Predicting Adaptive Performance in Multicultural Teams: A Causal Model

The present study examines a causal model of adaptive performance in multicultural team settings. Dispositional traits are posited to influence adaptive performance through the mediating mechanisms of stress appraisals and self-efficacy. Beyond examining the causal paths associated with predictors, the study includes a commensurate focus on adaptability as an outcome and addresses the measurement issues that surround adaptive performance. Results support the posited causal model across divergent measurement methods for adaptive performance. Implications for the selection and training of multicultural teams, and directions for future research are discussed. Multicultural teams have become the standard in business and governments around the world (Connaughton & Shuffler, 2007; Ilgen & Pulakos, 1999). This cultural diversity has created new work demands and placed greater emphasis on the need to be adaptable. Although adaptability is recognized as a crucial aspect of effective teamwork (Burke, Stagl, Salas, Pierce, & Kendall, 2006; Chen, Thomas & Wallace, 2005), particularly multicultural teamwork, a consistent definition and understanding of adaptability is absent in the literature (for a review see Stokes, 2008). Consequently, researchers often operationalize adaptability in terms amenable to their study, paying less attention to construct validity and diminishing our ability to compare results across studies. For example, adaptability can be construed as a predictor or an outcome. Further Predicting Adaptive Performance in Multicultural Teams: A Causal Model 19 2 RTO-MP-HFM-142 complicating matters, researchers examining adaptability as an outcome have used differing measurement methods (e.g., Griffin & Hesketh, 2003; Kozlowski et al., 2001; LePine, 2005; Pulakos, Schmitt, Dorsey, Arad, Hedge, & Borman., 2002; Zaccaro & Banks, 2004). Considering the divergent approaches adopted, it is important to clarify the prediction of adaptability on both sides, predictor and outcome, and across measurement methods. Therefore, the intent of the present research is to examine a causal model of adaptive performance, combining the two most prominent outcome measures (subjective ratings and objective task scores), in an effort to ensure generalizability of results and offer clear guidance in terms of the selection and training of multicultural teams. 1.0 PREDICTING ADAPTIVE PERFORMANCE There appear to be three general categories associated with predicting adaptive performance: cognitive, dispositional traits such as personality variables, and situational influences. General and specific cognitive abilities (e.g., Allworth & Hesketh, 1999; LePine, Colquitt, & Erez, 2000; Pulakos et al., 2002) and personality factors (e.g., Allworth & Hesketh, 1999; Griffin & Hesketh, 2003; LePine et al., 2000; Pulakos et al., 2002) have been consistently related to adaptive performance. Of the Big Five, the personality factors Openness to Experience and Neuroticism are the best predictors of adaptive performance (e.g., Allworth & Hesketh, 1999; LePine et al., 2000). Conscientiousness and Extraversion have received marginal support as predictors of adaptive performance (e.g., Griffin & Hesketh, 2003; Pulakos et al., 2002). Beyond the traditional, global constructs of cognitive ability and personality factors, unique predictors of adaptive performance have been identified, including change-related self-efficacy and prior experience with adaptive situations (e.g., Allworth & Hesketh, 1999; Griffin & Hesketh, 2003; Pulakos et al., 2002). Griffin and Hesketh (2003) have explored situational influences, finding that job complexity and management support influence adaptive performance. 1.1 An Adaptive Profile of Dispositional Traits Although numerous predictors have been examined, it is not clear why they predict adaptive performance. A NATO research team attempted to determine the profile of an adaptive worker, revealing a three-factor indicator of adaptive performance (Svensson, Lindoff, Anderson, Norlander, & Sutton, 2005). Data collection sites were chosen based on their high need for adaptive performance in workers, although a criterion measure of adaptive performance was not assessed. The intent was to identify latent factors denoting an adaptive worker profile, not predict adaptive performance. This research examined numerous indicators of adaptability, including personality and cognitively-oriented variables. The data reduction and modelling efforts revealed that most indicators loaded on three factors: 1) Instability, 2) Adaptability, and 3) Need for Structure. Instability was composed of Fear of Invalidity and Neuroticism. Adaptability was composed of Emotion Regulation and Cultural Adjustment. Need for Structure was composed of Personal Need for Structure and Need for Cognitive Structure. (Each indicator is explained in more detail below.) Note that in these research efforts, adaptability has been designated and operationalized as a predictor variable (as opposed to an outcome variable). The present research expanded upon research by Svensson et al., (2005) by including a criterion measure of adaptive performance to provide predictive validity to the identified adaptive profile. In the present research adaptability as a predictor and adaptive performance as an outcome are operationalized as separate constructs, each with their own measurement scales. Adaptability is assessed by emotion regulation and cultural adjustment, and adaptive performance is assessed as an outcome measured by subjective performance ratings and objective performance scores. 1.2 Proximal Predictors of Adaptive Performance Identification of predictive traits does not explain how or why such traits affect adaptive performance. To gain such an understanding, proximal predictors must be examined. Task specific stress appraisals, which Predicting Adaptive Performance in Multicultural Teams: A Causal Model RTO-MP-HFM-142 19 3 have yet to be examined in the context of adaptive performance, and adaptive specific self-efficacy are two such proximal predictors. 1.2.1 Stress Appraisals As opposed to viewing the notion of ‘fit’ as determined solely by an observer, individuals themselves evaluate whether their skills and abilities are commensurate with the requirements of the situation. Such evaluations are referred to as stress appraisals (Lazarus & Folkman, 1984). Stress appraisals are comprised of two evaluative components: primary and secondary. Primary appraisals evaluate the personal relevance of a situation in terms of the potential threat it presents in relation to the individual’s goals, values, and beliefs. Secondary appraisals evaluate one’s resources for responding to the demands of the situation. The primary and secondary evaluative components combine to result in a continuum of appraisal outcomes where individuals range from being challenged to threatened (Blascovich & Mendes, 2000). Threat appraisals occur when individuals believe their resources, such as skills and abilities, are disproportionate to the demands of the situation. Challenge appraisals occur when individuals construe their resources as proportionate to or exceeding situational demands. Threat and challenge appraisals have been found to differentially affect performance and physiological responses (Schneider, 2004; Tomaka, Blascovich, Kelsey, & Leitten, 1993) and affective outcomes (Schneider, 2004). Dispositionally adaptive individuals will likely appraise highly demanding and complex situations as a challenge, whereas non-adaptive individuals should appraise the situation as a threat. In accord with the notion of person-environment fit, adaptive individuals will have the appropriate abilities and other characteristics needed to respond to a complex situation. For example, adaptive individuals tend to be low in need for structure, embracing the uncertainty and spontaneous nature of changing situations (Svensson et al., 2005), and they are typically higher in cognitive ability. This low need for structure decreases the potential threat of adaptive situations (primary appraisal), and the higher cognitive ability serves as a coping resource (secondary appraisal). Thus, such individuals may appraise adaptive situations as a challenge, contributing to higher adaptive performance. 1.2.2 Self-Efficacy Although self-efficacy has often been identified as a significant predictor of adaptive performance (Allworth & Hesketh, 1999; Griffin & Hesketh, 2003; Kozlowski et al., 2001; Pulakos et al., 2002), it has yet to be examined as a mediator of effects on adaptive performance. Research has indicated that self-efficacy often serves as a proximal predictor of general performance while other individual attributes tend to be distal, or antecedent to self-efficacy (e.g., Gist & Mitchell, 1992; Mathieu, Martineau, & Tannenbaum, 1993). The present research construed self-efficacy as a mediator of the effects of dispositional traits and stress appraisals on adaptive performance. As presented in Figure 1, dispositional adaptability (i.e., adaptive profile) is expected to influence individuals stress appraisals, which in turn influence self-efficacy, and ultimately adaptive performance. In summary, the hypothesized model presented in Figure 1 clarifies the casual mechanisms through which adaptive performance is likely influenced. It expands upon previous research on dispositional adaptability, and posits the mediating variables (i.e., stress appraisals and self-efficacy) through which adaptive performance is enhanced. Further, by including two prominent measures of adaptive performance (subjective ratings and objective performance scores), we can investigate convergence of results. Hypothesis 1: Stress appraisals and self-efficacy will mediate the relationship between dispositional adaptability and adaptive performance (subjective and objective) in multicultural teams. Predicting Adaptive Performance in Multicultural Teams: A Causal Model 19 4 RTO-MP-HFM-142 2.0 METHOD 2.1 Participants As part of a larger study, a total of 275 people (59% female) from a midwestern university population volunteered to participate in the study in exchange for course credit (200 subjects) or for monetary remuneration (75 subjects) in the amount of $30. The sample was culturally diverse with 64% Caucasian, 16% African American, 17% international students primarily from India, and 3% of other nationalities. The age distribution of the sample ranged from 18 to 49, mean = 21. Due to computer malfunctions, data involving objective adaptive performance were only available for 150 participants. As the focus of the present research is on multicultural teams, analysis of the hypothesized model presented in Figure 1 was limited to mixed culture teams only (N = 25 teams, 125 individuals).