Performance

Introduction

Performance assessment is a class of testing that purports to measure individual and team ability to display complex knowledge and skills. This entry will describe attributes and uses of performance assessments, discuss some technical issues, and point to likely approaches using technology to strengthen their utility. Performance assessment can take many forms, including observed actions (either live or recorded) of the examinee(s), the evaluation of products and related processes created by the respondent(s), or the conduct and outcome of multiple-stage projects. Performance assessment can focus on the output, or end result, or attend to the procedures by which a goal is accomplished. The stimuli eliciting a performance can be many, including simple verbal directions, printed text, computer-supported stimuli, or other agreed-upon signals.

Although performance assessment has sometimes been defined in the negative – as ‘not paper and pencil’ or not multiple-choice test format – it is very possible that a given performance assessment might incorporate many types of response formats. The main idea in performance assessment is to get an integrated sample of student accomplishment, usually one that requires substantial time to produce. Performance assessments, like other tests, can be administered on a stand-alone basis [Page 681]at a particular interval, such as the end of the year, or can be integrated more directly into instruction. A collection of performances, creating a portfolio of a student's performance, can be constructed. These portfolios could either document emerging expertise over a period of time or contain a sample of best pieces to show the highest level of performance of which the student is capable. Common attributes of performance assessments are:

• Extended format
• Applied domain
• Complex task
• Constructed response
• Rated or judged scores

While predominantly used to draw conclusions and make decisions about individuals, performance assessments have been used to measure team processes in school, military, professional, and business settings. One such approach that has received much attention in team research is low-fidelity networked simulations (Bowers, Salas, Prince & Brannick, 1992; Weaver, Bowers, Salas & Cannon-Bowers, 1995). These assessments enable teams of individuals to perform a collaborative task, with individually defined responsibilities, using a networked computer environment. Examples of this type of assessment include evaluations of aircrew coordination (Salas, Bowers & Cannon-Bowers, 1995) and team negotiation practices and processes (O'Neil, Chung & Brown, 1997).

The advantages of this form of team performance assessment are many. Notably, these assessments can be achieved at a relatively low cost, particularly in comparison to full-scale simulations or live administration. In addition, such assessments provide for increased experimental control of independent variables in team research. They provide a means to facilitate the breadth and depth of research in the area of team performance assessment. Moreover, these assessments provide a method for the generation and testing of team performance theory (Weaver et al., 1995).

Finally, team performance assessments provide an effective platform for the investigation of the psychometric properties of various team effectiveness measures. There is converging research to support the reliability and validity of low-fidelity networked simulations as a research tool in the investigation of assessing team performance (Bowers et al., 1992).

...