|
on Big Data |
| By: | Jana Kim Gutt (Paderborn University); Kirsten Thommes (Paderborn University) |
| Abstract: | Employees are frequently evaluated on a numerical scale by their supervisors. These numerical assessments inform far-reaching managerial decisions, such as promotions, training opportunities, and dismissals. Yet, they often lack accuracy, are subject to supervisor bias, and do not provide justification for the ratings. In this paper, we address the limitations of numerical ratings by letting individuals provide spoken assessments of others and use a Random Forest algorithm to convert the spoken assessments into numbers (algorithmic ratings). Through this method, we combine the advantages of qualitative feedback and numerical ratings while potentially mitigating common biases. Our results suggest that the algorithmic ratings more accurately reflect the distribution of competencies (as measured by psychometric tests) than assigned numerical ratings (assigned ratings). The algorithmic ratings are considerably more nuanced and less skewed compared to the assigned ratings. Our findings highlight the potential of combining spoken comments with a machine learning model to enhance the accuracy of employee assessments in organizational settings. |
| Keywords: | performance appraisals, rating prediction, machine learning, spoken comments |
| JEL: | J24 M51 D91 |
| Date: | 2025–02 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:dispap:130 |
| By: | Jana Kim Gutt (Paderborn University); Kirsten Thommes (Paderborn University); Miro Mehic (Paderborn University) |
| Abstract: | Performance appraisals are subject to recent debates with one common denominator: most discussions point to their lack of accuracy. In theory, performance appraisals aim to reflect an employee’s performance over a certain period of time. However, recent research shows that appraisals fall short in reaching this goal. Although many studies acknowledge the benefits of performance comments over ratings on a scale, research has paid little attention to the potential of performance comments to achieve higher accuracy in performance evaluations. To approach this issue, we conducted a laboratory experiment and collected objective performance data as well as numerical and verbal performance appraisals. In particular, we compile numerical ratings, written comments, and spoken comments on performance from independent evaluators. To make the numbers (assigned ratings) and the comments comparable, we applied a Random Forest algorithm to transfer the comments into numerical ratings (algorithmic ratings). By analyzing each rating (assigned and algorithmic) in relation to the performance, we find evidence that spoken comments reflect performance differences most accurately within a team. Our results offer important insights into how performance appraisals may be approached to reflect objective performance more accurately. |
| Keywords: | performance appraisal, rating accuracy, rating format, performance appraisal comment, rating scale |
| JEL: | J24 M51 D91 |
| Date: | 2025–02 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:dispap:132 |
| By: | Bokan, Nikola; Lenza, Michele; Araujo, Douglas; Comazzi, Fabio Alberto |
| Abstract: | Word embeddings are vectors of real numbers associated with words, designed to capture semantic and syntactic similarity between the words in a corpus of text. We estimate the word embeddings of the European Central Bank’s introductory statements at monetary policy press conferences by using a simple natural language processing model (Word2Vec), only based on the information and model parameters available as of each press conference. We show that a measure based on such embeddings contributes to improve core inflation forecasts multiple quarters ahead. Other common textual analysis techniques, such as dictionary-based metrics or sentiment metrics do not obtain the same results. The information contained in the embeddings remains valuable for out-of-sample forecasting even after controlling for the central bank inflation forecasts, which are an important input for the introductory statements. JEL Classification: E31, E37, E58 |
| Keywords: | central bank texts, embeddings, forecasting, inflation |
| Date: | 2025–04 |
| URL: | https://d.repec.org/n?u=RePEc:ecb:ecbwps:20253047 |
| By: | Sophie M. Behr; Till Köveker; Merve Kücük |
| Abstract: | Russia’s invasion of Ukraine in 2022 was accompanied by a significant reduction of its gas supply to Europe, causing sharp energy price surges and prompting governments to respond with public appeals and programs aimed at reducing consumption. This paper investigates the effects of price increases and non-monetary factors, such as public appeals and saving programs, on residential energy savings during the crisis. Using a unique building-level dataset on residential energy consumption and prices in Germany, we identify price-driven savings and energy price elasticities with a DiD-PSM approach. By comparing buildings that faced price increases to buildings with constant prices, we can isolate price-driven savings from contemporaneous non-monetary effects. Our findings reveal that while increased prices led to moderate short-run energy savings, the majority of observed savings were driven by non-monetary factors. Consequently, we identify a relatively low short-run price elasticity of residential heat energy demand of -0.07. Going beyond average effect estimation, we use two machine learning methods to calculate building-level price-driven and non-price-driven savings, then analyzing their variation with socio-economic characteristics using census data. |
| Keywords: | Energy crisis, Energy policy, Causal inference, Double machine learning |
| JEL: | Q41 Q48 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:diw:diwwpp:dp2112 |
| By: | Jana Kim Gutt (Paderborn University) |
| Abstract: | When conducting performance appraisals, evaluators largely depend on their subjective perception. This makes performance appraisals particularly vulnerable to biases, especially along the lines of warmth and competence – the two primary dimensions of social judgment. Warmth reflects the ability to build and maintain social relationships, while competence refers to achieving goals and completing tasks. Although both dimensions have been extensively studied in the past, there is limited understanding of how they influence observation-based assessments, particularly in relation to the evaluation format and the impact of the rater and ratee gender. To address this gap, the study employs a laboratory experiment to investigate how warmth- and competence-related behaviors in a task setting translate into performance appraisals, namely numerical ratings, written comments, and spoken comments. The evaluation comments are converted into numerical ratings using a machine learning algorithm, which allows for comparison with the assigned numerical ratings. Findings reveal that the consideration of warmth and competence depends not only on the appraisal format but also on the rater (evaluator) and ratee (task-solver) gender. This study enhances the understanding of how evaluations differ across formats and examines the role of gender in shaping perceptions of warmth and competence. |
| Keywords: | performance appraisal, evaluation formats, social judgment, machine learning, gender stereotypes, quantitative text analysis |
| JEL: | J24 M51 D91 |
| Date: | 2025–02 |
| URL: | https://d.repec.org/n?u=RePEc:pdn:dispap:131 |
| By: | Maddalena Honorati; Céline Ferré; Tomasz Gajderowicz |
| Keywords: | Social Protections and Labor-Labor Markets Information and Communication Technologies-ICT Applications |
| Date: | 2023–11 |
| URL: | https://d.repec.org/n?u=RePEc:wbk:wboper:40628 |
| By: | Brent, Daniel A.; Wietelman, Derek (Resources for the Future); Wichman, Casey (Resources for the Future) |
| Abstract: | Using price incentives to allocate scarce resources is a core tenet of economics but may result in unpalatable distributional outcomes. We analyze the efficacy of prices as a means of inducing water conservation during severe drought by studying the introduction of surcharges enacted within existing nonlinear rate structures. Embedding machine learning counterfactual prediction methods within a demand framework to isolate exogenous price variation, we find evidence that households exhibit a significant demand response despite the temporary nature of surcharges. However, further investigation reveals that surcharges alone cannot explain a majority of the conservation observed despite steep price increases. “Budget-based” rates undercut scarcity signals by shielding large users from binding price increases, and surcharges themselves do little to reduce the regressivity of water expenditures. Simpler rate structures can dominate along equity dimensions, and their progressivity can be enhanced via lump-sum transfers within the rate structure. |
| Date: | 2025–03–12 |
| URL: | https://d.repec.org/n?u=RePEc:rff:dpaper:dp-25-07 |
| By: | Samantha M. Treacy; Alexandra B. Moura |
| Abstract: | As reliance on solar photovoltaic (PV) generation grows, particularly in Alberta, accounting for the impact of wildfire smoke on solar energy production is crucial. This is particularly relevant in regions with high PV generation potential, such as Alberta, as they are often more vulnerable to frequent and intense wildfires. This study quantifies PV energy losses and financial impacts due to wildfire smoke in Alberta, using fine particulate matter 2.5 (PM2.5) as a proxy for smoke pollution. Historical weather and PM2.5 data, along with simulated PV production from actual completed, proposed, and under-construction projects, are used to train and test the model. The simulated data is validated against real production data. The six-year study (2018–2023) covers major wildfire years and employs machine learning techniques, particularly random forest regression, to isolate the effects of PM2.5 on solar production. Financial losses are estimated in Canadian dollars, adjusted for inflation to December 2023. Results show a PV production decline of up to 6.3% at a single solar site over six years, with an overall average reduction of 3.91% under severe conditions. The cumulative impact led to a 0.19% average generation loss, equating to over $4.5 million in financial losses. Higher smoke levels consistently correlate with greater solar energy losses, aligning with findings from other regions. The results of this study enhance our understanding of climate change impacts on solar energy, highlighting wildfire smoke as a relevant factor. As PV adoption expands, these findings offer valuable insights for decision-makers and operational planners, emphasizing the need for strategies to mitigate smoke-related disruptions and ensure energy reliability. |
| Keywords: | Photovoltaic production; Wildfires; PM2.5; Financial impact; Random forest; Solar power. |
| JEL: | C55 N72 P18 Q42 Q54 |
| Date: | 2025–03 |
| URL: | https://d.repec.org/n?u=RePEc:ise:remwps:wp03732025 |