A Qualitative Study Investigating the Sensemaking Process of Engineers Performing Windstorm Risk Surveys. (September 2018)

• Record Type: Journal Article
• Authors: Agnisarman, Sruthy; Khasawneh, Amro; Ponathil, Amal; Lopes, Snowil; Madathil, Kapil Chalil
• Is Part Of: Proceedings of the Human Factors and Ergonomics Society ... Annual Meeting. Volume 62:Part 1(2018)
• Page Start: 1776
• Page End: 1776
• Publication Date: 2018-09
• Subjects: Human engineering -- Congresses; 620.8205
• DOI: 10.1177/1541931218621402
• Languages: English
• ISSNs: 1541-9312

Abstract

Insurance loss prevention survey, specifically windstorm survey, is the process of investigating potential damages associated with a building or structure in the event of an extreme weather condition such as a hurricane or tornado. Windstorm inspection is a type of visual risk assessment survey performed to investigate and identify the risk factors that might result in severe damages in the event of extreme weather conditions such as hurricanes or tornados. This survey involves an engineer going to a site and inspecting the property according to a standard protocol. Though they follow the protocol, the inspection process is never straightforward. The engineers have to exercise their judgement and analytical skills while evaluating the property. This process depends highly on the skills and expertise of the engineer. This could result in certain biases and mistakes. This exploratory qualitative research investigated the sensemaking process of insurance risk engineers while performing windstorm surveys. A combination of convenient sampling and maximum variation strategy was used to recruit participants from a specific insurance company. Ten risk engineers with experience ranging from less than one year to 20 years (M = 4.7) were interviewed. Additionally, we identified a subject matter expert rom this insurance company. The subject matter expert performed a mock windstorm loss preventions survey. The mock inspection activity was video recorded. Following the mock inspection, a semi-structured interview protocol was developed to understand the sensemaking process of risk engineers. The recruited participants were interviewed via phone for 90-120 minutes. The interview responses were audio recorded and the recordings were de-identified (used numbers to identify the recordings). The responses were then transcribed by an external agency for analysis purpose. The qualitative data was analyzed following an inductive thematic approach, one of the most common qualitative data analysis methods (Padgett, 2011). The first author led the inductive coding process. The coding process was completed in two steps. In the first step, the researchers identified open codes from the transcripts and then the transcripts were coded individually and the percentage of agreement was calculated (38.4% across all transcripts). However, the coders reached complete consensus after discussion. In the next step the coding schema was updated based on the results of the first step of coding. The researchers coded the transcripts again and the percentage agreement was calculated (54% across all transcripts). The transcripts were then imported to ATLAS.ti, a qualitative data analysis software. The relationships among codes were identified using the querying capability available in the software platform. Upon completing the analysis, the SME was approached to discuss the validity of our findings. The windstorm loss prevention survey is a skill-based inspection process requiring the physical presence of the risk engineers. One of the main challenges the engineers face is the environmental uncertainty. Forecasting risk in an extreme weather condition requires the knowledge of various factors including, but not limited to, the wind speed, building dimensions, building age, roof type, roof material, building occupancy and surface roughness. This is often overwhelming to the engineers leading to certain biases and errors. Moreover, the required information about these factors is not always available to the engineers. This further complicates the risk inspection process. The engineers will have to resort to their internal guidelines and assumptions to make recommendations in such conditions. However, the validity of such recommendations/findings is questionable because it is based on various unknown or uncertain factors. The insights from this study can be used to develop automated technologies that assist risk engineers while performing the inspection task. The primary objective of the design will be to minimize information overload and to reduce the cognitive demand on the risk engineers. The next step will be to develop a cognitive task analysis report to understand the needs of risk engineers in order to develop a system that best caters to their needs.