Hypothesis Formalization: Empirical Findings, Software Limitations, and Design Implications

Data analysis requires translating higher level questions and hypotheses into computable statistical models. We present a mixed-methods study aimed at identifying the steps, considerations, and challenges involved in operationalizing hypotheses into statistical models, a process we refer to as hypothesis formalization . In a formative content analysis of 50 research papers, we find that researchers highlight decomposing a hypothesis into sub-hypotheses, selecting proxy variables, and formulating statistical models based on data collection design as key steps. In a lab study, we find that analysts fixated on implementation and shaped their analyses to fit familiar approaches, even if sub-optimal. In an analysis of software tools, we find that tools provide inconsistent, low-level abstractions that may limit the statistical models analysts use to formalize hypotheses. Based on these observations, we characterize hypothesis formalization as a dual-search process balancing conceptual and statistical considerations constrained by data and computation and discuss implications for future tools.

The Intergenerational Mobility of Income: A Study Applied to the Spanish Case (2005–2011)

Intergenerational income mobility has attracted the interest of many economists for—among other reasons—its role as a mechanism for reducing inequalities and achieving equal opportunities. In this paper, we analyse the intergenerational mobility of income in Spain in the years 2005 and 2011, located at different phases of the economic cycle. We use proxy variables (the economic situation of the household during the adolescence of the informant and the educational level achieved by parents) to study intergenerational income mobility, because there are not extant surveys with income information from parents and their descendants when they are part of a different household. With these variables, we try to verify the existence and degree of mobility by analysing different methodologies. The results suggest the existence of mobility in the two studied years, although a trend towards a reduction in intergenerational mobility is confirmed, already detected by other authors.

The determinants of corporate FX speculation – Why firms increase risk

PurposeEmpirical evidence on the determinants of corporate FX speculation is ambiguous. We note that the conflicting findings of prior studies could be the result of different methodologies in determining speculation. Using a novel approach to defining speculative activities, we seek to help solve the puzzle of the determinants of speculation and examine which firms engage in such activities and why they do so.Design/methodology/approachThis paper examines an unexplored regulatory environment that contains publicly reported FX risk data on the firms' exposures before and after hedging per year and currency. This unprecedented data granularity allows us to use actual reported volumes instead of proxy variables in defining speculation and to examine whether the convexity theories are empirically supported in FX risk management.FindingsWe find that frequent speculators are smaller, have more growth opportunities and possess lower internal resources, which indicates unprecedented empirical evidence for the convexity theories in FX risk management. Further, we provide evidence that corporate speculation might be linked to the application of hedge accounting.Practical implicationsWe help solve the questions of which and why firms engage in speculative activities. This can provide valuable information to various stakeholders such as financial analysts, investors, or regulators, which can help prevent imperiling corporate losses and curb excessive speculative financial activities.Originality/valueIn order to question the unresolved issue of the determinants of speculation, this paper is the first to use openly available accounting data with actual reported FX exposure information before and after hedging in defining speculation, instead of relying on proxy variables for FX exposure and derivative usage with potential estimation errors.

A kognitív készségek hatása a diplomás bérekre

Összefoglaló. Ebben a cikkben Magyarországon elsőként vizsgálom a kognitív készségek szerepét a felsőoktatási végzettség megtérülésében. Az adatok a KRTK Adatbank Kapcsolt Államigazgatási Paneladatbázisából származnak (Sebők 2019). Az adatforrás lehetővé teszi a különböző államigazgatási adatbázisok együttes vizsgálatát a magyar lakosság 50%-os mintáján. Az elemzésben a 2008-ban 10. osztályos középiskolások kompetenciaeredményeit mint a korai kognitív készségek proxy változóját használom a hosszú távú diplomás pályakövetéses vizsgálatomban. A tanulmányban az oktatás hozamszámítási megközelítései közül a kereseti függvények módszerét alkalmazom. Summary. This paper investigates the role of cognitive skills in the return to higher education (HE) in Hungary. It makes use of linked Hungarian administrative data, which contains labor market and educational information of about 50 percent of the Hungarian population, for the period of 2003 and 2017 (Sebők 2019). The estimates are focused on the early carrier path of HE graduates who completed their National Assessment of Basic Competencies 10th class tests in 2008. The paper uses Mincer-type regression models with the test scores as the the proxy variables of cognitive skills.

Measurement validity in historical conflict data: Comparing datasets from Spain

Conflict research usually suffers from data availability problems, which sometimes motivates the use of use of proxy variables for violent events. But since they are usually the only alternative to measure violence patterns, there is not ground-truth data to compare them to. This limitation explains why there are no studies assessing their validity. This research note exploits a case where there are two sources on political violence: the Spanish Civil War. Comparing georeferenced mass graves and direct records of victimization, I show that the differences between these two datasets are not random but respond to different data generation processes, introducing important biases. Results highlight the need for a more careful assessment when using proxy variables for political violence.

Cloud drop number concentrations over the western North Atlantic Ocean: seasonal cycle, aerosol interrelationships, and other influential factors

Cloud drop number concentrations (Nd) over the western North Atlantic Ocean (WNAO) are generally highest during the winter (DJF) and lowest in summer (JJA), in contrast to aerosol proxy variables (aerosol optical depth, aerosol index, surface aerosol mass concentrations, surface cloud condensation nuclei (CCN) concentrations) that generally peak in spring (MAM) and JJA with minima in DJF. Using aircraft, satellite remote sensing, ground-based in situ measurement data, and reanalysis data, we characterize factors explaining the divergent seasonal cycles and furthermore probe into factors influencing Nd on seasonal timescales. The results can be summarized well by features most pronounced in DJF, including features associated with cold-air outbreak (CAO) conditions such as enhanced values of CAO index, planetary boundary layer height (PBLH), low-level liquid cloud fraction, and cloud-top height, in addition to winds aligned with continental outflow. Data sorted into high- and low-Nd days in each season, especially in DJF, revealed that all of these conditions were enhanced on the high-Nd days, including reduced sea level pressure and stronger wind speeds. Although aerosols may be more abundant in MAM and JJA, the conditions needed to activate those particles into cloud droplets are weaker than in colder months, which is demonstrated by calculations of the strongest (weakest) aerosol indirect effects in DJF (JJA) based on comparing Nd to perturbations in four different aerosol proxy variables (total and sulfate aerosol optical depth, aerosol index, surface mass concentration of sulfate). We used three machine learning models and up to 14 input variables to infer about most influential factors related to Nd for DJF and JJA, with the best performance obtained with gradient-boosted regression tree (GBRT) analysis. The model results indicated that cloud fraction was the most important input variable, followed by some combination (depending on season) of CAO index and surface mass concentrations of sulfate and organic carbon. Future work is recommended to further understand aspects uncovered here such as impacts of free tropospheric aerosol entrainment on clouds, degree of boundary layer coupling, wet scavenging, and giant CCN effects on aerosol–Nd relationships, updraft velocity, and vertical structure of cloud properties such as adiabaticity that impact the satellite estimation of Nd.

Inferring Grassland Drought Stress with Unsupervised Learning from Airborne Hyperspectral VNIR Imagery

The 2018–2019 Central European drought had a grave impact on natural and managed ecosystems, affecting their health and productivity. We examined patterns in hyperspectral VNIR imagery using an unsupervised learning approach to improve ecosystem monitoring and the understanding of grassland drought responses. The main objectives of this study were (1) to evaluate the application of simplex volume maximisation (SiVM), an unsupervised learning method, for the detection of grassland drought stress in high-dimensional remote sensing data at the ecosystem scale and (2) to analyse the contributions of different spectral plant and soil traits to the computed stress signal. The drought status of the research site was assessed with a non-parametric standardised precipitation–evapotranspiration index (SPEI) and soil moisture measurements. We used airborne HySpex VNIR-1800 data from spring 2018 and 2019 to compare vegetation condition at the onset of the drought with the state after one year. SiVM, an interpretable matrix factorisation technique, was used to derive typical extreme spectra (archetypes) from the hyperspectral data. The classification of archetypes allowed for the inference of qualitative drought stress levels. The results were evaluated using a set of geophysical measurements and vegetation indices as proxy variables for drought-inhibited vegetation growth. The successful application of SiVM for grassland stress detection at the ecosystem canopy scale was verified in a correlation analysis. The predictor importance was assessed with boosted beta regression. In the resulting interannual stress model, carotenoid-related variables had among the highest coefficient values. The significance of the photochemical reflectance index that uses 512 nm as reference wavelength (PRI512) demonstrates the value of combining imaging spectrometry and unsupervised learning for the monitoring of vegetation stress. It also shows the potential of archetypical reflectance spectra to be used for the remote estimation of photosynthetic efficiency. More conclusive results could be achieved by using vegetation measurements instead of proxy variables for evaluation. It must also be investigated how the method can be generalised across ecosystems.

Migraine and stress—an exploratory cross-country study of external stress factors

Objective The data collected by the Global Burden of Disease 2016 project indicate that migraine ranks second in high-income countries with very competitive and flexible labour markets, and first in low- and middle-income countries suffering from civic unrest and conflict. This raises the question whether external stress factors may be correlated with migraine years lived with disability per 100,000 inhabitants (YLD). The objective of this exploratory study is to test the hypothesis that external stress factors are correlated with the prevalence and severity of migraine at the country level. The analysis uses two country groups: developed and developing countries. For the first group, the proxy variables for stress are labour productivity and unemployment rate. For the second group, the proxy variables measure conflict-related deaths and share of migrant/refugee population. Results The results show a positive relationship between the stress variables on the one hand and migraine YLD on the other hand for both country groups. Almost all results are statistically significant at p < 0.01. These exploratory findings suggest that societal stress factors may be potential candidates for modifiable factors for the prevalence and/or severity of migraine at the country level.