However, the assumption is that readers will access the other papers in order to understand their definition of choice. It can be argued that their use of a limited ‘subset’ of 137 semi-structured interviews over a period of 39 months (Ball, Oboe and Geekier, 1996, p. 89) can be problematic in that it raises issues of validity and generalization on two accounts: Firstly, the sample size and make-up (that is the number of parents and type of parents interviewed) limits its validity for national or global generalization.
In addition, there appears to be an under- representation of the ‘working class’ in their final year of study. Dooley (1 997, p. 20) points out that only 27 parents were interviewed in the last year of the study. However, it can be argued that the depth of the study provided through evidence from ‘lived markets’ (p. 89) could make generalization more acceptable in this case. Secondly, the changing nature of ‘social settings and participants’ in ‘lived markets’, as Ball, Oboe and Geekier (1996) describe it, can change over time and could produce a different set of results.
For instance, educational reforms at the later stages of the study might have led to interviewees at the end of the study having had very different experiences f the schooling system to those earlier in the study. Agree with Tootles (1997, p. 21 9) assertion that ‘the Education Reform Act  had had the most time to become established. ‘ Thus, supporting the notion that the changing nature of social settings and participants make validity and generalization difficult. Ball and Geekier (1 997, p. 580) counter-argue this: … He relationship between choice and class appeared consistent over the three years despite the passage of time, provides triangular elating and evidence which would appear to strengthen rather than distort our conclusions about he inter-connectedness between choice and class (Ball & Geekier, 1997, p. 580). Nevertheless, the localized nature of the study set in London boroughs can be problematic in terms of generalization across other types of settings, for example: rural areas in England and Wales which do not necessarily have the distinctive features of inner-city London.
Furthermore, the locations of these boroughs are not made clear; as Dooley (1997) suggests they are ‘camouflaged’ (Dooley, 1 997, p. 21 7). This has implications for sampling; assumptions could be made about the class of implies (and access) prevalent in certain boroughs; for example, the population is mainly privileged class in Chelsea and Kensington, and working class in Tower Hamlets, therefore, it would have been more helpful if researchers have included a less ‘camouflaged’ account of, and clearer justification for, their choice of sampling.
The limitations of their qualitative research and sampling could also have been more explicitly acknowledged and perhaps a more clear explanation of their sampling strategy provided. Nevertheless, the overall validity of using a qualitative approach in itself for he purposes of this particular study is not questioned; it is through this qualitative approach that a number of interesting nuances, relations pips and complexities between social class and choice of school has been highlighted throughout the study.
Therefore, I agree with Ball and Geekier (1997) that: Dooley wants to ignore the complexities, contradictions and ambiguities which are embedded in the lives, perspectives and actions of families. (Ball and Geekier, 1 997, p. 584) To further support this notion, Flick et al. (2004) assert that: its [qualitative] approach to the phenomena under investigation is frequently more open and ‘more involved’ than other research strategies…. Take into account the views of those involved and the subjective and social constructs of their world. (Flick et al. 2004) Ball, Oboe and Geekier (1996) provide a table with the breakdown of occupation, education and housing of their sub-sample of parents, suggesting that this might help the reader ‘gain some grasp of the terrain Of schooling across which the choices examined are being made. ‘ (up. 89-91 With this analysis of the ‘lived’ (p. 89) markets, I would argue that the insertion of the table at the point where it is goes little to add to the clarity of the analysis or findings, and it is only later on that explicit links and references are made to the table.
Therefore it appears to add to the complexity of the reading at this point. It relies heavily on assumptions that readers can make sense of how it fits in with the ‘analysis of three ideal-types of school chooser (p. 92). In addition, it would have been helpful if a succinct description or explanation of what they mean by ‘lived’ markets had been included, and perhaps the links of how the information in the table fitted to the rest of the study could have been lealer.
Related to this, it appears that Ball, Oboe and Sewerage’s (1996) categorization of families has led to an almost ‘forced-fit’ of certain social classes into three neatly-defined categories as indicated in Figure 1 (p. 93). The sample of families appears to fit neatly into predetermined categories rather than a naturalistic fit. Dictionary. Com offers a definition for the term naturalistic: Naturalistic refers to the imitating of nature or natural surroundings. Dooley (1997) raises this as an ‘personalization of concepts’ issue in his response to this study.
This is discussed further on. Furthermore, it can be argued that the neatly-fitted categories, and indeed the sampling over a three-year period, disregards social mobility, and the dynamics of the complex biographies and lives of families is ignored. Examples of this include upward mobility, increase in household wealth or incomes, upward movement in social status and, on the other end of the scale, loss of earnings or wealth, loss of status. If social mobility and other changes (for example: environmental and external influences) in the ‘lived markets’ (Ball, Oboe and Geekier, 1 996, p. ) over time were taken into account and included in the findings, then perhaps very different conclusions could have been drawn. It could be assumed, and perhaps viewed with some suspicion, that the researchers deliberately avoided this so that their case for relationship between social class and school choice could be validated. In addition, the specific labeling of each category as ‘privileged’, ‘semi-skilled’ and ‘disconnected’ (Ball, Oboe and Geekier, 1996, p. 92) raises some questions about their use.
Each label comes with its own connotations that could be easily misinterpreted. These labels may give a succinct picture of the type of people that would fit into that particular category, but perhaps only in a very simplistic way. The researchers do little to acknowledge or clarify this. It seems that Ball, Oboe and Geekier (1996) only justify the use of their chosen category titles as a means to an end, by which I mean being able to fit families of different classes (and their findings) neatly into each category.
For example, some families that fit into the privileged category, although of professional standing or middle class, might not necessarily have the ‘expert’ coding skills required to make effective choices, whereas someone in the ‘disconnected’ class, although it is assumed they are insufficiently educated, might display a level of decoding skills that transgress what is assumed of their class. A notion of arrogance can be inferred from Ball, Oboe and Gigahertz’s (1996) description of the four qualities of the privileged class as qualities solely belonging to that particular class; for example: ‘inclination to choice’ (p. 3); ‘capacity to engage’ (p. 93); ‘desires and concerns’ (p. 93) and ‘response to schools’ (p. 94). It is not sufficiently acknowledged or explained that these could transgress the social classes and do not necessarily only feature strongly within the privileged class – there is insufficient data presented to support this. It is also problematic to assume that ‘working class’ families are indeed ‘disconnected’ – the word itself has negative connotations.
The evidence provided suggests otherwise, for example ‘child-matching’ (Ball et 1996, p. 94) would Suggest a deep connectedness to the process Of choice of schools, and this appears to be evident in both privileged and disconnected categories. The idea of ‘plant and facilities’ (p. 1 06) as characteristics that only impress the ‘disconnected’ and that they do not have the ‘language of secondary educational meanings’ (p. 106) is presumptuous and is not evident in their findings.
Some of the commentary made within the ‘disconnected’ category does not seem to be backed by evidence in the findings, but rather appears to be inductive on the part of the researchers; for example, there is no data to show: ‘disconnected families typically see schools as much the same’; ‘there are limits to disconnected families’ ability to see (Ball et al. , 1996, p. 08). It is true that the extracts from three ‘working class’ mothers do mention plant and facilities as impressive, however it can be argued that this is not sufficient to generalize the above statements to all ‘working class’ families.
Furthermore, there is no acknowledgement that there is a possibility that ‘plant and facilities’ (p. 1 06) could be an important factor in choice for privileged choosers. The fact that the data does not show these factors or other variables could call into question what questions were asked at the interview and how they Were asked. There is also some disparity in the reservation of the findings which seems to favor the privileged chooser. It appears the bulk of the findings and commentary is devoted to the privileged class.
Whether this is intentional or a natural outcome of the analysis of data is questionable. Sample size in favor of the privileged class is discussed further on. In addition, the majority of interviewees appear to be mothers. There is limited evidence to reflect the views of fathers within the ‘working class’ group, or indeed any other family members or careers. It also excludes any influence that other family members, including fathers, may have on the sews that mothers have or present.
Linked to this is the issue of a definition of what ‘family’ means; there is no clarity on what is meant by ‘family. A question arises as to whether ‘family’ is used in the traditional sense Of the word, or whether it includes family in any other sense, for example: does it include single-parent families, extended families or foster parents. Furthermore, I agree with Dooley (1997, p. 219) that their typology is strongly class-related and that the ‘disconnected’ fall ‘overwhelmingly’ into the irking class’ and the privileged are ‘overwhelmingly’ middle class.
Categorizing families into these three simplistic categories is contradictory to Boride’s descriptions; even though it is acknowledged that some parallels are drawn with Broodier (Ball, Oboe and Geekier, 1996, p. 92), this has implications for generalizing views from each category and increases the complexity of understanding of the interplay of ‘human actions’ in the ‘social field’ where choice is located. It is interesting to note that the researchers point out: ‘these are ideal types rather than naturalistic categories’ and that these present trends patterns in the data. (Ball, Oboe and Geekier, 1996, p. 2). However there is no clarification about what are naturalistic categories or explicit justification as to why the data could not be presented in these naturalistic categories. Ball, Oboe and Geekier (1996) rely heavily on Boride’s (1986) work on social judgment and taste. It could be argued that there is almost an expectation that the reader is accepting of Boride’s ideas or definitions of: cultural goods, social origin, educational capital, choice, competition, social ‘field’, ‘disposition to recognize’ and ‘capacity to adopt the esthetics disposition’ amongst others, as ‘truth’.
Although the researchers’ suggestion that ideally this paper should be read alongside their other research in this series is acknowledged, it is essential in particular studies to ensure concepts such as ‘taste’ and ‘social judgment’ are used with such clarity that it does not leave it open to any interpretation that can affect the scope of understanding of the study.
It can also be argued that relying heavily on one particular researcher (in this case, Broodier) or body of evidence (Boride’s work) can lead to bias on particular viewpoints or perceptions of incepts, ideas and findings or evidence that fits into what the researcher aims to present as ‘truth’, in this case choice is influenced by and perpetuates social class. Any limitations or contradictions to Boride’s views have not been acknowledged in this study. In addition, concepts used such as market forces (p. 89), economy of cultural goods (p. 91 ), social capital and cultural capital (p. 3) have not been clearly defined. There is an assumption that readers will have an adequate understanding of these concepts in relation to the study. Dooley (1997) challenges Ball, Bow and Geekier (1996) on two rounds: firstly on methodological grounds, and secondly on the ‘personalization of their concepts’ (Dooley, 1997, p. 217) in his aims to discredit their research based on these grounds, and it appears also to discredit qualitative research. Although a valid point is made, that qualitative research should not be immune from the scrutiny of the research community (p. 1 8), Toeless suggestion that ‘reasonable judgments’ be made by applying ‘certain criteria and that heuristic judgments be made on the plausibility and creditability of evidence supplied’, in a way, devalues qualitative research. Toeless (1997) claim that his paper ‘fits squarely’ into a normative framework set out by Foster et al. (1996) and Hammerless (1992) (p. 21 8) suggests that he finds it unproblematic to question the validity of the qualitative approach taken by Ball, Bow and Geekier (1996) without reproach.
However, Dooley (1997) does acknowledge that their (Ball et al. ‘s. , 1996) research gives us an insight into the complexities of the school choice process: ‘There is a clear need and value for this kind of qualitative work . Insights into school choice process. ‘ (Dooley, 1 997, p. 220). In fact, it might be en as a contradiction that Dooley is questioning the qualitative approach but chooses to include very substantial evidence from Severity et al. , (1 993, p. 7); Oboe et al. (Bibb, p. 69) and Oboe et al. , (Bibb, p. 5) to support a disregard for the use Of statistical methods. Nevertheless, there is acknowledgement that ‘eschewing’ statistical methods is legitimate (Dooley, 1997, p. 220). Dooley (1997) offers five limitations as to why Ball, Bow and Gershwin’s (1996) sample selection tends to ‘inhibit (p. 220) generalization: the first is the localized nature of the sample: sample is based in London only treasures of that city not generalize beyond the capital’ (Dooley, 1997, p. 220) – it is claimed that there are features of the city that cannot be generalized beyond London.
It can be argued that although this might be true of some of the characteristics of the city (characteristics which Dooley does not specify), there are significant features, for example, common characteristics found in different classes, which can be generalize nationally or globally. One of the key notions presented in Ball, Bow and Sewerage’s (1996) study is that there are distinct features that are prevalent in ACH type of class, for example, employment status and type of employment does tend to fit in with certain class groups.
This is a picture that can be seen nationally or globally, therefore this tends to weaken Tootles argument that the findings cannot be generalized. The second has to do with the social mix of the sample. I agree to an extent with Tootles (1997, p. 220) claim that the sample might not necessarily be representative of a social mix that could be found in particular London boroughs. In addition, the ‘camouflaging’ of the sample makes it difficult to ascertain the exact social mix and demographic of he sample. Ball and Geekier (1997, p. 79) dismiss Tellers ‘misgivings’ about the social mix of the sampling and attribute it to a typesetting error. As suggested earlier, evidence of a more transparent, explicit and rigorous sampling strategy in Ball, Bow and Sewerage’s (1996) study could have increased acceptance of the validity and generalization of their findings. The third limitation is to do with sample size. A sample that is 58% ‘middle class’ and only 21 % ‘working class’ (Dooley, 1997, p. 220) appears to be disproportionate and unrepresentative of typical London boroughs. Dooley 1997) argues that the sample is too small to make it generalize.
It would appear that this criticism, together with the overall critique of Ball, Bow and Sewerage’s (1996) methodological approach, stems from a positivist perspective, where there is greater emphasis on statistical values and data, and higher values perhaps indicate more reliability in the findings. Dooley (1997, p. 221) suggests that Ball, Bow and Geekier (1996) have avoided any statistical machinery. From a qualitative standpoint, I disagree with Dooley (1997) that it might have been necessary to use any ‘statistical machinery’ Dooley, 1997, p. 21) to validate the findings of the study. Qualitative researchers, and in this case Ball and Geekier (1997, p. 579), would argue that particular themes from small localized samples can be generalized, especially if the main focus was on ‘theoretical saturation’ (Strauss, 1 987, up. 25-26) rather than sample size. Morse (2004) defines theoretical saturation as ‘the phase of qualitative data analysis in which the researcher has continued sampling and analyzing data until no new data appear and all concepts in the theory are well developed. The fourth limitation is to do with an ‘inductively constructed’ sample (Dooley, 1997, p. 220). Dooley (1997) claims that Ball, Oboe and Geekier (1996) ‘sought to find members of groups so as to ‘match as far as possible the socio-economic and ethnic three cluster areas. ‘ (Geekier et al. , 1995, p. 1 3). This is a contradiction, as Dooley (1997) found the lack of social mix of the sample earlier as problematic. Now he finds it problematic that Ball, Oboe and Geekier sought to address the issue of social mix.