scholarly journals Quantile Regression Models and the Crucial Difference Between Individual-Level Effects and Population-Level Effects

2020 ◽  
Author(s):  
Nicolai T. Borgen ◽  
Andreas Haupt ◽  
Øyvind N. Wiborg
Keyword(s):  
Quantile Regression ◽  
Regression Models ◽  
Treatment Effects ◽  
Population Level ◽  
Conditional Quantile ◽  
Individual Level ◽  
Quantile Treatment Effects ◽  
Crucial Difference ◽  
Practical Guidelines ◽  
Conceptual Distinction

The unconditional quantile regression (UQR) model – which has gained increasing popularity in the 2010s and is regularly applied in top-rated academic journals within sociology and other disciplines – is poorly understood and frequently misinterpreted. The main reason for its increased popularity is that the UQR model seemingly tackles an issue with the traditional conditional quantile regression (CQR) model: the interpretation of coefficients as quantile treatment effects changes whenever control variables are included. However, the UQR model was not developed to solve this issue but to study influences on quantile values of the overall outcome distribution. This paper clarifies the crucial conceptual distinction between influences on overall distributions, which we term population-level influences, and individual-level quantile treatment effects. Further, we use data simulations to illustrate that various classes of quantile regression models may, in some instances, give entirely different conclusions (to different questions). The conceptual and empirical distinctions between various quantile regression models underline the need to match the correct quantile regression model to the specific research questions. We conclude the paper with some practical guidelines for researchers.

scholarly journals Flexible and fast estimation of quantile treatment effects: The rqr and rqrplot commands

2021 ◽  
Author(s):  
Nicolai T. Borgen ◽  
Andreas Haupt ◽  
Øyvind N. Wiborg
Keyword(s):  
Quantile Regression ◽  
Fixed Effects ◽  
Regression Models ◽  
Treatment Effects ◽  
The Other ◽  
High Dimensional ◽  
Fast Estimation ◽  
Quantile Treatment Effects

Using quantile regression models to estimate quantile treatment effects is becoming increasingly popular. This paper introduces the rqr command that can be used to estimate residualized quantile regression (RQR) coefficients and the rqrplot postestimation command that can be used to effortless plot the coefficients. The main advantages of the rqr command compared to other Stata commands that estimate (unconditional) quantile treatment effects are that it can include high-dimensional fixed effects and that it is considerably faster than the other commands.

scholarly journals Examining the Effects of the Sacramento Dockless E-Bike Share on Bicycling and Driving

2021 ◽  
Vol 13 (1) ◽  
pp. 368
Author(s):  
Dillon T. Fitch ◽  
Hossain Mohiuddin ◽  
Susan L. Handy
Keyword(s):  
Travel Behavior ◽  
Regression Models ◽  
Population Level ◽  
Panel Survey ◽  
Multilevel Regression ◽  
Combine Data ◽  
Cross Sectional ◽  
Individual Level ◽  
Before And After ◽  
Bike Share

One way cities are looking to promote bicycling is by providing publicly or privately operated bike-share services, which enable individuals to rent bicycles for one-way trips. Although many studies have examined the use of bike-share services, little is known about how these services influence individual-level travel behavior more generally. In this study, we examine the behavior of users and non-users of a dockless, electric-assisted bike-share service in the Sacramento region of California. This service, operated by Jump until suspended due to the coronavirus pandemic, was one of the largest of its kind in the U.S., and spanned three California cities: Sacramento, West Sacramento, and Davis. We combine data from a repeat cross-sectional before-and-after survey of residents and a longitudinal panel survey of bike-share users with the goal of examining how the service influenced individual-level bicycling and driving. Results from multilevel regression models suggest that the effect of bike-share on average bicycling and driving at the population level is likely small. However, our results indicate that people who have used-bike share are likely to have increased their bicycling because of bike-share.

Testing exogeneity in nonparametric instrumental variables models identified by conditional quantile restrictions

2020 ◽  
Author(s):  
Jia-Young Michael Fu ◽  
Joel L Horowitz ◽  
Matthias Parey
Keyword(s):  
Quantile Regression ◽  
Instrumental Variables ◽  
Regression Models ◽  
Quantile Function ◽  
Structural Function ◽  
Conditional Quantile ◽  
Explanatory Variables ◽  
Monte Carlo Experiments ◽  
Nonparametric Estimates ◽  
Conditional Quantile Function

Summary This paper presents a test for exogeneity of explanatory variables in a nonparametric instrumental variables (IV) model whose structural function is identified through a conditional quantile restriction. Quantile regression models are increasingly important in applied econometrics. As with mean-regression models, an erroneous assumption that the explanatory variables in a quantile regression model are exogenous can lead to highly misleading results. In addition, a test of exogeneity based on an incorrectly specified parametric model can produce misleading results. This paper presents a test of exogeneity that does not assume that the structural function belongs to a known finite-dimensional parametric family and does not require estimation of this function. The latter property is important because nonparametric estimates of the structural function are unavoidably imprecise. The test presented here is consistent whenever the structural function differs from the conditional quantile function on a set of nonzero probability. The test has nontrivial power uniformly over a large class of structural functions that differ from the conditional quantile function by $O({n^{ - 1/2}})$. The results of Monte Carlo experiments and an empirical application illustrate the performance of the test.

scholarly journals A New Framework for Estimation of Unconditional Quantile Treatment Effects: The Residualized Quantile Regression (RQR) Model

2021 ◽  
Author(s):  
Nicolai T. Borgen ◽  
Andreas Haupt ◽  
Øyvind N. Wiborg
Keyword(s):  
Quantile Regression ◽  
Regression Model ◽  
Fixed Effects ◽  
Treatment Effects ◽  
Continuous Treatment ◽  
High Dimensional ◽  
Treatment Variable ◽  
Quantile Regression Model ◽  
Quantile Treatment Effects ◽  
New Framework

The identification of unconditional quantile treatment effects (QTE) has become increasingly popular within social sciences. However, current methods to identify unconditional QTEs of continuous treatment variables are incomplete. Contrary to popular belief, the unconditional quantile regression model introduced by Firpo, Fortin, and Lemieux (2009) does not identify QTE, while the propensity score framework of Firpo (2007) allows for only a binary treatment variable, and the generalized quantile regression model of Powell (2020) is unfeasible with high-dimensional fixed effects. This paper introduces a two-step approach to estimate unconditional QTEs where the treatment variable is first regressed on the control variables followed by a quantile regression of the outcome on the residualized treatment variable. Unlike much of the literature on quantile regression, this two-step residualized quantile regression framework is easy to understand, computationally fast, and can include high-dimensional fixed effects.

scholarly journals What Quantile Regression Does and Doesn't Do

2018 ◽  
Author(s):  
Sebastian Ernst Wenz
Keyword(s):  
Linear Regression ◽  
Quantile Regression ◽  
Regression Models ◽  
Simulated Data ◽  
Quantile Function ◽  
Linear Regression Models ◽  
Conditional Quantile ◽  
Correct Understanding ◽  
Mean Function ◽  
Conditional Quantile Function

Petscher and Logan (2014)’s description of quantile regression might mislead readers to believe it would estimate the relation between an outcome, y, and one or more predictors, x, at different quantiles of the unconditional distribution of y. However, quantile regression models the conditional quantile function of y given x just as linear regression models the conditional mean function. This article’s contribution is twofold: First, it discusses potential consequences of methodological misconceptions and formulations of Petscher and Logan (2014)’s presentation by contrasting features of quantile regression and linear regression. Secondly, it reinforces the importance of correct understanding of quantile regression in empirical research by illustrating similarities and differences of various quantile regression estimators and linear regression using simulated data.

Quantile Treatment Effects in the Presence of Covariates

2020 ◽  
Vol 102 (5) ◽  
pp. 994-1005 ◽  
Author(s):  
David Powell
Keyword(s):  
Quantile Regression ◽  
Instrumental Variable ◽  
Treatment Effects ◽  
Simultaneous Equations ◽  
Quantile Treatment Effects ◽  
Simultaneous Equations Models ◽  
Special Cases ◽  
Instrumental Variable Quantile Regression

This paper proposes a method to estimate unconditional quantile treatment effects (QTEs) given one or more treatment variables, which may be discrete or continuous, even when it is necessary to condition on covariates. The estimator, generalized quantile regression (GQR), is developed in an instrumental variable framework for generality to permit estimation of unconditional QTEs for endogenous policy variables, but it is also applicable in the conditionally exogenous case. The framework includes simultaneous equations models with nonadditive disturbances, which are functions of both unobserved and observed factors. Quantile regression and instrumental variable quantile regression are special cases of GQR and available in this framework.

scholarly journals Blood Pressure Trajectories Across the Life Course

2021 ◽  
Vol 34 (3) ◽  
pp. 234-241
Author(s):  
Norrina B Allen ◽  
Sadiya S Khan
Keyword(s):  
Blood Pressure ◽  
Measurement Errors ◽  
Clinical Care ◽  
Population Level ◽  
Cumulative Exposure ◽  
Older Adulthood ◽  
Individual Level ◽  
The Life Course ◽  
Level Group ◽  
Improve Health

Abstract High blood pressure (BP) is a strong modifiable risk factor for cardiovascular disease (CVD). Longitudinal BP patterns themselves may reflect the burden of risk and vascular damage due to prolonged cumulative exposure to high BP levels. Current studies have begun to characterize BP patterns as a trajectory over an individual’s lifetime. These BP trajectories take into account the absolute BP levels as well as the slope of BP changes throughout the lifetime thus incorporating longitudinal BP patterns into a single metric. Methodologic issues that need to be considered when examining BP trajectories include individual-level vs. population-level group-based modeling, use of distinct but complementary BP metrics (systolic, diastolic, mean arterial, mid, and pulse pressure), and potential for measurement errors related to varied settings, devices, and number of readings utilized. There appear to be very specific developmental periods during which divergent BP trajectories may emerge, specifically adolescence, the pregnancy period, and older adulthood. Lifetime BP trajectories are impacted by both individual-level and community-level factors and have been associated with incident hypertension, multimorbidity (CVD, renal disease, cognitive impairment), and overall life expectancy. Key unanswered questions remain around the additive predictive value of BP trajectories, intergenerational contributions to BP patterns (in utero BP exposure), and potential genetic drivers of BP patterns. The next phase in understanding BP trajectories needs to focus on how best to incorporate this knowledge into clinical care to reduce the burden of hypertensive-related outcomes and improve health equity.

Medication Adherence: Expanding the Conceptual Framework

2021 ◽  
Author(s):  
Marie Krousel-Wood ◽  
Leslie S Craig ◽  
Erin Peacock ◽  
Emily Zlotnick ◽  
Samantha O’Connell ◽  
...  
Keyword(s):  
Older Adults ◽  
Medication Adherence ◽  
Conceptual Framework ◽  
Clinical Outcomes ◽  
Implicit Attitudes ◽  
Explanatory Power ◽  
Population Level ◽  
Time Preferences ◽  
Adherence Behavior ◽  
Individual Level

Abstract Interventions targeting traditional barriers to antihypertensive medication adherence (AHMA) have been developed and evaluated, with evidence of modest improvements in adherence. Translation of these interventions into population-level improvements in adherence and clinical outcomes among older adults remains suboptimal. From the Cohort Study of Medication Adherence among Older adults (CoSMO), we evaluated traditional barriers to AHMA among older adults with established hypertension (N=1544; mean age=76.2 years, 59.5% women, 27.9% Black, 24.1% and 38.9% low adherence by proportion of days covered (i.e., PDC<0.80) and the 4-item Krousel-Wood Medication Adherence Scale (i.e., K-Wood-MAS-4≥1), respectively), finding that they explained 6.4% and 14.8% of variance in pharmacy refill and self-reported adherence, respectively. Persistent low adherence rates, coupled with low explanatory power of traditional barriers, suggest that other factors warrant attention. Prior research has investigated explicit attitudes toward medications as a driver of adherence; the roles of implicit attitudes and time preferences (e.g., immediate versus delayed gratification) as mechanisms underlying adherence behavior are emerging. Similarly, while associations of individual-level social determinants of health (SDOH) and medication adherence are well-reported, there is growing evidence about structural SDOH and specific pathways of effect. Building on published conceptual models and recent evidence, we propose an expanded conceptual framework that incorporates implicit attitudes, time preferences and structural SDOH, as emerging determinants that may explain additional variation in objectively and subjectively measured adherence. This model provides guidance for design, implementation and assessment of interventions targeting sustained improvement in implementation medication adherence and clinical outcomes among older women and men with hypertension.

Mindful of Personality Trait Change: Are Treatment Effects on Personality Trait Change Ephemeral and Attributable to Changes in States?

2021 ◽  
Author(s):  
Mirjam Stieger ◽  
Mathias Allemand ◽  
Brent Roberts ◽  
Jordan Davis
Keyword(s):  
Substance Use ◽  
Personality Trait ◽  
Treatment Effects ◽  
Intervention Study ◽  
Measure Change ◽  
Conceptual Distinction ◽  
State Changes ◽  
Highly Correlated ◽  
Changes Over Time

Objective: Are treatment effects on personality trait change ephemeral and attributable to change in clinical states? Data of an intervention study was used to examine if change in clinical states (e.g., stress or depression) accounts for change in personality traits and to test whether both changes in traits and clinical states were independently associated with substance use. Method: Seventy-nine substance use patients (Mage = 25.3, SD = 2.7; 35 % female) took part at a 4-week intervention and completed a total of 15 bi-monthly assessments across 28 weeks to measure change in traits and states during and after treatment. Results: The results suggest that participants showed large trait and state changes over time, which happened rapidly with the majority occurring during the first four weeks. Trait and state changes were highly correlated, but not perfectly redundant. Significant variance in personality trait change remained after controlling for change in clinical states. Moreover, both trait and state change independently predicted substance use. Conclusion: Personality trait change occurred relatively fast and was maintained until the last follow-up assessment six months after the end of the intervention. Also, the findings point to the notion that the conceptual distinction between traits and states may not be as important as originally thought.

Sign in / Sign up

Export Citation Format

Share Document