Individual-specific Preferences Research Articles

Habitat use strategies of African elephants under different seasonal and ecological constraints

Context Habitat selection is a fundamental process that shapes animal spatial ecology. Species with wide geographic distribution that occupy diverse habitats have to adapt their resource acquisition strategies to maximise their effectiveness under local ecological constraints, leading to intraspecific behavioural variability. Identifying environmental determinants of habitat use pattern and regional intraspecific differences advances our understanding of the ecological underpinnings of animal behaviour and is important in strategising effective conservation and management of free-ranging populations. Aims The aim of this study was to assess individual heterogeneity of habitat selection and use by African elephants under different seasonal and ecological constraints, in order to better understand the processes underlying their spatial behaviour. Methods We investigated the habitat selection pattern of 19 African elephants equipped with satellite-linked GPS-collars in two different ecosystems, resource-rich bushveld bordering Kruger National Park, South Africa (six individuals) and arid savannah of Etosha National Park, Namibia (13 individuals). By constructing individual-specific and population-level resource selection functions (RSFs), we examined seasonal differences of elephant habitat use pattern to identify the underlying ecological mechanisms. Key results Elephants were attracted to surface water in both study areas; but when water availability decreased in arid environment, they showed individual-specific preference in using natural vs artificial water sources. Road networks enabled efficient travel among resource patches, but its use differed between individuals. Areas with higher and more predictable vegetation productivity were generally preferred by elephants in dry season, but in more competitive arid savannah system there were individual/group-specific seasonal differences in resource selection patterns, likely reflecting the social dynamics among individuals. At population-level, the habitat selection pattern was less apparent due to considerable intra-population variability. Conclusions The substantial differences in model coefficients within and between our study populations demonstrate the spatio-behavioural plasticity of elephants under various environmental conditions and suggest that population-level RSFs may over-simplify elephants’ socio-ecological complexity. Implications Within the resource competition paradigm, individual-specific habitat selection may be as essential in maintaining population resilience as is the population-level pattern of resource use. Consequently, spatio-behavioural heterogeneity within and between populations should be accounted for in resource selection studies and all subsequent conservation management policies.

Do employers’ offers of paid maternity leave increase women’s job satisfaction?

ABSTRACT The importance of non-wage compensation like paid maternity leave may vary across workers in that the contribution of paid maternity leave to job satisfaction is more discernible for women of childbearing age than for men and older women. Using data from the National Longitudinal Survey of Youth (NLSY97), this study investigates the relationship between paid maternity leave offers and job satisfaction for women of childbearing age. However, individual-specific preferences for benefits may contribute to job satisfaction and having access to paid maternity leave. This study, therefore, uses an instrumental variable approach to address the presence of endogeneity in the relationship between paid maternity leave and job satisfaction. After instrumenting paid maternity leave variable appropriately, results from the bivariate probit model show that paid maternity leave offers positively contribute to women’s job satisfaction; women with paid maternity leave offers are more likely to be satisfied with their jobs than those with unpaid or no maternity leave offers. Furthermore, the findings remain consistent after conditioning the sample age at and above 24 years and firm size between 20 to 500 employees, respectively.

Heterogeneity in the WTA-WTP disparity for irrigation water reliability

This paper assesses the WTP and WTA for improvements and deteriorations, respectively, in irrigation water supply reliability. The assessment relies on a double-sided discrete choice experiment valuation using latent-class modeling accounting for preference and scale heterogeneity. This valuation approach is empirically implemented using a case study of a Spanish irrigated district significantly impacted by climate change. The results obtained show individual-specific preference heterogeneity in the WTA-WTP disparity, primarily driven by the different impacts of water reliability on farmers' utility (changes in business revenues and costs, uncertainty in business performance, and farm income effects) and interindividual differences in loss aversion (different degrees of endowment effect). Additionally, the significant scale heterogeneity and ordering effects found suggest that it may be advisable to use modeling approaches that account for them. Several policy-relevant implications can be drawn, including the non-neutrality of the initial allocation of property rights, repercussions on the cost-benefit of climate change adaptation measures, and the need to account for irrigators’ preference heterogeneity in order to design successful market-based instruments.

Is female mate choice repeatable across males with nearly identical songs?

Individual-specific mate preferences are thought to be widespread, but they are still poorly understood in terms of mechanisms and function. Earlier work on a songbird (the zebra finch, Taeniopygia guttata ) showed predominantly individual-specific mate preferences and signs of behavioural incompatibility with certain partners. However, the phenotypic target of preference or aversion remains unclear. A previous study suggested that female preferences may be related to variation in male song. Male zebra finches have individually distinct songs and together with individual-specific female song preferences they could function according to a ‘key–lock principle’. Here we report on a preregistered study in which we tested the hypothesis that individual females respond to males with highly similar songs in a similar way. We selected 18 duos of males with nearly identical songs and released them with females in communal breeding aviaries. We scored female preference as the female's responsiveness to male courtship and confirmed that females indeed had strong individual-specific mate preferences. However, female responsiveness scores were not positively correlated across males closely matched for their song, thereby rejecting the idea of a song-based ‘key–lock principle’ as an explanation for the observed individual-specific preferences. We suggest researchers should focus on details of the interactions between potential partners to elucidate the cause of behavioural incompatibility. • Individual-specific mate preferences are common but poorly understood. • Mate preferences differ strongly between individual female zebra finches. • Individual males differ strikingly in their songs. • Females respond differently to males with exceptionally similar songs. • Song traits are hence not the target of individual-specific preferences.

How Well Targeted Are Soda Taxes?

Soda taxes aim to reduce excessive sugar consumption. We assess who is most impacted by soda taxes. We estimate demand using micro longitudinal data covering on-the-go purchases, and exploit the panel dimension to estimate individual-specific preferences. We relate these preferences and counterfactual predictions to individual characteristics and show that soda taxes are relatively effective at targeting the sugar intake of the young, are less successful at targeting the intake of those with high total dietary sugar, and are unlikely to be strongly regressive especially if consumers benefit from averted internalities. (JEL D12, H22, H25, H71)

Investigating personalized exit choice behavior in fire accidents using the hierarchical Bayes estimator of the random coefficient logit model

Understanding how people behave during the fire evacuation is fundamental to predict the required time to evacuate buildings and transportation systems and enhance their safety. To date, several studies have been carried out to investigate the impact of several social and environmental factors on how evacuees choose exits while moving towards safe places. However, none of the existing studies has identified individual-specific taste preferences for exit choice behavior. As such, they do not allow the analyses of individual-specific preferences across heterogeneous decision makers. This work overcomes this limitation through a Hierarchical Bayes estimation approach, which allows the estimation of individual-specific parameters. The methodology is applied to the data of non-immersive Virtual Reality-based stated preference experiments. With the individual-level knowledge, analyses show more granular insights of decision makers’ trade-off of different factors, including two distinct groups of people as “followers” or “non-followers” and the impact of age, nationality, and education on the herding behavior during the fire evacuation.

Individual-specific point and interval conditional estimates of latent class logit parameters

Within the realm of logit-type random parameter models to address unobserved heterogeneity in preferences there are two dominant approaches: the mixed logit model, which assumes parametric and continuous heterogeneity distributions, and the latent class logit model, which is a discrete and semiparametric counterpart of mixed logit. In addition to offer flexibility benefits, random parameter models allow researchers to make conditional (posterior) inference on preference parameters at the individual-specific level. In this paper we extend the individual-specific experimental approach, that was conducted by Revelt and Train (2000) for the continuous heterogeneity distributions of a mixed logit, to the discrete case of the latent class logit model. Our Monte Carlo study results confirm the expectation that for a given number of individuals, the density of the conditional means converges to the conditional population as the number of choice situations increases. We also add to the analysis the behavior of interval estimates using two methods for the derivation of standard errors of the individual-specific estimates. In general, as we have more information of the choices made by the individuals, we are in better shape to identify individual-specific preferences. Our main conclusion is that accurate individual-specific estimation is possible – including correct assignment to classes, but a large number of choice situations is needed to correctly approximate the true underlying distribution.

Individual-Specific Point and Interval Conditional Estimates of Latent Class Logit Parameters

Within the realm of logit-type random parameter models to address unobserved heterogeneity in preferences there are two dominant approaches: the mixed logit model, which assumes parametric and continuous heterogeneity distributions, and the latent class logit model, which is a discrete and semiparametric counterpart of mixed logit. In addition to offer flexibility benefits, random parameter models allow researchers to make conditional (posterior) inference on preference parameters at the individual-specific level. In this paper we extend the individual-specific experimental approach, that was conducted by for the continuous heterogeneity distributions of a mixed logit, to the discrete case of the latent class logit model. Our Monte Carlo study results confirm the expectation that for a given number of individuals, the density of the conditional means converges to the conditional population as the number of choice situations increases. We also add to the analysis the behavior of interval estimates using two methods for the derivation of standard errors of the individual-specific estimates. In general, as we have more information of the choices made by the individuals, we are in better shape to identify individual-specific preferences. Our main conclusion is that accurate individual-specific estimation is possible – including correct assignment to classes, but a large number of choice situations is needed to correctly approximate the true underlying distribution.

Testing the External Validity of a Discrete Choice Experiment Method: An Application to Latent Tuberculosis Infection Treatment.

To explore the external validity and predictive power of stated preferences obtained from a discrete choice experiment (DCE) by comparing the predicted behavior of respondents to their actual choices at an individual level. A DCE was performed in patients before being offered treatment for latent tuberculosis infection. A mixed logit model was estimated using hierarchical Bayes. The individual-specific preference coefficients were used to calculate the expected probability of choosing the treatment by each patient. The predicted choice using this probability was compared with their actual decision. We used a receiver-operating characteristic curve and different thresholds to convert probabilities into the predicted choices. The comparability of different distributions for the random parameters was also examined. Our results identified significant heterogeneity in preferences for all attributes among respondents. The best model correctly predicted actual treatment decisions for 83% of the participants. The results from using different thresholds and a receiver-operating characteristic curve also confirmed the compatibility between predicted and actual choices. We showed that individual-specific coefficients reflected respondents' actual choices more closely compared with the aggregate-level estimates. The results of this study provided support for the external validity of DCEs on the basis of their power to predict actual behavior in this setting. Future investigations are, however, required to establish the external validity of DCEs in different settings.

Consumer Preference Elicitation of Complex Products Using Fuzzy Support Vector Machine Active Learning

As technology advances, new products (e.g., digital cameras, computer tablets, etc.) have become increasingly more complex. Researchers often face considerable challenges in understanding consumers’ preferences for such products. This paper proposes an adaptive decompositional framework to elicit consumers’ preferences for complex products. The proposed method starts with collaborative-filtered initial part-worths, followed by an adaptive question selection process that uses a fuzzy support vector machine active learning algorithm to adaptively refine the individual-specific preference estimate after each question. Our empirical and synthetic studies suggest that the proposed method performs well for product categories equipped with as many as 70 to 100 attribute levels, which is typically considered prohibitive for decompositional preference elicitation methods. In addition, we demonstrate that the proposed method provides a natural remedy for a long-standing challenge in adaptive question design by gauging the possibility of response errors on the fly and incorporating the results into the survey design. This research also explores in a live setting how responses from previous respondents may be used to facilitate active learning of the focal respondent’s product preferences. Overall, the proposed approach offers new capabilities that complement existing preference elicitation methods, particularly in the context of complex products. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mksc.2015.0946 .

Spatial preference heterogeneity in forest recreation

In this study, we analyze the preferences for recreational use of forests in Lorraine (Northeastern France), applying stated preference data. Our approach allows us to estimate individual-specific preferences for recreational use of different forest types. These estimates are used in a second stage of the analysis where we test whether preferences depend on access to recreation sites. We find that there is significant preference heterogeneity with respect to most forest attributes. The spatial analysis shows that preferences for forests with parking and picnic facilities are correlated with having access to such forests while for the other attributes considered (dominant tree species, trekking paths and presence of lake and rivers) we find no correlation between stated preferences and accessibility.This implies that the problem of endogenous distances in the travel cost method (Parsons, 1991) may be present in the estimation of welfare economic values for parking and picnic facilities in the analyzed model. The results underline the importance of considering spatial heterogeneity of preferences carrying out economic valuation of spatial-delineated environmental goods and that the spatial variation in willingness to pay for such goods is not only explained by the users' transport costs of accessing the sites.

Combining Individual-Level Discrete Choice Experiment Estimates and Costs to Inform Health Care Management Decisions about Customized Care: The Case of Follow-Up Strategies after Breast Cancer Treatment

ObjectiveCustomized care can be beneficial for patients when preferences for health care programs are heterogeneous. Yet, there is little guidance on how individual-specific preferences and cost data can be combined to inform health care decisions about customized care. Therefore, we propose a discrete choice experiment–based approach that illustrates how to analyze the cost-effectiveness of customized (and noncustomized) care programs to provide information for hospital managers. MethodsWe exploit the fact that choice models make it possible to determine whether preference heterogeneity exists and to obtain individual-specific parameter estimates. We present an approach of how to combine these individual-specific parameter estimates from a random parameter model (mixed logit model) with cost data to analyze the cost-effectiveness of customized care and demonstrate our method in the case of follow-up after breast cancer treatment. ResultsWe found that there is significant preference heterogeneity for all except two attributes of breast cancer treatment follow-up and that the fully customized care program leads to higher utility and lower costs than the current standardized program. Compared with the single alternative program, the fully customized care program has increased benefits and higher costs. Thus, it is necessary for health care decision makers to judge whether the use of resources for customized care is cost-effective. ConclusionsDecision makers should consider using the results obtained from our methodological approach when they consider implementing customized health care programs, because it may help to find ways to save costs and increase patient satisfaction.

Activity Travel Planning and Rescheduling Behavior

Travel behavior research of the past few years has shown an increasing interest in dynamic activity scheduling. Research was done to analyze factors influencing the actual activity scheduling process by using detailed activity travel data from an extensive data set collected in Flanders (Belgium). A model examines the attributes that influence activity planning; a second model analyzes the factors that affect activity rescheduling. The explanatory variables considered in both models are individual, household, activity, and schedule attributes, and their impact is analyzed by using mixed logit models. Random effects are added to the models to test for within-individual variance. The results reveal that activity and schedule characteristics considerably affect activity planning. The rescheduling model also has several highly significant activity and schedule attributes. Most individual and household attributes considered do not influence activity planning and rescheduling behavior, although strong statistical evidence indicates that individual-specific preferences for planning and rescheduling determine the process of activity scheduling.

Recovery Preference Exploration

To present a new assessment approach, referred to as recovery preference exploration (RPE), for exploring the personal meaning of functional loss and recovery. RPE determines how people would choose to recover from profound disability if they could control that recovery. Twenty-six patients with a variety of medical conditions and one or more limitations in the functions being addressed were recruited from an inner-city ambulatory care clinic. The patients imagined recovery from 15 functional limitations, beginning with severe problems in all functions and ending with complete independence. Individual-specific preferences for recovery in each function were calculated relative to the other 14 and were submitted for principal components analyses. Imagined difficulty in toileting and with depression were most troubling. Principal components analyses identified trade-off choices among domains of physical, psychological, and social functioning. Some people valued physical independence above psychological well-being or social abilities. Others had opposite patterns. Patients' narrative explanations, when triangulated, were consistent with their preferences. RPE makes visible the highly personal nature of feelings about ability and disability. Our results may help guide the selection of rehabilitation interventions in ambulatory care.

Memory decay and cache site preferences in hoarding coal tits — a laboratory study

Summary Many animals use hoarding as a long-term strategy to ensure a food supply at times of shortage. This study explores the respective roles of memory and site preferences in cache recovery by coal tits (Periparus ater). We compared the retrieval accuracy or foraging efficiency of the cacher itself (‘caching coal tit’), a naive conspecific (‘pilfering coal tit’) and a non-hoarding heterospecific (‘pilfering great tit’; Parus major) after six different retention intervals. Our experiment shows that the persistence of the coal tits’ memory is up to 4 weeks in the laboratory. Species specific storage and foraging site preferences enhance cache recovery after longer intervals. We find no evidence for individual-specific preferences. Pilfering great tits are capable of learning the coal tits’ hoarding preferences. This ability may affect coal tits’ hoarding behaviour in more natural conditions as they frequently forage in mixed flocks with great tits.