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.

A Situational Judgment Test of Supervisory Knowledge in the U.S. Army1<br />

Mary Ann Hanson<br />

Personnel Decisions Research Institutes, Inc.<br />

Walter C. Borman<br />

The University of South Florida and Personnel Decisions Research Institutes, Inc.<br />

A situational judgment test involves presenting respondents with realistic job situations, usually<br />

described in writing, and asking them to respond in a multiple choice format regarding what should be<br />

done in each situation. Situational judgment tests have been developed by other researchers to predict<br />

job performance, especially for management and supervisory positions (e.g., Motowidlo, Carter, & Dunnette,<br />

1989; Tenopyr, 1969).<br />

This paper describes the development, field test, and preliminary construct validation of a situational<br />

judgment test designed to measure supervisory skill for non-commissioned officers (NCOs) in the U.S.<br />

Army. In contrast with most previous research the Situational Judgment Test (SJT) is a criterion measure<br />

of job performance. It is targeted at first line supervisors (ranking E-5), and is intended to evaluate<br />

the effectiveness of their judgments about what to do in difficult supervisory situations. Thus, the SJT is<br />

somewhat like a job knowledge test in the supervisory part of the job. Although no research is available<br />

on the use of situational judgment tests as criterion measures, there is research available on the usefulness<br />

of written simulations - which are similar to situational judgment tests - as measures of professional<br />

knowledge in fields such as law and medicine. Researchers have found that scores on written simulations<br />

differentiate between groups with differing levels of experience or training and are often related to other<br />

measures of professional knowledge or performance (see Smith, 1983 for a review).<br />

Development of the SJT<br />

Method<br />

Development of the SJT involved asking groups of soldiers similar to the target NCOs (i.e., E-4s and<br />

E-5s) to describe a large number of difficult but realistic situations that Army first-line supervisors face<br />

on their jobs. Once a large number of these situations had been generated, a wide variety of possible<br />

actions (Le., response alternatives) for each situation were gathered, and ratings of the effectiveness of<br />

each of these actions were collected from both experts (senior NCOs) and the target group (E-5 NCOs in<br />

beginning supervisory positions). These effectiveness ratings were used to select situations and response<br />

alternatives to be included in the SJT. The effectiveness ratings from the senior NCOs (i.e., experts) were<br />

also the basis for the development of SJT scoring procedures. Each of these steps is described in more<br />

detail below.<br />

Participants in the workshops to develop situations and response alternatives were 52 NCOs from<br />

nine different Army posts. Some were NCOs from the target sample and some supervised target NCOs<br />

(ranks ranging from E-5 to E-6). A variation of the critical incident technique (Flanagan, 1954) was used<br />

to collect situations to be used as the item stems. Workshop participants were asked to write descriptions<br />

of difficult supervisory situations that they or their peers had experienced as first-line supervisors in the<br />

Army. This resulted in a pool of about 300 situations. Response alternatives were primarily generated by<br />

presenting participants in later workshops with the situations that had been collected and asking them to<br />

write, in two or three sentences, what they would do to respond effectively in that situation. This resulted<br />

in about 15 possible responses for each situation, These responses were content analyzed and grouped to<br />

reduce redundancies. The final result was four to ten response alternatives for each situation, with a<br />

mean of about six response alternatives.<br />

1 This research was funded by the U.S. Army Research Institute for the Behavioral and Social Sciences,<br />

Contract No. MDA903-82-C-053 1. All statements expressed in this paper are those of the authors and do not<br />

necessarily reflect the official opinions or policies of the U.S. Army Research Institute or those of the Department<br />

of the Army.<br />

268<br />

._<br />

---r<br />

I

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

Saved successfully!

Ooh no, something went wrong!