chuyên đề

Qualitative Data Analysis: Establishing typologies

06/01/2018    hoangthuphuong   PHD Journey

TH E N A T U R E OF T Y P O L O G I E S

Typologies have two important characteristics:

  • - they are multidimensional or multifactorial classificationsThat is, they combine two or more different dimensions so that a more refined or complex portrayal of a position or characteristic can be identified.
  • - they offer a classification in which categories are discrete and independent of each other. In other words, a feature or individual can only be assigned to one category. It is this latter property that gives them particular value in 'dividing' or 'sectoring' the social world.

 

D E T E C T I N G A N D D E F I N I N G T Y P O L O G I E S

There are a number of steps to be taken in the detection of a typology:

1) To identify the relevant dimensions of a typology. Lazarsfield and Barton (1951) describe this as the 'dimensions which underlie the discrimination' made by the typology. For this, it is important for the analyst to have a strong familiarity with the data set and that tasks further down the analytic hierarchy, such as identifying the elements of a phenomenon and refining categories, have been completed. This process is illustrated in Box 9.9 that without having fully understood a single phenomenon and then its relationship with other phenomena, It will also be worth investigating hunches developed in earlier analytical tasks or those that emerged during the course of fieldwork. Through such investigations, it should be possible to identify the dimensions that 'discriminate' in the typology and to rule out others.


2) To ensure that all the cases can be assigned to each of the dimensions being used in the typology. Unless the sample fits into each of the dimensions, and fits uniquely, the dimensions will not operate effectively within the typology. Once the dimensions of the typology have been checked in this way, then their cross-fertilisation into typographical categories can be made. Once this has been done the whole process of testing needs to start again to ensure that all cases can now be allocated to one, and only one, of the typological categories. The power of a typology lies in its ability to locate all cases in a series of related but independent categories. If there are people or cases that cannot be uniquely assigned, then the conception of the typological construction, or the dimensions on which it is based, need to be revisited.

 

The investigation of 'fit' to each of the typological categories is clearly a very iterative process. It will involve constant checking and tweaking of the assumed dimensions of the typology and the typological categories developed.

It is the cases that fall between categories or do not seem to fit neatly into any of the categories that are the key to developing a strong typology. 

Interrogation of these may bring to light a missing dimension from the 
typology or suggest an extra typological category. The analyst should continue to allocate and reallocate cases to the available typological categories until all cases have been assigned. This process undoubtedly requires adherence to an 'analytical conscience' in ensuring that all cases do fit and fit one category only. If they do not then the typology will need to be rejected as a method of sectoring although it still may be useful as a form of classification.


A final point about typologies concerns the internal coherence of the categories. While the categories themselves need to be discrete from each other, some variations within categories may occur in relation to the dimensions used for construction. Indeed, capturing any variation within categories will form part of the descriptive account of the typology. 


Although typologies are very powerful analytical tools, it is important to remember that they are not always appropriate or required. Not every qualitative study will lend itself to the creation of a typology, and it is possible to waste a lot of precious analysis time searching in vain for tenuous links between groupings of phenomena. Hammersley and Atkinson (1995) argue that to be effective, a typology should give good purchase on the data, and help explain differences - rather than be a purely conceptual exercise.

Developing a typology using charted data may be carried out within one chart or across two or more. It will involve cross-case analysis so that dimensions on which the study population divides have been identified. These will form the main dimensions of a typology.

Having identified the dimensions, the next step is to test them across the data set. It has to be possible to apply one, and only one, category to each participant for that dimension to work within the typology. Once the dimensions have been tested in this way, and if necessary revised - the next step is to decide how they combine to form typological categories. Multidimensional typological roupings are then constructed, and usually assigned some working title.

Once these newly constructed, multidimensional, categories have been devised, they then need to be tested across the whole sample. Each case is then inspected as a whole (that is looking horizontally through the set of charts) checking each of the phenomena that form the typology. On the basis of this, one of the typological categories is assigned and then clearly annotated on the chart

Source: Miles, M. B., Huberman, A. M., & Saldana, J. (2013). Qualitative data analysis. Sage.

 

CÁC BÀI VIẾT KHÁC CÙNG CHUYÊN ĐỀ

Qualitative data analysis: Analytic hierarchy
Qualitative data analysis: Analytic hierarchy
The analytic process requires three forms of activity: data management in which the raw data are reviewed, labelled, sorted and synthesised; descriptive accounts in which the analyst makes use of the ordered data to identifying key dimensions, map the range and diversity of each phenomenon and develop classifications and typologies; and explanatory accounts in which the analyst builds explanations about why the data take the forms that are found and presented.
Computer-assisted qualitative methods
Computer-assisted qualitative methods
Weitzman (2000) provides the most up-to-date categorisation of CAQDAS software, building upon earlier work with Miles (cited above), which categorises software into five types: