Neighborhood Density Research Articles

Neighborhood Food Environment and Birth Weight Outcomes in New York City

Infants born with unhealthy birth weight are at greater risk for long-term health complications, but little is known about how neighborhood characteristics (eg, walkability, food environment) may affect birth weight outcomes. To assess whether neighborhood-level characteristics (poverty rate, food environment, and walkability) are associated with risk of unhealthy birth weight outcomes and to evaluate whether gestational weight gain mediated these associations. The population-based cross-sectional study included births in the 2015 vital statistics records from the New York City Department of Health and Mental Hygiene. Only singleton births and observations with complete birth weight and covariate data were included. Analyses were performed from November 2021 to March 2022. Residential neighborhood-level characteristics, including poverty, food environment (healthy and unhealthy food retail establishments), and walkability (measured by both walkable destinations and a neighborhood walkability index combining walkability measures like street intersection and transit stop density). Neighborhood-level variables categorized into quartiles. The main outcomes were birth certificate birth weight measures including small for gestational age (SGA), large for gestational age (LGA), and sex-specific birth weight for gestational age z-score. Generalized linear mixed-effects models and hierarchical linear models estimated risk ratios for associations between density of neighborhood-level characteristics within a 1-km buffer of residential census block centroid and birth weight outcomes. The study included 106 194 births in New York City. The mean (SD) age of pregnant individuals in the sample was 29.9 (6.1) years. Prevalence of SGA and LGA were 12.9% and 8.4%, respectively. Residence in the highest density quartile of healthy food retail establishments compared with the lowest quartile was associated with lower adjusted risk of SGA (with adjustment for individual covariates including gestational weight gain z-score: risk ratio [RR], 0.89; 95% CI 0.83-0.97). Higher neighborhood density of unhealthy food retail establishments was associated with higher adjusted risk of delivering an infant classified as SGA (fourth vs first quartile: RR, 1.12; 95% CI, 1.01-1.24). The RR for the association between density of unhealthy food retail establishments and risk of LGA was higher after adjustment for all covariates in each quartile compared with quartile 1 (second: RR, 1.12 [95% CI, 1.04-1.20]; third: RR, 1.18 [95% CI, 1.08-1.29]; fourth: RR, 1.16; [95% CI, 1.04-1.29]). There were no associations between neighborhood walkability and birth weight outcomes (SGA for fourth vs first quartile: RR, 1.01 [95% CI, 0.94-1.08]; LGA for fourth vs first quartile: RR, 1.06 [95% CI, 0.98-1.14]). In this population-based cross-sectional study, healthfulness of neighborhood food environments was associated with risk of SGA and LGA. The findings support use of urban design and planning guidelines to improve food environments to support healthy pregnancies and birth weight.

Predicting Confrontation Naming in the Logopenic Variant of Primary Progressive Aphasia

ABSTRACT Background Naming difficulties are prominent and pervasive in the logopenic variant of primary progressive aphasia (lvPPA) and may reflect its underlying deficits in phonological processing. Importantly, some words appear to be more vulnerable to deterioration than others. We hypothesize that these differences can be explained, in part, by the unique psycholinguistic properties of words. Aims Our study investigated the role of psycholinguistic properties of words, along with their underlying psycholinguistic factors, in predicting confrontation naming performance in individuals with lvPPA. Methods & Procedures Naming accuracy data were collected from 14 individuals with lvPPA using the Boston Naming Test (BNT). For each test item, values were extracted for frequency, age of acquisition (AoA), familiarity, word length, phonological neighborhood density (PND), semantic neighborhood density (SND), arousal, and valence. We examined the effects of these psycholinguistic properties on naming accuracy using logistic regression analyses at the individual level and multiple linear regression analysis at the group level. Further, we performed mediation analyses to examine the mediating contributions of related psycholinguistic properties on naming accuracy. Finally, principal component analysis (PCA) was performed to extract fundamental factors of the psycholinguistic properties. These factor scores were entered into regression analyses to investigate their relative contribution to naming accuracy in lvPPA. Outcomes & Results Age of acquisition emerged as the strongest psycholinguistic predictor of naming accuracy in lvPPA at both the individual and group levels. The influence of AoA on naming accuracy was only partially mediated by frequency. Three principal psycholinguistic factors were extracted from the PCA. These were interpreted as “lexical-semantic usage”, “phonological complexity”, and “emotional intensity”. Results indicated that “lexical-semantic usage” and “emotional intensity” predicted naming performance at both the individual and group levels. Conclusions Age of acquisition has been associated with the phonological level in models of word naming. The observed psycholinguistic effects on naming accuracy are discussed in the context of phonological deficits in lvPPA and models of word naming.

Segmentation Clustering Algorithm for Point Cloud Processing of Targets on The Ground

After the ground extraction algorithm, there are multiple objects, including the target in the non-ground point cloud. To effectively extract the features of individual objects and overcome the dependence of the traditional algorithm based on the normal vector and curvature of the region growing point cloud segmentation method on the initial point, the normal vector angle in the neighborhood, and the curvature threshold selection. In this paper, we propose a fast clustering segmentation method for region-growing point clouds based on neighborhood density. The test results show that the point cloud segmentation algorithm proposed in this paper has better performance than the traditional Euclidean clustering point cloud segmentation algorithm and region growing point cloud segmentation algorithm in terms of timeliness and correct segmentation rate. This new segmentation clustering algorithm can quickly implement automatic point cloud clustering segmentation of ground objects.

Sap flow as a function of variables within nested scales: ordinary least squares vs. spatial regression models

Understanding scale-dependent influential drivers of sap flow variability can help managers and policymakers to allocate resources within a particular scale to improve forest health and resiliency against water-stress stimuli such as drought and insects, e.g. bark beetle infestations. We defined a daily measure of sap flow as a function of variables within nested scales of landscape, stand, and tree, using ordinary least squares (OLS), spatial lag and error regression models. Model covariates were elevation, latitude (Y-coordinate), longitude (X-coordinate), neighborhood tree density, tree diameter at breast height, and bark temperature for 40-surveyed Norway spruce (Picea abies) in the Czech Republic, Central Europe. Trees were spatially distributed within 19-established subplots across five plots, with distances ranging 2–9 km, at which variations in soil water potential and temperature were limited. The daily measure of sap flow within the regional scale allowed us to avoid the temporal and spatial variability of climate effects on sap flow. A relatively flat terrain across subplots also allowed us to control the effects of slope, aspect, and topography-related solar incidence angle on sap flow. Sap flow was strongly spatially autocorrelated, so OLS models failed to take spatial autocorrelation into account unless to some extent, depending on the spatial distribution of samples, by including latitude and/or longitude in the models. Among spatial regression models, spatial error models performed better than lag models, allowing to capture the effects of unmeasured independent variables. Sap flow variability for the most part (∼70%) was explained by the landscape-level variable of elevation followed by the stand-level variable of tree density, and the remaining part by variables related to tree characteristics; a nested down-scaling function, defined and visualized for the first time. Therefore, thinning forest stands and future plantations with optimum distances, based on the elevation gradients, may be required to counterbalance the allocation of resources, e.g. water, nutrients, and light, among trees, leading to enhance forest health and resiliency against water-stress stimuli.

Disentangling the Role of Biphone Probability From Neighborhood Density in the Perception of Nonwords.

In six experiments we explored how biphone probability and lexical neighborhood density influence listeners' categorization of vowels embedded in nonword sequences. We found independent effects of each. Listeners shifted categorization of a phonetic continuum to create a higher probability sequence, even when neighborhood density was controlled. Similarly, listeners shifted categorization to create a nonword from a denser neighborhood, even when biphone probability was controlled. Next, using a visual world eye-tracking task, we determined that biphone probability information is used rapidly by listeners in perception. In contrast, task complexity and irrelevant variability in the stimuli interfere with neighborhood density effects. These results support a model in which both biphone probability and neighborhood density independently affect word recognition, but only biphone probability effects are observed early in processing.

Second dialect acquisition and phonetic vowel reduction in the American Midwest

Geographic mobility can lead to the acquisition of new regional dialect features. This second dialect acquisition is highly variable across individuals and is affected by a range of linguistic and social factors. The realization of dialect-specific features is also affected by linguistic variables related to phonetic reduction, but this interaction has been primarily examined with a mix of mobile and non-mobile participants. In the current study, second dialect acquisition by Midwestern American young adults and its interaction with phonetic reduction processes was examined. Relative to lifetime residents of the Northern and Midland regions of American English, some Northern transplants to the Midland region exhibited second dialect acquisition and others exhibited maintenance of Northern dialect features. All talkers showed phonetic reduction due to lexical frequency, phonological neighborhood density, discourse mention, semantic predictability, and speaking style. These phonetic reduction processes only weakly interacted with dialect variation, such that less phonetic reduction was observed overall when it was potentially in conflict with dialect-specific vowel features. Taken together, the results provide additional evidence for substantial individual variation in second dialect acquisition, but limited evidence of an effect of second dialect acquisition on the interaction between dialect variation and phonetic reduction processes.

How Do Persons Found NCRMD and Identified as Indigenous Differ from Other Persons Found NCRMD: Profiles, Trajectories, and Outcomes

Indigenous individuals are vastly over-represented among people incarcerated in Canada. We collected extensive clinical information and outcome data from Review Board (RB) files and obtained lifetime criminal records for 1800 individuals found Not Criminally Responsible on Account of Mental Disorder (NCRMD) in BC (n = 222), ON (n = 484), and QC (n = 1094). Indigenous and non-Indigenous people were compared on (a) socio-demographic, clinical, and criminal histories; (b) index offenses; (c) processing by the RB; and (d) recidivism. Compared to published rates of the disproportionate incarceration of Indigenous people in prisons in Canada (30%), just 3.9% of people in custody with an NCRMD finding were identified as Indigenous. Compared to non-Indigenous people, Indigenous people had higher rates of substance use disorders, personality disorders, and lower rates of mood disorders at verdict and came from low population density neighborhoods but high population density homes. Indigenous individuals were detained in custody longer and remained under supervision longer than non-Indigenous individuals but recidivated at similar rates. Criminal histories, mental health characteristics, and index offenses of Indigenous people found NCRMD were similar to non-Indigenous people found NCRMD. Further research is required to determine if seriously mentally ill Indigenous people who come into contact with the justice system are considered for the NCRMD defense similarly to non-Indigenous people and to explore why Indigenous individuals receive more restrictive dispositions, yet time to reoffending is similar.

An indicator preselection based evolutionary algorithm with auxiliary angle selection for many-objective optimization

Many-objective evolutionary algorithms (MaOEAs) have received significant achievements in recent years. Maintaining a balance between convergence and diversity becomes a key challenge for many-objective evolutionary algorithms when the number of optimization objectives increases. To address this issue, we propose a many-objective evolutionary algorithm using the indicator preselection and auxiliary angle selection (PSEA). In PSEA, a unit vector-based indicator is proposed to pre-select the population region for increasing selection pressure and maintaining diversity simultaneously, which is utilized to identify a promising region in the objective space. Due to the poor quality of individuals outside the promising region, these individuals in the current population can be temporarily discarded. Then, to ensure the diversity of the population, a new strategy based on the second auxiliary angle strategy is designed to calculate the neighborhood density. Finally, in the environmental selection, these strategies are employed for selecting individuals with good convergence and diversity from the candidate set one by one to enter the next generation. The experimental results on commonly used benchmark test problems and many-objective traveling salesman problems with objectives varying from 5 to 20 have demonstrated that PSEA outperforms some state-of-the-art approaches.

Iconicity ratings for 14,000+ English words.

Iconic words and signs are characterized by a perceived resemblance between aspects of their form and aspects of their meaning. For example, in English, iconic words include peep and crash, which mimic the sounds they denote, and wiggle and zigzag, which mimic motion. As a semiotic property of words and signs, iconicity has been demonstrated to play a role in word learning, language processing, and language evolution. This paper presents the results of a large-scale norming study for more than 14,000 English words conducted with over 1400 American English speakers. We demonstrate the utility of these ratings by replicating a number of existing findings showing that iconicity ratings are related to age of acquisition, sensory modality, semantic neighborhood density, structural markedness, and playfulness. We discuss possible use cases and limitations of the rating dataset, which is made publicly available.

Regularity bounded sensor clustering

A non-empirical, unsupervised learning, and graph analytic techniques pipeline investigated the earthquake sensor station proximity (<30 km) and neighborhood density (2–5 sensor) constrained geometric properties of New Zealand’s GeoNet seismic network. They are specific to the Propagation of the Local Undamped Motion (PLUM) earthquake detection algorithm. The experimental design considered 14 cloud and graph clustering processes, with variable combinations of 10–50 km distance and 2–5 sensor constraints. We prove that a clustering process alone is inadequate, and a strict graph r-regularity constraint is essential for forming valid sensor clusters. The goodness of fit measurements revealed HDBSCAN was best for initiating the pipeline. On average 44 % of the sensors fit well within the geometric bounds with a 45 % cluster quality score. With the regularity relaxed, 60 % of the sensors resulted in a 42 % node-degree weighted mean absolute error. The method can be applied to any proximity and density-constrained sensor network.

Exploring changes in residential preference during COVID-19: Implications to contemporary urban planning.

The advent of the COVID-19 pandemic and the subsequent lockdown has reshuffled our daily routines and activity spaces. The home and its immediate environment have attained a critical role in coping with the confinement both as living, working and recreational space. Drawing on a longitudinal survey from greater Oslo, we analyze shifts in residential preferences amidst COVID-19. Given the pandemic induced movement restrictions, we pay special attention to the mediating role of perceived accessibility on the link between several sociodemographic/locational/housing characteristics on preference shifts. Amidst a drop in perceived accessibility, outdoor assets like gardens or balconies, and motility by car and public transport are amongst the residential and locational amenities that have gained most in people's considerations for a new dwelling peri-pandemically. Dwelling satisfaction and preferences with regard to neighborhood density, dwelling size, and work-proximity, have remained largely unchanged despite a clear trend towards more teleworking. Our results can be understood as a nod of approval to contemporary compact urban planning. Nonetheless challenges as to bolstering its resilience still linger and more needs to be done to mitigate the inequalities in (perceived) access to in-home, near-home, and mobility assets that we have also observed to have (re)-immerged in the wake of the pandemic.

Towards the development of energy sharing methodology between different buildings for high efficient neighborhood design

Devising methods to recover waste heat and to share it within a neighborhood can lower the energy footprint of this neighborhood and reduce its dependence on the utility grids. This paper presents a novel methodology for capturing waste energy from individual buildings and sharing the recovered energy with an entire neighborhood. A survey of the current state of research on energy sharing is presented. A mixed-use neighborhood consisting of retail, residential, and greenhouse buildings is modeled to develop a holistic design of energy sharing strategy. Neighborhood-level waste heat recovery and distribution network is modeled in the EnergyPlus-DesignBuilder co-simulation platform. The energy sharing methodology is applied to different neighborhood configurations and neighborhood densities. The results show that with the application of energy sharing methodology integrated with energy efficient design and on-site solar energy, a net reduction of 74% energy in the mixed-use neighborhood is achieved compared to a design complying with the minimum requirement of the applicable energy codes. Similarly, a net 90.46% reduction in greenhouse gas (GHG) emissions is realized by energy sharing and on-site carbon capturing using neighborhood greenhouse.

Lexical Characteristics of the Speech Intelligibility Test: Effects on Transcription Intelligibility for Speakers With Multiple Sclerosis and Parkinson's Disease.

Lexical characteristics of speech stimuli can significantly impact intelligibility. However, lexical characteristics of the widely used Speech Intelligibility Test (SIT) are unknown. We aimed to (a) define variation in neighborhood density, word frequency, grammatical word class, and type-token ratio across a large corpus of SIT sentences and tests and (b) determine the relationship of lexical characteristics to speech intelligibility in speakers with multiple sclerosis (MS), Parkinson's disease (PD), and neurologically healthy controls. Using an extant database of 92 speakers (32 controls, 30 speakers with MS, and 30 speakers with PD), percent correct intelligibility scores were obtained for the SIT. Neighborhood density, word frequency, word class, and type-token ratio were calculated and summed for each of the 11 sentences of each SIT test. The distribution of each characteristic across SIT sentences and tests was examined. Linear mixed-effects models were performed to assess the relationship between intelligibility and the lexical characteristics. There was large variability in the distribution of lexical characteristics across this large corpus of SIT sentences and tests. Modeling revealed a relationship between intelligibility and the lexical characteristics, with word frequency and word class significantly contributing to the model. Three primary findings emerged: (a) There was considerable variability in lexical characteristics both within and across the large corpus of SIT tests; (b) there was not a robust association between intelligibility and the lexical characteristics; and (c) findings from a study demonstrating an effect of neighborhood density and word frequency on intelligibility were replicated. Clinical and research implications of the findings are discussed, and three exemplar SIT tests systematically controlling for neighborhood density and word frequency are provided.

Construct validity of the physical activity neighborhood environment scale-Oman.

This study aims to examine the construct validity of Physical Activity Neighborhoods Environment Scales, Oman (PANES-O), and compare the subjective perceptions with objective measures in Muscat, the capital area of Oman. Walkability index scores using GIS maps were calculated for 35 study areas in Muscat based on which five low and 5 high walkable study areas were randomly selected. A community survey was then conducted in November 2020 in each study area using the 16-item PANES-O instrument to measure the participants' perception of neighborhood density, land use mix, infrastructure, safety, aesthetics, and street connectivity. Due to pandemic restrictions, a social media-based purposive sampling strategy was utilized to reach community-based networks and complete digital data collection. Significant differences between the low and high walkablehigh-walkable neighborhoods were observed for 2 of 3 macroenvironment subscales, density and land use. Respondents in high walkable neighborhoods perceived their areas as having more twin villas (P = 0.001) and apartment buildings (P < 0.001), greater access to destinations (like more shops, and places to go within walking distance; P < 0.001), easy access to public transport (P < 0.001), and more places to be active (P < 0.001); than their counterparts in low walkable neighborhoods. In terms of microenvironmental attributes, respondents in high walkablehigh-walkable neighborhoods perceived their areas to have better infrastructure, better aesthetic qualities, and better social environment than their counterparts in low walkablelow-walkable neighborhoods. Significant differences in perceptions across 12 of the 16-item PANES tool confirmed that 6 of the 7 subscales were significantly sensitive to built environment attributes between the low and high walkable study areas. Respondents in high walkable neighborhoods perceived their areas as having greater access to destinations (like more shops, places to go within walking distance; P ≤ 0.001), easy access to public transport (P ≤ 0.001), more places to be active (P ≤ 0.001), better infrastructure (like more sidewalks, facilities to bicycle; P ≤ 0.001), and better aesthetic qualities (P ≤ 0.001). PANES-O also was able to rate high walkable neighborhoods to be higher in residential density and land-use mix compared to the low walkable neighborhoods demonstrating its sensitivity to the GIS maps' objective measures. These results provide preliminary strong support for the construct validity of PANES-O, suggestingconfirming that it is a promising tool for assessing macroenvironmental perceptions related to physical activity in Oman. Further research using objective measures of microenvironments and device-based physical activity scores is needed to confirm the criterion validity of the 10 micro-environmental attributes of PANES-O using objective measures for the microenvironment. PANES-O could be used to generate and develop the needed evidence on the most appropriate approaches to improving the built environment to promote physical activity and urban planning in Omanthe country.

Cognitive Factors Related to Metaphor Goodness in Poetic and Non-literary Metaphor

ABSTRACT In this paper we examine the effect of two cognitive variables, Semantic Neighborhood Density and Interpretive Diversity, in first, distinguishing between literary (poetic) and nonliterary metaphor, and second, in determining what makes for a good metaphor. Analyses of items taken from a widely used set ofmetaphor norms indicated that while literary and nonliterary metaphor did not differ in many ways, the poetic items tended to 1) contain concepts that came from a more dense semantic space, 2) contain topic and vehicles that came from equally dense semantic space, 3) suggest a greater number of possible interpretations as the topic and vehicle became more semantically dissimilar, and 4) evoke more emergent interpretations (i.e., less likely to be a characteristic of the topic or vehicle when considered separately). In addition, we found one way that the two variables were related to metaphor goodness: better metaphors were those with vehicles that came from increasingly less dense semantic space. This correlation was only reliable for literary, poetic items, presumably because these items were taken from a richer semantic environment suggesting many more alternative possibilities.

Do age and language impairment affect speed of recognition for words with high and low closeness centrality within the phonological network?

Purpose Speed and accuracy of lexical access change with healthy ageing and neurodegeneration. While a word's immediate phonological neighbourhood density (i.e. words differing by a single phoneme) influences access, connectivity to all words in the phonological network (i.e. closeness centrality) may influence processing. This study aimed to investigate the effect of closeness centrality on speed and accuracy of lexical processing pre- and post- a single word-training session in healthy younger and older adults, and adults with logopenic primary progressive aphasia (lvPPA), which affects phonological processing. Method Participants included 29 young and 17 older healthy controls, and 10 adults with lvPPA. Participants received one session of word-training on words with high or low closeness centrality, using a picture-word verification task. Changes in lexical decision reaction times (RT) and accuracy were measured. Result Baseline RT was unaffected by age and accuracy was at ceiling for controls. Post-training, only young adults' RT were significantly faster. Adults with lvPPA were slower and less accurate than controls at baseline, with no training effect. Closeness centrality did not influence performance. Conclusion Absence of training effect for older adults suggests higher threshold to induce priming, possibly associated with insufficient dosage or fatigue. Implications for word-finding interventions with older adults are discussed.

English vowel production conditioned by probabilistic accessibility of words: A comparison between L1 and L2 speakers*

This study investigated the influences of probabilistic accessibility of the word being produced – as determined by its usage frequency and neighborhood density – on native and high-proficiency L2 speakers’ realization of six English monophthong vowels. The native group hyperarticulated the vowels over an expanded acoustic space when the vowel occurred in words with low frequency and high density, supporting the claim that vowel forms are modified in accordance with the probabilistic accessibility of words. However, temporal expansion occurred in words with greater accessibility (i.e., with high frequency and low density) as an effect of low phonotactic probability in low-density words, particularly in attended speech. This suggests that temporal modification in the opposite direction may be part of the phonetic characteristics that are enhanced in communicatively driven focus realization. Conversely, none of these spectral and temporal patterns were found in the L2 group, thereby indicating that even the high-proficiency L2 speakers may not have developed experience-based sensitivity to the modulation of sub-categorical phonetic details indexed with word-level probabilistic information. The results are discussed with respect to how phonological representations are shaped in a word-specific manner for the sake of communicatively driven lexical intelligibility, and what factors may contribute to the lack of native-like sensitivity in L2 speech.

Multilingual spoken word recognition: A megastudy approach

Bilingualism research has primarily focused on the perception and processing of individual sounds or word learning. Many studies have investigated how bilingual listeners perceive sound contrasts that don't create lexical distinctions in their native language. There is substantially less research that has investigated how word-level properties impact L2 auditory processing. The present study examines how auditory lexical processing differs between monolingual and bilingual listeners with different language backgrounds. We collected lexical decision accuracies and latencies for 26,793 words and 9600 pseudowords from the Massive Auditory Lexical Decision database from native and non-native listener responses. We compare the language backgrounds of our 1028 listeners and group them into four groups: native speakers, early, early-late, and late bilinguals. We report the findings of an analysis investigating how language background and proficiency modulate lexical effects to understand how language background influences spoken word recognition. We find differences in the effects of lexical frequency, phonological neighborhood density, and phonological uniqueness point between the different listener groups. We discuss the impact of bilingualism in a word recognition task and how these results inform models of spoken word recognition and second language acquisition.

Fast density estimation for density-based clustering methods

Density-based clustering algorithms are widely used for discovering clusters in pattern recognition and machine learning. They can deal with non-hyperspherical clusters and are robust to outliers. However, the runtime of density-based algorithms is heavily dominated by neighborhood finding and density estimation which is time-consuming. Meanwhile, the traditional acceleration methods using indexing techniques such as KD-tree may not be effective when the dimension of the data increases. To address these issues, this paper proposes a fast range query algorithm, called Fast Principal Component Analysis Pruning (FPCAP), with the help of the fast principal component analysis technique in conjunction with geometric information provided by the principal attributes of the data. Based on FPCAP, a framework for accelerating density-based clustering algorithms is developed and successfully applied to accelerate the Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm and the BLOCK-DBSCAN algorithm, and improved DBSCAN (called IDBSCAN) and improved BLOCK-DBSCAN (called BLOCK-IDBSCAN) are then obtained, respectively. IDBSCAN and BLOCK-IDBSCAN preserve the advantage of DBSCAN and BLOCK-DBSCAN, respectively, while greatly reducing the computation of redundant distances. Experiments on seven benchmark datasets demonstrate that the proposed algorithm improves the computational efficiency significantly.

Refugee secondary migration from small cities: evidence from Utica, New York

ABSTRACT Refugees in the US have increasingly been resettled in small cities, many of which have viewed refugee resettlement as a means of economic development. Using a unique geo-coded data set of almost 600 refugees from over 30 countries resettled in Utica, New York, our study offers primary evidence on the determinants of refugee intentions to migrate from their resettlement community. We find that human capital, co-ethnic enclaves, length of time in the US, and multigenerational integration play a key role in refugees’ intentions to migrate. While the neighbourhood density of refugees has no impact, increases in the density of co-ethnic enclaves have a strong negative effect on the intent to leave in the near future. This effect is particularly strong during refugees’ first five years in the US. Generation 1.5 refugees are significantly more likely to plan to move than those who arrived to their resettlement community at an older age. Our findings highlight the effects of country of origin, education, age, past family migration, and religious affiliation. For small-city governments and resettlement agencies seeking to retain resettled refugees, our results suggest a focus on building stronger co-ethnic enclaves over an emphasis on the diversity of the resettled refugee population.