Point-wise confidence interval estimation by neural networks: A comparative study based on automotive engine calibration.

In developing neural network techniques for real world applications it is still very rare to see estimates of confidence placed on the neural network predictions. This is a major deficiency, especially in safety-critical systems. In this paper we explore three distinct methods of producing point-wise confidence intervals using neural networks. We compare and contrast Bayesian, Gaussian Process and Predictive error bars evaluated on real data. The problem domain is concerned with the calibration of a real automotive engine management system for both air-fuel ratio determination and on-line ignition timing. This problem requires real-time control and is a good candidate for exploring the use of confidence predictions due to its safety-critical nature.

Range of confidence scale and consumer knowledge calibration

The study examined the effect of range of a confidence scale on consumer knowledge calibration, specifically whether a restricted range scale (25%- 100%) leads to difference in calibration compared to a full range scale (0%-100%), for multiple-choice questions. A quasi-experimental study using student participants (N = 434) was employed. Data were collected from two samples; in the first sample (N = 167) a full range confidence scale was used, and in the second sample (N = 267) a restricted range scale was used. No differences were found between the two scales on knowledge calibration. Results from studies of knowledge calibration employing restricted range and full range confidence scales are thus comparable. © Psychological Reports 2014.

Parental confidence in managing food allergy:development of the Food Allergy Self-Efficacy Scale for parents (FASE-P)

Background: Food allergy is often a life-long condition that requires constant vigilance in order to prevent accidental exposure and avoid potentially life-threatening symptoms. Parents’ confidence in managing their child’s food allergy may relate to the poor quality of life anxiety and worry reported by parents of food allergic children. Objective: The aim of the current study was to develop and validate the first scale to measure parental confidence (self-efficacy) in managing food allergy in their child. Methods: The Food Allergy Self-Efficacy Scale for Parents (FASE-P) was developed through interviews with 53 parents, consultation of the literature and experts in the area. The FASE-P was then completed by 434 parents of food allergic children from a general population sample in addition to the General Self-Efficacy Scale (GSES), the Food Allergy Quality of Life Parental Burden Scale (FAQL-PB), the General Health Questionnaire (GHQ12) and the Food Allergy Impact Measure (FAIM). A total of 250 parents completed the re-test of the FASE-P. Results: Factor and reliability analysis resulted in a 21 item scale with 5 sub-scales. The overall scale and sub-scales has good to excellent internal consistency (α’s of 0.63-0.89) and the scale is stable over time. There were low to moderate significant correlations with the GSES, FAIM and GHQ12 and strong correlations with the FAQL-PB, with better parental confidence relating to better general self-efficacy, better quality of life and better mental health in the parent. Poorer self-efficacy was related to egg and milk allergy; self-efficacy was not related to severity of allergy. Conclusions and clinical relevance: The FASE-P is a reliable and valid scale for use with parents from a general population. Its application within clinical settings could aid provision of advice and improve targeted interventions by identifying areas where parents have less confidence in managing their child’s food allergy.