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...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

––––––––––––––––––––––––––––––––––––––––– THE TWENTY STATEMENTS TEST –––––––––– 89APPLICATION OF THE REFERENTIAL FRAMES METHOD ––––––––––––––––––––––––––––––––––––––––––Our own <strong>in</strong>itial <strong>research</strong> with the TST focused on the changes <strong>in</strong> self identity whichaccompany the career entry of a group of graduate entrants <strong>in</strong><strong>to</strong> a large technically basedorganization (Arnold and Nicholson, 1991; Nicholson and Arnold, 1989a, 1989b, 1990). Indevelop<strong>in</strong>g our methodology for the <strong>research</strong>, we sought measures which would reflecttransformations <strong>in</strong> aspects of self identity over the course of the entry transition. We assembleda battery of methods for this purpose, <strong>in</strong>clud<strong>in</strong>g reper<strong>to</strong>ry grids, <strong>in</strong>terviews, and standardizedquestionnaires. Many of these methods tend <strong>to</strong> bias <strong>to</strong>wards stability (‘retest reliability’ <strong>in</strong>test<strong>in</strong>g jargon), and we wanted <strong>to</strong> be sure that, even over a short period of time, we would<strong>in</strong>clude <strong>in</strong>strumentation highly sensitive <strong>to</strong> subtle shifts <strong>in</strong> identity, through which we wouldbe able <strong>to</strong> test differential predictions about personal change.We hypothesized that, over time, there would be an <strong>in</strong>crease <strong>in</strong> frequency and saliency ofoccupational self-identifications; for example, reference <strong>to</strong> their specialist roles <strong>in</strong> eng<strong>in</strong>eer<strong>in</strong>g,comput<strong>in</strong>g or market<strong>in</strong>g, and <strong>in</strong>creas<strong>in</strong>g endorsement of <strong>organizational</strong> values and norms. We alsoexpected a decreas<strong>in</strong>g proportion of C-type/reflective statements, and an <strong>in</strong>crease <strong>in</strong> B-type/socialstatements. We were especially attracted by the possibilities the TST offered for reveal<strong>in</strong>g degreesof change <strong>in</strong> free-float<strong>in</strong>g reflective constructions versus more socially anchored self perceptionsthrough this period of early career role adoption and professional identity development.The group of 97 graduate entrants was made up of 33 1982 entrants, all of whomcompleted the TST with<strong>in</strong> two weeks of jo<strong>in</strong><strong>in</strong>g the organization, and a further 16 (four fromeach of four departments) who had jo<strong>in</strong>ed the organization <strong>in</strong> each of the years 1978 <strong>to</strong> 1981.Of the 97 who completed the standard TST <strong>in</strong> autumn 1982, 94 completed it aga<strong>in</strong> a yearlater, 32 women and 62 men. We felt that an atmosphere of trust, develop<strong>in</strong>g over the courseof <strong>in</strong>tensive <strong>in</strong>terviews, was important for the successful adm<strong>in</strong>istration of an <strong>in</strong>strument suchas the TST, which requires respondents <strong>to</strong> be unconstra<strong>in</strong>ed <strong>in</strong> their choice of self def<strong>in</strong>itions.Individuals completed the form <strong>in</strong> privacy, and then handed it <strong>to</strong> us for analysis.Confidentiality was assured and guaranteed by a system of unique codes. At the first phase,respondents produced an average of 19.7 responses; and at the second phase, 19.5 responses.This yielded a <strong>to</strong>tal of 3,774 statements <strong>to</strong> classify and analyse.RATING SYSTEM ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––The system we used for rat<strong>in</strong>g statements was multi-level, provid<strong>in</strong>g three classifications:response type (referential frame); for C-mode/reflective statements only, we constructed ourresponse content level (the SICV – Skills, Interests, Character and Values System); andresponse value (self evaluation). The purpose here was <strong>to</strong> evolve a system of rat<strong>in</strong>g whichcould tell us <strong>in</strong> a standard form someth<strong>in</strong>g about the content of people’s responses. The authorsdeveloped the system after many iterations with the data and ad hoc rat<strong>in</strong>g trials, before it wasfully developed for test by <strong>in</strong>dependent raters. The earliest systems we used proved unwieldywith <strong>to</strong>o many overlapp<strong>in</strong>g categories, with which we failed <strong>to</strong> achieve a reasonable level ofreliability. Return<strong>in</strong>g <strong>to</strong> this issue some time later, another social psychologist jo<strong>in</strong>ed us as arater, us<strong>in</strong>g the system as described below. All three of us tra<strong>in</strong>ed <strong>to</strong>gether for one week, us<strong>in</strong>gsix pro<strong>to</strong>cols for the purposes of illustrat<strong>in</strong>g categories. The second author <strong>to</strong>ok on the roleof arbiter for the <strong>in</strong>dependent rat<strong>in</strong>gs of the other two.

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

Saved successfully!

Ooh no, something went wrong!