Elizabeth Whitmore©

In a recent evaluation of a prenatal program for single expectant mothers, conducted by a group of programme participants, we discussed why they felt they could obtain better information than I, a university trained evaluator. One of the mothers explained:

"You're dealing with a lot of people on social assistance and welfare. You're dealing with real hard to reach low self esteem people. And when they see anybody coming in that they think is high class or has anything to do with welfare and you working with them, they are scared to death that you're going to squeal on them... They (the respondents) are just scared that you work with those people (social workers), you deal with them, you're high up there so that they can't trust you cuz you're right in with them. But we're not in with them (social workers) and we're not in there to tear them apart. And I think they really know that."

This says a great deal about oppression and about 'truth.' It highlights what I learned in working with this team of co-evaluators about participation and power, about trust and class barriers, about language and meaning and their relationship to the quality of research or evaluation. By class, I am referring to oppressed groups of people in our society who have internalized a view of themselves as worthless, as less intelligent, and as somehow undeserving of the privileges which others assume as given. These groups have less education and fewer opportunities, are poor and/or dependent on social assistance. Many are single parent women or members of marginalized racial or ethnic groups in a predominantly white society.

In this discussion, I'd like to explore the gap which exists between university trained researchers and evaluators and members of oppressed groups from whom we collect data and interpret its meaning. This has profound implications for how we define and identify what is "good" research or evaluation, and how we know when we see it, especially as it relates to participation and to working with people who are marginalized. I use the term quality rather than validity. As Reason suggests, the term validity is too ideologically laden, whereas 'quality' "allows more space for us to formulate new standards and to draw on widely different fields of thought.." (Reason, 1991).


The prenatal education programme

In the mid-1980's, a community resident (who later became programme Coordinator) approached the Director of the Community Services Association, to inquire about assistance in implementing a prenatal programme for single women. Over 20% of the households in the community were headed by single women, many of whom received social assistance as their primary source of income. This group often experienced serious health, social and financial problems during their pregnancies. For the most part, they did not attend prenatal classes sponsored by the Department of Health or at the local hospital. These traditional prenatal programmes usually assume a (male) partner, are held in a downtown hospital or health unit, do not provide transportation and do not address personal or social concerns of participants. Public health nurses reported that the most under-represented groups in their prenatal classes were teens, single parents, low income women and working women.


Local social workers, public health nurses, nutritionists, librarians and church officials held a series of meetings about supporting such a programme and as a result a Prenatal Advisory Committee was formed. This group drafted a proposal to Health and Welfare Canada and in early 1986, the Committee received funding for three years to develop a prototype prenatal programme. Later, two programme participants were added to the Committee. The Coordinator researched, designed and implemented the programme with the support and advice of the Advisory Committee.

This prenatal education programme reached out to single mothers in a variety of ways. Classes, held at the local Single Parent Centre, were small and informal, boyfriends, parents and relatives were welcome, the Coordinator wore casual clothes (a very important signal to participants that she wasn't "putting on airs"). Participants were encouraged to share their experiences with each other and ask questions. A wide variety of personal and social topics of interest to participants (including abuse) were raised, along with the immediate physical concerns of labour and delivery.

Recruitment was done with sensitivity and patience. Initially, women were referred by public health nurses and nutritionists who worked in the community and were well known. The Coordinator then contacted each potential participant personally and visited her in her home. She told the mother-to-be about the programme, answered questions and reassured her that no one would pressure her or tell her what to do. There was also provision for concrete needs and incentives (the Coordinator provided transportation for anyone needing it, which was most of the participants, child care was provided for those who had other children, nutritious snacks were served, the Centre often had used clothing to give away). Since many women came into the programme very late in their pregnancy and would deliver before the 8 week programme was completed, the Coordinator developed a crash course (of 1-3 weeks) so that everyone could participate, even if only briefly.

Advocacy was formally recognized as an important aspect of the programme and was structured into the Coordinator's role. She spent many hours assisting the women negotiate their way through the welfare bureaucracy, for example. She played the role of advisor, confidante and even mother, responding to individual needs as they arose. For many, she was the only person they could count on in times of crisis. The Coordinator also made a point of visiting each mother in the hospital and followed up afterwards, keeping track of their progress (and that of the new baby) as they struggled with the pressures of single motherhood.

The Coordinator hired one of the participants as an assistant; other participants were asked to lead discussions (on breast feeding, for example). The fact that their peers were an active part of the programme offered role models for women who are so often isolated and whose self esteem is fragile. In discussions, everyone was encouraged to share their knowledge and experience; everyone's input was valued.

Once the programme had run for a year, the word had spread and young women themselves contacted the Single Parent Centre to sign up. The Coordinator was, by then, well known and liked by the women. Participants brought their friends, who were welcome to try out a class or two. Classes were soon overflowing and the programme was reaching women earlier in their pregnancy. Within two years of operation, the programme had attracted 97 participants. Most of these women were single, and about half were in their teens. 81% received social assistance as their primary source of income.

The Single Parent Centre also developed postpartum classes and a follow up parenting programme, both in close collaboration with the Coordinator.


The evaluation

In 1988, the programme was nearing the end of its three year grant. The Community Advisory Committee issued a call for evaluation proposals, specifically requesting designs in keeping with the informal nature of the programme and also sensitive to the needs and vulnerability of such a marginalized group. The Advisory Committee stated up front in its evaluation plan, for example, that "overzealous testing of these women for the sake of evaluation will defeat its entire objective. Since many of these women have limited skills in reading and writing, extensive questionnaires and testing will only prove threatening." Those representing the funder locally were very supportive of a community based approach, both for the programme itself and for the evaluation.

In response to their call for proposals, I suggested a participatory approach in which I would hire a group of programme participants as co-evaluators and train them to do the evaluation. This approach was based on assumptions inherent in the participatory research and evaluation (PR/PE) model. These include the assumption that experts are not the only ones who can create valid knowledge. Ordinary people are capable of generating knowledge which is as important and as valid as that produced by more highly structured and scientific processes. This model also recognizes that science is a cultural product and both the process and results reflect an historical and cultural context.

The proposal was accepted and together with the Advisory Committee, we worked out a recruiting procedure which resulted in four of the women who had taken the programme being hired. Four Committee members agreed to serve as an evaluation subcommittee which worked closely with me throughout the process.




We began by drafting a letter to all programme participants describing the evaluation and inviting them to apply. The teaching assistant then "translated" it into language which would be easily understood and accepted by the women. We included a simple application form, asking for basic information (age, schooling, etc.), one reference, and a brief statement on why they were interested in the job. These letters were sent to all women who had participated in the programme over the 3 years (at least all those we had current addresses for). We also posted notices at the Single Parent Centre, a place many of them gathered for a variety of activities.

This was followed up with personal contact (with those women still living in the area) by the nutritionists and others (the Director of the Single Parent Centre, for example) who knew most of the women personally and were trusted by them. They encouraged the women to apply and supported them in following through. Some women already had jobs, others did not want to work while their children were infants. A number, however, welcomed part time work as a chance to get out of the house, to earn some money and to have the challenge of an interesting job.

A nutritionist, specifically one known to the applicants (which we hoped would make them feel more comfortable) and I formed an interviewing team. We interviewed all (6) who applied. We were interested primarily in determining whether the applicant would be reliable (were there obstacles to her attending regularly and sticking with it? did she have back up supports in time of crisis? was she planning to move? had she moved frequently over the past few years?, etc.). We also tried to assess how comfortable she would be contacting strangers, how well she could explain the evaluation to interviewees and take down notes of what was said. We reassured her that she would not be penalized on her social assistance because she was working and that we would provide for transportation and child care.

Of the six, one was eliminated when she seemed only marginally interested in doing the work. We decided to hire the other five (even though the budget allowed for four) on the assumption that there would probably be at least one drop out. This actually happened in the first week and we ended up with four, as planned. (I'll call them Jane, Sue, Nancy and Rhonda.)

The evaluation team (the four programme participants and myself) worked together for six months (September to March), first designing the evaluation, then collecting the data, analyzing the information and finally, reporting it. We met regularly two mornings a week as a group, doing other tasks individually in between, such as interviewing or analyzing specific data.


One day at a time

I knew that we needed to begin slowly, for they needed to get to know me and each other and they also needed to build some confidence that they could, in fact, do this. I knew it would take time for them to trust me, for as a university professor I lived in another world. They had a long history of negative experiences with social workers and as far as they were concerned, I was "one of them." Somehow I had to become a real person to them and this wasn't going to happen overnight.

We began by getting to know each other and sharing a lot of experiences with the programme and information about our lives as single mothers. We also did some group exercises, learning how to give feedback and just having fun together. We learned what an evaluation was all about and different ways we could go about doing it. Each got a brightly coloured binder for notes and information. This was a concrete representation of our work, reinforcing not only the teamwork, but also the importance of the task at hand. I tried to build task and process into each meeting, so that we got on with our work but we also had space to share feelings and concerns. At the beginning, I took responsibility for facilitating each meeting, setting out the next tasks to be done and offering ideas of how to do them. As time went on, we decided this more as a team.

For data collection, we decided on a mixed method approach, gathering both qualitative and quantitative information from a variety of sources. This included interviewing as many programme participants as possible, at least the ones we could get hold of. (Over the three year period of the programme, many had moved and we had no way of contacting them.) Other sources of information included a mailed survey of professionals in the community who were likely to have contact with programme participants (health professionals, nutritionists, social workers, etc.), a survey of all members of the Advisory Committee, and in-depth interviews with the Coordinator and her assistant.

We began designing the participant questionnaire by first taking the original goals of the programme and developing categories of questions related to them. So, for example, one goal was to provide a support system during childbearing and the immediate postnatal period; another, to give programme participants the tools (information, skills, confidence) to help them deal effectively with social service and other systems (hospital, other health services, etc.). In responding, they could draw on their own experience in the programme. We brainstormed possible questions (Did you make friends with anyone in the class? Did the classes help you know your rights as a mother in the hospital? Please explain) and spent several weeks refining them and wording them so that they were clear and easily understood. We wanted it to be down to earth, not full of big words.

We also formulated the questionnaire so that it would be easy to write down their answers and to show them that no names (only an ID number) would be used. Confidentiality was very important, for we knew that they would be very reluctant to talk to strangers and fearful that what they said might be used against them, especially in relation to welfare. Before we actually began contacting the women, we practiced listening and did some interviewing with each other. They each did two pre-test interviews with people they knew and we made final revisions to the questionnaire on the basis of their feedback.

For the next 6-8 weeks, the women conducted interviews with programme participants. I interviewed the Coordinator and her assistant. We met once a week during that period, sharing our adventures and noting preliminary impressions. They found that the interviews took many hours, for the respondents wanted to talk. From my reflexive log:

All reported that the respondents want to TALK and TALK. That's a real characteristic of all their contacts. And they say, well, we just can't cut people off in the middle of a sentence and say we have to pay attention to the interview questions only. They are sensitive to the respondents' need to tell their stories and also to get information.

Jane reported that one respondent wanted to ask her a lot of questions about getting on welfare, the rights of a mother, etc. My log that day quotes Jane:

"I feel good them asking me (questions about welfare, etc.) I guess I am kind of an expert since I have a child and have been on welfare."


Rhonda reported that one of hers said she asked the nurse "what are you doing?" (even though she knew) just to test the nurse. The respondent felt that the nurse ought to ask permission to do any medical procedure, even though she was prepared to give her consent. She also asked other questions and when the Dr. asked "where did you learn that?", she answered proudly, "in the prenatal classes!"

The women gained confidence as they exchanged stories about their interviews. They also supported each other. Jane, for example, felt quite comfortable going into the housing project, while others didn't, so she did all the interviews with respondents living there. They knew the community and who was around. From my log:

Nancy reported that one person's mother said she hadn't seen her daughter for months. Sue's response was, "Bullshit. I seen her just the other day walking down the street!" Sue agreed she'd interview her and switched a name with Nancy.

Sometimes they had to make two or three visits before they found the respondent home and able to talk to them. A few refused to be interviewed. This was partly because they were scared of answering questions from a stranger, or perhaps because, at first, they told them that the evaluation was being funded by Health and Welfare and the respondents thought they were from welfare. Once they left that out of their introduction, they had very few refusals.

As we proceeded, I met regularly with the Evaluation Subcommittee, keeping them up to date on our progress. About half way through the process, the Advisory Committee asked to meet with us, for they were intensely interested in the process and wanted to meet and talk with the women themselves. At first, the women were quite intimidated, and we prepared carefully. We put our information on flip charts (sample, timetable) so that they could talk from the information on the chart. I also urged them to talk about their experiences in their own words and reassured them that Committee members wanted to know what we were doing, not because they were "checking up" but because they were interested. We agreed that I would begin and then divided up the rest of it. I recorded what happened in my log:

They were quite scared at first, but then found their voices and took off. I started by presenting some of the material on the flip chart and gradually they began to fill in the gaps. Then (one committee member) asked them directly about their impressions so far. Sue is particularly good at speaking in front of a group and had no trouble finding her words. Her sense of humour is wonderful and she really projects personality as she described what she is doing. Jane also began to speak up, especially when asked a direct question. Nancy reported that one of the nutritionists was mentioned by name several times as especially helpful. They noted how verbal the respondents are and want to talk about their experience and also about welfare...

At the end, a committee member exclaimed: "You are REALLY on top of it. I'm impressed. You are proving that this kind of research can work, and its very different from most evaluations."

The group was elated! The meeting had been a big boost to their confidence and sense of purpose. We continued interviewing and recording our progress as we went along, and by early November, the interviews had been completed. (The other data - survey of professionals and of the Advisory Committee - were still in process but continuing apace.)


Once most of the interviews were done, we began the analysis. We did this by cutting up a copy of the each questionnaire and putting the answers to each question in a separate envelope. We then divided up the envelopes, recorded all responses and developed preliminary categories. From my log:

It's hard for them to deal with this new task. It's a difficult cognitive one and they need a lot of guidance on how to sort the answers into piles. But at the same time, they were the ones who developed clear ways to record the answers on a piece of paper so that it summarized everything but no information is lost. I'm impressed with how thorough and careful they are. They don't want to make any mistakes.

Each of us was responsible for "presenting" a set of responses and then we would discuss it as a group. It was especially helpful to have the original interviewer present, for she could fill in the context when an answer was unclear. From my log:

Slowly, we're reducing the data - first by recording on one sheet all information on the slips of paper, then categorizing or trying to get the essence of what is in the data. They still have to work on reflecting the data and not their own experience or opinions. In some cases, the response tells us something quite clear. In others, it's more a matter of interpretation. And here their own knowledge of the programme, the respondents and the context helps.

We had talked earlier about the importance of their asking the questions as we had framed them, recording the answers accurately, and not commenting (or making any nonverbal gestures) on anyone's responses. We had discussed their own opinions openly and the fact that other participants not only had a right to different ideas, but that the evaluation report would not be credible if we said that everyone loved the programme and it had no problems. We had practiced how they would handle a respondent who had a different opinion than their own. As we analyzed the data, all of us kept this in mind and would often check this out with each other. We all wanted very badly to produce a report which would be believed by others, for our credibility, theirs and mine, was on the line.

When we had a pretty good sense of what the participant interviews told us, we compared the themes with the information from other sources (which we had analyzed in a similar way). Slowly, a set of conclusions and recommendations began to take shape.

About the time we were completing the analysis, we were asked to take part in a women's action research conference being held at a local university. They were intrigued, but very wary. At first, just as with the Advisory Committee, they said they would not talk. I reassured them that they could discuss their own experience and we agreed to set the whole next meeting aside to prepare for it. They had never been to a university before and were terrified at the prospect of speaking in front of such a group. My log recorded the action:

They were very nervous and clung to their "scripts" rather than let go with the spontaneity they showed at the Advisory Committee meeting. They were just beginning to open up when time ran out. The sea of strange faces was intimidating. This was their first time at a university, let alone speaking to women they perceive as "above them." The next speaker had us break into small groups, so they were forced to let go of each other and interact with others. Of course, they did just fine.


At lunch afterwards, they were on an incredible high. "All those important people actually listened to us! They liked what we said!" Jane was tickled pink to see her name on the programme. Having her name in a "book" made her feel so important. She took a handful of copies to take home. She just couldn't get over it.

As we developed our conclusions and recommendations, I began to put all of it on my word processor. We listed the main points we wanted to make as a group and then I put it into sentences. I brought my computer to our meeting place and we worked together, editing and especially purging "all them professor words." My log notes:

When I used big words, Sue said "This is supposed to be our report, too!" She keeps me honest.

They were also beginning to react to the fact that we were coming to the end of our work together. We had been working quite closely for six months now and our relationships had become intense. Leaving all this was going to be hard.

Over the last few weeks, we had spent increasing amounts of time discussing what each was planning to do after this project was finished. I quote Jane in my log:

"I like this kind of work. Any idea about jobs after this? I'm going to hate it when it ends. I look forward to coming to work. I wish it was full time. I love doing this stuff!"

Later, I comment:

They talked a lot today about what they could do after this project was finished and would very much like to find a way to use their new skills. They really enjoy what they are doing.

As we neared the finish line, there was simply no time to teach them how to use the computer, so I did all the typing. The tension mounted as our time began to run out and we were scrambling to complete the final report. We got it done by the deadline, but it was close. My log notes were brief and to the point, reflecting my own lack of time and fatigue, too:

Finished! Concentrated well. It's really going to be hard for all of us to let go.

The following week, they met with the Advisory Committee to give their final report. This time, they were much less intimidated. (Shortly thereafter, they also were asked to be speakers in a university class. By this time, our preparation consisted of a mere five minute huddle outside the classroom.)


The process of working together

As we proceeded with our task, the group dynamics proceeded apace. Already, by the third meeting, serious conflicts had arisen within the group. Three out of the four were on social assistance and were extremely sensitive to others' attitudes towards them. They had been irritated at Nancy, the fourth member of the group, who had reiterated several times that she was not on assistance (in the context of our discussions of the budget and of being careful not to jeopardize their welfare cheques). This was interpreted as snotty and a put-down by those who were. My log picks up on this:

These meetings are both productive and tumultuous!... After we formulated some questions and tried them out on each other, Jane and Sue asked if they could interview in pairs. Nancy (and Rhonda) expressed some reservations, feeling the respondents might be intimidated by two people. This led to a heated discussion, with Nancy trying to explain why this might be the case and Sue and Jane becoming more and more upset. The issue was clearly not interviewing in pairs, but their anger at Nancy.

This issue would haunt us throughout the process. There were days when we simply adjourned to the local coffee shop to talk it out. I remind myself, in the log that these times are important:

I must be flexible and allow time for personal issues to be discussed and to give support when needed. Today we accomplished very little of the 'task' but a lot of process. Work gets done intermittently, but personal needs must be attended to.

At one point, Rhonda got a full time job and dropped out, and we discussed the possibility of replacing her. The group felt that it would be difficult for a new person to become part of the group: "We've had all our fights and know each other now and how to get along. It would be hard for another person to understand that." The remaining three decided to divide up the extra work and saw no problem. The women's very strong commitment to the task and their conviction of its importance kept them going, however, even when emotions were high.

Each of us also had many other pressures in our lives (particularly illness) which influenced not only how much time we could devote to the task, but also our moods and how well any one of us could tolerate frustration. But, as time went on, they had their own way of working things out, during smoke breaks and on the phone afterwards, which was usually far more effective than any intervention by me. We also built in some fun, going out to dinner to celebrate holidays for example. These are women who have little opportunity to go out alone (without their children) and just let loose.

One major mistake we made was to decide to contact the press. Giving the programme and this kind of evaluation some publicity seemed like a good idea at the time. We decided that our approach would be to talk with a reporter as a group.

Nancy agreed to contact the local newspaper to set up an interview. The reporter insisted that he could speak with only one person and demanded that she do the interview that very evening. Naively, she gave him her address as she tried to explain to him that we wanted to be interviewed as a group. He then showed up at her home and basically bullied her into telling him about the evaluation.

The next day, an article appeared in the local newspaper. Jane and Sue were furious, accusing Nancy of doing it on purpose because she wanted to be the only one in the paper. (To make matters worse, the article highlighted issues of drug and alcohol use by expectant mothers, which, though a concern, was only one small part of our report.) All the old feelings came to the surface again and there was much bitterness. Though we contacted the paper's editor and followed it up with a very sharp letter, we could not undo the damage to the group. An undercurrent of tension pervaded our remaining work together.


Our world is different from yours

One day, as we were working on the questionnaire, the Programme Coordinator came into the room. At the time, she was doing a Master's degree and had recently interviewed two of the team members for her thesis. They commented on how difficult her questions had been. "I really had to think a lot in answering those questions. They were hard!" one woman blurted. Yet, when I asked what her questions were, they were almost exactly the same as the questions we were going to ask. Something else was going on.

Over the next six months, as we worked together, I became more and more aware of the implications of this incident. Their reactions to the Coordinator's questions had to do with how they viewed themselves in relation to her (a university educated professional), what they thought they ought to say or not say, and, above all, protecting themselves from people they saw as having power to harm them. Even though they liked and trusted her very much, there was still a barrier. It could be that they didn't want to hurt her feelings by being critical of the programme. Or perhaps, in spite of genuine feelings of friendship and caring, differences in class were indeed a major issue. This, in turn, had important implications for the question of the quality of data (both its gathering and analysis). These issues were evident in two ways: in our use of language and meaning, and in the question of trust.

Toward the end of the project, the women allowed me to tape a discussion pursuing these questions. We recalled that incident with the Coordinator and asked why they thought they could produce better information than I, a professional evaluator.

"Our world is different from yours. The fear is that people always want to humiliate you, put you down (for being on welfare)... We have a different lifestyle from you. We just don't trust people the way you do."

"The girl I interviewed said she never would have answered all them questions if it hadn't been me."

As we worked together, I became acutely aware of how different our worlds really were.


After the women had been hired, but before we began our work together, I interviewed them individually, pursuing my own research interest in participatory evaluation and the process of empowerment. I asked them about their expectations, what they hoped to get out of their participation in this process. I recorded my thoughts in a reflexive log:

This interview is not only an opportunity for me to gather information on my topic.. it's a chance to get to meet each person individually, in their own home. The atmosphere is informal and it's on their turf.

Later, I concluded in my log that

… the interviews went well. J. and S. were most talkative; N. was brief and to the point. R. talked about her feelings and asked me a lot of questions about social work."

Their recollections six months later were quite different. If I hadn't appreciated the barrier, they certainly did.

"But we didn't know you, OK? We had met you cuz we were applying for this job. We didn't know anything about you. We knew that you were a professor from the school of social work. That automatically says, 'whoa, dude!'"

So they told me what they thought I wanted to hear, and assumed that there was a 'right' answer somewhere (in spite of my assurances to the contrary at the beginning).

"I was scared I'd answer the wrong way… I was trying to give you answers that you wanted to hear.. that would make you feel that I was smart."

In spite of all my professional training and attempts to put them at ease, then, it didn't work very well. What was the difference between what they did in the interviews (with respondents) and what I did?

"I went in there and told them I was nobody... so they were willing to give it (information). It's people that are higher up that they're more scared of... they have a great fear of… being squealed on."

"It comes from different lifestyles… Things are so different. It's the way people live and the way they act, the way they are. You can't help it. You can't change yourself just because you do something.."

They instinctively know "one of their own" from outsiders, and one thing that is meant by outsider is someone with more power, more education, someone who is "up there", as they put it. With such people, they use subtle but very effective survival skills, telling "people up there" what they think they want to hear. Beyond that, however, is also a profound difference in style. They could share a space, a reality which I could not (no matter how nice, easy going, or "cool" I might be). The team members knew how to approach the respondents, putting them at ease.

"You talk about their lives, make them feel comfortable… that's when you're going to start getting information from them... They knew you were on the same level, their wave length… We were participants in the programme, just the same as them… "

"We would just get them going, praise them a bit… When you ask a question, you (meaning this researcher) just stop… and wait for an answer. That's scary! You can't just do that. They need a little bit of encouragement..."

I asked for an example of how they made the respondents feel comfortable.

"(The respondent) didn't have a phone, so on a Sunday afternoon, I had to take my daughter with me cuz I didn't have a baby-sitter. I just went to her door... and I told her who I was and why I was there and if she had a few minutes... So she invited me in and gave me coffee and fed my kid... And we talked about an hour directly about the programme and I spent four and one half hours there. I almost couldn't get away. It took about two hours and she really started to loosen up... If I talked about my labour and delivery, she'd go right into hers. She told me things about her sister and her boyfriend and herself that I'm sure her best friend doesn't know."

"I laughed and carried on and joked with them, made them feel comfortable. I told them stories about my labour… and they just wanted to get in there and tell me everything… They just started talking cuz they got so comfortable in what you were saying."

From the beginning, I knew that we would have to spend time building trust. While they initially quite easily found things in common among themselves, my world was clearly different. They were polite but wary. They waited (and expected) me to say or do something which would confirm their worst suspicions. In many ways, I was the enemy until proven otherwise and I found myself walking on eggs much of the time. Again referring to my initial interview with each of them (in which I took notes as they talked):

"You didn't tell us what you were writing, why you were writing, what was in our interview. We didn't know you. You're in touch with welfare... Our personal life, you were in our home... You just didn't understand.."

All of these things (what, why, etc.), I had indeed explained to them, of course. The issue was far more complex. Later, in my reflexive notes, I concluded that

it's easier to teach these women the methodological/technical skills than it is to teach middle class people about another culture."

The women felt enormously vulnerable. This sense of insecurity and being controlled by outsiders, especially those in authority, was pervasive (and not unjustified, in my view). Yet, they had their own ways of surviving. They told me, for example, that before applying for the job, they had checked me out pretty carefully. In explaining their own fears, one woman described the feelings of a respondent:

"It's people that are higher up that they're more scared of, and that's why... it's because they have a great fear being squealed on… One woman was scared to light up a cigarette. She thought that if I was with them up there that she wasn't allowed to have that cigarette because she wasn't allowed to take enough money out of that budget they gave her to buy herself some smokes."

Somehow they expected my life to be perfect and felt inadequate in comparison. My self disclosure helped break down barriers. On one occasion, I drove one of the women downtown and on the way, she asked if we could stop by my house.

"I didn't want to get too personal, like do you make enough money to live. I'm interested to know how other people live. Other people like in your profession. You are a professional and I'm not. I'm, like, down here. I just wanted to know and I thought you wouldn't mind."

In the course of conversation, she asked me about my daughter (who is the same age as she). I told her about some of the difficulties she'd had. The fact that a child of a professor would drop out of school astounded her. A bubble had been burst and she began to see me, not as a fantasy, but as another human being.

"See, everybody's normal. Everybody makes mistakes! (I thought your kids would be) 'eugenes', snobs."

"What's a 'eugene?'" I asked (confirming her suspicion that I was pretty out of touch).

"Oh man, you don't know what a 'eugene' is? Nerds, smart nerds..."

Gradually, we built up trust, so that in the end, they felt comfortable sharing their feelings with each other and with me. I could make mistakes and they were amused (and even gratified) rather than angry. With the respondents, however, there was almost instant trust, for the class barriers were not there. They understood each other and knew that they had a world in common. Consequently, information and feelings were shared which never would have been discussed with an outsider like me. But their interviews with programme participants were more than information gathering; this was a time, however brief, that they "connected" with another person and were treated with respect.

One might wonder if our work together wasn't just a way of using the women to get information from the programme participants which I couldn't get. Part of my own research work was to look at what they got out of participating in this project, using a framework involving empowerment at an individual, a group and at a community or environmental level (see Whitmore, 1990). I conclude that the women did benefit, especially at the individual and group levels. As individuals, they gained a great deal of confidence, in addition to the specific skills and knowledge learned. The group experience helped overcome the isolation felt by single mothers trying to bring up children alone, giving them an chance to interact with others in an atmosphere of support. We learned to work with our conflicts rather than just walk away; we learned to get beyond the anger and recognize that our common commitment to the task was also a commitment to each other, whatever the disagreements. At the community or environmental level, the benefits were less clear, for this assumes collective action by a more or less cohesive group. The benefits noted- speaking to outside groups, educational and employment opportunities and challenging the social service system - were ultimately tied to each women as an individual. The group was simply not cohesive enough to continue working together toward some kind of broader action.

So, while I would say that they were not simply "used" for my research ends, it is an important question to keep in mind when people like myself seek to involve others in participatory work. Participants should benefit, of course. But the underlying principles of a participatory approach are much more profound than that, and in this case, "empowerment" was limited to the individual and group levels.


Language and meaning

In spite of the fact that I was acutely aware of talking in "academise" and tried hard not to use jargon or big words, my small words were often their big words. What I assumed was "normal talk", they saw as "professor words." My choice of vocabulary, how I put my thoughts together, how I phrase a sentence - all reflect a level of education they didn't have. My language was simply different from theirs, no matter how hard I tried to bridge the gap. In referring to the initial (hiring) interview with them, one assured me that "you did good in the interview, but you just didn't know. You're not used to... talking with (us)." Our common vocabulary was limited.

Words are one thing, of course. Even more important, however, is who asks them and how they are asked. The meaning - what is understood and conveyed - is part and parcel of who and how, as well as what. Here is where their common ground with the respondents made such a huge difference.

"It doesn't matter how much you're trained. We were all brought up different and how to speak, and all these things. But you have to look at who you're talking to, not what you've been taught..."

The way we organized our thinking, how we expressed ourselves, both cognitively and emotionally, were different. The verbal and non-verbal meanings were simply not understood in the same way. They could insult or tease each other and laugh about it; the same things said by me were not funny. Other feelings (hurt, vulnerability, etc.) tended to be expressed non-verbally, not surprising when one realizes that being on welfare makes people feel so vulnerable. On one occasion, I built in a "get it out on the table" exercise to try to get at some underlying conflicts that were interfering with our work. To social workers, this may be good "communication skills", but they viewed it with enormous suspicion and the barriers went back up. We spent hours undoing the damage, which they did much more effectively by getting at the deeper issues in their own way (over the telephone or during smoke breaks).

Another issue here is one of ownership. The words, both in the questionnaire and in the report, needed to be theirs. It was easier with the questionnaire, for this was the tool for communicating with other programme participants. The analysis was also theirs, for it was they who analyzed and interpreted what the respondents meant. With the final report, however, our audience was the Advisory Committee and ultimately a funder. Of the four women, only one had completed high school and the reality was they could not write for such audiences. Such an expectation would be unfair and unrealistic. In the report to the government, we had to use certain language and for the most part, it could not be theirs. Our compromise was to decide as a group what we wanted to say and then I would format it and put it into sentences. Predictably, my sentences and words were not theirs, however, and we spent many hours purging "them big professor words." Even then, they felt the final report was more mine than theirs. How to reconcile the need to write a formal report for government departments and the importance of ownership continues to be a dilemma for me.

The oral presentation to the Advisory Committee was theirs, however. This was a place where they could discuss their experience in their own words and they did it extremely well. The presentation at the action research conference (and later to a university class) were enormously gratifying, as they realized that "they were really interested in what we had to say!" They also enjoyed listening to other presentations at the conference. Their confidence blossomed, as they realized that they had connected with a world which otherwise intimidated and often disrespected them.



What have we learned about oppression, about participation and about quality from this experience? One thing is clear to me: as a middle class, university educated researcher, I could never entirely share the meanings of those from less privileged groups, especially those in the most marginalized sectors of society. The verbal barriers are difficult enough. Beyond the verbal - affective, sense-making, one's experience of the world - understanding is class based, as it is also gender and race based. Our experience of the world is very very different.

The implications of this are being felt in the women's movement, for example. Until recently, the discourse has been dominated primarily by white middle class women, who have just assumed that their experience and priorities reflect that of all women. Working class women, women of colour and Native women are forcefully challenging this, for their reality is certainly not the same.

The barriers are real and very high, even if seemingly invisible on the surface. Passivity or quiescence is often mistaken for apathy; resistance is subtle but well understood by those without power, however. One well known example is the song "Steal away", sung by slaves in America to let others know when the coast was clear to escape. Gaventa, when asked in 1985 what he would change about his book Power and powerlessness: Rebellion and quiescence in an Appalachian Valley, remarked that he would reinterpret what appeared to be quiescence. He later realized that the people had their own, very subtle ways of resisting, which were deliberately invisible to outsiders.

In addition, interpretation of phenomena reflect one's experience and understanding of behaviour and events, which is certainly bound by class, race, gender, culture, etc. While I might conclude that the lack of a birth partner (someone, usually a family member or friend, assisting during labour and delivery) indicated social isolation, the women quickly recognized it as a reluctance to acknowledge the presence of their boyfriends for fear of being cut off welfare.

We think that respondents are telling us the truth, that we are collecting information that is valid. We think that we know the 'true' meaning of what we hear and see. This is a sad illusion. The reality is that the economic, cultural, racial and gender differences among people are profound and extremely complex. To ignore these is to create knowledge which is deeply flawed. (Yet much of our social policy is based on just such knowledge.) Furthermore, it is a fallacy to think that gender alone can overcome these barriers, as Kohler Riessman's (1987) analysis certainly illustrates.

Though these barriers can never be entirely eliminated, they can be reduced/permeated through a participatory process, however. The question of control and ownership is fundamental, of course. Who benefits from the research? Who is in charge of decision making? Who is this really for? The role of the professional changes dramatically, with the focus on process as well as outcome. Such a process calls for skills well beyond the technical ones expected of traditional researchers and evaluators. These include first of all a commitment to the empowerment of others and a clarity of class, race and gender analysis (and age, ability/disability, and sexual orientation as well). It also involves communication skills, an understanding of individual and group dynamics, an ability and willingness to self disclose and share personal feelings and experiences. Such skills are well known in many schools of social work.

This takes time, however, more time than traditional research. It involves a commitment not only to attend to the process but also includes a willingness to go beyond the immediate task in helping people cope with poverty. This means helping to provide such basics as transportation, child care and sometimes money. It means mediating with the welfare department on occasion. It often calls for hours spent listening to personal concerns and responding as best one can. In a nutshell, it means recognizing all participants as human beings, with all our attendant needs, concerns and joys. It also means working at a broader level, taking action to try to change the larger political, social and economic structures which oppress us.

What does this have to do with 'truth'? Truth here refers to "somehow getting it right" (Reason, 1981, p. 241) though I hasten to add that this must not imply that there is one "right" way. More recently, Reason has indicated that "

we can no longer argue that our inquiry is in any sense a search for 'truth.' We can very clearly accept the post-modern statement that we are in a situation 'after truth.' So within this field of emerging practice, we need a methodological inquiry into the question of quality: what is good research. (Reason, 1991)

As Guba and Lincoln (1989) suggest, the real question is not "is it true?" but rather, "does it increase understanding?"

I would argue that the participatory process by itself is important but insufficient to enhance quality. A thorough analysis of the dynamics of oppression and attention to the issues raised here are key to that process if one is to "somehow get it right."



The author wishes to thank Karla Firth and Sharon Hollett for their contribution to the research as co-evaluators. The other two members of the evaluation team could not be reached personally, so their contribution must be acknowledged anonymously.