20.02.2013 Views

Qualitative_data_analysis

Qualitative_data_analysis

Qualitative_data_analysis

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

SPLITTING AND SPLICING 155<br />

Although we noted that not all fixed images are stereotypes, we didn’t spell out<br />

any criteria for distinguishing between ‘fixed’ and ‘unduly fixed’ images; nor did we<br />

figure out how to categorize any <strong>data</strong> which fitted the former rather than the latter.<br />

Also, many of the <strong>data</strong>bits assigned to temperament and to occupational<br />

characteristics ‘other than task’ do seem stereotypical—our dentists are poor, moody<br />

and volatile. As our <strong>data</strong>bits often relate to temperament as well as task, perhaps the<br />

stereotypical element in the former has coloured our interpretation of the latter?<br />

We can see now how first impressions combined with indirect support from<br />

related evidence could lead to our interpretation of the category ‘task’ as an aspect of<br />

stereotyping. By retracing our steps, reviewing our definitions and reassessing our<br />

category assignments, we give ourselves the space to reflect critically upon this<br />

interpretation, and if need be to modify or discard it.<br />

This reinterpretation of the category ‘task’ as dealing with occupational<br />

differences rather than occupational stereotypes has wider implications for our<br />

<strong>analysis</strong>. It shifts our attention from substance to style. Instead of invoking or<br />

depending on stereotypical images, the <strong>data</strong>bits assigned to the category ‘task’ rely<br />

rather on the absurd results of transposing characteristics from one occupation to<br />

another. It is through the incongruity of these images that Woody Allen achieves his<br />

humorous effects.<br />

Before we consider the implications of this reassessment, let us consider how we<br />

can choose between rival interpretations of the <strong>data</strong>. Our first interpretation was<br />

sufficiently plausible that we could work with it over a period, and yet we now want<br />

to discard it in favour of one which may give a quite different tenor to our results.<br />

Perhaps at this point we might be tempted to curse qualitative <strong>data</strong> <strong>analysis</strong> and<br />

invoke a plague on all interpretations. On the other hand, we may derive some<br />

confidence from the fact that we have been able to discriminate between different<br />

interpretations. By confronting the evidence critically, by making our categories and<br />

decisions as explicit as possible, and by retaining scepticism even with regard to<br />

categories central to our <strong>analysis</strong>, we have given ourselves the space to review and<br />

recant. We have not simply looked for confirmation of our initial categories by<br />

accumulating as much evidence as possible in their support. In short, when we<br />

review our categories in the light of our retrievals, we should be looking for<br />

confrontation with the <strong>data</strong> rather than confirmation of our categories.<br />

If we accept this reinterpretation, then we have to modify our <strong>analysis</strong><br />

accordingly. First we have to review all the <strong>data</strong>bits assigned to the sub-categories of<br />

stereotype, to distinguish those which are stereotypical from those which are not.<br />

This will require us to be more precise about what we mean by a ‘stereotype’ and<br />

what criteria we can use for assigning the <strong>data</strong>bits to the different categories. Thus we<br />

may distinguish between stereotypical and non-stereotypical images in terms of<br />

whether or not the assumption involved is reasonable. For example, it is not

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!