09.12.2012 Views

I__. - International Military Testing Association

I__. - International Military Testing Association

I__. - International Military Testing Association

SHOW MORE
SHOW LESS

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

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

the behavioral statements for guiding interview ratings, and orient the<br />

interviewer to use the questions, the worksheet, and the interview rating<br />

form.<br />

&lot Testinq the Structured Interview<br />

The interview protocol and rating form were pilot tested in two wave:<br />

with a total of 31 officer interviewers and 93 applicants in seven<br />

different locations. Means and standard deviations of the ratings provid.<br />

data on their spread and overall distribution.<br />

As part of the pilot testing, an interrater reliability study was<br />

conducted. One way to assess the quality of data emerging from the ne:q<br />

structured ,interview is to determine how closely two interviewers agree i<br />

their independent ratings of the same interviewees. Thus, we initiated a<br />

interrater reliability study with 10 officers interviewing a total of 24<br />

applicants. All interviewers were trained to do the structured intervie,.+<br />

and to use the rating form. Each applicant was interviewed by two office<br />

recruiters.<br />

After each interview session, the interviewer completed the rating<br />

form and provided a copy to the researcher. Officers interviewing the s:<br />

applicant never discussed that applicant before making their ratings, so<br />

the interview judgments were generated totally independently. Intraclaorcorrelations<br />

were computed for each dimension separately and for the sum<br />

the dimension ratings. This provides an estimate of the across-intervie-consistency<br />

of ratings made using the new interview protocol and rating<br />

form.<br />

RESULTS<br />

Means and standard deviations for Wave 2 interview ratings, gathered<br />

after major revisions of the interview protocol (after Wave 1 pilot<br />

testing), appear in Table 1. For these 59 applicants, means are close TV<br />

4.0 (on a 5-point scale) and standard deviations are approximately 1.0.<br />

Further, these means compare favorably with data for the previous inter:;;<br />

(M=4.68 in 1985). Of course, this is not a very fair comparison because<br />

ratings on the new format were gathered for research, whereas the 4.68 r:t;<br />

is based on operational ratings. Nonetheless, applicant ratings using th<br />

new protocol appear to provide reasonable spread for the interview rati.n!:l.<br />

of these typically high quality NROTC applicants.<br />

Table 1 also contains the interrater reliability coefficients for<br />

ratings made of the 24 applicants evaluated by two independent<br />

interviewers. These are very high reliabilities, with considerable<br />

agreement shown on the part of the interviewers.<br />

495

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

Saved successfully!

Ooh no, something went wrong!