Manipulations and Measures
Organizational stress research proposes many physical (e.g., heat, noise, fatigue) and nonphysical stressors. Cannon-Bowers and Salas (1998) identified the following list of stressors observed among naval crews, which have been used as realistic stress manipulations in many team contexts: multiple information sources, incomplete or conflicting information, rapidly changing and evolving scenarios, requirements for team coordination, adverse physical conditions (e.g., heat, noise, and sleep deprivation), performance pressure, time pressure, high work/information load, and threat (p. 19). From an organizational viewpoint, stress may consist of simultaneous reactions to external conditions (e.g., time pressure), as well as internal states (e.g., feelings of social inadequacy). For example, Driskell and Salas (1991) operationalized stress as both a threat to the individual’s well being as well as an increase in personal responsibility for team task outcome.
By stimulating an individual with a story, song, or a memory associated with a particular emotional state, it is possible to put an individual into a suggested mood temporarily. The intensity and persistence of these mood-induction procedures varies depending on the target mood desired, yet they have proven to be very successful when properly implemented. According to Westermann, Spies, Stahl, and Hesse (1996), presenting a short film or story is generally the most effective method to manipulate a subject’s mood, either positively or negatively.
These emotion and stress manipulations could be applied uniformly across a team under study. Alternatively, they could be selectively applied to specific team members to examine how a particular mood or stress state affects a team member in a different state (e.g., emotion contagion), or whether team members recognize and respond to another team member in stress.
Emotion and stress can be measured in multiple ways, including measures of physiological responses, as well as self-report survey measures. Physiological techniques provide real-time measurement over the duration of the experimental task, whereas survey measures are limited to snapshots in time, usually before and after the task. Physiological measurements such as heart rate, skin conductance, and skin temperature provide highly sensitive and real-time measures of stress and emotional responses, which are difficult for the subject to fake or consciously control. However, the intrusiveness of physiological measurement equipment needs to be considered carefully with respect to the tasks and setting of the study. For a comprehensive review of psychophysiological measurement techniques, see Cacioppo, Tassinary, and Berntson (2007).
Lazarus’ transactional perspective (Lazarus & Folkman, 1984) focuses on the individual’s perception of stress, which can be assessed via surveys. The short stress state questionnaire (SSSQ), provides a rapid and reliable assessment of three primary dimensions of stress (engagement, distress, and worry) with only 24 questions (Helton & Garland, 2006). Several context-specific stress assessments exist, such as the combat exposure scale (CES) that quantifies combat-related stressors (Keane et al., 1989). In cases where workload is used to manipulate stress, the NASA-TLX (Hart & Staveland, 1988) is a popular workload assessment tool for high-demand settings that provides a self-reported score from 0 to 100 computed from the weighted average of six subscales. However, the relationship between task load and stress is often an indirect one. Shorter, more contextually targeted surveys such as the CES have shown high construct validity within their intended domain when care is taken to ensure that the experimental context is aligned with the assumptions of the survey instrument.
Mood assessments are also self-reported using a range of survey instruments calibrated for specific emotional conditions. These may be used to assess feelings at the present moment (state measurements), or over longer periods of time (trait measurements). One such instrument is the Spielberger State-Trait Anxiety Inventory (STAI) (Spielberger, 1983), a popular and reliable measure of both state and trait anxiety. Respondents indicate their agreement with short phrases like I feel pleasant, I tire quickly, and I feel like crying. The widely used positive and negative affect schedule (PANAS) (Watson, Clark, & Tellegen, 1988) comprises two brief 10-item surveys that provide two independent measures of positive-(PA) and negative affectivity (NA). This instrument collects agreement with terms such as interested, irritable, excited, ashamed, strong, and nervous. Studies on emotion contagion often jointly use self-report and observational methods to capture perceived emotion as well as observable events or behaviors (e.g., facial expressions and posture) that reveal changes in emotional states, as neither alone is sufficient to measure effects that may be subtle and subconscious.
The generally negative connotation of stress makes it difficult to experimentally discriminate stress from negative emotion, as similar effects are sometimes seen in the presence of either one. As an example, manipulations of mood (Gasper & Clore, 2002), stress (Braunstein-Bercovitz, Dimentman-Ashkenazi, & Lubow, 2001), and anxiety (Derryberry & Reed, 1997) all demonstrated similar cognitive consequences. Therefore, a concern in this research is that measures of stress and measures of mood must not confound each other. There is some occasional overlap in the self-report instruments used for stress and mood measurement. In experimental trials, the SSSQ distress factor has appeared to be exclusively measuring negative effect, whereas the engagement factor measures motivation, and the worry factor measures cognitive effects (Helton & Garland, 2006). One could argue that there is little difference between I feel sad on the SSSQ distress factor and I feel upset that appears on both the positive affect scale of the PANAS and short form STAI. However, the SSSQ engagement factor shares two items with the PANAS: I feel alert and I feel active. This suggests that there is the potential for these instruments to confound stress and mood measurements, which recommends techniques such as factor analysis when multiple self-report instruments are used in conjunction with each other.