10.07.2015 Views

essential-guide-to-qualitative-in-organizational-research

essential-guide-to-qualitative-in-organizational-research

essential-guide-to-qualitative-in-organizational-research

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

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

–––––––––––––––––––––––––––––––––––––––––––––––––– ATTRIBUTIONAL CODING –––––––––– 239similar <strong>to</strong> that achieved us<strong>in</strong>g structured questionnaires. Good reliability requires clearlydef<strong>in</strong>ed dimensions and coders who have had sufficient tra<strong>in</strong><strong>in</strong>g and practice. Reliability iscalculated us<strong>in</strong>g Cohen’s kappa (see Fleiss, 1971) for approximately 20–30 per cent of theattributions extracted for a study. In general, Fleiss (1971) suggests that kappas above 0.4 areconsidered acceptable, whereas those above 0.6 are good.A further advantage of cod<strong>in</strong>g spoken attributions rests with the number of attributionsit is possible <strong>to</strong> generate. In comparison with questionnaires which may rely on six positiveand six negative attributions, the focus on spoken attributions for real events means that severalthousand attributions can be generated from a s<strong>in</strong>gle <strong>research</strong> study. Moreover, as attributionsare coded on each dimension, the result<strong>in</strong>g data set can prove large enough <strong>to</strong> permit<strong>in</strong>vestigation of both nomothetic and idiographic patterns. By enter<strong>in</strong>g the data <strong>in</strong><strong>to</strong> an SPSSdata sheet, it is possible <strong>to</strong> select certa<strong>in</strong> types of attributions, for example where the speakeris Agent and outcomes are positive.ADVANTAGES AND DISADVANTAGES OF THE METHOD ––––––––––––––––––––––––––––––––––––––––––Researchers need <strong>to</strong> weigh the advantages and disadvantages of us<strong>in</strong>g any <strong>research</strong> methodbefore determ<strong>in</strong><strong>in</strong>g which one best suits their particular <strong>research</strong> question. The potentialdisadvantages of attributional cod<strong>in</strong>g lie with its complexity – it can be more costly <strong>in</strong> termsof time and money than traditional questionnaire methodology. Moreover, it will take timebefore the newcomer feels entirely confident with the method, and able <strong>to</strong> achieve adequatelevels of reliability. Attributional cod<strong>in</strong>g also depends upon transcribed material and, thereforethe time or funds necessary <strong>to</strong> acquire such transcripts. However, a grow<strong>in</strong>g body oftranscribed material is now available through outlets such as Qualidata sponsored by the ESRC<strong>in</strong> the UK (see Chapter 30). This should help <strong>to</strong> reduce costs and <strong>in</strong>crease opportunities for<strong>qualitative</strong> <strong>research</strong>ers. Such issues aside, attributional cod<strong>in</strong>g can be a very reward<strong>in</strong>g methodfor explor<strong>in</strong>g how <strong>in</strong>dividuals make sense of their world.However, as mentioned previously, the analysis of public or spoken attributions sitssomewhat uneasily between the two powerful camps of constructionist and reductionist<strong>research</strong>. This means that <strong>research</strong> is potentially open <strong>to</strong> challenge by advocates of bothapproaches and, unfortunately, there is still a tendency <strong>to</strong> mistake the method for a particularepistemological approach. I believe that a strong advantage of attributional cod<strong>in</strong>g is that ithas the potential <strong>to</strong> be used by both <strong>qualitative</strong> and quantitative <strong>research</strong>ers. However, this doesmean that the <strong>research</strong>er needs <strong>to</strong> be absolutely clear about his or her assumptions when us<strong>in</strong>gthe method and report<strong>in</strong>g their f<strong>in</strong>d<strong>in</strong>gs. Quantify<strong>in</strong>g <strong>in</strong>dividual attributions accord<strong>in</strong>g <strong>to</strong> apre-specified cod<strong>in</strong>g framework fits with a reductionist perspective <strong>in</strong> that it enables the<strong>research</strong>er <strong>to</strong> explore consistency and track change <strong>in</strong> the patterns of attributions producedby <strong>in</strong>dividuals or groups of <strong>in</strong>dividuals. But, as Marshall (1994) po<strong>in</strong>ts out, the positivisticstandpo<strong>in</strong>t which accepts language as a transparent medium through which cognitions aretransmitted unproblematically, and without dis<strong>to</strong>rtion, is very much open <strong>to</strong> question. Indeed,when study<strong>in</strong>g public or spoken causal attributions, it is difficult (and no doubt unwise) <strong>to</strong>ignore the importance of context <strong>in</strong> contribut<strong>in</strong>g <strong>to</strong> the type of attributions produced.Similarly, rather than assume that discourse is divorced from cognition, the greater challengewill be <strong>to</strong> determ<strong>in</strong>e how context, cognition, personality and skill <strong>in</strong>teract when <strong>in</strong>dividualscommunicate their causal understand<strong>in</strong>g of the world <strong>to</strong> one another.

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

Saved successfully!

Ooh no, something went wrong!