prepared for the 3rd Annual ARCUS Award for Arctic Research Excellence.

COMLEY, P. (1996) 'The Use of the Internet as a Data Collection Method', <http://www.sga.co.uk/esomar.html>This link was correct as at the 7/7/99, but the connection no longer exists.

COOMBER, R. (1997) 'Using the Internet for Survey Research'. Sociological Research Online, Vol. 2 No. 2 <http://www .socresonline.org.uk/2/2/2.html>

DILLON, J. (2000) 'Battered women to tell their stories via email'. The Independent on Sunday, January 30 2000, p.3.

DYCK, I. (1993) 'Ethnography: a feminist method?'. The Canadian Geographer, 37 (1) 52-57.

ENGLAND, K. (1994) 'Getting Personal: reflexivity, positionality, and feminist research'. The Professional Geographer, 46 (1), 80-89.

FEATHERSTONE, M. AND BURROWS, R. (1995) Cyberspace/Cyberbodies/Cyberpunk: Cultures of Technological Embodiment. London: Sage.

FINCH, J. (1993) 'It's Great to have Someone to Talk to': Ethics and politics of Interviewing Women, Chapter 14 in HAMMERSLEY, M (Ed.) Social Research. Philosophy, Politics and Practice. London: Sage.

FONTANA, A. and FREY, J.H. (1994) 'Interviewing: The Art of Science'. Chapter 22 in DENZIN, N. and LINCOLN, Y. (Eds.) Handbook of Qualitative Research, London: Sage.

GAISER, T. (1997) 'Conducting Online Focus Groups: A Methodological Discussion'. Social Science Computer Review, Vol. 15, No.2, pp.135-44.

GIBSON-GRAHAM, J.K (1994) 'Stuffed if I know': reflections on post-modern feminist social research. Gender, Place and Culture, 1 (2) pp. 205-224.

GLESNE, C. and PESHKIN, A. (1992) Becoming Qualitative Researchers. New York: Longman.

GRAHAM, S. and MARVIN, S. (1996) Telecommunications and the City: Electronic Spaces, Urban Places. London: Routledge.

HINE, C. (2000) Virtual Ethnography. London: Sage.

HODKINSON, P. (1999) The Goth Scene as Trans-local Subculture. Unpublished PhD. Thesis. Centre for Urban and Regional Studies, University of Birmingham, UK.

HUGHES, J. (2000) Low-Investment Internet- Based Distance Learning Solutions: Systems and Procedures, TechKnowLogia, July/August pp.23-25 <http://www.TechKnowLogia.org >

JONES, S.G. (1995) Cybersociety: Computer- Mediated Communication and Community. London: Sage.

JONES, S.G (Ed) (1997) Virtual Culture: identity and communication in cybersociety. London: Sage.

JONES. S.G. (Ed) (1998) Cybersociety 2.0: Revisiting Computer-Mediated Communication and Community. Thousand Oakes, CA: Sage.

KING, S. (1994) 'Analysis of Electronic Support Groups for Recovering Addicts' Interpersonal Computing and Technology 2 (July) pp. 47-56.

KITCHIN, R. (1998) 'Towards geographies of cyberspace'. Progress in Human Geography, 22, 3, pp. 385-406

LAMP, J.M. AND HOWARD, P.A. (1999) 'Guiding Parent's Use of the Internet for Newborn Education'. Maternal- Child Nursing Journal, 24 (1) pp.33-36.

LEE, D. (1997) Interviewing Men: Vulnerabilities and Dilemmas, Women's Studies International Forum, Vol. 20, No. 4 pp. 553-564.

LOADER, B. (Ed.) (1997) The Governance of Cyberspace. London: Routledge.

MADGE, C. (1994) The ethics of research in the 'Third World'. In ROBSON, E. and WILLIS, K. (Eds.) DARG Monograph No. 8: Postgraduate Fieldwork in Developing Areas (Developing Areas Research Group of the Institute of British Geographers) pp. 91-102.

MANN, C. AND STEWART, F. (1999) Internet Methodologies and Qualitative Research, Unpublished conference paper, BSA, Glasgow, April 1999.

MANN, C. AND STEWART, F. (2000) Internet Communication and Qualitative Research, New Technologies and Social Research Series. London: Sage Publications.

MCCRACKEN, G. (1998) The Long Interview, Qualitative Research Methods, Volume 13. Newbury Park: Sage Publications.

MCDOWELL, L. (1992) 'Doing gender: Feminism, feminists and research methods in human geography'. Transactions of the Institute of British Geographers, 16 (4), pp. 400 - 419.

MICHAELSON, K.L. (1996) 'Information, Community, and Access'. Social Science Computer Review, Vol. 14, No.1, Spring 1996, pp. 57-59.

MILLER, T. (1998) Shifting Layers of Professional, Lay and Personal Narratives: Longitudinal Childbirth Research. Chapter 4 in RIBBENS, J. and EDWARDS, R. (Eds.) Feminist Dilemmas in Qualitative Research: Public Knowledge and Private Lives, London: Sage.

MOORHEAD, J. (2000) 'A Parent's Directory'. The Guardian, Lifeonline, <http://lifeonline.g.uardianunlimited.co.uk/week/story/0,6488,124257,00 .html>

MORAN, C., HOLT, V., AND MARTIN, D. (1997) 'What do women want to know after childbirth?' Birth, 24 (1) pp. 27-34.

MORGAN, D.L. (1988) Focus Groups as Qualitative Research. Sage University Paper Series on Qualitative Research Methods Series 16. Beverly Hills, CA: Sage.

MOSER, C.A. AND KALTON, G. (1971) Survey Methods in Social Investigation. Aldershot: Gower.

NIELSEN//NET RATINGS (2000) Internet Usage Statistics for the month of September 2000 UK. <http:// 209.249.142.57/uk/web/Nrpublicreports.usagemonthly> accessed October 23 2000.

NUA SURVEYS (2000) How many online? <http://w ww.nua.ie/surveys/how_many_online/europe.html> accessed October 23 2000.

NGUYEN, D.T. AND ALEXANDER, J. (1996) The Coming of Cyberspacetime and the End of Polity. Chapter 6 in Shields, R. (Ed.) Cultures of Internet: Virtual Spaces, Real Histories, Living Bodies. London: Sage.

OAKLEY A. (1981) 'Interviewing Women: a contradiction in terms'. Chapter 2 in ROBERTS, H. (Ed) Doing Feminist Research. London: Routledge.

O'CONNOR, H. AND MADGE, C. (forthcoming) Using the Internet as a Medium for Research, Working Paper, Centre for Labour Market Studies, University of Leicester.

O'LEAR, R.M. (1996) 'Using electronic mail (e-mail) surveys for geographic research: Lessons from a survey of Russian environmentalists', Professional Geographer, 48 (2) pp. 209-217.

PACCAGNELLA, L. (1997) 'Getting the Seat of your Pants Dirty: Strategies for Ethnographic Research on Virtual Communities'. Journal of Computer Mediated Communication 3 (1) June 1997 <http://jcmc.indiana.edu/vol3/issue1/paccagnella.html>

POSTER, M. (1995) Postmodern Virtualities, in FETHERSTONE, M. and BURROWS, R. (Eds.) Cyberspace/Cyberbodies/Cyberpunk: Cultures of Technological Embodiment, London: Sage.

RHEINGOLD, H. (1994) The Virtual Community:Finding Connection in a Computerised World. London: Secker and Warburg.

RIBBENS, J. (1989) Interviewing - An 'Unnatural Situation'? Women's Studies International Forum, Vol. 12, No.6, pp.579-592.

RIESSMAN, C.K. (1991) When Gender is Not Enough: Women Interviewing Women. Chapter 11 in LORBER, J. and FARRELL, S.A. (Eds.) The Social Construction of Gender, Newbury Park: Sage.

RIPPEN, H.E. (1997) White paper: Criteria for assessing the quality of health information on the Internet. Available at: <http://hitiweb.mitretek.org/docs/criteria.html >. Accessed March 18 1998.

ROBSON, C. (1993) Real World Research. A Resource for Social Scientists and Practicioner-Researchers. Oxford: Blackwell.

RODWAY, A. (ed) (1997) Bounty Infant Health and Feeding Guide. Diss, Norfolk: Bounty Publications Ltd.

SELWYN, N and ROBSON, K. (1998) 'Using email as a research tool'. Social Research Update, Issue 21, University of Surrey. Available at: <http://www.soc.surrey.ac.uk/sru/SRU21.html> SHARF, B. (1997) 'Communicating Breast Cancer Online: Support and Empowerment on the Internet'. Women and Health, Vol 26 (1) pp. 65-84.

SILKE, R. BRUCE, J. AND MAFFULLI, N. (1998) 'Accessing the Internet for patient information about orthopedics'. Journal of the American Medical Association, 280, p.1309.

SMITH, C. (1997) 'Casting the net: Surveying an Internet population', Journal of Computer Mediated Communication, vol. 3, no. 1. <http://jcmc.indiana.edu/vol3/issue1/smith.html>

SMITH, M. A. (1996) Voices from the WELL: The Logic of the Virtual Commons, <http://netscan.research.microsoft.com/csoc/papers/voices>

SMITH, S. (1998) Uncovering Key Aspects of Experience: The Use of In-depth Interviews in a Study of Women Returners to Education. Chapter 4 in STINA LYON, E. and BUSFIELD, J. (Eds.) Methodological Imaginations. London: Macmillian Press.

SPENDER, D. (1995) Nattering on the Net. Women, Power and Cyberspace. Melbourne: Spinifex Press.

ward, k.j. (1999) 'The Cyber-Ethnographic (Re)Construction of Two Feminist Online Communities'. Sociological Research Online, vol.4 no.1 < http://www.socresonline.org.uk/4/1/ward.html>

WELLMAN, B AND GULIA, M. (1999) Virtual Communities as Communities: Net surfers don't ride alone. Chapter 7 in SMITH, M.A and KOLLOCK, P. (Eds.) Communities in Cyberspace, London: Routledge.

WILLIAMS, C. (1999) 'Question: Who killed off Dr. Spock? Answer: The Net.' Guardian, 13 January 1999 pp.8- 9.