|
|
Publications
Santos, F.P., Jacinto, C., Silva, S.A., Fialho, T. and Guedes Soares, C. (2014), “Intercoder reliability of accidents at work for four variables of the ESAW methodology” (In Portuguese), International Symposium on Occupational Safety and Hygiene (SHO 2014), 13-14 February, Guimaraes, Portugal, pp. 386-389
This paper presents a study on the intercoder reliability of the coding process of work accidents within the Portuguese official system, which follows the European Statistics on Accidents at Work (ESAW) methodology of Eurostat (2001). It comprises two main objectives: 1) to assess the current situation, i.e., quantify the coding reliability performed by a group of “experts” of the Cabinet for Strategy and Planning (GEP) of the Portuguese Ministry of Solidarity and Social Security, and 2) to assess the “loss in reliability” (impact) when the coding is carried out by a group of Health and Safety Senior Technicians (H&SST), defined herein as the “non-experts” group. The study followed a structured methodology and besides using those two groups of coders it involved the following: the assessment of four nominal ESAW variables; the use of Krippendoff’s á through the macro KALPHA developed to estimate this reliability coefficient in SPSS software; 100 accident notifications (coding units) of year 2007, randomly selected from the GEP’s historical archive. The results showed that all variables held acceptable levels of reliability only when coded by the GEP group, and two of them even were classified as reliable data; by contrast, there was a considerable “loss in reliability” when the coding was transferred from the GEP group to the non-experts (H&SST) one. This proved that special training and expertise knowledge on the ESAW coding scheme has impact on the coders’ performance. Given that Portuguese authorities are developing a new electronic form for notifying accidents at work, it is expected that, in the near future, the coding task be accomplished online by non-experts, i.e., by the H&S professionals who currently fill in the notifications forms for company's insurers. Therefore, this study is important to outline a nationwide training plan in companies and for monitoring the quality of accident data (reliability).
If you did not manage to obtain a copy of this paper: Request a copy of this article
For information about all CENTEC publications you can download: Download the Complete List of CENTEC Publications
|