As discussed in Chapter One, our approach to studying skills at work assumes that skill needs must be examined in the context of work from the perspective of individuals engaged in that working community. The conceptual underpinnings of this approach stem from sociocultural theories, which argue that the social setting in which cognitive activity takes place is an integral part of that activity, not just the surrounding context for it (Lave 1988, 1991; Resnick, 1991; Rogoff and Charajay, 1995; Scribner, 1984, 1988; Vygotsky, 1978). The knowledge, attitudes, and abilities needed for a particular job can be understood only within a particular working context, from the perspective of individuals in the social setting. The context of the social setting can include other actors, the task at hand, the organization of the work, the physical or symbolic systems that comprise the job, and so on (Hart-Landsberg et al., 1992; Martin and Beach, 1992; Resnick, 1991; Scribner, 1984, 1988).
Within a social setting, work is often situated in communities of practice that share preferred ways of doing a task, establish standards for performance, and shape a newcomer's introduction to the working group (Lave and Wenger, 1991). The kind of information that a community of practice shares may lie outside the scope of officially designed jobs and include information about the group itself, such as informal status hierarchies and hidden communication networks (Levine and Moreland, 1991). Individuals and communities reside in workplaces defined by technology, organization, and activities that can constrain or support use and development of skills (Darrah, 1992).
This conceptual approach suggests a multilevel analysis of skills that takes at least three perspectives into account--individuals performing the work, the communities of practice, and the broader organizational setting.
As outlined in Chapter One, the discussion about workforce skills emphasizes the need to improve generic skills and work-related dispositions over technical, job-specific skills. Employers and many policymakers seem comfortable with the notion of generic skills and competencies and are able to describe what they desire of workers--abilities to solve problems, communicate effectively, work with others, take responsibility, work without supervision, and so on.
At present, there is no standard definition of generic skills, and the roles of general and context-specific knowledge in thinking is still a puzzling issue in social science. Studies of expertise from a cognitive science perspective support findings from sociocultural studies and suggest that general skills do not take the place of domain-specific knowledge, nor do they operate exactly the same way from domain to domain. Rather, specific applications of the general need to configure to the context (see Perkins and Salomon, 1989).
In this study, we began with a conception of generic skills developed in our previous research (Stasz et al., 1990; Stasz et al., 1993). We defined two broad categories of generic skills: basic or enabling skills, such as reading and simple mathematics; and complex reasoning skills, used to solve both formal and everyday problems encountered at school or work. We also included work-related attitudes or dispositions, such as cooperative skills or personal qualities (e.g., responsibility, sociability) that can affect learning and performance on the job. This conceptualization is similar to the widely accepted three-part foundation skills identified by SCANS (1991).
The focus of the analysis reported here is on problem solving, communications, teamwork, and work-related dispositions. We emphasize these areas over others because of the general consensus that these capabilities are lacking in the workforce and that improving them requires public policy action. Several new studies indicate that these general skills are important to employers and workers. The NCEQW survey (1995) shows that employers place high value on general skills: employers rate applicant attitude and communication skills as the two most important factors in hiring new nonsupervisory production workers. Cappelli and Rogowsky (1995) surveyed workers and supervisors about the importance of skills (as defined by SCANS, 1991), their contribution to job performance, and the relationship between new systems of work organization and skill requirements. Employees ranked thinking skills (problem solving) first, followed by "ability to work with others," communication skills (speaking, listening, writing), and "ability to work in teams." Supervisor ratings largely overlapped those of workers.
This emphasis builds on our previous research but also takes some new directions. Our initial studies examined how generic skills and work-related attitudes could be taught to high school students in academic and vocational classrooms. It focused on how teachers can design learning environments that would, for example, help students learn to solve problems, collaborate with others, and take responsibility for their learning (Stasz et al., 1990, 1993). The present research examines these capabilities in the context of work, from the perspective of workers and employers, and draws more on recent literature on sociocultural aspects of work and on noncognitive factors in learning and performance.
Discussions of skill needs in the changing workplace predict a shift in decisionmaking and problem solving from the supervisory level to the shop floor, where workers must cope on the spot with a growing number of unpredictable problems (Berryman and Bailey, 1992). Knowledge and skills are useful to the extent that workers can apply them to real problems and situations that they face at work.
Studies of cognition from a symbolic processing approach examine how problems are symbolically represented and manipulated; they often yield detailed analyses of problem characteristics, such as start states, goal states, constraints, and operators (e.g., Anderson, 1983; Newell and Simon, 1972). These studies are limited for our purposes because they typically study well-defined problems in laboratory or artificial settings rather than in real work contexts. Problems at work tend to be ill-defined, often unrecognized as problems, and have many possible solutions and solution methods (Lave, 1988). The research also typically focuses on individuals, and ignores social aspects of problem solving.
For the purposes of this study, we are interested in characterizing problem solving as it broadly defines work practice in each job, rather than providing a detailed breakdown of knowledge or procedures used to solve particular problems encountered (e.g., repairing a machine or diagnosing a patient's symptom). These broad themes should reflect the collaborative or interactive aspects of the work and the situated nature of problem-solving activities.
Many discussions of new skill requirements in the workforce mention teamwork as a necessary skill. The SCANS (1991, 1992) reports, for example, list "participates as a member of a team" as an interpersonal competency. Others argue that the changing workplace puts a premium on teamwork and the ability of team members to cope with unpredictable problems (Berryman and Bailey, 1992).
As Darrah (1992) points out, teamwork is not a "skill" but a description of how work is organized. What constitutes a team is subject to local definition and thus must be defined in relation to the working context. Thus, the "skill" or "interpersonal competency" needed to participate in a team will depend on the work organization.
The organizational behavior literature provides some relevant definitions. Hackman and Oldham's (1980) classical work, for example, distinguishes between self-managing and co-acting work groups. A self-managing work group is an intact and definable social system, with a defined piece of work and authority to manage the task on its own. Self-managing work teams are also called autonomous work groups, semi-autonomous work groups, self-regulating work teams, or simply work teams (Levine and Moreland, 1991). In co-acting groups, individuals may report to the same supervisor and work close to one another, but they have individually defined tasks (Hackman and Oldham, 1980).
Sociocultural literature has looked at the culture of work groups and the reasons why work is often socially distributed, rather than individual. An important research question concerns the optimal distribution of knowledge in work groups and how the conditions in which the group works and the nature of the actions it must take shape knowledge distribution (e.g., Levine and Moreland, 1991; Hutchins, 1991). Other studies examine the processes by which newcomers become members of work groups or communities of practice (e.g., Levine and Moreland, 1992; Lave and Wenger, 1991).
Our analysis first examines the organization of work and then discusses implications for participation in particular work organizations.
Communications skills are widely cited as among the most important skills needed by today's workers. As mentioned above, a recent national survey of employers identified communication as an important factor in making hiring decisions (NCEQW, 1995). A similar emphasis on communications skills appears in the SCANS framework. Four of the five foundation "basic skills" identified by SCANS are communications skills--the linguistic communications skills of reading, writing, speaking, and listening. Moreover, other components of the SCANS framework, including "personal qualities" such as amicable self-presentation and "workplace competencies" such as interpersonal skills and information utilization, also imply the need for strong communication skills (SCANS, Learning a Living,p. xiv).
Communications is a broad term that can be ambiguous. Unfortunately, this ambiguity is seldom taken into account when discussing communication skills on the job. Few would disagree, for example, that individuals who deal with the public (e.g., salespersons, flight attendants) may need different communications skills than more solitary workers, yet few discussions about communication skills make specific distinctions about how communication needs might vary from job to job.
Here we adopt a traditional analysis of communications that focuses on four axes: audience,or who is communicated with; purpose,why they are communicated with; style,the way in which the communicator presents himself or herself;[7] and mode, the means by which the communication is accomplished.[8] In addition to being familiar to instructors and trainers who may want to draw on our findings, this framework has the advantage of focusing on the situated nature of skills.
As we describe how the employees in our study actually communicated on the job, we will attend to the common distinction between using speech and text, but we will still consider both together, as does the SCANS framework. This is appropriate from several perspectives. In practice, these two kinds of skills are often used for communicating with similar purposes and audiences, and they may be used both in combination with and as a substitute for each other. (That is, communication through writing may often be used to supplement or to substitute for speaking, and vice versa.) Moreover, skills with spoken and with written language share a core of linguistic expertise that makes them amenable to related theoretical description and similar instructional strategies.
In the skills debate, much attention has been paid to the dispositions and attitudes needed to succeed on the job. Some studies suggest that the skills gap identified by employers may be more about attitudes than academic or technical skills (Cappelli, 1992; Cappelli and Ianozzi, 1995). While survey data indicate that employers value "attitudes" (Natriello, 1989; NCEQW, 1995), it is not clear what they mean by this. Some employers may seek workers who have initiative, whereas others might want workers who follow orders.
Other studies of employers' perceptions of noncognitive skills attempt to distinguish different characteristics, such as "personal traits" and "social skills" (Bikson and Law, 1994) or "motivation" and "prosocial behavior" (Cappelli, 1992). Similarly, surveys of workers indicate that workers perceive noncognitive factors (e.g., dedication, resourcefulness) as essential for skilled work (Billett, 1993). Although surveys asking about skills--cognitive or noncognitive--provide evidence of general trends, the answers do not reveal what characteristics workers really have or how they play out in actual work situations (Darrah, 1992).
Theoretical work on the interplay of cognitive and other factors in learning and performance is still in development. Relevant psychological theories examine such factors as volition (Corno, 1993), motivation (Dweck and Leggett, 1988; Dweck and Elliot, 1983), and dispositions (Prawat, 1989; Perkins, Jay, and Tishman, 1993a, b) as individual traits, yet recognize that situational context plays a role in shaping them. If context plays a role and traits are not static, then understanding noncognitive factors is important for public policy. Actions to develop positive dispositions toward work through education and training make sense only if they can be shaped (Cappelli, 1992).
Volition--paying attention to and working toward appropriate goals--is described by adjectives such as conscientiousness, disciplined, self-directed, resourceful, and striving. Volition directs intellectual and emotional energy to achieving goals, especially when the situation calls for it (e.g., if the task is difficult and there are distractions) (Corno, 1993).
Motivations account for the discrepancy between what individuals can do and what they actually do. Research distinguishes between "mastery" or "performance" orientations toward learning (Dweck and Leggett, 1988; Dweck and Elliott, 1983). A mastery approach seeks challenging tasks and persists under failure; it correlates with constructive views of ability, feelings of efficacy and confidence, and efficiency in complex learning situations. Individuals with a performance orientation are more concerned with how they might look to others than with what they might learn; this can influence how they value a task and the effort they put into it.
Dispositions or "habits of mind" are individuals' tendencies to put their capabilities into action and are thought to influence how individuals deal with various situations (Prawat, 1989; Perkins, Jay, and Tishman, 1993a). Dispositions are essential for performance because "unless one has the inclination to use it, ability will lay fallow" (Perkins, Jay, and Tishman, 1993a).
Perkins and his colleagues propose a dispositional theory of good thinking--defined as flexible, insightful, or productive thinking--which complements current views of the kind of thinking skills called for in the changing workplace. Dispositions have three components: inclination (the felt tendency toward behavior X), sensitivity (alertness to X occasions), and ability (ability to follow through with X behavior) (Perkins, Jay, and Tishman, 1993b). This perspective acknowledges that dispositions are grounded in belief systems, values, and attitudes, are culturally based, and are thus acquired through a process of acculturation. By observing and living within a particular culture, individuals gradually start to adopt the behavior and belief systems of the culture (Perkins et al., 1993; Brown, Collins, and Duguid, 1989).
Theoretical constructs like volition, motivation, and disposition are primarily psychological, have focused on learning in academic or laboratory settings, and have barely explored sociocultural influences. Thus they provide only an initial starting point for examining the meaning of dispositions in work settings. Because the concept of dispositions appears to subsume both motivation and volition, we used this term in our analysis.
Because a primary motivation for studying skills at work is the widespread belief that skills are changing, it is important to assess employers' views on skill demands and their responses to changing demands, including the local forces and economic conditions that shape them (Grubb et al., 1992; Darrah, 1992). At the institutional or workplace level, research on organizational change and organizational productivity suggests several themes to examine. Employers can respond to perceived skill demands in several ways, including hiring employees with certain skills, producing skills in their own training programs, or creating incentives that encourage workers to invest in their own skills. In addition to creating new skill capabilities, appropriate work design and human resource policies must be in place to achieve better performance results (Finegold, 1991; Bailey, 1993b). Once an organization adopts a strategy for acquiring the skills it needs, the quality of the implementation process--the decisions and actions that translate ideas and policies into day-to-day practices--is key to reaching successful outcomes (Bikson, Gutek, and Mankin, 1981; Bikson and Eveland, 1991).
To determine employer responses to changing skill needs, we gathered data to answer several questions. How do employers view their skill needs, given the changes their organization or industry faces? What policies do they establish to meet their needs, including selection, hiring, and training practices? Are their decisions linked to the firm's business strategy or organizational mission? Or are they influenced by union rules or other factors? Are their policies effective? Our analysis includes informal and structured on-the-job training in addition to formal training and, in particular, the role of the community of practice in assimilating newcomers and providing training. We also examine other policies that can affect skill development, such as the existence of career paths and unionization.
We conducted the field study in three overlapping phases--site selection, worksite observations and interviews, and analysis. We used a multisite, replicated case study design in which similar sets of criteria were used to select participating firms and individuals within them, and in which common data gathering procedures were employed across the sites. This conceptual approach lends itself to a case study research design that is particularly appropriate for examining and interpreting ongoing processes in real-world contexts, especially when the processes to be studied (work processes or approaches to training or recruitment) are not sharply separable from their contexts and when the variables of interest are likely to outnumber the units of study (Yin, 1994). We next detail our research and analysis methods and define the variables of interest. Table 2.1 illustrates the data sources.
We were interested in sites that were experiencing changes affecting the skill demands of frontline workers in their subbaccalaureate labor force and were willing to cooperate with the research demands of the study (interviewing several levels of management; observing and
Data Sources
| ___________________Source______________________ | |||||
| Classes of Variables | Telephone Interview | In-Person Interview | Cognitive Task Analysis | Observation | Document |
| Worker skills and
dispositions | X | X | X | X | |
| Community of
practice/work organization | X | X | X | X | |
| Organization
characteristics (e.g., size, product, ownership) | X | X | X | ||
| Organization
mission/vision/ management strategy | X | X | X | ||
| Perceptions of
skill needs | X | X | X | X | |
| Responses to skill
needs (e.g., hiring, selection, training, change strategies) | X | X | X | ||
| Worksite connec-
tions to schools | X | X | |||
interacting with frontline workforce on the job). The subbaccalaureate labor force--the labor market for those with less than a baccalaureate degree but at least a high school diploma--was of particular interest for several reasons: It constitutes about three-fifths of the labor market; it has been growing steadily in the last decade, as have the community colleges and technical institutes that prepare occupational students; and it is poorly understood with respect to the relationship between formal schooling and subsequent employment (Grubb et al., 1992). It is also intensely local--subject to cyclical variation in the local economy, which can undermine incentives to accumulate extensive training and experience, and dominated by informal methods for searching for employment and hiring new workers (Grubb and McDonnell, 1991; Grubb et al., 1992). While the subbaccalaureate market encompasses a wide range of jobs across all job categories, our focus is on craft and technical jobs, rather than low-paying jobs often associated with a service economy (e.g., in fast-food restaurants and clerical pools).[9]
To identify sites and occupational categories in the greater Los Angeles area labor market, we sought information and examined data from several sources. We obtained local data on job demands to identify occupations currently in demand and likely to grow in the coming decades. Data from the Economic Development Department on projections of employment in Los Angeles County from 1990-1997 indicated an average growth rate for all occupations of about 5.7 percent. Since growth was expected in virtually every sector of the economy--the exception being manufacturing in the high tech/aerospace segments--we examined projections for the fastest growing occupations to help define the potential sample. Another reason to study "in demand" jobs is that these jobs are likely to have organizational support because of their value to a firm.
We employed a "snowballing" or "chain-sampling" approach to locate potential sites (see Patton, 1980). We contacted firms throughout metropolitan Los Angeles based on leads from a school-business collaborative operating locally and other professional contacts. In addition, we visited several community colleges that engage in a variety of workforce development programs, such as contract training for local employers and retraining for displaced workers. From these sources, we developed a list of potential candidate firms and began a process of first contacting firms to gauge their interest in being part of the study. Altogether, we contacted nine firms over six months.
The selection process involved several steps, including a telephone interview, meetings with management to negotiate study approval, and meetings to identify target departments and workers. Within the targeted department, we identified experienced and less-experienced workers to observe and interview. The final study sample included four sites and seven jobs, as shown in Table 2.2. The sites represent both service and manufacturing sectors, and vary by size, unionization, and other characteristics. Chapter Three describes sites and jobs in more detail.[10]
Site visits occurred over about one week, with some returns to accommodate study participants' schedules. Worksite observations covered three aspects of work common across all jobs: a start-up period, everyday routines, and everyday relations with others. We
Study Sites and Target Jobs
| Site | Target Job |
| Transportation agency (TA) | Construction inspector Survey inspector |
| Health care agency (HA) | Licensed vocational nurse Home health aide |
| Microprocessor manufacturing (MPM) | Test technician Equipment technician |
| Traffic management (TM) | Traffic signal technician |
employed Spradley's (1980) framework for understanding social settings as an organizing guide for fieldworkers' questions, observations, and fieldnotes. The social setting framework includes the following dimensions: space, actors, acts, activities, events, objects, goals, time, and feeling. Trained fieldworkers were asked to include holistic observations about everyday routines, relations, and work start-up in their daily fieldnotes.
Since jobs are socially organized and physically situated in particular ways, we developed study plans for jobs that fit particular conditions. For jobs that convene experienced and new workers in a common space and assign specific tasks, we expected that one observer could accomplish observation in two days. Only the equipment technician job (at MPM) fell within this class. On the first day, the fieldworker had a brief tour and work orientation, followed by observation of everyday routine. For fairly routine assigned tasks, each task was observed on the average of 8-10 times. Then the fieldworker focused on everyday relations involved to accomplish tasks (i.e., the people involved, their positions in the firm, and their relationship to tasks.) On day two, the fieldworker observed tasks and relationships required to initiate work at the beginning of the work day.
When jobs involved teaming, we revised the approach to include an additional half-day to observe the team's routines and relations. The test cell technicians and survey inspectors fell into this class of work.
When jobs required workers to be highly mobile or itinerant, we studied the job over three days, traveling with frontline workers during that time. The travel time permitted informal interviews with frontline workers about tasks and work context. Construction inspector, home health aide, licensed vocational nurse, and traffic signal technician jobs fell within this classification.
For all observations, we sought to observe enough workers to cover the critical tasks assigned to a job and to witness the job performed by at least one newcomer and two to three experienced workers. After the observation phase, fieldworkers conducted formal interviews with frontline workers and department supervisors (discussed below).
Because of our relatively short time at worksites and because the jobs we selected are technical and in some cases require licenses and certifications to perform, we did not undertake participant observation. Rather, we depended on our worker informants to articulate their understanding of the work and tasks.
We conducted two types of formal semi-structured interviews. The first interview covered dimensions of the work context, including the organization, vision, and mission, products and services, staffing and training investments, skill and attitude requirements, productivity improvements, change strategies, and perceptions about workers and their concerns. We probed these aspects of work context across classes of participants, including top managers, high-level contractors and consultants, human resources officers, trainers, middle managers, department supervisors, frontline workers, and key professionals who support department activities.[11] Supervisor and frontline worker interviews also asked about respondents' education and training histories. Each audiotaped interview required one to one and one-half hours. Table 2.3 shows interviews conducted at each worksite.
The second type of semi-structured formal interview inquired about tasks involved in the frontline job. This cognitive task analysis (CTA) was meant to elicit specific task dimensions: production rules, production processes, mental models, and images. Results from the CTA are presented separately (see unpublished manuscript by Black et al., 1995) and includes data from the CTA about worker attitudes and dispositions.
To gather additional data, we attempted to collect the following documents systematically across firms: organization charts, employee
Interviews by Worksite
| Level | TA | MPM | TM | HA |
| Executive | 2 | 2 | 1 | 1 |
| Human resources | 2 | 2 | 1 | |
| Manager | 4 | 1 | 1 | 1 |
| Trainer | 1 | 1 | 7a | 1 |
| Staff professional | 1 | 1 | ||
| Supervisor | 2 | 2 | 1 | 1 |
| Frontline workerb | 4 | 4 | 4 | 3 |
aIncludes one individual interview and two group interviews, representing three agencies overall. bInterview and cognitive task analysis questions for each. | ||||
benefit packages, annual reports, employee evaluation forms and procedures, company newsletters, syllabi for training programs, and procedures used by frontline workers to perform tasks.
Following procedures developed in our earlier research (Stasz et al., 1993), the analytic phase of the field study involved an iterative process of indexing observational data, domain analysis, and generation of themes. We developed index categories that corresponded with the dimensions of work context used to structure the formal interviews, and on other study goals (e.g., worksite connection to schools). As fieldwork commenced, we built on the initial domains by adding categories that emerged at different sites.
Fieldworkers coded interviews and observation notes, which were entered into a computer program for managing text data. To achieve reliability in coding, each fieldworker indexed several sets of fieldnotes and interviews written by other fieldworkers. As a group, the study team compared results, clarified definitions, and identified any missing categories.
Once our preliminary analysis was completed, we returned to each worksite to debrief the study participants on our interpretations of the results and preliminary findings. The debriefing provided a validity check for our findings and gave study participants the opportunity to discuss the implications of the findings for their worksite.
The perceptions and behaviors of the respondents in this study's sample may not generalize to all job incumbents, managers, and worksites in similar occupational areas. Indeed, because of the situated nature of skills and dispositions and because of expected variations in organizational responses to skill development needs and issues, we did not aim for generalizability in the statistical sense. However, the triangulation method and debriefings to the firms did help provide validity checks of respondents' answers to surveys and fieldworkers' observations.
The jobs and worksites were selected because they represent technical work in the subbaccalaureate labor market in organizations facing changing skill demands. Such organizations are thus likely to have grappled with the questions at hand: the relevance and importance of generic skills and dispositions in work, organizational responses to acquiring and developing skills, and employers' connections with education providers as a way to identify potential technical workers and to assist incumbent workers' skill development. The study can provide rich information to other organizations facing changes that affect their skill demands and to educational providers attempting to improve youth preparation for work.
[7]In rhetorical theory, this axis is often referred to as ethos.
[8]On the history and applications of this framework, see James L. Kinneavy, A Theory of Discourse: The Aims of Discourse, New York: W. W. Norton, 1971, p. 18 ff.
[9]Service occupations have remained relatively stable: from 1900-1988 service occupations grew only 4 percent (Barley, 1995). Furthermore, skill requirements for these jobs are less likely to be affected by the changes we have discussed, except to further differentiate them from occupations where technical skills are at a premium.
[10]Because the jobs finally selected do not necessarily match available public data, and because firms do not always back up their perceived job demands with hard data (e.g., the transportation agency), we do not have demand data for all the jobs in the sample. The available data on target occupations with the largest absolute growth show 15.8 percent growth from 1990-1997 for licensed vocational nurses and 39.1 percent growth for home health care workers. Industry data for the same period show 2.6 percent growth for electronic equipment manufacturing (test cell and equipment technicians) and 7.7 percent growth for all construction, with construction trade occupations increasing 6.3 percent (construction and survey inspectors belong to this subgroup). (Source: Occupational Outlook Handbook, 1992-93 Edition, U.S. Department of Labor, Bureau of Labor Statistics, May 1992.)
[11]Although our study design called for participation by union leadership, we were not successful in accomplishing this. Three of the seven frontline jobs we studied were represented by a local union. In two cases, the workers described their union as bargaining units to preserve pay and benefits for a wide range of workers. In the third case (home health), the union leadership did not return our calls requesting interviews, even though they had expressed interest in participating.