Adaptability in Coalition Teamwork Leader and Team Adaptability in Multi-National Coalitions (LTAMC)

This chapter presents the initial analyses of the data from the LTAMC experiments on organization, culture, communication, trust, language and group processes. Both experimental and survey data was collected. It was shown that the VSM/Hofstede measurement of culture gave more variance between national cultures than the GCS/Matsumoto measurement, and was found the most useful for analyses. When comparing the participating nations’ scores on the cultural dimensions, some of them were rather surprising compared to previous research on this; on the Individualism/Collectivism dimension the current scores as well as the rank order were quite different from what Hofstede found in his original IBM study. In the original study, the USA and the Netherlands scored the most individualistic, while Sweden and Norway scored the most collectivistic. In the current study this was the other way around; Sweden and Norway scored the most individualistic and the USA and the Netherlands the most collectivistic. Reasons suggested were sample differences (military versus civilian/IBM), and/or that one or more questions in the survey may be less suited for military samples. There was found higher levels of trust within the national than within the international teams. This result supports existing research, and suggests a need to invest more time and training in order to build trust within culturally diversified teams in NATO coalitions. The more culturally different the team composition, the higher will the need be to focus on this. Amount of communication varied greatly between the groups. Some of the variation was linked to English language capability; the better in English, the more they communicated. This underlines the importance of language proficiency for personnel in NATO work. Flat organizational structure was shown to be related to flexibility. In this game environment, respondents also seemed to have a more positive view of the team organization when they experienced it as more hierarchic and centralized. Suggested reasons for this finding: simple task situation (has previously been linked to making a centralized organization advantageous), game communication and information management systems (e.g. it took more time to communicate and share information with all in a decentralized manner than in a centralized manner if the team was dispersed), and military and cultural sample preference (people tend to prefer and work most efficiently in systems to which they are accustomed). These findings implicates the importance of having the organization fit both the task and the personnel (both in terms of their cultural makeup and what they are trained for), as well as the information management and collaborative systems being aligned to support the organizational structure and processes. Subjects from low Pd cultures tended to rate the team organization more positively than those from high Pd cultures. There was no relationship found between rating the organization and Uncertainty avoidance (Ua). ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT RTO-TR-HFM-138 3 29 3.2.1 Background [LTAMC SABRE Study] In the following, there is a brief description of background, theory, method, participants, materials and procedure; please be referred to chapter 2 and 3.1 for a more comprehensive presentation of these subject areas. The LTAMC project members collected experimental data in the period April 2006 – May 2007. Crosscultural cooperation was the main focus in the experiments. This chapter presents the initial descriptive analyses of the data from these experiments on organization, culture, communication, trust, language and group processes. A total of 56 experiments were conducted and 5 nations contributed. The experiments were conducted in Norway (16+6), Sweden (9+6), Bulgaria (8+6), the Netherlands (8+8) and the USA (7+6). 48 experiments (the national experiments) were conducted using local networks while 8 experiments (the international experiments) were conducted on the internet. Many researchers in the field of cross-cultural psychology have shown how nations vary across various aspects of culture, e.g. [54][55][56][49][61]. The most well-established and researched theory of crosscultural differences that we have are Hofstede’s dimensions of culturally based values – especially as concerns organizational and work related issues. His work has been corroborated and expanded through numerous other studies, also with military samples (e.g., [63] [64] [65] [66] [67]). The culturally based value-dimensions developed by Hofstede [55] are called: Individualism/Collectivism, Power distance, Uncertainty avoidance, Masculinity/Femininity and Long-term/Short-term Orientation. Individualism/Collectivism (I/C) refers to a cultural difference in group as opposed to individual orientation. Group orientation is linked to tight ties between people, whereas individual orientation is linked to loose ties between people. High score indicate individualism (I). Power distance (Pd) is defined as a difference in the actual and experienced distribution of power between people in a hierarchy. High scores indicate high Pd. Uncertainty avoidance (Ua) refers to a difference in need for predictability and rule orientation. High scores indicate high Ua. Masculinity/Femininity (M/F) refers to whether the culture values toughness, assertiveness and a focus on material success as opposed to modesty, concern for others and a focus on the quality of life. High scores indicate masculinity (M). Long-term/ Short-term Orientation (Lt/St) refers to a difference in focus; the present versus distant future. The former indicates a propensity for action whereas the latter indicates a propensity for planning. High scores indicate long-term orientation (Lt). But there are also newer developments in the field, like MeridianGlobal and Matsumoto’s six dimensions of culture [47], which nevertheless bare some resemblance to Hofstede’s dimensions. They have built on existing research in the field and developed the dimensions: Independence/Interdependence, Egalitarian/Status, Risk/Restraint, Direct/Indirect Communication, Task/Relationship, and Short-term/Long-term Orientation. Two of the many areas culture affects are organizational and team behavior. As also indicated in 3.1, organization structure as well as culture, have an impact on team processes. For instance, how is organizational and team structure affecting processes as well as being interpreted differently in different 31 The Norwegian focus has mainly been on organizational and social factors, team cooperation and culture. This reflects the focus in the FFI projects NBD in operations and Collaboration in Networks, within which the Norwegian LTAMC work has been conducted. 32 The first number in the parenthesis indicate the number of national experiments in each country, with a nationally homogenous subject composition, while the latter number indicate the number of international experiments, with a nationally heterogeneous subject composition, that each country participated in. In Norway, there were conducted a total of 16 Norwegian national experiments, with samples from two different Norwegian military populations, plus 6 international experiments. ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT 3 30 RTO-TR-HFM-138 cultures? Does English language ability affect cooperative processes? And is the level of trust the same in national and international groups? We wanted to explore such themes in the experiments conducted. For more theoretical aspects, please be referred to chapter 2 and 3.1. 3.2.2 Method 3.2.2.1 Participants There were a total number of 224 subjects participating in the experiments, 4 in each experiment. The experiments were conducted with participants from 5 nations; 48 of the groups were culturally homogenous (i.e. same nationality) while 8 of them were culturally heterogeneous (i.e. different nationalities). All participants were military officers, 52 male and 4 female. In Norway, there were conducted two series of national experiments, with samples from two different Norwegian military populations (one from a graduate and one from an undergraduate military college), which are also intended to function as a control when making comparisons across national cultures. 3.2.2.2 Materials In order to study the different issues related to cooperation in multinational teams in a controlled environment, the LTAMC chose to employ a computer based role play (NeverWinter Nights, NwN), adapted for our research purposes (Situation Authorable Behavior Research Environment, SABRE), as the main instrument, the method being both explorative and innovative. In addition, there were 6 computerized surveys distributed before, under and after the experimental game session. 4 pre-game computerized surveys were administered on background, personality (NEO-PI) and culture (GCS and VSM-94). During the experiment session, the subjects were at three different times in the game interrupted by a prompt to answer questions measuring their Situation awareness (SA). 2 surveys were administered after the experimental game session was over, called debriefing questionnaires 1 and 2. These were constructed for the purpose of the LTAMC experiments. Debriefing questionnaire 2 is an organization questionnaire containing 36 organization related questions constructed on the basis of the organizational survey employed in the field studies (AW04, BG05, presented in chapter 3.1). The questions covered topics such as organization (hierarchy, centralization, leader behavior), decision-making, work-load, trust, information-sharing, communication and language. 3.2.2.3 Procedure In the experimental set-up, the 4 subjects in each experiment were randomly assigned to a role in the game. In the game scenario, the subjects were given an assignment as a team and could cooperate to solve the 33 Including demographics, language and computer use/knowledge. 34 Short version of Costa & Mc Crae’s NEO Personality Inventory [44], [45]. 35 MeridianGlobal and Matsumoto’s Globesmart Commander Survey [47]. 36 Hofstede’s Value Survey Module [55]. The VSM survey does not measure culture at an individual level – it has been validated to measure culture at the level of countries or regions. The GCS, on the other hand, is intended to be used on an individual level, but has not yet been validated. 37 Debriefing questionnaire 1 is a general survey asking 51 questions from the game. 38 Questions were reviewed by peers and revised on the basis of feed-back from the subjects in two pilot studies conducted in Norway prior to the first experiments. ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT RTO-TR-HFM-138 3 31 assignment. They communicated through “chat”. The national experiments there were conducted within local networks while the international experiments were done over the internet. The experiments were timed to one hour, but subjects were in for 5-6,5 hours on the experiment day. This included: getting an introductory brief, game learning and planning sessions, completing surveys, and receiving a debrief in the end. (For more details, see chapter 2). 3.3 ASSESSMENT #3: CULTURE AND TEAMWORK IN AN OPERATIONAL ENVIRONMENT Dr. Joan Johnston, Research Psychologist, Naval Air Training Systems Division, U.S.A. 3.3.1 Background [AW05 Study] The original purpose of the AW05 research initiative was to extend development of: (1) the model of organizational and interpersonal factors (Chapter 4), (2) the model of cultural adaptability (Chapter 4), and (3) training cultural adaptability (Chapter 5). Allied Warrior 05 (AW05) was a SHAPE/Allied Command Operations (ACO) Computer Assisted Exercise designed to certify the NATO Response Force VI (NRF-6) capability, under the command of Joint Force Command Lisbon (JFC Lisbon), for the six month period starting January 2006. The planned experimental design was based on the whole headquarters concept with the intent of studying J2 and J3 staffs. Similar to AW04, AW05 data collection was expected to take place with the same participants in two waves; at two different times and two different sites. Objectives 1 and 2 were addressed in a first wave of data collection at the Joint Force Command Headquarters in Lisbon before the actual AW05 exercise. The first wave results are reported in Chapter 4. The second wave of data collection with the same participants was expected to take place at the AW05 exercise (CPX) at Montijo, Portugal Air Base. The plan was to collect anecdotal data from the wave 1 participants for: 1) content development of the GlobeSmart® Commander culture and teamwork training tool, the primary deliverable for the LTAMC project, and 2) assessing the impact of three GlobeSmart® Commander training modules on multicultural teamwork at the operational level. However, two external events resulted in a completely revised data collection plan at the CPX. First, a natural disaster in Pakistan resulted in deployment for many of the AW05 DJTF before the CPX, who therefore could not participate in the second wave. In addition, inadequate internet bandwidth at the CPX resulted in the inability of researchers to present the computer-based GlobeSmart® training. Therefore, the first objective was revised to provide a presentation to a new group of participants on the GlobeSmart training program and obtain their observations and feedback on the potential for using GlobeSmart with DJTF staff. The second objective also had to be revised. The extensive national diversity of NATO officers in this study enabled us to explore, through semi-structured interviews, propositions based on a model for understanding cultural diversity in cognition and teamwork developed by Sutton and Pierce (see Chapters 2 and 5 for further discussion). Past research has found that normative scores on the GlobeSmart cultural dimensions strongly distinguish country of origin or nationality. Past research has also established that effective communication strategies for teamwork (e.g., information exchange, initiative, and support), are considered important for successful team performance outcomes. Therefore, we proposed that this widely diverse sample of experienced NATO officers would be expected to have extensive knowledge about what makes multi-cultural teams effective. Therefore, based on the Sutton and Pierce model's expectations for effective teamwork, we proposed the NATO officers would tend to describe important teamwork behaviours with: ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT 3 32 RTO-TR-HFM-138 a) An egalitarian perspective (e.g. being "self-directed," using "flexible roles," and "challenging opinions of others in power") vice a "status" perspective (e.g., team members follow and enforce rules/guidelines, expect to use appropriate behaviours for specific roles, and respect status and position power) (Proposition 1); b) A "low need for certainty" perspective (e.g., demonstrate quick results, value flexibility and initiative, and speed is valued more than thoroughness) vice a "high need for certainty (e.g., spend time on background research, establish proper processes and systems, and take time before making a change) (Proposition 2); and c) an "interdependent" perspective (e.g., focus more on cooperation and group goals, use group decision making styles, and reward and recognize the group) vice an "independent" perspective (e.g., take more individual initiative, use individual decision making styles, and reward/recognize individuals) (Proposition 3). 3.3.2 Method The study was exploratory in kind and used both qualitative and quantitative methods of analysis. A semistructured interview strategy was designed to identify trends in NATO officer attitudes about teamwork in the context of the GlobeSmart culture dimensions. 3.3.2.1 Participants Twenty-two NATO officers volunteered to answer a series of interview questions on the three dimensions of teamwork. These officers represented a good diversity in NATO countries: Canada, Denmark, France, Germany, Greece, Netherlands, Norway, Portugal, Romania, Spain, Turkey, and the United Kingdom. 3.3.2.2 Materials Interview questions were created to elicit officer attitudes toward three dimensions of teamwork that have been found to be effective team member behaviors. The interview questions were designed to take an unobtrusive approach to gaining insight out the potential influence of the cultural dimensions. Therefore, we attempted to avoid obtaining opinions about cultural biases or effective teamwork. The questions were designed to obtain responses about how officers did their job in the context of teamwork dimensions that were known to be effective. The officers were asked the following questions about how they expected to work with other people, both at their NATO Headquarters and during the Allied Warrior exercise: Information Sharing • How do you share information with others? • Do you modify your approach depending on the nationality of the person you are dealing with? Initiative/Leadership • How do you provide guidance? For example, directing someone to take action or instruct them on how to perform a task. How does that change depending on nationality of the individual? • How do you communicate your priorities for others? Does it change depending on nationality? Backup/Support • How do you bring an error to a team member's attention and see that it is corrected? • How does this procedure change depending on nationality of the team members? • What do you do when you see that a team member is overloaded or is having difficulty performing a task? ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT RTO-TR-HFM-138 3 33 3.3.2.3 Procedure The Montijo command post was established in an air craft hanger with dividers separating work groups and teams as designated by the commander. LTAMC experimentation was conducted in an eight foot by four foot area enclosed on three sides by dividers. Privacy was not an option. The area contained two tables, two unclassified computers, and several chairs. Observation of the command post was unobtrusive and occurred constantly throughout the day and some evenings. On the afternoon of 28 Nov, we identified 30 individuals for training based on their role in the CPX, specifically individuals playing FOps, JMC, SYNK, COM, and J 2. Survey feedback was available to all participants who completed the surveys at Joint Force Command Lisbon, but not all persons who completed the surveys were members of the DJTF and not all members of the DJTF had completed the surveys. That afternoon and the following morning, the experiment coordinator went to each of the identified players individually and let them select an available 30-minute training time from 0830 to 1500 on 28 or 29 Nov, These training dates had been agreed upon at the FPC in July 05. Two researchers were available to train on each of those dates, and except time off for lunch, DJTF staff had 40 time slots from which to choose. It was the intent of the researchers to administer the GlobeSmart Commander prototype cultural awareness training to select DJTF staff. Unfortunately, there was insufficient unclassified bandwidth at the Montijo site to support the on-line training venue. Therefore, researchers had to make do by providing 30 minutes of verbal cultural/teamwork training on an individual basis to experiment participants. Feedback and reactions to the presentation were obtained from the participants. It is likely that receiving training in this manner frustrated the staff. All training was completed as planned by 1500 on 30 Nov. Planned follow-up interviews were conducted on 4-5 Dec. Four researchers conducted the approx. 1-hour interviews between the hours of 1200 and 1700 on the 4th and 0900 to 1500 on the 5th, giving targeted interviewees (i.e., those individuals who had received cultural/teamwork training) 44 time slots from which to choose. Again, the experiment coordinator went to each of the interviewees individually to schedule their interviews. Each interview was conducted with the interviewer asking the questions for each of the teamwork dimensions. In addition participants were asked whether or not they modified their approach based upon the nationality of the person they were speaking to. Specifically, they were asked if there were any workarounds in which the participants modified their approach based upon the nationality and language ability of the person they were dealing with. Interviewers hand wrote interviewee responses to each question. Each interview was begun with the questions about information exchange, but in keeping with the natural flow of the interview, the progress of the remaining interview questions depended on the direction the interviewee's answers took. In the end, however, each interviewer addressed all of the questions. Two raters, blind to interviewee nationality, independently reviewed transcripts of each of the participant interviews. They each rated the interview responses to the teamwork dimensions using a six-item Likert-type scale for each cultural dimension (egalitarian/status; risk/restraint; independent/interdependent; direct/indirect; task/relationship; and short-term/long-term. The scales were provided by Aperian Global © with their permission and have been used extensively in prior work within the GlobeSmart Commander training they built. Each scale item allowed for a single rating of 1, 2, 3, 4, or 5. For example, a score of 1 or 2 on the egalitarian/status scale indicated the interviewee's response had phrases that supported a somewhat (2) or strong (1) egalitarian perspective. Whereas, a score of 4 or 5 on the same scale indicated the interviewee's response was had phrases that supported a somewhat (4) or strong (5) status perspective. A "3" indicated the interview response had equal numbers of phrases that supported a "balanced" egalitarian and status perspective. A rating of “Not Applicable” (“N/A”) was applied if no response was given or when a response did not include sufficient detail to allow for a rating. For responses in which there was initial disagreement ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT 3 34 RTO-TR-HFM-138 between raters, a discussion led to a consensus rating. The first type of rater disagreement occurred when a number of responses did not provide very much information and were subsequently rated by one rater (but not the other) as “N/A.” Upon discussion, it was often the case that one rater was able to provide an adequate description of the observed behaviour to the second rater, and consequently a consensus rating was made. The second type of rater disagreement occurred when one rater suggested, for example, a rating of “3” and the other suggested a rating of “4.” The raters then discussed the interview response in detail and came to a consensus on a single rating. 3.3.3 Results The purpose of the analyses in this section was exploratory and not subjected to statistical analysis due to the small dataset. Therefore, propositions, rather than hypotheses were tested to identify if trends in the results would justify further analyses. Frequency analyses were converted to percentages for the cultural ratings on each of the three teamwork dimensions. For each culture dimension, the rating frequencies were reduced to three categories. For example, a score of 1 or 2 was converted to a percentage representing an "egalitarian orientation," a score of 3 was converted to a percentage representing a balanced egalitarian/status orientation, and a score of 4 or 5 was converted to a percentage representing a "status orientation." Results of cultural dimension ratings for each teamwork dimension are presented as percentages in Tables 6, 7, and 8 the tables are formatted with the percent rating results listed in the shaded columns under the headers "orientation." Percentages are listed in the columns next to each of the six pairs of cultural dimensions. The "balanced" orientation percentages are listed between each dimension. Results will be described as a tendency toward an orientation, but significance tests have not been made. Therefore, the findings are descriptive rather than confirmatory. The response rate is listed in the last column and was calculated as the sum of the response rates for the teamwork category. The total number of interviews included in the analysis is listed in parentheses. A high response rate might indicate a cultural dimension has some relevance to the officer's perception of a particular teamwork dimension. Conversely, if a low response rate was obtained from the ratings, then it might indicate the cultural dimension had little relevance to the teamwork dimension. Table 6 presents the percent of the leadership/initiative interview responses rated for degree of orientation toward the six culture dimension. The response rate was very good for the majority of the ratings except it was fair (57%) for the short term/long term dimension. Proposition 1 was partially supported. The ratings tended toward an egalitarian (32%) and interdependent (27%) orientation, but there was also a tendency toward the balanced orientation on both, 23% and 32%, respectively. In contrast, the responses were almost evenly split on the risk (27%)/restraint (32%) orientation, with only a few "balanced" (9%) responses. For the remaining cultural dimensions, the responses tended to have a direct (67%), task (42%), and short-term (33%)/balanced (19%) orientation. ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT RTO-TR-HFM-138 3 35 Table 6: Percent of Leadership/Initiative interview responses rated for degree of orientation on the six culture dimensions. Orientation Balanced Orientation Response Rate (N) Egalitarian 31.82 22.73 18.18 Status 73 (22) Risk 27.27 9.09 31.82 Restraint 68 (22) Independent 13.64 31.82 27.27 Interdependent 73 (22) Direct 66.67 4.76 9.52 Indirect 81 (21) Task 42.86 14.28 23.81 Relationship 81 (21) Short-Term 33.33 19.05 4.76 Long-Term 57 (21) Table 7 presents the percent of the backup/support interview responses rated for degree of orientation toward the six culture dimensions. The response rate ranged from fair (57%) to good (73%) across the dimensions. Proposition 1 was supported. The responses tended toward an egalitarian (45%), restraint 38%), and interdependent (41%) orientation. For the remaining dimensions, the responses tended to have a direct (45%) and relationship (33%) and short-term (29%) orientation. Table 7: Percent of Backup/Support interview responses rated for degree of orientation toward the six culture dimensions. Orientation Balanced Orientation Response Rate (N) Egalitarian 45.45 18.18 9.09 Status 73 (22) Risk 14.28 4.76 38.09 Restraint 57 (21) Independent 9.09 13.64 40.91 Interdependent 64 (22) Direct 45.45 4.54 22.73 Indirect 72 (22) Task 19.05 14.28 33.33 Relationship 67 (21) Short-Term 28.57 9.52 19.05 Long-Term 57 (21) Table 8 presents the percent of the information sharing interview responses rated for degree of orientation toward the six culture dimensions. The response rate was fair (57%) on two dimensions, but very low on the remaining dimensions. Therefore, any conclusions drawn from these results are very questionable. Proposition 1 was partially supported. The responses tended toward an egalitarian (23%) and interdependent (23%) orientation, but were evenly split on risk (24%)/restraint (24%) orientation. For the remaining dimensions, the responses tended to have a direct (23%) and short-term (24%) orientation. The responses for the taskrelationship dimension were mainly balanced (24%), with the remaining responses split between task (19%) and relationship orientation (14%). ASSESSMENT OF ORGANIZATIONAL AND INTERPERSONAL FACTORS IN A SIMULATED MISSION AND IN AN OPERATIONAL ENVIRONMENT 3 36 RTO-TR-HFM-138 Table 8: Percent of Information Sharing interview responses rated for degree of orientation toward the six culture dimensions. Orientation Balanced Orientation Response Rate (N) Egalitarian 22.73 13.64 9.09 Status 45 (22) Risk 23.81 9.52 23.81 Restraint 57 (21) Independent 4.54 0 22.73 Interdependent 27 (22) Direct 22.73 0 4.54 Indirect 27 (21) Task 19.05 23.81 14.28 Relationship 57 (21) Short-Term 23.81 14.28 9.52 Long-Term 48 (21) 3.3.3.1 Language English language proficiency level is deemed to affect team processes. It was measured through the average of 4 questions in the background questionnaire. An internal consistency test of these questions showed very good reliability; Cronbach's Alpha = .895. See table 9 for details. Table 9: Language ability scale items: Item reliability. Questions Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted Is English your native language? ,754 ,886 How would you rate your ability to read and write in English? ,729 ,897 How often do you feel that you get more easily stressed when working in an English-speaking environment rather than in your native language? ,860 ,831 How often do you become more reserved about presenting your point of view in English than in your native language? ,842 ,836