Latent Variable Model Research Articles

Urban Air Mobility Passengers’ Profiling: Evidence from Milan Airports, Italy

Urban air mobility (UAM) is expected to offer new travel options for passengers to and from airports in the near future, despite uncertainties associated with regulatory issues, environmental concerns, and societal impacts. This paper analyzes the socioeconomic and behavioral factors that could influence users’ modal choices (including UAM services) for accessing and egressing airports. Using revealed and stated preference data collected at Milan airports (Italy), mixed logit and hybrid choice modeling specifications are estimated and compared with profile potential UAM passengers. Our findings suggest that the level of service, socioeconomic factors, and trip-related variables explain passengers’ choices better than latent traits such as fear of flying, propensity for technological advances, and expectations about the safety of UAM services. In other words, the additional complexity of hybrid choice modeling is not justified by the slight gain in likelihood compared with the estimated mixed logit model. The results also indicate that high-income individuals traveling for business purposes are the most likely demand segment to use UAM services, at least initially. Moreover, highly educated individuals and employees who have their travel expenses reimbursed for work trips are less likely to choose UAM services for airport access or egress, preferring traditional ground taxis.

Development and validation of a measure for father involvement during early childhood in a resource-limited context

BackgroundGlobally, there is mounting evidence about the importance of father involvement for improving child and family wellbeing. However, there are limited measurement tools available for assessing father involvement in low-resource settings globally. The objective of this study was to develop and validate a scale for measuring fathers’ engagement in caregiving activities for young children and families in rural Kenya.MethodsWe used baseline data collected in February–March 2023 from a cluster-randomized controlled trial of a parenting intervention aimed at improving early child development in Nyamira and Vihiga counties in Western Kenya. The analytic sample comprised 460 primary caregivers of children under 18 months of age (91.3% mothers) who were in a relationship with a male caregiver of the young child (i.e., father). The primary caregiver reported on the 25-item father involvement scale. We conducted exploratory and confirmatory factor analyses to determine the dimensionality of the measure and estimated latent variable regression models to assess construct and predictive validity in terms of associations with sociodemographic factors, other paternal and maternal psychosocial measures, and early child development outcomes.ResultsThe scale had a four-factor structure representing father involvement in (1) childcare activities, (2) play and affection, (3) household chores, and (4) early learning activities. Each factor showed good internal consistency reliability and evidence of multiple forms of validity. The four factors were associated with child and household sociodemographic characteristics and more positive paternal and maternal psychosocial outcomes. Father involvement across all four domains was associated with higher early child cognitive, language, motor, and socioemotional development scores.ConclusionsOur findings highlight the multidimensional nature of fathers’ caregiving roles in the lives of young children and their families. We establish a valid tool for assessing father involvement in rural Western Kenya that has potential for use in other similar contexts.

On the Selection of Workplace in the COVID-19 Era Using Attitudinal Factors

After two years of living with the threat of COVID-19 in Ontario, Canada, pre-pandemic circumstances returned. During the pandemic, we relied on ICT-based tools to carry out our daily tasks, and now we have reached a tipping point. Should we keep our new routines to benefit us in the future? Or should we return to our routines before the pandemic? This study utilized a travel survey to examine the impact of the COVID-19 pandemic on work activity-travel behavior and the persistence of new traditions in the post-pandemic era. The data for this study comes from a sample of 1,000 Greater Toronto Area residents who participated in a web-based survey in July, 2021, when Ontario began the third phase of reopening. This paper investigates work activity and workplace selection in more depth. Using factor analysis, a collection of latent attitudinal variables was identified. An integrated choice and latent variable model estimated the influential systematic and latent variables on the perceived workplace selection in the post-pandemic condition. The results showed that 71% of responders want to continue working from home at least once weekly after the pandemic. The preferred frequency of telecommuting in the post-pandemic period was positively correlated with education level, positive feelings about telecommuting experience, and certain occupation types, while negatively associated with age.

A comparison of joint species distribution models for percent cover data

Abstract Joint species distribution models (JSDMs) have gained considerable traction among ecologists over the past decade, due to their capacity to answer a wide range of questions at both the species‐ and the community‐level. The family of generalised linear latent variable models in particular has proven popular for building JSDMs, being able to handle many response types including presence‐absence data, biomass, overdispersed and/or zero‐inflated counts. We extend latent variable models to handle percent cover response variables, with vegetation, sessile invertebrate and macroalgal cover data representing the prime examples of such data arising in community ecology. Sparsity is a commonly encountered challenge with percent cover data. Responses are typically recorded as percentages covered per plot, though some species may be completely absent or present, that is, have 0% or 100% cover, respectively, rendering the use of beta distribution inadequate. We propose two JSDMs suitable for percent cover data, namely a hurdle beta model and an ordered beta model. We compare the two proposed approaches to a beta distribution for shifted responses, transformed presence‐absence data and an ordinal model for percent cover classes. Results demonstrate the hurdle beta JSDM was generally the most accurate at retrieving the latent variables and predicting ecological percent cover data.

Costs and benefits of item reduction: The abridgment of the Emotion Regulation Questionnaire (ERQ)

Shortening existing instruments is a highly required procedure, as short scales may have several advantages over the long versions, especially in time and/or resources restrictions. However, abbreviated forms may be weaker than their parent versions from both content coverage and psychometric robustness. Also, the abridgment of instruments is often lacking in methodological strictness, and the potential drawbacks of the shortened scales are rarely reported. The current study aims to describe the whole process of scale shortening, emphasizing the potential costs and benefits, in terms of balance between time-resource savings and loss of validity and reliability. We shortened the Emotion Regulation Questionnaire (ERQ), involving a sample of 459 participants (53.2% males). Item reduction was driven by searching to preserve the content breadth of the construct and scale's psychometric quality. Our results supported the two-factor structure of the measure (Cognitive Reappraisal and Expressive Suppression), χ2(8) = 11.357 ns, CFI = 0.995, TLI = 0.990, RMSEA = 0.030 (0.000–0.067), SRMR = 0.031, and three items were selected for each subscale. The two intended factors showed good levels of reliability (α > 0.710). A latent variable model was performed to evaluate how the original ERQ and our proposed short version (ERQS) were related to depression, anxiety, and stress: A similar pattern of associations was found, with Cognitive Reappraisal (negatively) and Expressive Suppression (positively) reporting significant but weak associations. The ERQ-S can be beneficial over the original version, as it effectively assesses the two emotion regulation strategies with a trivial loss in reliability and predictive validity.

A Note on the Asymptotic Posterior Normality of Multivariate Latent Traits in an IRT Model for Polytomous Items of Mixed Format

The asymptotic posterior normality (APN) of the latent vector in an item response theory model is an important argument in modeling and inference. For a single latent trait, Chang and Stout proved its APN for binary items under general conditions, which generalized Chang for polytomous data and Kornely and Kateri for multivariate latent traits (MLT) and binary items. As MLT and polytomous items are nowadays common in psychometry, an APN-theory covering both simultaneously remains an open and ongoing problem. We generalize the APN-theory accordingly, providing thus a broader foundation for developments relying on APN. We also prove consistency of common estimators for the MLT, extending the according results of Chang and Kornely and Kateri.

The Improved EMS Algorithm for Latent Variable Selection in M3PL Model

One of the main concerns in multidimensional item response theory (MIRT) is to detect the relationship between items and latent traits, which can be treated as a latent variable selection problem. An attractive method for latent variable selection in multidimensional 2-parameter logistic (M2PL) model is to minimize the observed Bayesian information criterion (BIC) by the expectation model selection (EMS) algorithm. The EMS algorithm extends the EM algorithm and allows the updates of the model (e.g., the loading structure in MIRT) in the iterations along with the parameters under the model. As an extension of the M2PL model, the multidimensional 3-parameter logistic (M3PL) model introduces an additional guessing parameter which makes the latent variable selection more challenging. In this paper, a well-designed EMS algorithm, named improved EMS (IEMS), is proposed to accurately and efficiently detect the underlying true loading structure in the M3PL model, which also works for the M2PL model. In simulation studies, we compare the IEMS algorithm with several state-of-art methods and the IEMS is of competitiveness in terms of model recovery, estimation precision, and computational efficiency. The IEMS algorithm is illustrated by its application to two real data sets.

Balancing the scales: assessing the impact of irrigation and pathogen burden on potato blackleg disease and soil microbial communities

BackgroundUnderstanding the interaction between environmental conditions, crop yields, and soil health is crucial for sustainable agriculture in a changing climate. Management practices to limit disease are a balancing act. For example, in potato production, dry conditions favour common scab (Streptomyces spp.) and wet conditions favour blackleg disease (Pectobacterium spp.). The exact mechanisms involved and how these link to changes in the soil microbiome are unclear. Our objectives were to test how irrigation management and bacterial pathogen load in potato seed stocks impact: (i) crop yields; (ii) disease development (blackleg or common scab); and (iii) soil microbial community dynamics.MethodsWe used stocks of seed potatoes with varying natural levels of Pectobacterium (Jelly [high load], Jelly [low load] and Estima [Zero – no Pectobacterium]). Stocks were grown under four irrigation regimes that differed in the timing and level of watering. The soil microbial communities were profiled using amplicon sequencing at 50% plant emergence and at harvest. Generalised linear latent variable models and an annotation-free mathematical framework approach (ensemble quotient analysis) were then used to show the interacting microbes with irrigation regime and Pectobacterium pathogen levels.ResultsIrrigation increased blackleg symptoms in the plots planted with stocks with low and high levels of Pectobacterium (22–34%) but not in the zero stock (2–6%). However, withholding irrigation increased common scab symptoms (2–5%) and reduced crop yields. Irrigation did not impact the composition of the soil microbiome, but planting stock with a high Pectobacterium burden resulted in an increased abundance of Planctomycetota, Anaerolinea and Acidobacteria species within the microbiome. Ensemble quotient analysis highlighted the Anaerolinea taxa were highly associated with high levels of Pectobacterium in the seed stock and blackleg symptoms in the field.ConclusionsWe conclude that planting seed stocks with a high Pectobacterium burden alters the abundance of specific microbial species within the soil microbiome and suggest that managing pathogen load in seed stocks could substantially affect soil communities, affecting crop health and productivity.BWvw4eepooU5poSUC2MUy8Video

Evidence for Greater Marking along Ethnic Boundaries.

The coordination of beliefs, norms, and behaviors is foundational to theories of group formation. However, because beliefs and norms are not directly observable, signaling mechanisms are required to build reliable signals of latent traits. Although the mathematical theory behind these signals is robust, there is very little testing of ethnic marker theory or of its key propositions that markers become more prevalent along ethnic boundaries and where more than two cultural groups are in contact. We present an ethnographic test of this theory with phonetic differences serving as potential group signals. The data derive from an ethnographic and linguistic investigation in two contrasting secondary school settings in Utah, one that is majority European American and one that is ethnically more diverse. Word list recordings were collected as part of interviews with teens from different backgrounds. We extracted acoustic data from the speech of European Americans (EAs) and Pacific Islanders (PIs), then analyzed differences in the pronunciation of the vowel in words such as "bit." We found evidence of greater phonetic marking at the more diverse school, along more prominent boundaries of ethnic interaction. These results align with predictions made by the theoretical models. This initial empirical test of model predictions provides justification for the development of more complex models that could account for more variables within the environment.

The structured assessment of violence risk in youth demonstrates measurement invariance between Black and White justice-referred youths.

Identification and implementation of effective methods for reducing racial/ethnic bias and disparities in legal settings are paramount in the United States and other countries. One procedure originally thought to reduce bias in legal decisions is the use of risk assessment instruments, which is now being heavily scrutinized. Measurement invariance, a latent trait technique, is a robust method for assessing one form of bias. Measurement invariance involves determining whether risk items in an instrument appear to be functioning the same between racial or other groups. Thus, the present study examined measurement invariance of the Structured Assessment of Violence Risk in Youth (SAVRY) between non-Latino Black and White youths to examine racial bias. We expected the SAVRY to be invariant (lacking measurement bias) between Black and White youths. The sample included 687 Black and 361 White youths, and the study used a large, multistate data set of SAVRYs conducted by probation officers. We conducted measurement invariance testing in a series of hierarchical steps including testing configural and scalar invariance. The SAVRY demonstrated scalar invariance (equal thresholds for ratings from "low" to "moderate" and "moderate" to "high") for all items except one-community disorganization. The findings lend further credibility to the SAVRY, and the structured professional judgment approach, as a method to assess violence risk and case planning needs among youths involved in the legal system. These findings provide more confidence that significant differences in SAVRY risk level or items between Black and White youths are not based in measurement bias, with the exception of the community disorganization item. Potential fixes discussed include eliminating or deemphasizing this item in final risk level ratings. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Modelling nonlinear moderation effects with local structural equation modelling (LSEM): A non-technical introduction.

Understanding the differential strength of effects in the presence of a third variable, known as a moderation effect, is a common research goal in many psychological and behavioural science fields. If structural equation modelling is applied to test effects of interest, the investigation of differential strength of effects will typically ask how parameters of a latent variable model are influenced by categorical or continuous moderators, such as age, socio-economic status, personality traits, etc. Traditional approaches to continuous moderators in SEMs predominantly address linear moderation effects, risking the oversight of nonlinear effects. Moreover, some approaches have methodological limitations, for example, the need to categorise moderators or to pre-specify parametric forms of moderation. This tutorial introduces local structural equation modelling (LSEM) in a non-technical way. LSEM is a nonparametric approach that allows the analysis of nonlinear moderation effects without the above-mentioned limitations. Using an empirical dataset, we demonstrate the implementation of LSEM through the R-sirt package, emphasising its versatility in both exploratory analysis of nonlinear moderation without prior knowledge and confirmatory testing of hypothesised moderation functions. The tutorial also addresses common modelling issues and extends the discussion to different application scenarios, demonstrating its flexibility.

Capturing autonomy, competence, and relatedness at work: further examining and validating an English language version of the work-related basic need satisfaction scale

Self-Determination Theory (SDT) maintains that the satisfaction of the basic psychological needs for autonomy, competence, and relatedness is associated with optimal individual functioning, including in the workplace. A self-report instrument, the Work-related Basic Need Satisfaction scale (W-BNS), has previously been developed and validated in Dutch and Italian. We aimed to validate an English version of the W-BNS. We also evaluated a bifactor model to assess the extent to which the item responses could be explained by a single, underlying general latent trait. A Rasch calibration was also carried out to assess the extent to which the scores conformed to the assumptions of ‘fundamental measurement’ and could be converted to a common metric. We used data from 141 staff employed by a large UK-based mental health service provider. The postulated three-factor structure provided a good fit to the data. However, a bifactor model, introducing an underlying general factor, provided a superior fit. The items generally conformed to the Rasch measurement model. Evidence of convergent/divergent validity was observed via the correlations between the W-BNS scores and those for the Basic Psychological Needs Satisfaction and Frustration Scale (BPNFS). Regarding construct validity, both the separate needs and total W-BNS scores statistically significantly predicted an individual’s reported intention to leave the current employer. Our findings foster research with the WBNS and have implications for how the W-BNS is optimally implemented in practice as a useful brief tool for assessing staff work-related need satisfaction.

Representations of the intrinsic value of information in mouse orbitofrontal cortex.

Animals are motivated to seek information that does not influence reward outcomes, suggesting that information has intrinsic value. We have developed an odor-based information seeking task that reveals that mice choose to receive information even though it does not alter the reward outcome. Moreover, mice are willing to pay for information by sacrificing water reward, suggesting that information is of intrinsic value to a mouse. We used a microendoscope to reveal neural activity in orbitofrontal cortex (OFC) while mice learned the information seeking task. We observed the emergence of distinct populations of neurons responsive to odors predictive of information and odors predictive of water reward. A latent variable model recapitulated these different representations in the low-dimensional dynamics of OFC neuronal population activity. These data suggest that mice have evolved separate pathways to represent the intrinsic value of information and the extrinsic value of water reward. Thus, the desire to acquire knowledge is observed in mice, and the value of this information is represented in the OFC. The mouse now provides a facile experimental system to study the representation of the value of information, a higher cognitive variable.

On a general theoretical framework of reliability.

Reliability is an essential measure of how closely observed scores represent latent scores (reflecting constructs), assuming some latent variable measurement model. We present a general theoretical framework of reliability, placing emphasis on measuring the association between latent and observed scores. This framework was inspired by McDonald's (Psychometrika, 76, 511) regression framework, which highlighted the coefficient of determination as a measure of reliability. We extend McDonald's (Psychometrika, 76, 511) framework beyond coefficients of determination and introduce four desiderata for reliability measures (estimability, normalization, symmetry, and invariance). We also present theoretical examples to illustrate distinct measures of reliability and report on a numerical study that demonstrates the behaviour of different reliability measures. We conclude with a discussion on the use of reliability coefficients and outline future avenues of research.

Entrepreneurial potential in digital transformation: scale development using item response theory

PurposeThis study aims to develop a scale using item response theory (IRT) to assess the entrepreneurial potential for digital transformation in Brazilian companies.Design/methodology/approachIRT was used to develop a scale for entrepreneurial potential in digital transformation. This scale was constructed from a questionnaire, covering the domains: (1) data-driven culture, (2) openness to knowledge and adaptation to change, (3) connectivity and (4) creativity and innovation. The questionnaire was administered to a sample of 216 entrepreneurs from small business enterprise (SBE) and startups in Brazil.FindingsA questionnaire was developed and validated to assess the latent trait of entrepreneurial potential within the context of digital transformation. Additionally, a three-level scale of entrepreneurial potential was established: low (level I), intermediate (level II) and high (level III). The interpretation of this scale provides valuable information on which domains, such as data-driven culture, innovation, among others, can be enhanced to improve the potential of entrepreneur for digital transformation.Research limitations/implicationsThe sample was limited to small Brazilian companies and startups, which may restrict the applicability of the results to other business or geographic contexts. Additionally, the items evaluated in the scale may not fully capture all nuances of entrepreneurial potential for digital transformation. Future research should consider including new items that cover a broader range of entrepreneurial characteristics.Practical implicationsThe findings of this study have significant practical implications for the Brazilian entrepreneurial ecosystem, the entrepreneurs themselves, public policy makers and entrepreneurship support institutions. These results can guide digital transformation strategies, adjustments in public policies and investments, thereby promoting economic development and innovation in the country.Originality/valueThis study stands out for using IRT as a robust methodology to develop an interpretative scale to assess entrepreneurial potential in the digital transformation era. By focusing on Brazilian SBEs and startups, the study offers an original contribution on how these companies are handling the challenges of digitalization and identifying areas for improvement to further promote digital transformation among entrepreneurs.

Psychometric evaluation of the Adelphi Adherence Questionnaire (ADAQ©) in adults with osteoarthritis

BackgroundMedication non-adherence is a common issue in chronic illness. The World Health Organization has recognized a need for a valid and reliable method of measuring adherence to understand and mitigate non-adherence. This study aimed to psychometrically evaluate the English version of the Adelphi Adherence Questionnaire (ADAQ©), a questionnaire designed to assess patient-reported medication adherence across multiple therapy areas, in patients with Osteoarthritis (OA).MethodologyData from the Adelphi OA Disease Specific Programme™, a survey of physicians and their consulting adult patients with OA conducted in the United States, November 2020 to March 2021, was used to assess the psychometric properties of the ADAQ. Patients completed the ADAQ, Adherence to Refills and Medication Scale (ARMS), Western Ontario and McMaster Universities Arthritis Index (WOMAC), and EQ-5D-3L. The measurement model of the 13-item ADAQ was assessed and refined using latent variable modelling (Multiple Indicator Multiple Cause, confirmatory and exploratory factor analyses, item response theory, Mokken scaling, and bifactor analyses). Correlational analyses (Spearman’s rank and polyserial as appropriate) with ARMS, WOMAC, and EQ-5D-3L scores assessed construct validity. Anchor- and distribution-based analyses were performed to estimate between-group clinically important differences (CID).ResultsOverall, 723 patients were included in this analysis (54.5% female, 69.0% aged ≥ 60). Latent variable modelling indicated a unidimensional reflective model was appropriate, with a bifactor model confirming an 11-item essentially unidimensional score. Items 12 and 13 were excluded from scoring as they measured a different concept. The ADAQ had high internal reliability with omega hierarchical and Cronbach’s alpha coefficients of 0.89 and 0.97, respectively. Convergent validity was supported by moderate correlations with items of the ARMS, and physician-reported adherence and compliance. Mean differences in ADAQ score between high and low adherence groups yielded CID estimates between 0.49 and 1.05 points, with a correlation-weighted average of 0.81 points.ConclusionThis scoring model showed strong construct validity and internal consistency reliability when assessing medication adherence in OA. Future work should focus on confirming validity across a range of disease areas.

Item response theory-based continuous test norming.

In norm-referenced psychological testing, an individual's performance is expressed in relation to a reference population using a standardized score, like an intelligence quotient score. The reference population can depend on a continuous variable, like age. Current continuous norming methods transform the raw score into an age-dependent standardized score. Such methods have the shortcoming to solely rely on the raw test scores, ignoring valuable information from individual item responses. Instead of modeling the raw test scores, we propose modeling the item scores with a Bayesian two-parameter logistic (2PL) item response theory model with age-dependent mean and variance of the latent trait distribution, 2PL-norm for short. Norms are then derived using the estimated latent trait score and the age-dependent distribution parameters. Simulations show that 2PL-norms are overall more accurate than those from the most popular raw score-based norming methods cNORM and generalized additive models for location, scale, and shape (GAMLSS). Furthermore, the credible intervals of 2PL-norm exhibit clearly superior coverage over the confidence intervals of the raw score-based methods. The only issue of 2PL-norm is its slightly lower performance at the tails of the norms. Among the raw score-based norming methods, GAMLSS outperforms cNORM. For empirical practice this suggests the use of 2PL-norm, if the model assumptions hold. If not, or the interest is solely in the point estimates of the extreme trait positions, GAMLSS-based norming is a better alternative. The use of the 2PL-norm is illustrated and compared with GAMLSS and cNORM using empirical data, and code is provided, so that users can readily apply 2PL-norm to their normative data. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Evaluating methods for estimating influence effects in social networks: a simulation study

The contagion effect, also known as the peer effect, plays a significant role in social networks. It refers to the phenomenon in which one person or group can influence the behavior of other individuals with whom they share social connections. However, accurately estimating the contagion effect is challenging due to the confounding impact of peer selection. Previous studies have shown that this can be viewed as a problem of omitted variable bias. In an attempt to address this issue, simulation studies were conducted to compare the effectiveness of various methodologies in estimating peer influence within social networks, including latent variable models and machine learning techniques. Our research demonstrates that the performance of various approaches vary depending on the social network scenario. Overall, the latent space approach exhibits the best performance. This analysis provides valuable insights for researchers studying estimation techniques and emphasizes the importance of understanding the strengths, limitations, and objectives of these methods before using them for inference-based estimation.

MCMC stopping rules in latent variable modelling.

Bayesian analysis relies heavily on the Markov chain Monte Carlo (MCMC) algorithm to obtain random samples from posterior distributions. In this study, we compare the performance of MCMC stopping rules and provide a guideline for determining the termination point of the MCMC algorithm in latent variable models. In simulation studies, we examine the performance of four different MCMC stopping rules: potential scale reduction factor (PSRF), fixed-width stopping rule, Geweke's diagnostic, and effective sample size. Specifically, we evaluate these stopping rules in the context of the DINA model and the bifactor item response theory model, two commonly used latent variable models in educational and psychological measurement. Our simulation study findings suggest that single-chain approaches outperform multiple-chain approaches in terms of item parameter accuracy. However, when it comes to person parameter estimates, the effect of stopping rules diminishes. We caution against relying solely on the univariate PSRF, which is the most popular method, as it may terminate the algorithm prematurely and produce biased item parameter estimates if the cut-off value is not chosen carefully. Our research offers guidance to practitioners on choosing suitable stopping rules to improve the precision of the MCMC algorithm in models involving latent variables.

Weighing user's privacy calculus on personal information disclosure: the moderating effect of social media identification

PurposeWith the widespread collection and utilization of user data, privacy security has become a crucial factor influencing online engagement. In response to the growing concern about privacy security issues on social media, this research aims to examine the key causes of social media users' privacy calculus and how the balance between perceived privacy risks and benefits affects users' privacy concerns and their subsequent willingness to disclose personal information.Design/methodology/approachThe characteristics of the privacy calculus were extracted through partially structured interviews. A research model derived from privacy calculus theory was constructed, and latent variable modeling was employed to validate the proposed hypotheses.FindingsInformation sensitivity, experiences of privacy violations, social influence and the effectiveness of privacy policies influence users' privacy calculus. Privacy risk positively influences privacy concerns. Personal information disclosure willingness is positively influenced by privacy benefits and negatively influenced by privacy concerns, with both paths moderated by social media identification.Originality/valueThis study explores the key antecedents of users' privacy calculus and how these factors influence privacy concerns and subsequent willingness to disclose information on social media. It offers new insights into the privacy paradox observed within social media by validating the moderating role of social media identification on users' information disclosure willingness.