Judith H. Heerwagen, Ph.D., J.H. Heerwagen and Associates
In the past decade, organizations have begun to look at their buildings not just as a way to house people but also as a way to fulfill strategic objectives (Becker and Steele, 1995; Horgen et al, 1999; Ouye and Bellas, 1999; Grantham, 2000). In part this is due to re-engineering and downsizing of the past two decades. Also, however, chief executive officers (CEOs) are beginning to think of their buildings as ways to achieve strategic goals such as customer integration, decreased time to market, increased innovation, attraction and retention of high-quality workers, and enhanced productivity of work groups.
Traditional post-occupancy evaluation (POE) methods do not provide the type of feedback needed to assess these organizational outcomes. POEs focus on individual-level assessment, most typically on satisfaction, use patterns, and comfort, rather than on organizational- or group-level outcomes associated with core business goals and objectives. Because organizations are increasingly asked to justify all of their major expenses, including facilities, evaluation methods that begin to address these higher-level issues would be of great value. At the present time, there are very few data to show linkages between facility design and business goals.
As a result, decisions are often made on the basis of reducing costs. Current cost-focused strategies include reducing the size of work stations, moving to a universal plan with only a few work station options, eliminating private offices or personally assigned spaces, and telecommuting. Evaluation methods that identify and measure the business value of facilities would be a highly valuable way to expand the current knowledge base and to provide a wider array of outcomes against which to measure facility effectiveness.
Ouye (1998; Ouye and Bellas, 1999) suggests that workspace design and evaluation can become more strategic by adopting the tools of business, specifically the Balanced Scorecard (BSC) approach proposed by Kaplan and Norton (1996). As applied to facilities, the BSC approach pioneered by Ouye and the Workplace Productivity Consortium is valuable not only for evaluation purposes, but also for design because it forces designers to think systematically about the relationship between the workplace and organizational effectiveness. Although the Balanced Scorecard was developed primarily with the private sector in mind, the approach is also applicable to the evaluation of government facilities. A core theme for both the private sector and the government is to provide facilities that are both efficient and effective. As noted in the General Services Administration’s (GSA’s) The Integrated Workplace (GSA, 1999, p.5):
By using the Integrated Workplace as part of your strategic development plan, matching business goals to workplace designs, you can consolidate and reconfigure the spaces where you work while providing people with the tools they need to support the organization’s mission.
Even more important for federal facilities is the strong link between the BSC approach and the requirements for strategic planning and performance assessment initiated by the Government Performance and Results Act of 1993 (GPRA). GPRA was enacted as part of the Clinton administration’s “Reinventing Government” initiative to increase the efficiency of federal agencies and to make them more accountable for achieving program results. GPRA requires federal
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 79
Appendix B A Balanced Scorecard Approach to Post-Occupancy Evaluation: Using the Tools of Business to Evaluate Facilities Judith H. Heerwagen, Ph.D., J.H. Heerwagen and Associates In the past decade, organizations have begun to look base and to provide a wider array of outcomes against at their buildings not just as a way to house people but which to measure facility effectiveness. also as a way to fulfill strategic objectives (Becker and Ouye (1998; Ouye and Bellas, 1999) suggests that Steele, 1995; Horgen et al, 1999; Ouye and Bellas, workspace design and evaluation can become more 1999; Grantham, 2000). In part this is due to re- strategic by adopting the tools of business, specifically engineering and downsizing of the past two decades. the Balanced Scorecard (BSC) approach proposed by Also, however, chief executive officers (CEOs) are Kaplan and Norton (1996). As applied to facilities, the beginning to think of their buildings as ways to achieve BSC approach pioneered by Ouye and the Workplace strategic goals such as customer integration, decreased Productivity Consortium is valuable not only for evalu- time to market, increased innovation, attraction and ation purposes, but also for design because it forces retention of high-quality workers, and enhanced pro- designers to think systematically about the relationship ductivity of work groups. between the workplace and organizational effective- Traditional post-occupancy evaluation (POE) methods ness. Although the Balanced Scorecard was developed do not provide the type of feedback needed to assess primarily with the private sector in mind, the approach these organizational outcomes. POEs focus on individual- is also applicable to the evaluation of government level assessment, most typically on satisfaction, use facilities. A core theme for both the private sector and patterns, and comfort, rather than on organizational- or the government is to provide facilities that are both group-level outcomes associated with core business efficient and effective. As noted in the General Services goals and objectives. Because organizations are Administration’s (GSA’s) The Integrated Workplace increasingly asked to justify all of their major expenses, (GSA, 1999, p.5): including facilities, evaluation methods that begin to By using the Integrated Workplace as part of your strategic develop- address these higher-level issues would be of great ment plan, matching business goals to workplace designs, you can value. At the present time, there are very few data to consolidate and reconfigure the spaces where you work while pro- viding people with the tools they need to support the organization’s show linkages between facility design and business mission. goals. As a result, decisions are often made on the basis of Even more important for federal facilities is the reducing costs. Current cost-focused strategies include strong link between the BSC approach and the require- reducing the size of work stations, moving to a universal ments for strategic planning and performance assess- plan with only a few work station options, eliminating ment initiated by the Government Performance and private offices or personally assigned spaces, and Results Act of 1993 (GPRA). GPRA was enacted as telecommuting. Evaluation methods that identify and part of the Clinton administration’s “Reinventing measure the business value of facilities would be a Government” initiative to increase the efficiency of highly valuable way to expand the current knowledge federal agencies and to make them more accountable for achieving program results. GPRA requires federal 79
OCR for page 79
80 LEARNING FROM OUR BUILDINGS departments and agencies to develop methods for mea- The term “balanced” refers to several factors. First, suring their performance against strategic goals and there is a balance across the four categories to avoid program objectives, an approach that is very consistent overemphasis on financial outcomes. Second, the with the Balanced Scorecard. Because the BSC focuses evaluation includes both quantitative and qualitative on evaluation as a means to enhance overall strategic measures to capture the full value of the design project. performance, the results from the BSC approach would And third, there is a balance in the levels of analysis– provide valuable input to the GPRA performance from individual and group outcomes to higher-level review for federal facilities. At the present time, mea- organizational outcomes. Figure B-1 shows some pos- sures of facility “success” include costs per square foot sible measures to use in each category. All of these of space and square foot per occupant. Such measures measures have logical links to the workplace environ- do not address the strategic issues of concern to GPRA ment, and in some cases there is empirical support also. and the Balanced Scorecard. This appendix draws on the framework developed The Balanced Scorecard assesses four categories of by Ouye (1998; Ouye and Bellas, 1999) but expands it performance: financial, business process, customer to include greater discussion of the links between relations, and human resource development (Kaplan physical space and business outcomes. The approach and Norton call this dimension “learning and growth”). and process described in this chapter also focus on dif- BUSINESS PROCESS OUTCOMES FINANCIAL OUTCOMES • • Process innovation Operating/maintenance costs Work process efficiency • • Costs of building related litigation • Product quality • Resale value of property • Time to market • Rentability of space STAKEHOLDER RELATIONS HUMAN RESOURCE DEVELOPMENT • • Quality of work life Public image and reputation • • Customer satisfaction Personal productivity • • Psychological and social well being Community relationships • Turnover • Cultural change FIGURE B-1 Building evaluation measures using the BSC approach.
OCR for page 79
81 APPENDIX B ferent categories of outcomes. Whereas Ouye’s “work- understanding of the entire building stock, while other place performance” process focuses on strategic measures would be unique to the goals and objectives performance, group performance, and workplace effec- of the particular organization, department, or division tiveness, this appendix links evaluation more directly (Ouye, 1998). to the four dimensions of the Balanced Scorecard. APPLYING THE BSC APPROACH ADVANTAGES OF USING THE BALANCED In applying the BSC approach, the following gen- SCORECARD APPROACH eral steps need to be taken: At the present time, POEs focus on the human resource dimension of the BSC and primarily on quality 1. Identify overall organization mission and specific of work life (which includes environmental satisfac- objectives for each of the four BSC dimensions. tion, comfort, functional effectiveness of space, access 2. Identify how the facility design is expected to help to resources, etc.) This is an important component, but meet each objective. only one indicator of the success of the facility. In fact, 3. Select specific measures for each of the organiza- it is possible for a facility to rate very high on these tional objectives based on links to the workplace characteristics, but to have a negative impact on the design. Set performance goals for each measure. other areas if the building costs considerably more to 4. Conduct evaluation “pre” and “post.” operate and maintain or if the design interferes with 5. Interpret findings in light of the mission and key work processes in some way. This may happen if a objectives. design emphasizes visual openness to enhance com- 6. Identify key lessons learned. munication at the expense of ability to concentrate (Brill and Weidemann, 1999). An example will help illustrate this process. The BSC approach, in contrast, begins by asking these kinds of questions: 1. Identify Mission and Objectives • How can workplace design positively influence Mission: Become a showcase government office of the outcomes that organizations value? future. • How can it reduce costs or increase income? • How can it enhance human resource development? Specific objectives for each BSC dimension: • How can the workplace enhance work processes and reduce time to market? • Financial—reduce the costs of modifying facilities. • How can the work environment enhance customer • Business process—reduce time for delivery of relationships and present a more positive face to products; create more collaborative working rela- the public? tionships within and between units • Stakeholder relationships—upgrade the image of By asking these questions at the beginning of the government workspace, increase customer satis- design project, the BSC approach provides an analyti- faction. cal structure to the entire process, from conceptualiza- • Human resources—improve overall quality of tion through evaluation and finally to “lessons learned.” work life; reduce turnover and absenteeism. For an organization or design firm, these lessons learned become the knowledge base for future design 2. Identify Potential Links to the efforts. Physical Environment One of the trade-offs inherent in using the BSC approach, however, is the difficulty of generalizing to This step of the evaluation process is the most different contexts. Because the evaluation methodology neglected in facility evaluation. It requires conscious is so closely linked to a unit’s own mission and objec- articulation of design hypotheses and assumptions tives, it is difficult to generalize findings to other spaces about expected links between the specific objectives and units. To deal with this difficulty, a core set of and the features of the environment. By making these measures could be used across facilities to gain a better potential links more explicit, it will be easier to inter-
OCR for page 79
82 LEARNING FROM OUR BUILDINGS pret results and to assess differences between spaces sions, presence of food nearby, some degree of separa- that vary on key physical features and attributes. tion from main traffic routes, acoustical separation for Furthermore, by testing specific design hypotheses, the nearby workers who are not participating, ability of BSC approach can be used to test and develop theory. team members to personalize the space, and ability to In contrast, most POE research is theoretically weak maintain information displays, group artifacts, and and does not contribute to either hypothesis testing or work in progress until these items are no longer needed theory development. (Allen, 1977; Sims et al., 1998; Brager et al., 2000). This step requires research on what is known already Once these potential features are identified, the baseline as well as logical speculations. The hypothesized links and new environments can be assessed to identify the form the basis for characterizing the baseline and new extent to which these features are present. environments. It includes physical measures of the A similar process would be carried out for each of ambient environment (thermal, lighting, acoustics, and the objectives. For many of the objectives, there is air quality) and characterization of other key features likely to be little research available. Nonetheless, the and attributes of the environment that are known or assumptions, hypotheses, and predictions should still suspected to influence the outcomes of interest to the be articulated and linkages to the environment should study. The specific features and attributes used in the be logically consistent. As another example, reduced characterization profiles are related to the objectives time to market could be influenced by factors such as and measures. For instance, if one of the objectives is co-location of people working on the task, easy access to reduce absenteeism, then features of the environ- to electronic groupware tools to coordinate work, suf- ment known to influence illness need to be catalogued ficient group space for spontaneous meetings, vertical pre and post. These include materials selection for car- surfaces for continual visual display of work in progress peting and furnishings; ventilation rates; ventilation and schedules, and central storage for materials associ- distribution; thermal conditions; cleaning procedures; ated with the task while it is ongoing. and heating, ventilation, and air conditioning (HVAC) maintenance (Fisk and Rosenfeld, 1997; Heerwagen, 3. Identify Specific Measures for Each Objective 2000). The profiles will enable the evaluation team to understand better what works and what does not and Many different kinds of measures are likely to be why. They can also be incorporated into a database that used in the evaluation process. Nonetheless, each can be integrated across facilities. potential measure should be assessed against general For example, one of the measures cited above for criteria to help decide whether or not it should be improved work process is increased collaboration both included in the evaluation process. These criteria follow: within and between units, especially support for spon- taneous interactions and meetings. A quick review of • Usefulness—the measure addresses the mission, research literature provides the basis for developing the goals, and objectives of the business unit and can potential links to the environment. Specific questions be used in strategic planning. to address in the literature review include: How do • Reliability—the measure produces consistent people use the environment for social interactions? results when applied again. What aspects of the environment encourage different • Validity—the measure is a good indicator of the kinds of interactions? How do groups work, and what outcome of interest (it measures what it purports resources do they typically use? How important are to measure). spontaneous meetings compared to planned meetings, • Efficiency—the overall measurement plan uses and how do they differ from one another? the minimal set of measures needed to do the job Recent research on informal communications and and enables conclusions to be drawn from the interactions in work settings shows specific features entire data set. and attributes of informal spaces that are likely to influ- • Discrimination—the measures will allow small ence the extent to which the spaces are used and their changes to be noticed. degree of usefulness to work groups. These features • Balance—the entire set of measures will include include comfortable seating to encourage lingering, both quantitative and qualitative measures and location in areas adjacent to private workspaces to direct and indirect measures. Quantitative data encourage casual teaming, white boards for discus- can be translated into numbers and used for
OCR for page 79
83 APPENDIX B statistical analyses. Qualitative data, on the other total costs of special equipment or services, and the hand, often include interviews and results from amount of time overall to carry out the change from the focus groups that are more difficult to translate initial request to completion. In order to demonstrate into numeric scales. Nonetheless, such data pro- reduced costs (the objective), pre and post comparisons vide a rich understanding of the context and pro- would focus on the total number of people involved in cesses that make it easier to interpret quantitative a move and the costs of their time; the overall time results. Further, qualitative approaches are often needed to carry out the change; and the total dollar costs used as a means to develop items for surveys and of the move. structured interviews or other data-gathering mechanisms. The second aspect of a balanced Delivery Time for Products. This would require family of measures is direct versus indirect mea- tracking the time to actually produce a product such as sures of performance. Direct measures include a report, starting with the initial assignment and ending outputs such as sales volume. Indirect measures with the date of final delivery. To identify where in the are often correlated with performance or are the overall process the efficiencies occurred, subtasks building blocks of performance rather than actual would also be timed. Data would include a written com- performance output. Examples are frequency of mentary on work process, number of people involved use, occupant satisfaction, or absenteeism. in the task and their responsibilities, and reasons for any unusual delays or work stoppages. The best tasks Setting Performance Goals. The organization needs for such an analysis are reports or other products that to decide prior to the evaluation what degree of are done on a repetitive basis and therefore are likely to improvement it is working toward for each of the iden- be very similar from year to year. If different types of tified measures. Does even the slightest increase in the products are selected pre and post, then any differences expected outcomes matter? Alternatively, should you in delivery times could be due to factors such as task aim for a 10 percent improvement, a 25 percent complexity rather than to increased efficiencies result- improvement? Setting performance guidelines will ing from changes in physical space. help in data interpretation and conclusions. Scientific research uses statistical significance as proof of Facility Image. Data on image would include sub- success. However, statistical analyses may not be as jective assessments through brief surveys completed useful in an applied context. The degree and direction by visitors, customers, job applicants, and staff. Spe- of change over time may be more relevant to organiza- cific questions would depend upon the nature of the tional performance. Very few performance evaluation facility and the type of work. Pre-post analysis would processes, including the Balanced Scorecard, use look for changes in perceptions. statistical analyses to judge whether organizational changes are “working.” Instead, managers look at the Customer Satisfaction. Data on customer satisfaction overall profile of outcomes and make a decision about could include surveys, analysis of unsolicited customer new policies or procedures based on how well the data messages (complaints, concerns, praise), customer match improvement goals (Kaplan and Norton, 1996). retention, and number of new customers. Objective The following examples identify some potential data could include the time needed to complete trans- measures for each of the stated objectives in our hypo- actions with different customers or stakeholders or the thetical example. number of requests for information processed per day. Costs of Modifying Facilities. This would involve Inter- and Intra-unit Collaboration. Communica- identification and calculation of all costs involved in tions and collaboration activities are notoriously diffi- relocating workers or reconfiguring office space, cult to document accurately unless logs are kept of all including costs associated with packing or unpacking, meetings, formal and informal. Furthermore, the value time and costs of facilities staff needed to reconfigure of collaboration is reflected not only in the frequency work stations, time associated with planning the move of the meetings, but also in the outcomes of the inter- and reporting, costs for any special equipment or ser- actions (e.g., new ideas, problems solved more quickly). vices, and lost work time. The data would include total To get as accurate a picture as possible of changes in costs of staff time (number of hours × hourly pay rate), meeting characteristics, multiple methods should be
OCR for page 79
84 LEARNING FROM OUR BUILDINGS used. First, motion sensors can be used to gather data of turnover rates should then be assessed against what on frequency of use for particular spaces. The sensors is considered a desirable level. would have to contain counters or other data-processing technology (e.g., a sensor that would measure how long 4. Conduct Evaluation Pre and Post the lights were on) that would ignore short-duration changes (e.g., someone walking into the room briefly Key issues in conducting the pre-post process and then leaving). The occupancy data would have to include gaining cooperation from managers and staff be supplemented either with behavioral observations who will be the study subjects, timing of the evalua- or surveys and interviews that gathered information on tion, and use of control or comparison groups. number and character of meetings attended within the past week (or some other limited time period to enhance Gaining Cooperation. The evaluation process will the potential for accurate recall). The survey-interview fail if the occupants are reluctant to participate or if process would also gather data on the attendees, the there are insufficient staff to help with the organiza- nature of the meeting (spontaneous versus planned; tional data gathering for some of the measures (such as focused on a specific problem, brainstorming, task turnover rates). Occupants are much more likely to con- integration, information exchange, and so forth), and tinue to be engaged in the process if they are involved the perceived value of the meeting (specific outcomes, in helping design the measurement plan and if they see usefulness, etc.). a benefit from participation. Having support from high- Data analysis would compare the number of meet- level organizational leaders is also critical because it ings, the participants (number from within the unit, signals the importance of the project. The facility occu- number from other units), the purpose, the outcome, pants also need to be informed of how the data will be and the perceived value. If the facility had an impact on used and they need to be assured that their own input collaboration, one would expect to find a wider range will be kept confidential. of participants, more meetings for problem-solving and brainstorming versus simple information exchange, Control Groups. Because so many other factors can more spontaneous meetings, and a higher perceived influence the outcomes being studied, it is difficult to value. know whether performance changes are due to the workplace itself or to other factors that may change Quality of Work Life. POEs traditionally focus on simultaneously. This is especially true when the design quality-of-work-life issues such as comfort, environ- is part of an organizational change effort, which is often mental satisfaction, work experiences and perceptions, the case. Confounding factors may be internal to the sense of place, and sense of belonging. Many design organization (changes in policies or markets), or they firms and research organizations have examples of may be external to the organization but nonetheless can surveys that are used in a pre-post analysis. affect business performance (such as economic condi- tions). The best way to avoid problems of interpreting Turnover. There is a great interest in retaining the success of a design is to use control groups along workers due to the high costs of turnover, in terms of with pre-post studies. An appropriate control group both the financial costs associated with hiring someone would be a business unit in the same building that does new and the knowledge costs that result from losing a similar kind of work but is not going through a work- valuable skills and knowledge when a worker leaves. place change. The control group should be as similar to Turnover is usually calculated as a rate of workers who the design change group as possible. voluntarily leave an organization divided by the total The control group is studied at the same time as the number of workers for the same time period. Turnover group experiencing the design changes, with both does not include retirements, dismissals, deaths, or loss groups studied during the pre and post design phases. of staff due to disabling illness. Some degree of staff Although the control group would not experience the turnover is important because it introduces new ideas design change, it would get the same surveys or other and new skills into an organization. Thus, for evalua- measures at the same time. If the design has an impact tion purposes, the organization needs to decide what independent of organizational issues, then the pre-post degree of turnover is desirable. Pre-post comparisons responses for those in the design change condition
OCR for page 79
85 APPENDIX B should show greater differences across time than for parison groups. It will also be important to use a high those in the control group. degree of logic in interpreting results, to look for con- sistency across facilities that share similar features, and Timing of Measurement. The pre-measurement to look at relationships between measures. should be done at least two to four months prior to the For instance, if absenteeism is of interest, then move into a new facility in order to avoid issues and absenteeism rates should be associated with other out- problems associated with the move itself. Ideally, the comes, such as symptom expression or low levels of existing facility should be evaluated before work motivation, both of which could lead to taking days off begins on the new facility, although this is very seldom due to illness or lack of desire to come to work. Assess- done due to the need to assemble a research team and ing patterns of absenteeism will further aid in interpre- develop a measurement plan. The post-measures tation of results. Absence associated with motivational should be done six to nine months after project comple- issues is likely to have a different pattern of days off tion to enable workers to adapt to the new setting. The than absenteeism due to illness. Because illnesses delay will help to diminish the “settling-in” phase when happen randomly and often last for more than one day, problems may be most obvious and the workplace absenteeism due to illness should be clustered and needs to be fine-tuned. It will also reduce the impact of randomly distributed over the days of the week. Moti- a “halo” effect associated with being in a new or reno- vational absenteeism, on the other hand, is more likely vated space. to occur on particular days of the week (especially Monday or Friday) and would be more likely to occur just one day at a time, not for several days. Another 5. Interpreting Results way to assess absenteeism is to look at its opposite— When data analysis is complete, the project recon- attendance. Attendance can be assessed by determin- siders the design hypotheses and asks: Do the data ing the percentage of workers with perfect attendance support the hypotheses? Do the results meet the perfor- or the percentage who used less than the allowed mance goals? number of sick days in the year prior to and the year There are very few scientific research studies that after the move to the new facility. In addition to look- show complete support for all hypotheses and predic- ing at relationships between measured outcomes, there tions. Thus, we would not expect to find perfect align- should also be a logical connection to the physical envi- ment in design evaluation. Where misalignments occur, ronment profiles. Absenteeism and illness symptoms it is important to try to understand why this happened. should be associated with factors such as poor indoor The design and evaluation teams will have a natural air quality and low maintenance of HVAC systems. tendency to focus on the positive and ignore the results Another problem for interpreting the results on that do not turn out the way they expected. However, it facility evaluation is that redesign often goes hand in is often more valuable to understand why things went hand with organizational change. Thus, positive (or wrong for several reasons. First, you do not want to negative) results could be due to organizational issues repeat the mistakes. Second, negative results often and not to the physical environment. This is where con- force a rethinking of basic assumptions and a search trol groups become very valuable. If the organizational for better links between the environment and the change is widespread, then similar units should also behavioral outcomes. experience the effects of these changes. Thus, differ- A problem with all facility evaluations, regardless ences between the control group and the group in the of specific methodologies, is the issue of causation. If a new space are more likely to be related to the environ- new facility is found to have positive outcomes, can ment. Again, the use of logical thinking is also impor- these be attributed unequivocally to the physical envi- tant. When organizational change occurs, some aspects ronment and not to other factors? The answer is clearly of behavior are more likely to be influenced than others. no. Causation can be determined only by carefully For instance, if staff perceive the change very nega- designed experiments that vary only one component at tively, then motivationally influenced absenteeism may a time. Since this is unrealistic in field settings, the go up in both the new and the control spaces. Other causation issue will always be present. The best that outcomes, such as the costs of “churn,” are less likely can be done is to minimize other explanations to the to be affected by organizational change policies. degree possible through the use of control and com- Greater assurance of a true connection between the
OCR for page 79
86 LEARNING FROM OUR BUILDINGS physical features of the space and the measured out- project, the specific measures used to test the hypothesis, comes can be gained also by using a geographical pre and post photos of the space, brief summaries of information system (GIS) approach to data analysis. the data, key lessons learned, connections to other Outcomes on various subjective measures can be studies, connections to the full research findings pre plotted on floor plans to gain a greater understanding and post, and recommendations for future designs. The of the spatial distribution of responses. For instance, a presentation of lessons learned should be as visual as GIS format used by the author to assess environmental possible to allow for maximum understanding and satisfaction and comfort in an office building in Cali- retention (Norman, 1993). Graphs, photos, and key fornia clearly showed that problems associated with words and concepts are much more likely to be useful distractions occurred primarily in particular locations, than long verbal explanations that can be accessed if regardless of whether people were in private offices or desired through links to full reports. cubicles. Similarly, thermal and air quality discomfort In addition, simple methods to display overall results tended to cluster more in some areas than others. At the would aid interpretation and lessons learned. For present time, most post-occupancy data analysis uses instance, results could be visualized using color-coded human characteristics as the primary unit of differen- icons to provide an easy visual interpretation: green tiation (e.g., different job categories, gender, age), with could be used to show strong support for the hypothesis comparisons in responses across job categories or age. and meeting or exceeding performance goals; yellow By supplementing the demographic data with geo- could be used to show minimal support or no change; graphical analysis, the evaluation will provide a more and red could be used for measures that did not support complete picture of the facility. A similar process, the hypotheses or showed negative change. called spatial modeling, has been suggested by Aronoff and Kaplan (1995). Both the GIS and the spatial model- SUMMARY AND CONCLUSIONS ing approaches allow for analysis of the variability and distribution of responses in a spatial format. This appendix has described an approach to post- occupancy evaluation that is more closely linked to business and workplace strategies than existing meth- 7. Identifying Lessons Learned odologies. Although the Balanced Scorecard approach An issue with lessons learned is: Where should the does not present any new measurement techniques or knowledge reside—in people’s heads or in the envi- breakthrough methodologies, it does provide a process ronment? Should the lessons be internalized and for more effectively linking facilities to an organiza- become part of one’s tacit operating knowledge, or tion’s overall mission and goals. An advantage to using should the lessons be located for anyone to access—in the BSC approach for federal facilities is its close rela- reports, databases, and so forth? Both should happen. tionship to the comprehensive performance assessment If people are going to work with the knowledge gained, required by GPRA. Traditional POEs provide an they need to incorporate it into their everyday ways of important source of input, but measures tend to be thinking and working. Internalization takes time and focused on a limited range of topics and on the occu- continued work with the knowledge and issues (Norman, pants’ perspective, rather than on the broader, strategic 1993; Stewart, 1999). Once internalized, knowledge is focus of the BSC. part of a person’s intellectual capital and leaves the For large real estate portfolios, such as those in fed- organization when the person does. This is why knowl- eral departments, the determinant of facility success is edge also needs to be made explicit so it can become an not only how well the overall building stock performs organizational asset, not just a personal asset (Stewart, with respect to core POE measures used across facili- 1999). Seminars and presentations on the results of ties, but also how well each design fits its particular facility evaluations—with both positive and negative context and how well it meets the business objectives results highlighted—should be an ongoing practice. of the unit. The Balanced Scorecard was developed Since a major purpose of evaluation is to apply the specifically for the purpose of providing data to assess knowledge gained to future projects, simple databases overall performance and to identify areas that need that could be accessed by key words would be espe- attention. cially valuable to future designs. The database should include the design hypotheses and assumptions for each
OCR for page 79
87 APPENDIX B ABOUT THE AUTHOR Aronoff, S., and Kaplan, A. (1995). Total Workplace Performance: Rethinking the Office Environment. Ottawa: WDL Publications. Judith H. Heerwagen is an environmental psycholo- Becker, F., and Steele, F. (1995). Workplace by Design: Mapping the High Performance Workscape. San Francisco, Calif.: Jossey-Bass. gist whose research and writing have focused on the Brager, G., Heerwagen, J., Bauman, F., Huizenga, C., Powell, K., Ruland, human factors of sustainable design and workplace A. and Ring, E. (2000). Team Spaces and Collaboration: Links to the ecology. Dr. Heerwagen currently has her own con- Environment. Berkeley: University of California, Center for the Built Environment. sulting and research practice in Seattle. Prior to start- Brill, M., and Weidemann, S. (1999). Workshop presented at the ing her own business, Dr. Heerwagen was a principal Alt.Office99 Conference, San Francisco, Calif. December. with Space, LLC, a strategic planning and design firm, Fisk, W., and Rosenfeld, A.H. (1997). Estimates of improved productivity and health from better indoor environments. Indoor Air 7:158-172. and a senior scientist at the Pacific Northwest National General Services Administration. (1999). The Integrated Workplace: A Laboratory. At Space Dr. Heerwagen was codirector of Comprehensive Approach to Developing Workspace. Washington, D.C.: research and helped develop metrics for the Workplace Office of Governmentwide Policy and Office of Real Property. Performance Diagnostic Tool. At the Pacific Northwest Grantham, C. (2000). The Future of Work: The Promise of the New Digital Work Society. New York: McGraw-Hill, Commerce Net Press. National Laboratory she was responsible for developing Heerwagen, J. (2000). Green buildings, organizational success and occu- research methodologies to assess the human and orga- pant productivity. Building Research and Information 28(5/6):353-367. nizational impacts of building design. Dr. Heerwagen Horgen, T.H., Joroff, M.L., Porter, W.L., and Schon, D.A. (1999). Excel- lence by Design: Transforming Workplace and Work Practice. New has been an invited participant at a number of national York: Wiley. meetings focused on workplace productivity. She was Kaplan, R.S., and Norton, D.P. (1996). The Balanced Scorecard. Boston: on the research faculty at the University of Washington Harvard Business School Press. Norman, D. (1993). Things That Make Us Smart: Defending Human in the College of Architecture and Urban Planning and Attributes in the Age of the Machine. Reading, Mass.: Addison-Wesley. at the College of Nursing. Dr. Heerwagen is a member Ouye, J.O. (1998). Measuring workplace performance: Or, yes, Virginia, of the American Psychological Association. She holds you can measure workplace performance. Paper presented at the AIA a bachelor of science in communications from the Uni- Conference on Highly Effective Facilities, Cincinnati, Ohio, March 12-14. versity of Illinois, Champaign-Urbana, and a Ph.D. in Ouye, J.O., and Bellas, J. (1999). The Competitive Workplace. Tokyo: psychology from the University of Washington. Kokuyo (fully translated in English and Japanese). Sims, W.R., Joroff, M., and Becker, F. (1998). Teamspace Strategies: Creating and Managing Environments to Support High Performance REFERENCES Teamwork. Atlanta: IDRC Foundation. Stewart, T.A. (1999). Intellectual Capital. New York: Doubleday. Allen, T. (1977). Managing the Flow of Technology. Cambridge, Mass.: MIT Press.