| Sumario: | Objectives. Self-reports boast extensive industrial implementations attributable to simplicity and higher applicability than other ergonomic assessment methods. Especially, small-scale industries and non-routine work analyses, amidst financial and technical constraints, rely largely on self-reports. However, self-reports exhibit low reliability. Existing reliability enhancement research is confined within the feedback phase, overlooking potential variabilities inherent in additional elements of the assessment process. This article aims to enhance reliability with an exhaustive variability exploration throughout the self-report assessment framework, developing targeted control measures. Methods. Self-report ergonomic assessment of assembly–disassembly tasks was conducted with 80 participants leveraging a 5-point unipolar Likert scale. Variability sources were identified through statistical hypothesis testing involving Kruskal–Wallis, Wilcoxon signed-rank and paired t tests. Modes of variability occurrence were elucidated through post-hoc analyses including Dunn's test and qualitative analysis. Targeted variability control measures were derived through qualitative causal inferences. Results. Four major reliability influencing factors and their modes of occurrence were identified. Response scales designed incorporating variability control measures corresponding to each of the identified factors showed an increase in Krippendorff's α reliability coefficients (from 0.67 < α < 0.87 to α > 0.98). Conclusion. The proposed approach holistically mitigates variability sources throughout the assessment process, facilitating reliable results without detriment to simplicity of implementation.
|
|---|