Positive Exemplar Exposure

In adults, implicit racial bias has been linked to prejudiced and discriminatory behavior. However, implicit racial biases emerge well before adulthood; as young as age six, children have already internalized the racial attitudes of their culture. Thus, it is critical for researchers to understand how to change implicit racial bias early in development, before its negative effects compound across the lifespan. The following chapter highlights one potential method of bias reduction in childhood: exposure to positive exemplars. As this method is both scalable and child-friendly, it has the potential to be used with young children on a broader cultural level. This chapter details child-friendly methods for measuring bias change and provides two examples of studies that have successfully employed positive exemplar exposure to reduce children’s implicit racial bias. I conclude the chapter with recommendations for future use of this intervention cross-culturally, as well as broader cultural applications.

Assessing Point Forecast Bias Across Multiple Time Series: Measures and Visual Tools

Measuring bias is important as it helps identify flaws in quantitative forecasting methods or judgmental forecasts. It can, therefore, potentially help improve forecasts. Despite this, bias tends to be under-represented in the literature: many studies focus solely on measuring accuracy. Methods for assessing bias in single series are relatively well-known and well-researched, but for datasets containing thousands of observations for multiple series, the methodology for measuring and reporting bias is less obvious. We compare alternative approaches against a number of criteria when rolling-origin point forecasts are available for different forecasting methods and for multiple horizons over multiple series. We focus on relatively simple, yet interpretable and easy-to-implement metrics and visualization tools that are likely to be applicable in practice. To study the statistical properties of alternative measures we use theoretical concepts and simulation experiments based on artificial data with predetermined features. We describe the difference between mean and median bias, describe the connection between metrics for accuracy and bias, provide suitable bias measures depending on the loss function used to optimise forecasts, and suggest which measures for accuracy should be used to accompany bias indicators. We propose several new measures and provide our recommendations on how to evaluate forecast bias across multiple series.

Assessing Point Forecast Bias Across Multiple Time Series: Measures and Visual Tools

Measuring bias is important as it helps identify flaws in quantitative forecasting methods or judgmental forecasts. It can, therefore, potentially help improve forecasts. Despite this, bias tends to be under-represented in the literature: many studies focus solely on measuring accuracy. Methods for assessing bias in single series are relatively well-known and well-researched, but for datasets containing thousands of observations for multiple series, the methodology for measuring and reporting bias is less obvious. We compare alternative approaches against a number of criteria when rolling-origin point forecasts are available for different forecasting methods and for multiple horizons over multiple series. We focus on relatively simple, yet interpretable and easy-to-implement metrics and visualization tools that are likely to be applicable in practice. To study the statistical properties of alternative measures we use theoretical concepts and simulation experiments based on artificial data with predetermined features. We describe the difference between mean and median bias, describe the connection between metrics for accuracy and bias, provide suitable bias measures depending on the loss function used to optimise forecasts, and suggest which measures for accuracy should be used to accompany bias indicators. We propose several new measures and provide our recommendations on how to evaluate forecast bias across multiple series.

Measuring and Controlling Bias for Some Bayesian Inferences and the Relation to Frequentist Criteria

A common concern with Bayesian methodology in scientific contexts is that inferences can be heavily influenced by subjective biases. As presented here, there are two types of bias for some quantity of interest: bias against and bias in favor. Based upon the principle of evidence, it is shown how to measure and control these biases for both hypothesis assessment and estimation problems. Optimality results are established for the principle of evidence as the basis of the approach to these problems. A close relationship is established between measuring bias in Bayesian inferences and frequentist properties that hold for any proper prior. This leads to a possible resolution to an apparent conflict between these approaches to statistical reasoning. Frequentism is seen as establishing figures of merit for a statistical study, while Bayes determines the inferences based upon statistical evidence.

Quantifying Social Biases in NLP: A Generalization and Empirical Comparison of Extrinsic Fairness Metrics

Measuring bias is key for better understanding and addressing unfairness in NLP/ML models. This is often done via fairness metrics, which quantify the differences in a model’s behaviour across a range of demographic groups. In this work, we shed more light on the differences and similarities between the fairness metrics used in NLP. First, we unify a broad range of existing metrics under three generalized fairness metrics, revealing the connections between them. Next, we carry out an extensive empirical comparison of existing metrics and demonstrate that the observed differences in bias measurement can be systematically explained via differences in parameter choices for our generalized metrics.

Anthropometric measures in predicting myocardial infarction risk. Do we know what we are measuring? Bias in research occurred worldwide when the true unhealthy body composition was not well compared

Obesity is a major risk factor for myocardial infarction (MI). However, how to measure whole-risk with simple baseline characteristics? Anthropometrically, association for metrics does not equate causation on incident MI. Besides, association may present effects of bias rather than the true putative risk may be responsible for all or much of the epidemiological causality, and a different body composition between groups with similar baseline confounding variables may provide false-positives in outcomes. Thus, in evaluating whole-risk by anthropometry all metrics are not enterely valid at all times, and the lack of balance between measurements will be particularly prone to the generation of false-positive results. The purpose of this article is to critically review key findings for association biases from different studies. From the INTERHEART, waist-to-hip ratio (WHR) has been deemed as an excellent MI risk predictor, and other results have conferred to WHR a greater excess risk in women than in men. Nevertheless, a novel insight have revealed that WHR-associated risk would appear biased if metrics to compare had no balance and equivalence relation. Baseline characteristics of thousands of MI cases are well known, but anthropometry, mathematics and epidemiology have taught us something, and comment on it below. To date, no method was used to address biases for balancing the distribution of measurements between groups to be compared. Thus, WHR and waist circumference as being mathematical fraction and unit of whole-length, repectivelly, presented association biases when true unhealthy body composition was not well compared by group and by sex. It occurred for unbalancing both measurements and unhealthy body composition when comparing strength of association for metrics. Only waist-to-height ratio as being measure directly associated to a volume of risk yields no biases and should be the metric used to compare the body composition of risk, either by age or by sex.