Monday, October 12, 2015

Building Occupant Surveys – Maximizing Their Effectiveness

Formulating surveys for building occupants as part of post occupancy evaluation (POE), master planning process, retrocommissioning effort, etc., is always a balancing act. You want to ask enough questions to gather the information you need relative to the assessment scope of work, but you also want to avoid making it so long that individuals quit half way through or refuse to even look at it. For some building populations, such as K-12 teachers, who are surveyed to death and typically stretched wafer thin, it’s all the more critical to strategically construct your questions.

Conducting ethnographic type fieldwork as part of the assessment – performing in context interviews and observations – generally allows you to minimize the survey questions asked. Interviews and observations provide opportunities in addition to surveys for “touching” building occupants. Planning for all three allows you to limit the number of survey questions asked compared to engaging occupants with only surveys.

Looking at the survey itself, one strategy is to minimize the questions asking occupants to provide their impressions of the facility as a whole, such as their perception of the facility’s overall impact on their performance, and instead focus on the specific facility aspects you’re really interested in. These might include ratings of thermal comfort, satisfaction with specific lighting control components, or their perception of how thermal comfort impacts their performance.

Our research has shown that questions asking about the facility as a whole, or questions asking about overall facility impacts on individual performance, have a greater degree of correlation to questions asking about an occupant’s level of engagement than when compared to more detailed facility related questions. Once the questions start drilling down further into specific facility elements’/systems’ impacts on individual performance, or ratings of the specific elements/systems in and of themselves, the degree of correlation decreases.

The table below provides the results of a series of Spearman’s Rho tests that were run as part of the POEs of eight K-12 schools to determine the degree of correlation between the primary question acting as a proxy for engagement and the questions asking to rate the various aspects of the school facilities. The number listed in each cell is the resulting correlation coefficient. It can range between -1.0 to 1.0, with a positive correlation indicating the ratings are increasing together while a negative correlation indicates the value of one rating increases while the other rating decreases; 0 indicates no correlation or association. Teachers/staff were surveyed at all eight schools; students were only surveyed at the four middle and high schools.

Looking at the eight set of teacher/staff correlations and the four set of student correlations, in 11 of the 12 sets of data the strongest correlations (cells highlighted in green) occur when comparing the engagement question with questions asking about the facility as a whole or asking about overall facility impacts on individual performance. When rating the whole facility or school, one is trying to mentally average everything together and it’s harder to keep aspects of engagement out of the internal mental equation than compared to when one is just rating individual facility elements.

So we’re better able to determine actual facility performance (as perceived by the teachers/staff and students) from the detailed questions as opposed to the general questions; engagement has more influence on the general question responses (and vice versa).

You can use this as justification for excluding questions more general in nature and instead focusing on the various individual aspects of the facility that are of primary relevance to the scope of work. While the resulting survey is longer than if only general questions are asked, it nevertheless is shorter than when both types of questions are included. It should also result in a more effective use of the occupant’s time, as you should get a better picture of the occupants’ perception of facility performance relative to their needs as opposed to their level of engagement.