Thomas (1997) presents and interesting, but somewhat flawed, qualitative study of disablism as it applies to attitudes towards mothers or soon to be mothers with disabilities. Thomas draws on data from in-depth interviews with 17 disabled mothers or disabled to be mothers. She makes her presentation and describes the manifestation of disablism applied to these women in terms of three themes that she admittedly subjectively chose: (1) the struggles, both personal and those placed on them from others regarding the risk of giving birth to children that may suffer disabilities themselves due to the medications the mothers take or due to genetic deformities; (2) doubts from others about their being able to adequately parent their children given their disability (the good mother theme); and (3) the experience of receiving unwanted help and from others due to their disabilities (fueled by the perception of their being inadequate mothers). The problem here is that Thomas has chosen two themes that are so closely related they are almost one in the same, the unwanted help theme and the ability to parent or good mother theme are highly related; in fact the unwanted help theme is a consequence of the good mother theme. In reading the paper I found her description of these two themes and the experiences of the participants regarding these somewhat redundant. After reading the paper I was hungry for broader view of the experiences of these women. Thomas admits that there were other themes that emerged from the data and it may have been more helpful to focus on more disparate themes from the data analysis to get a better picture of these women's overall struggles and experiences. With respect to the data collection and analysis itself, Thomas does not spend a great deal of time explaining how the data was coded and analyzed. Her description of the sample and data collection is more than adequate; however, we are not explained as to how the interviews were coded or the criteria for selecting themes from the analysis. We do not know what themes were most prevalent, how severity or intensity of perceived disablism was rated and coded (surely some descriptions were not as florid as others), how data was clumped into themes, etc. We are told that the data was cross-indexed and there were 19 categories grouped into themes, but we are not given a table of the categories or the themes they were grouped under. Moreover, we are not given the criteria that were utilized to group them into themes. This were tables can be useful. All it would have required is a table of the categories and a brief description of how these were grouped into themes. It is important in any study to describe the analysis and data collection in such a manner as that readers could reproduce the procedures utilized in the study based on reading it, otherwise there are no grounds to accept the findings as trustworthy. Even in qualitative research, any hint of subjectivity is detrimental to the research conclusions.

This brings up a discussion of trustworthiness in this data. Thomas reports that the themes in her study could be argued to be manifestations of disablism. What are the criteria for this argument? She offers none. Is there any research indicating that these themes are observed to a greater extent in disabled mothers than nondisabled mothers? We are not given the answer. For instance, Thomas states that her study suggests that disabled women are particularly vulnerable to the "at risk" theme. But there is no evidence in the study at all that this is true. She assumes it. Is it then possible that these experiences might be common to those women experiencing motherhood in general? Thomas acknowledges this to some extent, but continues to assert that disablism is responsible for the experiences of these women. Maybe they are, maybe they are not. We really have no way of knowing based on this study. Additionally, Thomas believes the even though the sample and procedure may not lead to generalziability, her theoretical argument is generalizable due to the sample not being atypical of disable women (p 627). I would disagree somewhat given these earlier concerns. However, transferability may occur if practitioners find that their situations are similar to the ones in the study (Bassey, 1981). If we assume that there are many reasons for the manifestation of these themes regarding motherhood including disablism, overbearing relatives or friends, marital strife, past experiences, ageism, etc. then perhaps the experiences of these women are transferable to a broader group, but are not indicative of...

In this respect the experiences of the women in this study may be quite valuable in many different contexts.
This leads me to issues with credibility. As mentioned above we do not know if the data is operationalized consistently or how it was operationalized at all. In addition, as mentioned above, do the experiences of these women represent disablism? We really have no way of knowing. But, we could have found out quite easily. Thomas had access to a large data set consisting of interviews with 1400 women. Could she have matched her 17 participants with other mothers from the data base and crosschecked to see if nondisabled mothers had similar experiences to the disabled mothers? Thomas could have easily done this. By comparing the experiences of disabled mothers to matched nondisabled mothers we would have received some valuable insight as to where disablism begins and where some of these experiences may be relatively common. While I do believe that many of the experiences these women had were fueled by disablism, a researcher should be able demonstrate empirically or cite research concerning this for credibility purposes. Thomas assumes that these experiences represent disablism because her sample is disabled and this not good research.

Moreover, the research method using only interviews of the patients might also lead to poor credibility (Yin, 1994). The use of triangulation, different methods of data collection used together, compensates for the individual limitations of each and takes advantage of their strengths. Perhaps other methods of data collection might have been entertained. There are also real checks of the data, no peer reviews of the collection and analysis which could have contributed to stronger creditability.

In terms of the dependability of the data, it is reasonably certain that the same participants would have responded in a similar manner at a later date or given a different researcher. But there are some other issues with dependability. For instance, we also know that several different disabilities are present in the sample and yet Thomas only discusses the subjective experiences of a select few of her participants. Do we believe that someone with asthma would experience the same levels of disablism as someone with a missing limb or with a psychiatric diagnosis? Should we believe that? Thomas does not delve into this issue at all. Could there be a relationship between a type or category of a disability and disablism? We would think that there might be and we should question the notion that a disability is just a disability that leads to disablism. Almost everyone in the study has a different disability. Four have back or arthritis related problems and two have different amputations (one an arm amputation the other has a leg amputation, thus this in itself may be a confound). The rest of the sample has a wide range of issues ranging from lupus to a mental disorder. Therefore, I would have some question as to the dependability of the data insofar as the manifestation of disablism is concerned. It would have been more prudent to at least have included more qualitative data from all patients.

Qualitative data analysis is tricky to say the least. No one study can do it perfectly. Likewise, no quantitative study is without flaws. But in this study there does not seem to be the attention to detail that one would expect from a study in a peer-reviewed journal. The real value of qualitative studies lies in their ability to provide what quantitative studies do not do, a focus of the subjective processes that occur in empirically demonstrated variables, constructs, or qualities. However, it does not make sense to perform a qualitative study on a construct that may not be empirically validated. Should physicists study the dimensions of a flat earth? Should psychologists study how certain phrenology constructs contribute to the current unemployment situation? Likewise, if disablism exists, and it probably does to some extent at least, we would need to empirically validate in what ways it exists before we assign qualitative measures to it. Qualitative and quantitative findings need to work hand in hand to support one another.

Nonetheless, this study is like a Rorschach analysis. While there may be some concerns regarding the methodology it still provides some very rich descriptions and insight into the subjective experiences of mothers who face significant barriers regarding their right to propagate and raise their children. The strongest data from this study with regard…