Multinomial Logit Research Articles

The impact of individual, scientific and organizational factors on the adoption of AR in university education

PurposeAugmented reality (AR) is an advanced version of the dynamic physical space that is perceived and received via visual, audio, digital and other sensory stimuli. The capabilities of virtual technologies change the field of university and education considerably. The necessity of using virtual technologies in the education field was revealed more in unforeseen disasters such as the COVID-19 pandemic. The adoption of a technology by its users is an important factor in the successful implementation of the technology. The present study evaluates several factors affecting the adoption of AR technology in the realm of tertiary education.Design/methodology/approachThis study is applied in nature, and the necessary data were gathered through a survey questionnaire. The opinions of 621 students were investigated using a simple random sampling method. The multinomial logit test was used in this research.FindingsIt was found that individual and social factors such as age, education level, major and economic characteristics such as one’s income in a month, expenses of a person in a month, the level of access to high-speed internet and access to a laptop or smartphone are effective in AR technology adoption in the field of academic education.Originality/valueThe theoretical contribution of this study is to identify the key factors that influence the adoption of AR technology and develop a model specifically applicable to the educational field. The results of this research can be used by university managers and educational policymakers for the efficient and effective use of this technology.

Public transport mode choice behavior of different-income passengers during large-scale public health incidents

Public transport ridership has been hit hard by the COVID-19 pandemic in many countries. Investigating passengers' public transport mode choice behavior during large-scale public health incidents can uncover the major influential factors and help propose policies and strategies to reduce the pandemic transmission and recover the public transport revenue. This study develops an integrated choice and latent variables (ICLV) model by income based on structural equation model to model passengers' public transport choice behavior during the normalized stage of the pandemic. The model considers passengers' socioeconomic attributes, travel attributes, and attitude-perception attributes, and can appropriately capture passengers' psychological latent attributes. Taking Beijing China as an example, we collect some revealed preference survey data online. The modeling results show that the risk perception as a mediator variable has a significant impact on mode preference. Moreover, the convenience of public transport has the largest influence on risk perception. These findings suggest that risk perception and the convenience of public transport play a major role in passengers' mode choice behavior. In addition, the impacts of the various influential factors on the public transport mode choices are significantly different across different income groups. Further, the ICLV model can achieve better performance and is superior to the traditional Multinomial Logit model. The modeling framework can help propose targeted and instructive strategies during the normalized stage of the pandemic by uncovering the major influential factors in passengers’ public transport mode choices, which is applicable to similar pandemics in the future.

Exploring factors affecting the injury severity of highway-rail grade crossing crashes during the COVID-19 pandemic

ABSTRACT The outbreak of COVID-19 pandemic could alter the injury severity pattern of highway-rail grade crossing (HRGC) crashes. To compare factors affecting injury severities before and during the pandemic, HRGC crashes occurring between 2018 and 2021 in the United States were collected. A random parameters logit model with heterogeneity in means was developed to investigate the impact of multiple factors during the pandemic (2020-2021). Besides, a multinomial logit model was developed for the pre-pandemic (2018-2019) crashes. The present study indicates that the existence of old drivers and illuminated crossings generated random parameters with heterogeneity in means during the pandemic, which were affected by the driver in vehicle, estimated train speed, and crossing AADT indicators. As for the pre-pandemic period, no significant random parameters and unobserved heterogeneity were captured. Significant variables also differed markedly between these two periods, with no significant driver-related or vehicle-related indicators identified from 2018 to 2019.

Comparing discrete choice and machine learning models in predicting destination choice

Destination choice modeling has long been dominated by theory-based discrete choice models. Simultaneously, machine learning has demonstrated improved predictive performance to other fields of discrete choice modeling. The objective of this research was to compare machine learning models and a multinomial logit model in predicting destination choice. The models were assessed on their predictive performance using metrics for both binary classification and probabilistic classification. The results indicate that machine learning models, especially a random forest model, could bring improvements in prediction accuracy. The more data was used in training the models, the better the machine learning models tended to perform compared to the multinomial logit model. With less data, the multinomial logit model performed comparatively well. The findings are relevant for the field of destination choice modeling, where evidence on the use of machine learning models is very limited. In addition, the unbalanced choice sets of destination choice models with multiple non-chosen alternatives increases the need for further research in model fit and parameter tuning.

Disputes of Adverse Possession on Forestland and the Determinants of Case Outcomes

The doctrine of adverse possession presents a risk to forestland ownership because timber management is infrequent. In this study, the extent of adverse possession claims on forestland in the United States is examined through a set of 243 published legal cases, and the determinants of the case outcomes are analyzed by a multinomial logit model. Those cases are from 32 states and distributed over 200 years, with an average of 12 cases per decade. Adverse possession claimants win 44% of the cases. The regression analysis reveals that whether an adverse possessor has a defective title is the leading factor in determining the case outcomes, and the periods required by statutes tend to negatively impact the winning probability of adverse possessors. The empirical findings have implications for landowners to defend their titles or remove a cloud on titles and for legislators to revise relevant laws in the future.

Weighting factors for LCA—a new set from a global survey

PurposeThis paper provides global weights (weighting factors) for the three endpoint impact categories (areas of protection (AoPs)) of the United Nations Environment Programme (UNEP) Life Cycle Initiative’s “Global Guidance for Life Cycle Impact Assessment Indicators and Methods” (GLAM) project, namely human health, ecosystem quality, and natural resources and ecosystem services.MethodsA discrete choice experiment (DCE) was conducted to elicit the preferences of respondents on the GLAM AoPs, and they were then used to calculate the respective weights. Responses were obtained from a subset of countries pertaining to each income level defined by the World Bank (i.e. low, lower-middle, upper-middle, and high). The adimensional (between 0 and 1) weights were derived using two different approaches: econometric and multiple criteria decision analysis (MCDA). The econometric approach obtained weights by transforming the estimated preference parameters from a multinomial logit model. The MCDA approach obtained weights representing the vectors that best reconstitute the choices of each individual, using linear programming to fit an additive value function.ResultsWhen considering responses from all income groups, the weights from the econometric approach are 0.42, 0.31, and 0.26 for human health, ecosystem quality, and natural resources and ecosystem services, respectively. Following the same order for the AoPs, the weights from the MCDA approach are 0.41, 0.32, and 0.27. For high-income countries, ecosystem quality has the highest weight; for upper-middle-income countries, ecosystem quality and human health have the same weights using the econometric approach, while in the MCDA approach, human health is weighted higher than ecosystem quality. For the two lower income country groups, the priority is given to human health with both approaches. Recommendations for the use of these weights are also provided, as well as a comparison with other existing weights.ConclusionThe two methods obtained similar weights overall, although with some differences when disaggregated by income groups. The weights proposed in this paper are suitable for decision-makers or users who want to use survey-derived weights for endpoint-based LCA when using the AoPs of GLAM. These weights can be used in projects where the decision-makers do not want to or have no resources to identify a set of weights themselves, or when decision-makers are not involved.

The impacts of combining incentives on carpooling for commuting in the Paris region

Carpooling is one solution to help reduce the negative externalities of solo driving. The aim of this study is to analyze the effect of incentives on solo drivers’ intention to carpool. A questionnaire survey was carried out among 1,329 participants, including 590 (38.39%) residents of the Paris metropolitan area who use a car to get to work. Modal choice estimates were modeled using multinomial logit and binomial logit econometric models for car drivers only. First, we analyzed the effect of non-financial incentives, such as nudges and trust in others during the carpooling trip. Second, we analyzed the effect of the combination of nudges with employer subsidies for sustainable mobility on the intention to carpool as a driver or passenger for the home-to-work (or home-to-school) commute. The results show that people aged 18–49, employees, and those with higher levels of education are more receptive to behavioral interventions such as nudges. Applying nudges in isolation is more effective at encouraging carpooling for commuting than combining nudges with employers’ subsidies for sustainable mobility.

Determinants of local food producer participation in state‐sponsored marketing programs: Evidence from Missouri

AbstractThis study examines producer participation choices considering a variety of potential benefits linked to state‐sponsored marketing programs, using a real choice dataset of farmers in Missouri. Multinomial logit models are employed to predict determinants of farmer enrollment in three tiers of the Missouri Grown local food marketing program: free, $50, and $100 memberships. We find that farmers who expect to increase sales or receive special services from the program are more likely to choose paid membership levels, while those who seek new opportunities are more associated with free participation. Furthermore, farms with annual sales above $150,000 and those with male or white decision‐makers are more likely to be $100 members, while young and beginning farmers tend to participate at a paid tier of $50 in the program.

Identifying the causal effects of driver distraction on hazardous actions at intersections based on propensity score weighting

Distracted driving is a threat to traffic safety that can result in more traffic crashes. Although previous studies have been conducted to explore the relationship between driver distraction and hazardous driving actions, few studies are available to identify the causation between them. Thus, the study intended to evaluate the causal effects of distraction on hazardous driving actions at intersections based on the crash data extracted from the Crash Report Sampling System (2021–2022). The multinomial logit model was employed to reveal the factors contributing to driver distraction. Then, propensity score weighting was adopted to balance the factor distributions between distraction and non-distraction cases to identify the causal effects on hazardous actions. Results indicated that 1) the propensity of distraction is relevant to factors such as the driver’s age, gender, vehicle type, speed limit, area, weather, and light condition, 2) driver distraction can significantly increase the probability of risky actions including speeding, running red lights, failing to obey stop signs, failing to yield, following too closely, and 3) the causal effects show great diversity for different distraction types. The findings serve to understand the influence mechanism of distraction on specific crash risks and develop countermeasures to reduce distraction and hazardous driving actions.

A novel choice model combining utility maximization and the disjunctive decision rules, application to two case studies

Most choice models, e.g. Multinomial Logit (MNL), rely on random utility theory, which assumes that a compensatory utility maximization decision rule explains an individual’s choice behaviour. Research has shown, however, that behaviour is sometimes better explained by non-compensatory decision rules. While some research has used Latent Class Choice Models (LCCMs) to account for multiple decision rules, many of them – such as the disjunctive rule – have yet to be explored. This paper formulates, estimates, and evaluates a LCCM that combines the MNL with a Generalised Random Disjunctive Model (GRDM), a new choice model we develop. Addressing deficiencies of existing disjunctive choice models, the GRDM allows for relative importance between attributes and is insensitive to irrelevant attributes. Unlike most non-compensatory models, it is tractable and incorporates random error terms for capturing unobserved heterogeneity across choice situations. The GRDM can be expressed as a Universal Logit (UL) model, which helps derive welfare metrics such as Marginal Rates of Substitution and elasticities and makes it possible to estimate the model with traditional software packages. The LCCM combining the GRDM and the MNL is estimated in two large-scale case studies: cyclists’ route choice and public transport route choice. Results are compared with other relevant LCCM specifications and the individual choice models, where it is found that the MNL + GRDM LCCM provides the best fit to the data. We also interpret the fitted parameters and calculate the Marginal Rates of Substitution, which align with behavioural expectations.

Rethinking cycling safety: The role of gender in cyclist crash injury severity outcomes

Given the ongoing climate crisis and the need for environmentally friendly communities, there has been an increasing interest in sustainable mobility solutions such as cycling. This study seeks to incorporate an equitable component to studying cycling safety and uses one full year’s data of 4,457 single bicycle-single motor vehicle crashes that took place in 2022 in the state of Florida to estimate a series of random parameters multinomial logit models with heterogeneity in the means and variances to capture gender differences in outcome severities. A comparison of advanced statistical models such as unconstrained and partially constrained approaches, that were previously employed in the literature to test for temporal stability, is undertaken in a new application. A partially constrained model is estimated to best identify gender specific factors and argue the need to evaluate and promote safety of female and male cyclists separately. The study finds substantial differences between how the contributing factors and crash circumstances impact the crash injury severity of women and men cyclists. It evaluates factors such as age, location, cyclist behavior, weather, and road design as well as performs out-of-sample simulation to gain additional insights. The findings of this research emphasize the need for targeted approaches in designing our cities and policy making that account for the collective differences in behavior and experience of women and men cyclists.

Dynamic pricing for perishable product in the presence of low-carbon preference

ABSTRACT As countries worldwide strive to implement energy conservation and emission reduction measures, modeling consumers’ low-carbon preferences and purchasing decisions has become a crucial factor in implementing dynamic pricing strategies for firms. This paper investigates the dynamic pricing problems for a single perishable product and substitutable perishable products, respectively, where consumer is affected by low-carbon preference. We first propose the decision framework of consumer choice with low-carbon preference based on the multinomial logit model. On this basis, we establish dynamic pricing models by the dynamic programming method and obtain the optimal pricing strategies. We analytically show that consumer’s low-carbon preference has a positive effect on the optimal price for a single perishable product. Meanwhile, numerical examples illustrate that consumer’s low-carbon preference has a positive effect on the optimal prices for substitutable perishable products. Finally, we give some suggestions to increase firm’s revenue when consumer’s low-carbon preference is present.

Understanding spatial inequalities and stratification in transportation accessibility to social infrastructures in South Korea: multi-dimensional planning insights

This research investigated spatial inequalities in transportation accessibility to social infrastructures (SIs) in South Korea, using a multi-dimensional methodological approach, including descriptive/bivariate analysis, explanatory factor analysis (EFA), K-Mean cluster analysis, and multinomial logit model (MNL). Our study confirmed pronounced spatial disparities in transportation accessibility to SIs, highlighting significantly lower access in rural and remote regions compared to urban centers and densely populated areas, consistent with existing literature. Building on prior findings, several additional findings were identified. First, we uncovered significant positive correlations among accessibility to different types of SIs in four critical categories: green and recreation spaces, health and aged care facilities, educational institutions, and justice and emergency services, revealing prevalent spatial inequality patterns. Second, we identified three distinct accessibility clusters (High, Middle, and Low) across the critical SI categories. Specifically, residents within the High cluster benefited from the closest average network distances to all SIs, while those in the Low cluster faced significant accessibility burdens (e.g., 22.9 km for welfare facilities, 20.1 km for hospitals, and 19.2 km for elderly care facilities). Third, MNL identified factors such as population density and housing prices as pivotal in spatial stratification of accessibility. Specifically, areas with lower SI accessibility tended to have a higher proportion of elderly residents. Also, decreased accessibility correlated with diminished traffic volumes across all transportation modes, particularly public transportation. This research contributes to enhancing our understanding of spatial inequalities in transportation accessibility to SIs and offers insights crucial for transportation and urban planning.

Developing mapping algorithms to predict EQ-5D health utility values from Bath Ankylosing Spondylitis Disease Activity Index and Bath Ankylosing Spondylitis Functional Index among patients with Ankylosing Spondylitis

BackgroundPreference-based measures of health-related quality of life (HRQoL), such as the EQ-5D or the SF-6D, are essential for health economic evaluation. However, they are rarely included in clinical trials of ankylosing spondylitis (AS). This study aims to develop mapping algorithms to predict EQ-5D-3L and EQ-5D-5L health utility scores from the Bath Ankylosing Disease Activity Index (BASDAI) and the Bath Ankylosing Spondylitis Functional Index (BASFI).MethodsPatients with AS were recruited from the largest tertiary hospital in Shandong province, China, between December 2019 and October 2020. Patients were selected by convenience sampling method according to the following criteria: (1) diagnosed with AS according to the New York criteria; (2) aged 18 years and above; and (3) without mental disorders; (4) able to understand the questionnaires; (5) without serious complications. There were 243 patients who completed the face-to-face questionnaire survey, and 5 cases with missing values in key variables were excluded. Ordinary least squares, censored least absolute deviations, Tobit, adjusted limited dependent variable mixture model and beta-mixture model (BM) in the direct approach and ordered logit and multinomial logit (Mlogit) model in the response approach were used to develop mapping algorithms. Mean absolute error, root mean square error, Spearman’s correlation coefficient and concordance correlation coefficient were used to access predictive performance.ResultsThe 238 patients with AS had a mean age of 35.19 (SD = 9.59) years, and the majority (74.47%) were male. The observed EQ-5D-3L and EQ-5D-5L health utility values were 0.88 (SD = 0.12) and 0.74 (SD = 0.27), respectively. The EQ-5D-5L had higher conceptual overlap with the BASDAI and BASFI than the EQ-5D-3L did. The Mlogit was the best-performing model for the EQ-5D-3L, and the BM showed better performance in predicting EQ-5D-5L than other direct and indirect mapping models did.ConclusionThis study demonstrates that the EQ-5D-5L, rather than EQ-5D-3L, should be selected as the target outcome measure of HRQoL in patients with AS in China, and the BM mapping algorithm could be used to predict EQ-5D-5L values from BASDAI and BASFI for health economic evaluation.

The interdependence between hospital choice and waiting time — with a case study in urban China

Hospital choice models often employ random utility theory and include waiting time as a choice determinant. When applied to evaluate health system improvement interventions, these models disregard that hospital choice in turn is a determinant of waiting time. We present a novel, general model capturing the endogeneous relationship between waiting time and hospital choice, including the choice to opt out, and characterize the unique equilibrium solution of the resulting convex problem. We apply the general model in a case study on the urban Chinese health system, specifying that patient choice follows a multinomial logit (MNL) model and waiting times are determined by M/M/1 queues. The results reveal that analyses which solely rely on MNL models overestimate the effectiveness of present policy interventions and that this effectiveness is limited. We explore alternative, more effective, improvement interventions.

Social Vulnerability and National Diabetes Prevention Program Recognition Status

IntroductionThe CDC National Diabetes Prevention Program (National DPP) aims to reduce the incidence of diabetes in the U.S. Organizations delivering the National DPP receive pending, preliminary, full, or full-plus recognition status based on specific program criteria and outcomes. Achieving full/full-plus recognition is critical for organizations to sustain the program and receive reimbursements to cover costs, but organizations in disadvantaged areas may face barriers to obtaining this level of recognition. This study examined the association between county-level social vulnerability and full/full-plus recognition status within the National DPP. MethodsUsing the 2022 National DPP registry and the 2018 CDC Social Vulnerability Index (SVI), a three-level categorical dependent variable was created (n=843): counties without organizations having full/full-plus recognition, counties with at least one organization not having full/full-plus recognition, and counties with all organizations having full/full-plus recognition. A multinomial logit model was analyzed in 2023 to examine the association between SVI and in-person full/full-plus recognition organizations at the county level, adjusting for confounders. ResultsCompared to counties with low social vulnerability, counties with higher social vulnerability had significantly higher odds of having no organizations with full/full-plus recognition. For example, counties with high SVI had 2.63 (95% CI: 1.55-4.47) times higher odds of having no organizations with full/full-plus recognition compared to having all organizations with full/full-plus CDC recognition. ConclusionsThe findings suggest disparities in the National DPP recognition status among organizations in vulnerable communities. Developing strategies to ensure organizations in high social vulnerability areas achieve at least full recognition status is critical for program sustainability and reducing diabetes-related health disparities.

Emotional and functional well-being in long-term breast cancer survivorship.

Emotional and functional well-being (EWB and FWB) are important components of mental health and quality of life. This study aims to evaluate long-term EWB and FWB in breast cancer (BC) survivors. The Carolina Breast Cancer Study Phase 3 oversampled Black and younger (< 50years in age) women so that they each represent approximately 50% of the study population and assessed participants' EWB and FWB with the Functional Assessment of Cancer Therapy-Breast (FACT-B) at 5- (baseline), 25-, and 84-months post diagnosis. Multinomial logit models were used to estimate adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for associations between demographic and clinical characteristics and well-being change relative to baseline. Among 2,781 participants with BC, average EWB and FWB improved with time since diagnosis. Persistent FWB decrements were associated with Black race [OR 1.4 (95% CI 1.2-1.7) and 1.3 (95% CI 1.1-1.6), at 25-months and 84-months respectively], older age [OR 1.4 (95% CI 1.1-1.7) and 1.5 (95% CI 1.2-1.8), respectively], no chemotherapy, and recurrence [OR 2.9 (95% CI 1.8-4.8) and 3.1 (95% CI 2.1-4.6), respectively]. EWB decrements were associated with advanced stage and recurrence. Decrements in combined (FWB+EWB) well-being were associated with recurrence at both follow-up survey timepoints [ORs 4.7 (95% CI 2.7-8.0) and 4.3 (95% CI 2.8-6.6), respectively]. Long-term well-being varies by demographics and clinical features, with Black women and women with aggressive disease at greatest risk of long-term decrements.

Gender preferences for multiple attributes of soil and water conservation in Northern Rwanda

Despite the dominance of female labor in agricultural production, female-controlled farm plots have lower efficiency compared to plot managed by male-headed households, which indicates a huge gender disparity in agricultural productivity. Overlooking gender preferences when designing interventions that promote the conservation of soil and water resources might face challenges in adoption and could result in ineffective policies to close the gender gap. This study seeks to analyze gender-specific preferences regarding attributes of soil and water conservation (SWC) in northern Rwanda. A best-worst survey was conducted among 653 respondents, comprising 253 males and 400 females, representing 422 households surveyed between September and December 2019. The analysis of BWS data involved assessing attribute-level relative importance, Pearson correlation, and maximum difference scaling using multinomial logit (MNL). Findings from attribute-level importance analysis revealed significant gender-based disparities in preferences across three important SWC attribute scenarios: the high scenario (between 65 % and 100 %), the moderate scenario (between 50 % and 65 %), and the basic scenario (with <50 % relative importance). The study identified heterogeneity in preferences regarding the relative importance of SWC attribute levels. Pearson correlation analysis revealed substantial synergies among attribute levels linked to land consolidation, improved land tenure, and joint SWC decision-making between genders. Additionally, the study identified trade-offs among multiple levels of SWC attributes, including households' SWC decision-making and physical and structural measures. The results from MNL regression show that both males and females exhibit positive preferences for multiple levels of SWC attributes, but show negative preferences when it comes to household decisions involving multiple SWC strategies. The study highlights the importance of equal opportunities for males and females’ participation in agricultural transformation through the adoption of SWC technologies as a fundamental step towards sustainable agricultural intensification. It advocates for gender transformational approaches to incentivize the scaling up of SWC practices and promote packages with lower uptake rates. Additionally, the study suggests enhancing knowledge and extension education in SWC to better understand diverse needs and preferences of female farmers.

The effects of free prescriptions on community pharmacy selection: A discrete choice experiment

BackgroundLittle is known about preferences for community pharmacies, particularly the influence of prescription co-payments, and for New Zealand's priority populations, Māori and Pacific Peoples. Improving understanding of community pharmacy preferences will enable tailoring services to meet community needs. ObjectivesThis study aims to investigate New Zealanders’ preferences for community pharmacies when collecting prescriptions. Additionally, variations in preferences for community pharmacy attributes between different latent and demographic groups were examined. MethodsFocus group discussions with various community groups were thematically analyzed to develop six attributes: location, wait time, customer service, prescription co-payments, parking availability, and nearby businesses. Participants were asked to complete an online survey involving 12 choice tasks, where they had to choose their preferred option of 3 unlabeled pharmacies along with demographic questions. A mixed multinomial logit model and latent classes analysis were used to assess and compare the participant preferences. ResultsThe sample of 553 participants, representative of the New Zealand population, resulted in 19,908 observations for analysis. The most attractive pharmacy attribute was its proximity, being within a 10-min travel distance from home or work. The importance of prescription co-payments is evident, with free prescriptions being the second most attractive attribute level and $15 NZD prescription co-payments being the least appealing. Different classes placed importance on different attributes, the largest of which prioritized prescription co-payments. Including demographic characteristics did not improve model accuracy nor predict class membership. ConclusionsUnder current policy, the most effective way for pharmacies to attract business is by offering free prescriptions. However, the trend of adopting lower-cost models may reduce the quality of care they deliver. Policy decision-makers must decide if they are comfortable with this potential impact.

Choice Consistency in Discrete Choice Experiments: Does Numeracy Skill Matter?

ObjectivesThis study investigated the relationship between numeracy skills (NS) and choice consistency in discrete choice experiments (DCEs). MethodsA DCE was conducted to explore patients’ preferences for kidney transplantation in Italy. Patients completed the DCE and answered 3-item numeracy questions. A heteroskedastic multinomial logit model was used to investigate the effect of numeracy on choice consistency. ResultsHigher NS were associated with greater choice consistency, increasing the scale to 1.63 (P < .001), 1.39 (P < .001), and 1.18 (P < .001) for patients answering 3 of 3, 2 of 3, and 1 of 3 questions correctly, respectively, compared with those with no correct answers. This corresponded to 63%, 39%, and 18% more consistent choices, respectively. Accounting for choice consistency resulted in varying willingness-to-wait (WTW) estimates for kidney transplant attributes. Patients with the lowest numeracy (0/3) were willing to wait approximately 42 months [95% CI: 29.37, 54.68] for standard infectious risk, compared with 33 months [95% CI: 28.48, 38.09] for 1 of 3, 28 months [95% CI: 25.13, 30.32] for 2 of 3, and 24 months [95% CI: 20.51, 27.25] for 3 of 3 correct answers. However, WTW differences for an additional year of graft survival and neoplastic risk were not statistically significant across numeracy levels. Supplementary analyses of 2 additional DCEs on COVID-19 vaccinations and rheumatoid arthritis, conducted online, supported these findings: higher NS were associated with more consistent choices across different disease contexts and survey formats. ConclusionsThe findings suggested that combining patients with varying NS could bias WTW estimates, highlighting the need to consider numeracy in DCE data analysis and interpretation.