09.12.2012 Views

I__. - International Military Testing Association

I__. - International Military Testing Association

I__. - International Military Testing Association

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

clear reference support and acceptable item statistics, thus most of the 117<br />

items were deactivated for valid reasons. It was surprising, however, that<br />

3% of the items in question had "low content validity" cited on the i tern<br />

record card as the reason for deactivation when originally all SMEs had felt<br />

the item content was essential for successful job performance. These results<br />

could be due to administrative error, SME reconsideration of the item con-<br />

tent, or a number of other possible reasons.<br />

The second objective of this research was to determine how SMEs and project<br />

psychologists perceived the value and usefulness of the CVR forms. A three<br />

question survey was administered to 21 project psychologists at USAFOMS. A<br />

similar survey was administered to 151 SMEs upon completion of the rating<br />

forms and selection of items for deactivation. A four-point rating scale was<br />

used for the responses: strongly agree, agree, disagree, and strongly. dis-<br />

agree. The questions and summary of responses are as follows:<br />

(11 When selecting previously used items to be reused, the Content Validity<br />

Rating forms hglped identify those items most essential to successful<br />

performance in the specialty.<br />

49% (74) SMEs answered positively (agree or strongly agree)<br />

52% (111 project psychologists answered positively<br />

(2) The Content Validity Rating forms helped bring out different points<br />

of view for discussion.<br />

56% (85) SMEs answered positively<br />

76% (16) project psychologists answered positively<br />

(3) The Content Validity Rating forms were a valuable tool in selecting<br />

Discussion<br />

items to be reused.<br />

39% (59) SMEs answered positively<br />

43% (9) project psychologists answered positively<br />

Through the analysis previously described, it became apparent that there was<br />

a significant impact of content validity ratings on subsequent identification<br />

of an item as acceptable or deactivated. For instance, with the 25 projects<br />

sampled, there were 52 individual SKTs examined. Of these 52 SKTs, 37 had<br />

higher average CVR values for acceptable items than the average CVR values of<br />

the deactivated items which is what would be expected -- a positive differ-<br />

ence between the two. It is also important to note that 6 of the 52 SKTs are<br />

not applicable in this analysis since in these cases all 100 items were designated<br />

as acceptable and thus, there was no average CVR value for deactivated<br />

items. The 9 remaining SKTs had higher average CVR values for the deactivated<br />

items than the average CVR values for the acceptable items. Of these 9<br />

SKTs, 2 were from projects with insignificant interrater reliabilities. Even<br />

though the expected effect did not hold true for every case, overall, items<br />

higher in content’ validity had a greater chance of being acceptable while<br />

items lower in content validity were more likely to be deactivated. This<br />

shows that on the whole, item content validity does play a role in the SMEs’<br />

evaluation of an item's testworthiness.<br />

The second objective of the research was to determine how SMEs and project<br />

psychologists perceived the usefulness of the CVR forms. It became evident<br />

that there was no universal agreement on the usefulness of the forms. Additionally,<br />

any project psychologist biases, either for or against the use of<br />

the forms, may have influenced how the psychologist administered the forms to<br />

239<br />

. .

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

Saved successfully!

Ooh no, something went wrong!